282 56 11MB
English Pages [615] Year 2019
E ntr e p r e n e u r ial Fi nan ce
This page intentionally left blank
Entrepreneurial Finance Venture Capital, Deal Structure & Valuation, Second Edition
Janet Kiholm Smith and Richard L. Smith
Stanford Business Books
An imprint of Stanford University Press Stanford, California
stanford university press
Stanford, California © 2011, 2019 by the Board of Trustees of the Leland Stanford Junior University. All rights reserved. No part of this book may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying and recording, or in any information storage or retrieval system without the prior written permission of Stanford University Press. Printed in the United States of America on acid-free, archival-quality paper Library of Congress Cataloging-in-Publication Data Names: Smith, Janet Kiholm, author. | Smith, Richard L., author. Title: Entrepreneurial finance : venture capital, deal structure & valuation / Janet Kiholm Smith and Richard L. Smith. Description: Second edition. | Stanford, California : Stanford Business Books, an imprint of Stanford University Press, 2019. | Includes bibliographical references and index. Identifiers: LCCN 2018060867 (print) | LCCN 2019001437 (ebook) | ISBN 9781503609129 (e-book) | ISBN 9781503603219 (cloth : alk. paper) Subjects: LCSH: Venture capital. | New business enterprises—Finance. | Entrepreneurship. Classification: LCC HG4751 (ebook) | LCC HG4751 .S597 2019 (print) | DDC 658.15/224—dc23 LC record available athttps://lccn.loc.gov/2018060867 Cover design by Kevin Barrett Kane Cover photograph by Anna Om Typeset by Newgen in Times New Roman MT Std 10/14
Contents
List of Illustrations xiii Abbreviations xix Preface xxiii Acknowledgments xxix About the Authors xxxi
Part 1 Getting Started Chapter 1 Introduction 3 1.1 What Makes Entrepreneurial Finance Different from Corporate Finance? 3 1.2
Entrepreneurship and the Entrepreneur 9
1.3
Hypothesis-Driven Entrepreneurship 17
1.4
The Stages of New Venture Development 23
1.5 Financial Performance and Stages of New Venture Development 25 1.6
The New Venture Business Model 28
1.7
Summary 30
Review Questions 31
Notes 32
References and Additional Reading 33
vi
Contents
Chapter 2 New Venture Financing: Considerations and Choices 36 2.1
The Sequence of New Venture Financing 37
2.2
Sources of New Venture Financing 39
2.3 What’s Different About Financing Not-for-Profit Ventures? 59 2.4
Organizational Form and Financing Choices 61
2.5
Regulatory Considerations 62
2.6
International Differences in Financing Options 66
2.7
Recap: Considerations in the Choice of Financing 67
2.8
How Financial Distress Affects Financing Choices 71
2.9
Summary 73
Review Questions 74
Notes 74
References and Additional Reading 77
Part 2 Financing of High-Risk, High-Growth Ventures Chapter 3 Venture Capital and Angel Investing 83 3.1
Development of the Venture Capital Market 84
3.2
The Organization of Venture Capital Firms 94
3.3
Investment Returns and Compensation 103
3.4
Impact of Compensation on Investment Selection 107
3.5
Aspects of the VC Industry Structure 108
3.6
How Venture Capitalists Can Add Value 110
3.7 Luck Versus Skill: What Accounts for Venture Capital Success? 112 3.8
The Role of Reputation in the Venture Capital Market 113
3.9
Angel Investing 114
3.10
Summary 118
Review Questions 119
Notes 119
References and Additional Reading 122
Contents vii
Chapter 4 Venture Deals 126 4.1
The Economic Framework for Financial Contracting 127
4.2
Essentials of Contract Design 132
4.3
Elements of VC Deal Structure 138
4.4
Analysis of Key Term Sheet Provisions 149
4.5
Deal Structures of Angel Investments 156
4.6
Summary 159
Review Questions 161
Notes 162
References and Additional Reading 163
Part 3 Financial Aspects of Strategic Planning Chapter 5 New Venture Strategy and Real Options 169 5.1
Product-Market, Financial, and Organizational Strategy 169
5.2
The Interdependence of Strategic Choices: An Example 171
5.3
What Makes a Plan or Decision Strategic? 172
5.4
Financial Strategy 172
5.5
Deciding on the Objective 173
5.6
Strategic Planning for New Ventures 175
5.7
Recognizing Real Options 177
5.8
Strategic Planning and Decision Trees 182
5.9
Decision Trees and Contract Negotiation 191
5.10
Rival Reactions and Game Trees 192
5.11
Real Options with Continuous Distributions 197
5.12
Summary 199
Review Questions 199
Notes 200
References and Additional Reading 201
Chapter 6 Developing Venture Strategy Using Simulation 204 6.1
Use of Simulation in Business Planning: An Example 205
viii
Contents
6.2
Who Relies on Simulation? 207
6.3
Simulation in New Venture Finance 207
6.4
Simulation: An Illustration 209
6.5
Simulating the Value of a Financial Option 214
6.6
Describing Risk 216
6.7
Using Simulation to Evaluate a Strategy 219
6.8
Valuing Real Options and Comparing Strategic Choices 228
6.9
Summary 239
Review Questions 239
Notes 240
References and Additional Reading 242
Part 4 Financial Forecasting and Assessing Financial Needs Chapter 7 Revenue Forecasting 245 7.1
Principles of Financial Forecasting 246
7.2
Forecasting Revenue 247
7.3
Estimating Uncertainty 254
7.4 Building a New Venture Revenue Forecast: An Illustration 256 7.5 Introducing Uncertainty to the Forecast: Continuing the Illustration 259 7.6 Calibrating the Development Timing Assumption: An Example 267 7.7
Summary 268
Review Questions 269
Notes 270
References and Additional Reading 270
Chapter 8 Financial Modeling 271 8.1
An Overview of Financial Statements 272
8.2
Working Capital, Growth, and Financial Needs 277
8.3
Developing Assumptions for the Financial Model 283
Contents ix
8.4
Building a Financial Model of the Venture 291
8.5
Adding Uncertainty to the Financial Model 303
8.6
NewCo: Building an Integrated Financial Model 308
8.7
Summary 318
Review Questions 319
Notes 320
References and Additional Reading 320
Chapter 9 Assessing Cash Needs 321 9.1
Cash Flow Breakeven Analysis 322
9.2
Sustainable Growth 327
9.3 Planning for Financial Needs When the Desired Growth Rate Exceeds the Sustainable Rate 332 9.4
Planning for Product-Market Uncertainty 334
9.5 Assessing Financial Needs with Sensitivity/Scenario Analysis 337 9.6
Assessing Financial Needs with Simulation 341
9.7
Summary 348
Review Questions 348
Notes 349
References and Additional Reading 350
Part 5 Valuation Chapter 10 Foundations of New Venture Valuation 353 10.1
Perspectives on the Valuation of New Ventures 354
10.2
Myths About New Venture Valuation 355
10.3
An Overview of Valuation Methods 359
10.4
Discounted Cash Flow Valuation 363
10.5
The Relative Value Method 374
10.6
Valuation by the Venture Capital Method 378
10.7
Valuation by the First Chicago Method 380
x
Contents
10.8
Reconciliation with the Pricing of Options 381
10.9
Required Rates of Return for Investing in New Ventures 384
10.10
The Entrepreneur’s Opportunity Cost of Capital 386
10.11
Matching Cash Flows and Discount Rates 392
10.12
Summary 396
Review Questions 398
Notes 398
References and Additional Reading 401
Chapter 11 New Venture Valuation in Practice 405 11.1
Criteria for Selecting a New Venture Valuation Method 405
11.2
Implementing the Continuing Value Concept 407
11.3
Implementing DCF Valuation Methods 413
11.4
New Venture Valuation: An Illustration 428
11.5
The Cost of Capital for Non-U.S. Investors 442
11.6
Some Practical Caveats on Implementation 443
11.7
Summary 444
Review Questions 445
Notes 445
References and Additional Reading 446
Chapter 11 Appendix: The Entrepreneur’s Perspective on Value 448 11A.1
Estimating the Entrepreneur’s Commitment to a Venture 448
11A.2
Valuing Partial-Commitment Investments 453
11A.3
Implementation: Partial Commitment 455
11A.4
Shortcuts and Extensions 459
11A.5
Summary 464
References and Additional Reading 465
Chapter 12 Designing and Valuing Staged Investment with Real Options 467 12.1
Staged Investment: The Venture Capital Method 467
12.2 Staged Investment: CAPM Valuation with Discrete Scenarios 474
Contents xi
12.3
Using Simulation to Design Financial Contracts 482
12.4
Valuing Different Types of Financial Claims 493
12.5
Summary 494
Review Questions 495
Notes 495
References and Additional Reading 496
Part 6 Harvesting and Beyond Chapter 13 Harvesting 501 13.1
Going Public 502
13.2
Acquisition 519
13.3
Valuing Private Transactions 522
13.4
Management Buyout 526
13.5
Employee Stock Ownership Plans 527
13.6
Roll-Up IPO 531
13.7
The Choice of Harvesting Method 532
13.8
Harvesting Choices of VC-Backed Firms 536
13.9
Summary 538
Review Questions 538
Notes 539
References and Additional Reading 545
Chapter 14 The Future of Entrepreneurial Finance 551 14.1
Completing the Circle 552
14.2 Public Policy and Entrepreneurial Activity: An International Comparison 557 14.3
Breaking New Ground 569
Notes 574
References and Additional Reading 576
Index 579
This page intentionally left blank
I l l u s t r at i o n s
Figures
1.1 1.2 1.3 1.4 1.5 1.6 2.1 2.2 2.3 2.4 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 3.10 4.1
Survival rates of new ventures 11 Private sector establishment births and deaths 12 Entrepreneurial involvement and motivation by country 14 Global differences in supportiveness for starting and doing business 15 Stages of new venture development 24 Financial performance and stages of new venture development 26 Sources of new venture financing 38 New VC investments by sector 46 Top 10 states for VC investing: 2016 46 Number of independent VC deals and deals with CVC involvement 49 New commitments of VC in the U.S.: 1969–2017 86 VC investments in the U.S.: 1970–2017 89 European new VC commitments: 2008–2017 90 New VC investments in Europe as percentages of country GDP 91 U.S. VC investments by stage of development: 1980–2017 92 Corporate VC involvement in U.S. VC deals: 2007–2017 93 Organizational structure of VC funds 95 The VC investment process 97 Net returns by quarter to limited partners of VC funds: 1981–2016 104 Depiction of a VC waterfall provision 105 Liquidation preference: $1MM investment with a 3X preference or convertible to 20% of common equity 150
xiii
xiv
Illustrations
4.2 Liquidation preference: $1MM investment with a 3X preference or convertible to 20% of common stock, versus an alternative contract of 25% of common stock 151 4.3 Comparison of four alternative terms for $1MM investment: 3X liquidation preference or convertible to 20% of common stock; 1X preference and full participation with 20% ownership; 2X liquidation preference with 20% participation up to a 4X cap; and 25% of common stock 152 5.1 Financial implications of product-market and organizational strategic choices 176 5.2 Expiration date values of call and put options on Spacely.com stock 178 5.3 Decision tree for accept/reject decision to invest in the venture 185 5.4 Decision tree for investing in a venture, with the option to delay investing until uncertainty about market demand is resolved 187 5.5 Decision tree for investing in a facility, with the option to expand the initial investment 189 5.6 Entry decision game tree 195 6.1 Venture NPV estimated based on the expected value of each input 210 6.2 Venture NPV estimated by simulation using the distribution of each input 212 6.3 Simulation model of stock price and put and call option expiration values 215 6.4 Simulation of 12 months of stock price performance 215 6.5 Illustrations of @RISK statistical distributions 218 6.6 Assumptions and statistical processes of the large-facility model 223 6.7 Unconditional simulation results 224 6.8 Distribution of market size estimates generated by simulation 226 6.9 Histogram of unit sales simulation results 227 6.10 Convergence of the entrepreneur’s average NPV 227 6.11 Small facility: NPV to entrepreneur 233 6.12 Large facility overlaid with small: NPV to entrepreneur 234 6.13 Subtree for venture learning option with simulated uncertainty 235 6.14 Subtree for venture expansion option with simulated uncertainty 237 7.1 NewCo revenue assumptions 257 7.2 NewCo revenue forecast 258 7.3 NewCo revenue simulation assumptions 263 7.4 Discrete probability approach to simulating development timing 264 7.5 NewCo simulated distribution of the start of revenue generation 266 7.6 NewCo revenue forecast: Sample trial results 267 8.1 Working capital policy template 279 8.2 Morebucks simulation of profitability and surplus cash 307
Illustrations xv
Tables
8.3 NewCo integrated financial model assumptions 309 8.4 NewCo expected financial performance 314 9.1 Net income and cash flow breakeven point, cash needs assessment, and investment value 324 9.2 Sustainable growth model template 328 9.3 Dynamic sustainable growth for an early-stage venture 331 9.4 iFree scenario analysis 340 9.5 NewCo total cash need as of quarter end 342 9.6 NewCo NPV results from simulation 344 9.7 NewCo simulation results for development completion timing (based on 1,000 trials) 344 10.1 Distribution of VC fund average annual IRRs (vintage years 1969–2013) 355 10.2 VC returns to limited partners by vintage year 357 10.3 How portfolio risk depends on the number of assets in the portfolio 368 10.4 The Capital Asset Pricing Model (CAPM) 369 11.1 Using continuing value to estimate the value of a new venture 407 11.2 Historical price/earnings ratios of the S&P 500 Index 412 11.3 Using arithmetic or geometric average returns 417 12.1 Staged investment decision model 477 13.1 IPO issue pricing process for a firm commitment underwriting 505 13.2 Average IPO underpricing by year (measured as initial return) 508 13.3 Average IPO underwriter spread by year 509 13.4 Structure of a private leveraged ESOP 529 13.5 Roll-up IPO 531 13.6 Numbers of exits of VC-backed firms via IPO and M&A and the NASDAQ Index by quarter 537 14.1 Early-stage entrepreneurial activity and GDP per capita 558 14.2 Gross expenditures on R&D by country (percentage of GDP) 559 14.3 Early-stage entrepreneurial activity for factor-driven, efficiency-driven, and innovation-driven countries 560 2.1 U.S. angel and VC investments: 2006–2016 44 2.2 Key U.S. securities market regulations for entrepreneurs and investors 63 3.1 Summary of terms in a VC limited partnership private placement memorandum (PPM) 99 3.2 Key covenants in a VC limited partnership agreement (LPA) 102 4.1 Standard provisions of a VC term sheet for convertible preferred investment 140 4.2 Simple capitalization table 148
xvi
Illustrations
4.3 Comparison of antidilution provisions: Full ratchet and weighted average ratchet 155 5.1 Examples of real options 181 7.1 Yardstick company data 251 8.1 Income statement (year ended 12/31/2018) 273 8.2 Balance sheet (year ended 12/31/2018) 275 8.3 Cash flow statement (year ended 12/31/2018) 276 8.4 Key business ratios for eating and drinking establishments 285 8.5 Coffee People, Inc. financials prior to IPO 287 8.6 Coffee People, Inc. common size statements 288 8.7 Coffee People, Inc. financial ratios 289 8.8 Step 1 of pro forma financial model for Morebucks: Income statement assumptions 292 8.9 Step 2 of pro forma financial model for Morebucks: Balance sheet assumptions 296 8.10 Step 3 of pro forma financial model for Morebucks: Investment assumption 300 8.11 Pro forma financial model for Morebucks with simulation 304 8.12 Morebucks: Summary of simulation results (based on 1,000 trials) 306 8.13 NewCo pro forma financial statements 311 8.14 NewCo simulation assumptions 316 9.1 NewCo cumulative NPV by first quarter of revenue generation 345 9.2 NewCo cumulative NPV by first quarter of revenue and revenue growth rate 347 10.1 Measures of expected cash flow 393 10.2 Matching cash flows to discount rates for various financial claims 395 11.1 Historical stock and bond returns 415 11.2 Average beta estimates for selected sectors 419 11.3 Beta estimates and market correlations for newly public, VC-backed firms 422 11.4 Standard deviation of market returns (S&P 500) 423 11.5 Valuation Template 1: Valuation by the RADR method based on discrete scenario cash flow forecast 431 11.6 Valuation Template 2: Valuation by the CEQ method based on discrete scenario cash flow forecast 433 11.7 Valuation at various discount rates by the Venture Capital Method 439 11A.1 Valuation Template 3: Present value of the entrepreneur’s wealth and commitment 453 11A.2 New venture simulation model ($000) 457 11A.3 Valuation Template 4: Partial commitment of the entrepreneur 458
Illustrations xvii
11A.4 Valuation Template 5: Entrepreneur’s cost of capital 463 11A.5 Underdiversification premium to required rate of return 465 12.1 Valuation Template 6: Single-stage investment—Venture Capital Method 468 12.2 Valuation Template 7: Multi-stage investment—Venture Capital Method 470 12.3 Investment and payoff assumptions 474 12.4 Venture Capital Method single-stage valuation 475 12.5 Valuation of unstaged investment: VC investor 478 12.6 Valuation of staged investment: VC investor 480 12.7 Valuation-based contracting model of VC ownership shares 481 12.8 New venture simulation with conditional second-stage investment 485 12.9 New venture simulation: Incremental effects of conditional second-stage investment 488 12.10 Summary of simulation investment and return results 489 12.11 Valuing the investor’s second-stage real option 490 12.12 Valuing the investor’s financial claim by discounting all expected cash flows 491 13.1 IPO direct costs and underwriter spread by issue size 509 13.2 Private placement discounts compared to equity market value 525
This page intentionally left blank
A b b r e v i at i o n s
A assets AP accounts payable AR accounts receivable BDC business development company BEP breakeven point BS balance sheet BV book value CAPM Capital Asset Pricing Model CEQ certainty equivalent CF cash flow CML capital market line COC cash on cash (net multiple) COGS cost of goods sold D debt or dividend per share (depending on context) D&A depreciation and amortization DCF discounted cash flow D/E debt to equity ratio D/V debt to value ratio DPO direct public offering E equity E/V equity to value ratio EBIT earnings before interest and taxes
xix
xx
Abbreviations
EBITDA earnings before interest, taxes, depreciation, and amortization EBT earnings before tax ERISA Employee Retirement Income Security Act ESOP employee stock ownership plan FDA Food and Drug Administration FY fiscal year g* sustainable growth rate GAAP generally accepted accounting principles GDP gross domestic product GEM Global Entrepreneurship Monitor GP general partner/partnership ICA Investment Company Act of 1940 ICO Initial coin offering IFRS International Financial Reporting Standards INT interest expense IPO initial public offering IRR internal rate of return IS income statement IT information technology JOBS Jump Start Our Business Startups Act of 2012 LBO leveraged buyout LP limited partner/partnership M&A merger and acquisition MBO management buyout MV market value NAICS North American Industry Classification System NAV net asset value NI net income NPV net present value NVCA National Venture Capital Association NWC net working capital NYSE New York Stock Exchange OCF operating cash flow OECD Organization for Economic Cooperation and Development OEM original equipment manufacturer OPM Black-Scholes Option Pricing Model OTC over-the-counter
Abbreviations xxi
P P&L P/E PEG PP&E PV R R&D RADR ROA ROR ROS RP RV SAFE SAFT SBA SBIC SBIR SEC SEO SG&A SIC SME SML TY VC WACC
price per share profit and loss price to earnings ratio price earnings to growth property, plant, and equipment present value retention ratio (1 − dividend payout ratio) research and development risk-adjusted discount rate return on assets rate of return return on sales risk premium relative value simple agreement for future equity simple agreement for future tokens Small Business Administration Small Business Investment Company Small Business Innovation Research Securities and Exchange Commission seasoned equity offering selling, general, and administrative expenses Standard Industrial Classification small and medium-size enterprise security market line tax year venture capital/capitalist weighted average cost of capital
This page intentionally left blank
P r e fac e
H i s to ry a b o u n ds w it h examples of extraordinary entrepreneurs whose
new ideas and products have changed the world. Many people are enamored with the idea of creating new products and starting businesses. Accompanying the interest in venture creation, there is broad interest in venture capital, investment banking, and careers related to new venture financing, deal structuring, and harvesting. Our primary motivation for writing this book is to empower students and practitioners to be more successful in developing and financing their ideas. Our overriding orientation is to apply the theory and methods of finance and economics to the rapidly evolving field of entrepreneurial finance. This book is unique in several ways. First, it builds on and significantly extends the tools and methods of corporate finance and financial economics to approach the difficult and important financial problems associated with starting and growing new ventures. Building on the foundations of financial economics makes the lessons more general and memorable and the applications easier to implement in varied settings. Mastery of the framework facilitates clearer and more defensible evaluation of opportunities and choices than is possible based on heuristics and intuition. With reliable and rigorous tools, you can be more confident that the decisions you make will be the right ones. Second, while many books address aspects of entrepreneurship (writing business plans, leading and managing new ventures, etc.), this book has a unique and direct focus on the question of how new venture financing choices can add value and turn marginal opportunities into valuable ones. We emphasize value creation as the objective of each financial choice that an entrepreneur or investor makes—including recent developments in new venture contracting, staging, deal xxiii
xxiv
Preface
structuring, real options, risk, diversification, and exit. Understanding the effects of each choice has the potential to add tremendous value to ideas and innovations. Third, in contrast to other books on entrepreneurial finance, we specifically address the influences of risk and uncertainty on new venture success, with particular attention to risky ventures with significant upside potential. We use discrete scenario and simulation analysis throughout the book to evaluate alternative strategies, assess financial needs, assess risk and expected cash flows as elements of valuation, and compare different deal structures and contract terms. Fourth, because both valuation and assessment of cash needs depend on projections of cash flow, we devote significant attention to methods of forecasting the pro forma financial statements of new ventures in an integrated fashion. Integrated financial modeling enables the entrepreneur or investor to use financial forecasting to conduct scenario analysis and to simulate such things as how growth rates that are faster or slower than expected affect cash needs. Fifth, we provide a comprehensive survey of approaches to new venture valuation with an emphasis on applications. Our approach to valuation is more comprehensive than most because we approach valuation from a contracting perspective that is affected by the different assessments of the entrepreneur and the investor. We recognize that the entrepreneur’s unique circumstances can lead to value conclusions that are different from those of a well-diversified investor.
Why Study Entrepreneurial Finance? Whether you see yourself as an entrepreneur, a corporate financial manager dealing with new projects, an asset manager, an investor, or a social entrepreneur, a solid understanding of entrepreneurial finance can help you make better decisions. Couple this with the estimate that over 50% of new businesses fail within a few years, and the value of understanding new venture finance becomes clear. Perhaps more telling is that even if a business survives, the entrepreneur may not. In a majority of venture capital-backed start-ups, nonfounders are appointed CEO within a few years of the start of operation. How can the hazards and pitfalls of forming new ventures be avoided or mitigated? The answer is to understand the financial economic foundations and use the best available decision-making tools and methods. A new venture should not be undertaken unless the expected reward is high enough to compensate for the value of forgone opportunities. Investing personal resources and time in a venture that should never have been pursued is just as serious an error as failing to invest in a good venture. Throughout the book, we reiterate
Preface xxv
and demonstrate through examples that this trade-off of risk and return is not easy to assess intuitively. Rather, this is an area where analytical rigor can add considerable value. Even the best initial projections, however, can prove to be overly optimistic as the future unfolds. It is important to base the decision to continue or abandon a venture on the same kind of rigorous analysis that was used in making the original decision. It is all too easy to continue investing time and resources in a venture that is destined for mediocre long-run performance or to give up on a venture that has experienced a temporary setback but still offers the potential for substantial gain. It is a rare individual who is good at both recognizing an opportunity and managing the venture to capitalize on that opportunity. Careful design of the organization at the outset helps ensure that a venture does not fail just because the visionary was not well suited to manage the day-to-day operations. Careful design also can help ensure that the entrepreneur does not lose control of the venture unnecessarily.
What’s New About This Book? This book builds on our previous books, Entrepreneurial Finance (2000, 2004) and Entrepreneurial Finance: Strategy, Valuation and Deal Structure (2011). Like the earlier books, it ties the applications to the underlying disciplines of finance and economics and makes clear in what ways the study of entrepreneurial finance is distinct from corporate finance. In contrast to other books, this book focuses less on small business entrepreneurship and focuses more on issues facing high-risk ventures that have significant growth potential and are candidates for external equity capital such as angel and venture capital investing. We emphasize key considerations such as new developments in financial contracting and negotiation, hypothesis-driven entrepreneurship, adding value through deal structuring and staging of investment, the choice and timing of financing, and valuation of risky ventures, including valuation of real options. Coming on the heels of the JOBS Act of 2012, there have been a number of innovations in the ways new ventures are financed and the contract terms that are used. Worldwide, there has been considerable growth in both angel and venture capital investing. There is more variety in the sources of financing available to entrepreneurs, ranging, as examples, from crowdfunding, to initial coin offerings (ICOs), to mini-IPOs. Technological changes have resulted in reduced costs
xxvi
Preface
of experimenting with new venture ideas, leading to an explosion of start-ups seeking early-stage funding. The result has been an increase in the prevalence of early-stage “nonpriced” financing in the form of sophisticated convertible notes and an increase in the number of funding rounds. We emphasize the usefulness of hypothesis-driven entrepreneurship by tying it to an intuitive discussion, using decision trees, of how real options can be created and exploited. The chapter on venture deals (Chapter 4) reflects the centrality of understanding how financial contacts can mitigate information costs and align incentives of the entrepreneur and investor. We examine venture contracting terms in detail (such as liquidation preferences, conversion rights, and antidilution terms) and analyze the terms of new-style convertible notes that now are commonly used by angels and some VCs in early stages. We reflect uncertainty of the real option values by combining decision trees and simulation (Chapter 5). The analysis is step-by-step in the context of specific examples. We devote significant attention to financial forecasting and the construction of integrated pro forma financial statements, as both are key to valuation, contracting, and assessment of cash needs. We break new ground in several areas by presenting material and analytical approaches that extend the frontier of academic research related to entrepreneurial finance. The book focuses on U.S. institutions but also provides international perspective by illustrating, throughout, the many similarities and the differences in analysis for U.S. ventures versus international ventures. The book concludes with a summary of major themes and a discussion of international differences in institutions and public policies aimed at encouraging innovation and entrepreneurship. We also identify open questions regarding the future of entrepreneurial finance. A number of technological changes have the potential to dramatically change the landscape for finance, including, as examples, blockchain and cryptocurrency, crowdfunding, artificial intelligence and machine learning, and internet connectivity (IoT).
Intended Audience This book can be used as a text in an advanced finance or entrepreneurship course and by scholars who are interested in research related to entrepreneurial finance. However, its intended audience is broader. The book is appropriate for students, entrepreneurs, practitioners involved with new ventures, and corporate financial managers looking for ways to encourage corporate venturing. We have designed the book for readers who are familiar with the basic concepts and tools of corporate finance, accounting, economics, and statis-
Preface xxvii
tics, and who seek a rigorous and systematic approach to adding value to new ventures through financing and deal structuring. On the companion website, we provide brief reviews for those who feel the need for a refresher on some key background concepts. The book is appropriate for graduate students, advanced undergraduates in business and economics, and executive MBA students. In our own teaching, we use the book and related materials with students at all levels. Because entrepreneurial finance builds on and integrates all areas of management, a course developed around the book can serve as a capstone integrative experience to the MBA or an undergraduate business degree. The book can be used effectively in an entrepreneurial finance course or in a course on venture capital and private equity. It is designed to be used either as a stand-alone resource or in conjunction with cases or a business planning exercise. Each chapter includes end-of-chapter review questions. The book’s website has end-of-chapter problems that are designed to give hands-on opportunities to apply the lessons of each chapter. The book can be used in a variety of different course formats: Some users like our use of simulation throughout the book. For those, we rely on readily available software (@RISK) and also provide files on the website that contain examples and problem solutions that are prepared using @RISK and also Oracle Crystal Ball, another popular software package. For those who are oriented to case method teaching, we have provided a series of our own interactive cases that correspond to the book chapters and have developed a list of commercially available cases that work well with the book. For those who see value in linking the coverage of entrepreneurial finance to a business planning exercise, the organization of the book follows the normal organization of the thought processes and financial contents of a business plan.
A Note About the Website and Internet Resources The book is designed to be used most effectively in conjunction with resources we provide on the book’s website, https://www.sup.org/entrepreneurialfinance. Much of the book relates to software, spreadsheets, templates, simulation applications, and interactive cases and tutorials that are available for download. For those teaching from the book, we also provide PowerPoint presentations by chapter. Instructor-specific resources are password protected. Instructors can gain access to teaching materials that accompany the book by contacting Stanford University Press at [email protected].
xxviii
Preface
The website contains problems and interactive cases, along with solutions, that complement the book material. All Excel figures in the book, including charts, are available as downloadable files so that the user can review the spreadsheet structure and cell formulas. When we use quantitative examples in discussion, the website normally includes a file that contains the backup spreadsheet analysis.
Simulation Because of the usefulness of simulation to evaluate risk and uncertainty, both prominent features of new venture development, we incorporate it into the book. Many readers will be familiar with commercial simulation packages such as @RISK and Oracle Crystal Ball. We primarily rely on @RISK in the book examples but provide both @RISK and Oracle Crystal Ball versions of the simulation analysis on the book’s website. When an Excel syntax is used in a simulated cell in a spreadsheet, we use the @RISK syntax. If you are a user of another commercial package you can study the parallel syntax by opening and modifying our example files.
Spreadsheets and Templates The website contains soft copies of the figures and tables in the book, including several templates that you can use to study your own new venture valuation questions (i.e., you can easily edit the template to study and value your own cash flow projections). We have provided copies of spreadsheets that are imbedded with simulation formulae as well as versions that do not require simulation.
Ac k n o w l e d g m e n t s
I n p r e pa r i n g t h e b o o k , we have benefited from numerous comments
from colleagues, students, venture capitalists, business angels, entrepreneurs, and friends. Over the years, an impressive group of reviewers and adopters have provided comments that have sharpened the presentation. These include Robin Anderson (University of Nebraska), Sanjai Bhagat (University of Colorado), Carol Billingham (Central Michigan University), Carol Marie Boyer (Clarkson University), Daniel Donoghue (Discovery Group), Fernando Fabre (Endeavor), Samuel Gray (New Mexico State University), Phil Greenwood (University of Wisconsin, Madison), Ilan Guedj (University of Texas), Thomas Hellmann (Stanford University), Glenn Hubbard (Columbia University), William C. Hudson (St. Cloud University), Steve Kaplan (University of Chicago), Cenk Karahan (Bogazici University), Jill Kickul (DePaul University), Frank Kerins (Montana State University), Sandy Klasa (University of Arizona), Kenji Kutsuna (Kobe University), Daniel McConaughy (California State University, Northridge), James Nelson (Florida State University), Bill Petty (Baylor University), Edward Rogoff (Baruch College), Chip Ruscher (University of Arizona), Bob Schwartz (Silver Fox Advisors), James Seward (University of Wisconsin, Madison), Jeffrey Sohl (University of New Hampshire), Howard Van Auken (Iowa State University), Nikhil Varaiya (San Diego State University), and Edward Williams (Rice University). We also thank our many colleagues at the Claremont Colleges and UC Riverside. Thanks also to those students in our classes who read and worked through early drafts of the book and website. These include undergraduate and MBA
xxix
xxx
Acknowledgments
students from our schools. We very much appreciate their goodwill and thoughtful suggestions. A number of practitioners, including venture capitalists, venture funds managers, entrepreneurs, and angel investors, were very generous in sharing their experiences and providing ideas for cases and other book material. Special thanks to Andy Horowitz, Richard Chino, Wayne Cantwell, Arielle Zuckerberg, Matthew Goldman, Russ Shields, John Sibert, Richard Sudek, Luann Bangsund, Luis Villalobos, John Jasper, John Kensey, Thomas Gephart, Kazuhiko Yamamoto, and Yoshi Bunya. In particular, their insights on valuation practices and deal negotiations and their willingness to challenge academic theory have added institutional richness to the book. We are grateful to our former co-author, Richard Bliss, Babson College, for many discussions and insights about venture finance and pedagogy. The team at Stanford University Press, including Steve Edward Catalano and Olivia Bartz, was a pleasure to work with. We are indebted to Margo Beth Fleming, formerly of Stanford University Press, who saw the value in developing an advanced applied book on entrepreneurial finance. Janet Kiholm Smith Claremont, California Richard L. Smith Riverside, California
About th e Autho rs
Ja n e t Kih o l m S m it h is the Von Tobel Professor of Economics and found-
ing dean of the Robert Day School of Economics and Finance, Claremont Mc Kenna College. She was formerly department chair and director of the Financial Economics Institute of CMC. She currently is director of the Center for Innovation & Entrepreneurship at CMC, which supports research, courses, and cocurricular programs related to innovation. She teaches courses on new venture finance, economics of strategy, industrial organization, and research methods. She is the author of numerous journal articles on topics ranging from IPO pricing to corporate philanthropy to managerial risk-taking behavior, including publications in finance and economics journals including Journal of Finance; Journal of Corporate Finance; Journal of Financial and Quantitative Analysis; Journal of Banking and Finance; Journal of Law and Economics; and Journal of Law, Economics, and Organization. Current research interests include the rise of private equity markets and the important role that working capital plays in financing new ventures. Smith consults for matters related to business plan advising, working capital management, contracts, and antitrust issues. She has served as a consultant for major corporations and the Federal Trade Commission on complex business litigation. Rich a r d L . S m i t h is the Philip L. Boyd Chair in Finance at the A. Gary
Anderson Graduate School of Management of the University of California, Riverside. He regularly teaches courses on new venture finance and strategic risk management. Before joining the UC Riverside faculty, Smith held the Leatherby
xxxi
xxxii
About the Authors
Chair in Entrepreneurship at Chapman University and prior to that, he was director of the Venture Finance Institute at Claremont Graduate University. He regularly teaches courses on new venture finance and strategic risk management. Smith has authored over 50 articles appearing in the Journal of Financial Economics, Journal of Finance, and Review of Financial Studies, among others. His research interests include venture capital, initial public offerings, and contracting and valuation issues related to private equity and entrepreneurial firms. He has served on the university investment committees of Arizona State University and Claremont Graduate University and is a former Chair of the Investment Advisory Council of the Arizona State Retirement System. He is a former member of Tech Coast Angels and has consulted extensively on valuation and deal structuring issues for venture capitalists, angel groups, and entrepreneurs. He has consulted or served as an expert witness on numerous cases involving financial contracting, securities litigation, valuation, and antitrust.
E ntr e p r e n e u r ial Fi nan ce
This page intentionally left blank
C h a p t e r One
I ntro du c tio n
T h o u s a n d s o f b u s i n e s s ventures are started every year. Many fail
within a short period. Of those that survive, most achieve only meager success, some achieve rates of return high enough to justify the initial investment, and a few achieve phenomenal success. What distinguishes the successes from the failures? There is no single answer. A new venture based on a good idea can fail because of poor implementation or bad luck. One that is based on a bad idea can fail despite excellent implementation. Many that survive but do not thrive should never have been undertaken. Sometimes, even when a venture is hugely successful, early financing mistakes limit the entrepreneur’s ability to share in the rewards. This is a book on financial decision making for new ventures. It provides the fundamentals for thinking analytically about whether a new venture opportunity is worth pursuing and about how to apply the tools of financial economic theory to enhance the expected value of the undertaking. Although our focus is on for-profit ventures, not-for-profit entrepreneurial ventures face similar challenges. Where there are important differences, we discuss the application of these tools to not-for-profit venturing.
1.1 What Makes Entrepreneurial Finance Different from Corporate Finance? The guiding principles of financial decision making—in both large, established companies and start-ups—can be stated succinctly: 3
4
Chapter One
• More of a good is preferred to less. • Present wealth is preferred to future wealth. • Safe assets are preferred to risky assets. These principles drive choices of resource allocation, including investment decisions and financing decisions. Investment decisions concern the acquisition of assets, which can be tangible, like a machine; intangible, like a patent; or simply an option to take some action in the future. The worth, or value, of an investment depends on its ability to generate cash flows for investors in the future and on the riskiness of those cash flows. Financing decisions concern the choice of financing (debt, equity, or some hybrid such as preferred stock), how much financing should come from investors, and how financial contracts should be structured. The range of decisions that can be approached using the finance paradigm is much broader than may be apparent at first glance. Financial considerations are applicable, for example, to the choice of organizational form (e.g., sole proprietorship, partnership, or corporation) and to the design of financial contracts to align the incentives of investors and entrepreneurs. Similarly, issues such as the choices of scale and scope of a venture can be analyzed as financial investment decisions (i.e., larger scale requires more outside financing). If the financial theories and decisions facing all ventures are similar, it is natural to wonder why entrepreneurial finance is worthy of special study—why aren’t the principles of corporate finance directly applicable in an entrepreneurial setting? After all, a basic course in corporate finance concerns investment and financing decisions of large public corporations and generally introduces valuation techniques such as discounted cash flow and cost of capital analysis. The limitation, however, is that corporate finance theory assumes away a number of issues that are of secondary importance in a large corporate setting but are critical to decision making for new ventures. These distinctions make entrepreneurial finance an intellectually challenging area worthy of special study. The focus on entrepreneurship and early-stage ventures dramatically changes the way the finance paradigm is applied. Moreover, certain techniques of entrepreneurial finance (such as thinking about investment opportunities as portfolios of real options), while they are particularly useful in a new venture setting, are also useful in the context of a large public corporation. They normally, however, do not receive much attention in corporate finance courses. We highlight 8 important differences: 1. The inseparability of new venture investment decisions from financing decisions
Introduction 5
2. The role of necessary underdiversification as a determinant of investment value 3. The extent of managerial involvement by investors in new ventures 4. The effects of substantial information problems on the firm’s ability to undertake a project 5. The role of contracting to resolve incentive and information problems in entrepreneurial ventures 6. The critical importance of real options in determining project value 7. The importance of harvesting (exit) as an aspect of new venture valuation and the investment decision 8. The distinction between maximizing value for the entrepreneur and maximizing shareholder value
Interdependence Between Investment and Financing Decisions In corporate finance, investment decisions and financing decisions are treated as independent of each other. The manager selects investments by comparing return on investment to the market interest rate for projects of equivalent risk. The manager typically does not need to convince investors of the merits of the investment and does not need to consider simultaneously how ownership of the assets will be financed or whether the firm’s shareholders prefer highdividend payouts or capital gains. For start-up businesses, however, the interdependencies between investment and financing decisions are much more complex and important. Among other things, the entrepreneur will not be able to pursue the venture without convincing an investor of the merits and will probably place a very different value on the new venture than will well-diversified investors. These differences are important because the entrepreneur cannot normally sell shares of a private venture to generate funds for current consumption (analogous to receiving a dividend). Therefore, simple adjustments to net present value (NPV) cannot be used to address the divergence of valuations between the entrepreneur and the investor. More generally, some investment choices are contingent on certain financing choices. For example, rapid growth may be possible only with substantial outside financing, whereas a large corporation may be able to finance the entire project with internally generated funds. The link between investment choices and financing choices creates complexity that does not arise in corporate finance.
6
Chapter One
Diversifiable Risk and Investment Value In corporate finance, the NPV of an investment (project) is determined by applying a discount factor to expected future cash flows.1 Corporate finance proposes that the discount factor depends only on nondiversifiable risk. But this proposition relies on the assumption that investors can diversify at low cost. Although the assumption holds for many investors in a new venture (e.g., the investors who participate in venture capital [VC] funds, wealthy angel investors, or shareholders of large lenders), it categorically does not hold for the entrepreneur. In fact, the entrepreneur often must invest a large fraction of his or her financial wealth and human capital in the venture. This difference between entrepreneurs and investors in ability to diversify results in the project value for entrepreneurs being different from the project value for investors. In corporate finance, because only nondiversifiable risk matters, the values of different financial claims are additive. At the firm level, value additivity implies that allocation of financial claims among different kinds of investors such as owners and managers does not affect the decision to accept a project. New ventures can also be considered “projects.” But for new ventures, because entrepreneurs and investors view risk differently, each ascribes a different value to the same risky asset. As a result, value additivity does not hold and the allocation of financial claims becomes important.
Managerial Involvement of Investors In public corporations, investors (stockholders and creditors) generally are passive and do not contribute managerial services. Nor do they normally have access to significant inside information. In contrast, some investors in new ventures (e.g., venture capitalists [VCs] and angel investors) frequently provide managerial and other services that can contribute to the venture’s success. Typically, these investors will have access to inside information that they gain as a result of their continuing investment in the venture and will be involved in important decisions about the venture.
Information Problems and Contract Design Separate from differences in valuation that arise from underdiversification (differences even when the entrepreneur and investors agree about the expected future cash flows and know that they agree), differences in value can arise because of information problems between the parties. Although information gaps also exist between insiders and outsiders of public corporations, the gaps need not materially affect the investment decisions of public corpora-
Introduction 7
tion m anagers. Public corporations generally can, and often do, make investment decisions without much immediate regard to the question of how investors perceive the value of the investment. Generally, corporate managers need not convince investors, lenders, or employees that a project is worth undertaking, at least in the short run. The situation is very different in the case of a start-up business that requires outside financing. In the latter case, investors look specifically to the venture to provide a return on their investments. Moreover, there is often no easy way for the entrepreneur to communicate her true beliefs about the potential for success of a new venture. From a financial perspective, this places considerable emphasis on finding ways to signal the entrepreneur’s confidence in the venture.
Incentive Alignment and Contract Design Incentive contracting clearly plays a role in the large public corporation. On the positive side, managerial stock options and performance bonuses are intended to align the interests of managers and investors. On the negative side, debt covenants and similar provisions are designed to discourage reliance on risky debt financing that can lead to inefficient investment decisions and other agency costs associated with heavy reliance on debt. The issues are similar for start-up businesses, but reliance on incentive contracts is, in some respects, more compelling. In contrast to the managers of public corporations, investors generally keep the entrepreneur on a short leash. There are compelling reasons to find ways to invest in the projects of unproven entrepreneurs. An investor who is good at identifying untested entrepreneurs who are likely to succeed can profitably participate in new ventures that would have been rejected by a less astute investor. The result is that investors use a variety of contractual devices to supplement their ability to identify and motivate high-quality entrepreneurs.
The Importance of Real Options Students and practitioners of corporate finance know to value projects by discounting expected future cash flows back to NPV. Even in the corporate setting, this approach is oversimplified, except with respect to the most basic independent investment projects. The more difficult valuation challenges are those for projects that include important real options. A real option is the right to undertake certain business initiatives, such as deferring, abandoning, or expanding a capital investment project after the initial investment has been made. In contrast to a financial option, such as a call or put on a share of stock, the underlying asset is tangible and the option reflects a change in how
8
Chapter One
the real asset is used. The reality is that most investing involves a process of acquiring, retaining, exercising, and abandoning real options. Nonetheless, the common practice of ignoring real options in corporate investment decisions suggests that they often are of secondary importance to the decision. The values of real options associated with an investment depend on the degree of uncertainty surrounding the investment. For projects such as an investment in research and development or an investment in a new industry, uncertainty levels are likely very high. This uncertainty adds to the importance of considering embedded option values. Nowhere is the importance of option values more central to investment decision making than for a start-up business. Staging of capital infusions, abandonment of the project, growth rate acceleration, and a variety of other choices all involve real options and contribute to the need for a process of investment decision making that focuses on recognizing and valuing the real options that are associated with the project.
Harvesting the Investment In corporate finance, investment opportunities are evaluated based on their capacity to generate free cash flow for the corporation. The investment decision does not depend on when the cash flows are distributed to investors, except that corporations generally will not retain cash that they cannot invest profitably. In their decisions to invest in the shares of a public corporation with liquid shares, investors normally are not focused on when they will sell or on the anticipated valuation at the time of sale or on the costs associated with selling. Investing in new ventures is different. New venture investments normally are not liquid and often do not generate any significant free cash flow for several years. Most investors in new ventures, and many entrepreneurs, have finite investment horizons. To realize returns on their investments, a public offering of equity or another kind of liquidity event must occur (e.g., an acquisition of the venture for cash or freely tradable shares of the acquirer). Such harvesting opportunities are among the main ways investors in new ventures realize returns on their investments. Because of the importance of liquidity events, they generally are forecasted explicitly. The forecasts are formally factored into valuation of the investment.
Value to the Entrepreneur The final difference between start-ups and public corporations is the focus on the entrepreneur. In the public corporation, the focus of decision making is on investment returns to shareholders. In a start-up, the true residual claimant is
Introduction 9
the entrepreneur. In the corporate setting, maximum shareholder value is the most frequently espoused financial objective. In contrast, the objective of the entrepreneur in deciding whether to pursue the venture and how to structure the financing is to maximize the value of the financial claims and other benefits that the entrepreneur is able to retain as the business grows. It is easy to envision cases in which an objective of maximizing share value would not be in the entrepreneur’s best interest. This is particularly true if the entrepreneur is unable to convince investors of the true value of the project and would therefore have to give up too large a fraction of ownership, or if the entrepreneur values other considerations besides share value. While the principles in this book are often developed and presented from the perspective of the entrepreneur (by structuring the investor’s claim to have an NPV of zero), they are no less valuable for investors. The analysis can easily be reformulated around the objective of maximizing NPV for an investor, such as a VC or angel investor. With regard to valuation and contracting, because both parties can benefit from knowing how the other views the venture, we study these topics from both perspectives.
1.2 Entrepreneurship and the Entrepreneur The term “entrepreneur” is of French origin. Its literal translation is simply “undertaker,” in the sense of one who undertakes to do something. In the early 1700s, the English banker Richard Cantillon coined the use of the word in a managerial context. He emphasized the notion of the entrepreneur as a bearer of risk, particularly with respect to provision of capital. This early usage, however, does not adequately characterize our current understanding of what it means to be an entrepreneur. In the early 1800s, the French economist J. B. Say described the entrepreneur as a person who seeks to shift economic resources from areas of low productivity to areas of high productivity. Although Say’s notion points us in a useful direction, it is too general. Most purposeful human activity can be described as shifting economic resources to higher-valued uses (or at least attempting to do so).2 The contributions of Cantillon and Say gained renewed attention in the early 1900s through the writings of two other economists. Joseph Schumpeter (1912) viewed the entrepreneur as actively seeking opportunities to innovate. In his view, the entrepreneur is the driver of economic progress, continually seeking to disturb the status quo in a quest for profits from deliberate and risky efforts to combine society’s resources in new and valuable ways. Current use of the term
10
Chapter One
“entrepreneurship” derives from these views and from more recent thinking by management scholars such as Peter Drucker. Drucker, who was a personal friend of Schumpeter, describes entrepreneurs as individuals who “create something new, something different; they change or transmute values.”3 Today, entrepreneurship is most often described as the pursuit of opportunities to combine and redeploy resources, without regard to current ownership or control of the resources. This notion clearly draws on the definition offered by Schumpeter but adds structure by recognizing that the entrepreneur is not constrained by current control of resources. Thinking of entrepreneurship in this way suggests a multidimensional process. The entrepreneur must: • Perceive an opportunity to create value by redeploying society’s resources • Devise a strategy for marshaling control of the necessary resources • Implement a plan of action to bring about the change • Harvest the rewards that accrue from the innovation This definition is broad enough to encompass entrepreneurship that arises in the for-profit sector, including extant corporations, as well as in the not-forprofit sector, including in universities and charitable foundations. The first step is to “perceive an opportunity to create value.” What are the characteristics of a good entrepreneurial opportunity? In general, the opportunity should: • Address a need or solve a problem for the customer • Identify a new product, service, process, or market • Demonstrate a sustainable competitive advantage • Offer the potential to be profitable and create value for the parties involved Our focus in this book is on assessing the last point, but the other factors are integral to the venture’s likelihood of success and to value creation for the entrepreneur.
Survival and Failure Rates of New Businesses To be successful, an entrepreneur needs to maintain a clear focus on how strategic choices and implementation decisions are likely to affect rewards. While this may increase the likelihood of success, many new ventures do not succeed and all entrepreneurial activity takes place in a dynamic economic environment.
Introduction 11 Fi g u r e 1 .1 Survival rates of new ventures https://w ww.bls .gov/b dm/entrepreneur ship/b dm_ chart3.htm.
100% 90%
The figure shows survival rates of new business establishments that were initiated in 1995, 2000, 2005, and 2010.
Percentage of ventures surviving
source:
80%
1995
70%
2000 2005
60%
2010
50% 40% 30% 20% 10% 0%
1
2
3
4
5
6
7
8
9 10 11 12 13 14 15 16 17 18 19 20 21 Years since starting
Figure 1.1 shows the survival rates of new ventures from a U.S. Department of Labor longitudinal study of business ventures that were launched from 1994 through 2016. The figure provides data for ventures started in 1995, 2000, 2005, and 2010. Based on the data, 50% of ventures survive for at least 5 years and about 35% survive for at least 10 years. There is almost no difference in survival rates across the four starting years in spite of the fact that they encompass varied periods of economic growth. Survival cannot be equated to success, nor does nonsurvival mean failure. In fact, in a Small Business Administration (SBA) study based on data compiled by the Census Bureau, one third of the entrepreneurs of businesses that did not survive reported that they considered the venture a success.4 Among other possibilities of successful closure, nonsurviving businesses may have been established to take advantage of transitory opportunities, may have been closed in one location and reopened in another, or may have been acquired. This would imply a “failure” rate significantly lower than suggested by the nonsurvival data in Figure 1.1. Although survival rates are consistent over time, the pace of business creation and termination can vary significantly with economic conditions. Figure 1.2 shows data on private sector establishment births and deaths between 2006 and 2017. Over the entire period, there were approximately 9.0 million firm births and 8.5 million deaths, for a net creation of about 500,000 additional establishments. Looking at subperiods reveals a different story. In the two years leading up to the recession of 2008, there was a net increase of approximately 175,000 new
12
Chapter One
Fi g u r e 1 . 2
260
Private sector establishment births and deaths
240 230 220 210 200 190 180
Births Deaths
170 March 2006 August 2006 January 2007 June 2007 November 2007 April 2008 September 2008 February 2009 July 2009 December 2009 May 2010 October 2010 March 2011 August 2011 January 2012 June 2012 November 2012 April 2013 September 2013 February 2014 July 2014 December 2014 May 2015 October 2015 March 2016 August 2016 January 2017 June 2017
The figure shows the number in thousands of births and deaths of establishments between 2006 and 2017.
Number of firms, thousands
source: https://w ww.bls .gov/news.release/c ewbd .t08.htm.
250
businesses. For the following two years, 2008 and 2009, deaths exceeded births by 212,000. As the economy began to pick up steam, births again surpassed deaths. For the last eight quarters of available data, there was a net creation of over 200,000 new establishments. The dynamism of the new venture landscape is expected. Entrepreneurs constantly come up with new and innovative ideas, but the pursuit and successful execution of the ideas is fraught with uncertainty. In addition, as firms evolve, decisions to merge, sell, or abandon the venture are a normal part of any evolving economy and, in fact, are necessary if assets are going to be consistently put to highest and best use.5
Types of Entrepreneurship Replicative versus innovative. There is a useful distinction between “replicative” and “innovative” entrepreneurship. Schumpeter wrote extensively about innovative entrepreneurs, who “act as destabilizing influences triggering ‘creative destruction’—the simultaneous creation of new industries through innovation and elimination of sectors of prior economies.” Innovative entrepreneurship has the potential to add huge value to economies and is often highrisk/high-return. Amazon.com, Netflix, Uber, and Airbnb were all founded by innovative entrepreneurs who challenged prevailing business models. William
Introduction 13
Baumol, a leading researcher in the area, calls them “bold and imaginative deviators from established business patterns and practices” (2002). Innovative entrepreneurship is the type most associated with Silicon Valley. Replicative entrepreneurs, on the other hand, function as efficient coordinators of resources. They start and maintain businesses that mimic predecessors. As population grows, the economy must provide more goods and services. Economies need more grocery stores, home improvement stores, dry cleaners, and donut shops. Many of these needs are filled by replicative entrepreneurs. Replicative businesses often stay small and don’t export their products or services outside the geographic area they serve. Many communities and local economies try to encourage this type of entrepreneurship and do quite well, as indicated by the numbers of new businesses started and people employed by them. Opportunity-based versus necessity-based. A related distinction is between “opportunity-based” and “necessity-based” entrepreneurship, terms coined by the Global Entrepreneurship Monitor (GEM) consortium.6 Innovative entrepreneurship is virtually all opportunity based, whereas replicative entrepreneurship is divided between opportunity and necessity. Necessitybased entrepreneurship represents people driven to entrepreneurship by lack of alternatives. Necessity-based ventures may be simple businesses, including microbusinesses that require negligible capital, have no employees other than the entrepreneur, and produce near-subsistence-level earnings. Figure 1.3 shows results of the 2017–2018 GEM survey, which assesses the overall level of entrepreneurial activity and the motivation behind it by country. Because opportunities vary depending on economic circumstances, perceptions of what constitutes opportunity-based entrepreneurship may vary across countries. There are some overall trends in Figure 1.3. The highest overall levels of engagement in entrepreneurship generally are found in countries that are less well-developed, such as some countries in Latin America and Africa. These countries also tend to report relatively high levels of necessitydriven entrepreneurship and are likely to view basic replicative activity as opportunity-driven. In contrast, many European countries and other countries that have focused economic development around large well-established businesses report low levels of total entrepreneurial activity and also low levels of necessity-based entrepreneurship. Countries such as the U.K., Ireland, and Luxembourg have overall rates of entrepreneurial activity that are close to the average across all reporting countries. In these countries, entrepreneurship is largely opportunity-driven and the levels of necessity-based entrepreneurship are low.
14
Chapter One
Fi g u r e 1 . 3
30
Entrepreneurial involvement and motivation by country Based on data from Global Entrepreneurship Monitor, 2017–2018; available at http://w ww .gemconsortium.org/d ata.
Necessity 25 Percentage of working-age population
20
15
10
5
0
Ecuador Guatemala Peru Lebanon Chile Vietnam Madagascar Malaysia Thailand Brazil Estonia Canada Colombia Panama Uruguay Latvia Mexico United States Egypt Iran Korea Israel Australia Slovakia Saudi Arabia Kazakhstan South Africa Puerto Rico China Netherlands India Luxembourg United Arab Emirates Croatia Ireland Poland Morocco Taiwan Switzerland United Kingdom Indonesia Qatar Cyprus Sweden Slovenia Spain Argentina Germany Greece Japan Italy Bosnia and Herzegovina France Bulgaria
The figure shows the percentage of a country’s workforce that is involved in entrepreneurial ventures. The workforce is defined as the working population between ages 18 and 64. Entrepreneurial involvement is classified as opportunity based or necessity based. Data are compiled by GEM consortium members through surveys of their local economies.
Opportunity Unknown
source:
Figure 1.3 shows clearly that entrepreneurial activity occurs everywhere, including once-closed economies like China and some of the world’s leastdeveloped countries. Competition for ideas and for financing has increased dramatically in recent decades. Several developments have contributed to the global growth of entrepreneurship. The personal computer, the wireless phone, and the Internet allow investors to reach markets that were inaccessible a few years ago. The importance of entrepreneurship in creating jobs and dynamic economies is now recognized by leading international entities like the European Union (EU), the United Nations (UN), and the World Bank, all of which sponsor initiatives and provide financial support for entrepreneurship. Governments worldwide are enacting policies and providing financial subsidies that they hope will encourage entrepreneurship. The World Bank ranks countries according to the supportiveness of their environments for starting and doing business, and this information provides additional context for the level and type of entrepreneurial activity. Figure 1.4 shows the most recent rankings of large-population countries. Ease of doing business (plotted in the figure) is based on 10 factors, including availability of financing, legal environment, and availability of employees. Ease of starting a business is one of the 10 factors; we plot it separately in Figure 1.4 because it is the factor most closely related to stand-alone entrepreneurial activity.
Introduction 15
80
Ease of doing business Ease of starting a business 70
Rank —Highest number is best (Max = 74)
60
50
40
30
20
0
New Zealand Canada Hong Kong SAR, China Singapore Australia Ireland Korea, Rep. Estonia Sweden United Kingdom Taiwan, China Belgium Norway Netherlands France Finland Lithuania Russian Federation Oman Serbia Denmark Thailand Israel Greece Panama Portugal United States United Arab Emirates Ukraine Iceland Uruguay Chile Italy Luxembourg Switzerland Hungary Turkey Czech Republic Spain Croatia Qatar Mexico China Colombia Iran, Islamic Rep. Zambia Egypt, Arab Rep. Jordan Japan Malaysia Germany Peru Kenya Austria Poland Vietnam Costa Rica Saudi Arabia South Africa El Salvador Pakistan Iraq India Argentina Yemen, Rep. Uganda Comoros Ecuador West Bank and Gaza Brazil Zimbabwe Somalia Haiti Venezuela, RB
10
Fi g u r e 1 . 4 Global difference in supportiveness for starting and doing business source:
http://w ww.doingbusiness.org/rankings.
The figure shows country ranks (highest being best) for ease of starting a business and ease of doing business based on ten equal-weighted factors that are assessed by the World Bank based on midyear 2017 data.
Countries that rank high on both dimensions, such as the first 4 in the figure, tend to be those that can support high-growth start-ups. Those that are high for ease of doing business but low for ease of starting a business, such as Germany, Japan, and Mexico, tend to be countries where high-growth entrepreneurial activity is conducted mainly through established corporations and business groups. Those where ease of starting a business is high but ease of doing business is low, such as Russia, Ukraine, and Iran, tend to be relatively unregulated environments with weak infrastructures. Those where both are low, such as Haiti, Somalia, and Venezuela, tend to be turbulent environments with political unrest and dictatorships or militaristic factions.
Corporate Venturing The international evidence highlights the point that in certain countries much entrepreneurial activity occurs within established businesses or in strategic partnerships among established businesses. Corporate venturing is
16
Chapter One
articularly common for ventures that require large and complex research p teams and use of generic testing equipment and where development times are long. For example, large pharmaceutical firms almost exclusively pursue pharmaceutical innovation. Similarly, introductions of new large commercial aircraft are normally pursued as strategic partnerships involving airframe manufacturers, engine manufacturers, and avionics manufacturers. Corporate venturing, while easier in some respects, is more challenging in others. One obvious difficulty is that of designing appropriate incentives to motivate entrepreneurial effort in large organizations without creating perceived equity imbalances among employees. Sometimes the venturing activities are so central to the core business that they are pursued as part of the company’s overall research and development (R&D) effort. Other times, corporations create separate entities that are wholly owned subsidiaries and operate much like independent VC firms.
Social Venturing There is considerable confusion surrounding the term “social venturing.” Most entrepreneurs and most VC and angel investors believe that what they are doing is beneficial to society—that they are “doing good” in the pursuit of “doing well.” Further, some aspects of the endeavors of many for-profit enterprises are explicitly viewed as intended to have some sort of social impact. Patagonia, for example, designates a small fraction of its revenues for environmental charities and promotes environmentally friendly clothing manufacturing, and TOMS Shoes provides free shoes to people living in poverty. Both of these companies are for-profit enterprises that have been certified as B Corporations by B Lab, which is a not-for-profit organization that was established in the U.S. to promote socially responsible business practices to for-profit corporations. The B Corporation designation has no separate legal standing in the U.S., but is recognized in most states. The B Corporation designation indicates that a venture adheres to certain principles established by B Lab but does not indicate that the venture is forgoing returns to investors for the pursuit of social objectives. This is one reason that the term “social venturing” can be confusing. For example, a for-profit business might decide to emphasize a social orientation or to seek B Corporation status because doing so appeals to a group of potential customers that the venture hopes to attract. The term “not-for-profit” can be similarly confusing. Narrowly, it means only that the venture does not have stockholders and is exempt from paying income taxes. Such a venture may still be seeking to maximize a residual of revenue, donations, or other inflows, minus expenses.
Introduction 17
Because of the potential confusion, we think of social venturing as involving entrepreneurial efforts where financial returns are traded off against social objectives—a true “double bottom line.” The distinguishing characteristic of a social venture versus a commercial one is the elevation of the objective to address social issues to a point where financial returns are sacrificed for a social objective. Profits are typically used partly to sustain the organization’s ability to provide a social benefit. Because assessing these trade-offs is often a matter of judgment, it can be unclear whether an enterprise is a true social venture or an effective marketer. An example is TOMS, a for-profit company founded by serial entrepreneur Blake Mycoskie in 2006 with $500,000 of proceeds from selling his online driver education business. TOMS (through TOMS Roasting Co.) began selling shoes using a “one-for-one” model, which meant donating a pair to children in poor countries for every pair sold. By 2012, the company had donated over two million pairs of shoes around the world. As of 2013, it had established manufacturing facilities in Haiti, Kenya, India, and Ethiopia, providing employment but also low-cost manufacturing. In 2014 TOMS expanded its “one-for-one” model to coffee, where each purchase results in a one-week supply of clean water for a person in need. It is clear that the enterprise is not purely social as later that year, Bain Capital acquired 50% of TOMS, of which Mycoskie was the sole owner, at a reported valuation of $625 million.7 The evolution of TOMS from opportunity identification to exit by acquisition parallels many successful entrepreneurial ventures. And the estimated $300 million that the founder received in the buyout shows that the financial returns to incorporating a social objective can be substantial. As with many successful innovations, the one-for-one business model has proliferated, with at least 40 companies adopting a similar approach as of 2016.8 Data suggest that social venturing will grow in importance. On the funding side, a large number of foundations and philanthropists are using more traditional for-profit benchmarks and analyses to assess the impact of their charitable giving. Finally, many established VC firms are creating funds specifically targeted to social impact ventures.
1.3 Hypothesis-Driven Entrepreneurship The previous discussion highlights that entrepreneurial activity is driven by many factors and takes many forms around the world. In this book, we include examples of replicative and innovative types of entrepreneurship but focus primarily on innovative entrepreneurship. New ventures that involve
18
Chapter One
innovations with uncertain potential pose significant challenges for strategy, forecasting, valuation, contracting, and financing choices—greater challenges than those faced by ventures that build on established business models and ventures that can be financed with small investments, traditional borrowing, and operating cash flows. By focusing on innovative entrepreneurship and high growth, we tackle the most challenging issues that entrepreneurs and investors may confront. The differences, however, are of degree rather than substance. All entrepreneurs and investors can benefit from better understanding of concepts like milestones, staging, and real options. These ideas fit well with the hypothesis-driven approach to evaluating and advancing entrepreneurial opportunities.9 Formulating, testing, and evaluating hypotheses is at the core of the scientific method. In recent years this approach to discovery has been successfully applied to new ventures. If done correctly and if the results are interpreted objectively, the approach can add tremendous value. Just as with any type of research project, the key is asking the right questions, formulating hypotheses based on those questions, systematically testing the hypotheses, and making evidence-based decisions to proceed, abandon, or pivot by modifying the hypothesis. As illustrated shortly, this modern approach to entrepreneurship breaks a business model into a set of hypotheses: design the method to test them, collect relevant data, and evaluate the results. Generally these hypotheses are sequenced in an economically sound way that is designed to quickly validate missioncritical assumptions and limit the resources that are devoted to ventures that may sound good but are destined to fail. You might, for example, first test the hypothesis that you can build a prototype with basic functionality in one month. This could make sense as an initial hypothesis if the potential for success depends critically on your ability to quickly develop a working prototype and if not being able to do so would be likely to leave the door open to a competitor who could execute more quickly. The question should be one that is falsifiable and quantifiable—either you find support for your hypothesis or you do not. The data must be of sufficient quality that you have confidence in the results and you must maintain objectivity when interpreting them.
Implementation of Hypothesis-Driven Entrepreneurship To illustrate the value of hypothesis-driven entrepreneurship, consider the early stages of Dropbox, a downloadable app that allows users to share, sync, and store files using the Internet rather than their hard drives. At the time of its introduction, Dropbox had several advantages over cloud storage competi-
Introduction 19
tors, including speed and simplicity. Users could access their files through office computers, laptop PCs, and mobile devices. As with most new ventures, there was considerable uncertainty surrounding the early stages of the company. Several crucial questions come to mind: Could a working prototype be developed in a reasonable amount of time? Who are the customers for the product—businesses, individuals, or both? What features would customers find most desirable? What would be the most effective way to attract customers? How much would it cost to acquire and retain a customer? How would the product make money? Each of these can be considered a research question where the underlying uncertainty can be reduced by hypothesis testing. As a first step, the Dropbox founders created a prototype for Windows PC users. Next, they needed to test the hypothesis that Dropbox would appeal to potential users. Rather than simply launching the product in an expensive rollout and taking the chance that it might underperform in some crucial way, they posted a brief product demo on the Hacker News website to solicit opinions and to identify a set of beta users for subsequent tests. After validating the product in this way, they moved on to the question of marketing. They were able to quickly rule out their initial hypothesis that they could benefit by a traditional B2B direct personal selling approach, and instead they turned to an innovative grassroots approach of marketing directly to individual users. The approach of marketing to consumers effectively turned consumers into marketing agents for the business applications of Dropbox. This process of testing and modifying and retesting allowed them to improve the prototype and successfully launch the product. The sequential approach to learning also creates opportunities to “stage” the venture, to set out observable milestones, and possibly to raise financing that is associated with achieving specific milestones rather than trying to fully fund the venture at the outset.
Measuring Progress with Milestones Hypothesis-driven entrepreneurship is closely linked to the notion of measuring new venture progress in terms of the attainment of important milestones. Rather than thinking of staging in terms of time intervals such as months or years, orienting around milestones is more useful. At early stages of venture development the uncertainty of future free cash flows (i.e., cash flows available to investors) is extremely high. As a result, raising capital at a very early stage is difficult and expensive for ventures that ultimately are successful. Relying on milestones enables the parties to engage in a structured program of hypothesis testing and to postpone financial commitments until they are needed. As milestones are achieved, uncertainty is reduced, so that investment valuations
20
Chapter One
are higher and the fraction of ownership that can be retained by the entrepreneur is increased. Each major milestone functions as a working hypothesis about the venture. In scientific terms, achieving a particular milestone means that the hypothesized outcome was not rejected and enables the entrepreneur to move ahead to the next. Falling short of a successful test is cause for concern but does not necessarily mean that the venture is a failure and should be abandoned. Understanding the reasons for failing to meet a milestone is important. If, for example, an entrepreneur believes that a prototype of the idea can be completed in 6 months but instead it takes 9, the delay indicates that some aspects of the venture need to be reexamined. There are several possible reasons for failing to achieve a milestone on time. Perhaps the entrepreneur underestimated the technical difficulties; alternatively, the entrepreneur may have mismanaged the project. The distinction is critical to deciding how, or even whether, the venture should continue. In any case, milestones enable the entrepreneur and investors to sharpen their expectations about ultimate success or failure. Milestones also help identify ways to enhance the expected benefits of the project. Appropriate milestones differ with circumstances. For some kinds of ventures, the first significant milestone is concept testing, such as what the Dropbox team did before actually rolling out the product. The objective of concept testing is to do enough fieldwork to determine whether a market opportunity exists and whether there is enough upside potential to warrant continued investment. Reaching this milestone helps resolve some of the uncertainty and contributes to convergence of expectations between the entrepreneur and investors. A second milestone might be the completion of a prototype. Normally, a prototype is an early-stage working model of the envisioned product—for example, a beta website for an Internet company. Completing the prototype forces the entrepreneur to anticipate and encounter a variety of issues related to product development. By doing so, the entrepreneur can gain increased understanding of technological bottlenecks, manufacturing costs, and materials availability. In addition to resolving uncertainties about the product, reaching this milestone may help establish a first-mover advantage for the entrepreneur so that it is easier to control development of the idea. It also provides early tangible evidence of the entrepreneur’s managerial ability. The Dropbox founders, as noted, actually pursued these two milestones in the opposite order—they first developed the prototype, then used the prototype in a video to test the concept that helped it identify beta testers and preempt possible competitors. The appropriate sequencing of milestone objectives is situationspecific. If a working prototype would be time-consuming and expensive, it could
Introduction 21
make sense to devote more initial effort to assessing market potential before committing resources to prototype development. Conversely, if a prototype can be developed quickly and inexpensively, the quality of information about the market could be enhanced by using the prototype to test the concept. The nature of the venture determines which milestones afford the most potential to assess progress and to facilitate the venture’s development. The following are some additional examples of milestones: Completing a minimally viable product (MVP). An MVP is one with enough features to appeal to beta testers/early adopters. Dropbox used the feedback from its demonstration video to design and limit the features of its launch product. Making a key hire. A key hire is a person who can help build the management team, bring essential expertise, and potentially signal to rivals, customers, and financial backers that the company is on track for success. In 2007, Facebook made a major stride toward becoming an economically viable company when it was able to hire Sheryl Sandberg as chief operating officer (COO), taking her away from a well-paying position at Google. Facebook became profitable under her leadership while other social media sites languished, had to pivot, or failed. Securing the commitment of an essential supplier. When Amazon .com launched in 1994 with an initial focus on book distribution, it did so with almost no book inventory and no significant facilities for storing or shipping books. Before it could do so, it needed commitments from major book wholesale distributors, especially Ingram, the gorilla of book distribution, to supply Amazon.com on terms comparable to those it offered to other retail distributors. Attraction of first independent financing. Oculus VR is a company specializing in virtual reality hardware and software. It received its first independent funding in 2012 from a Kickstarter crowdfunding round that raised $2.4 million, almost 10 times the $250,000 objective for the campaign. The contributors did not receive financial claims, but rather, if the contribution was large enough, they were promised a “dev-kit” version of the product. However, even though there was not financial commitment from investors, the round did demonstrate the intensity of demand for virtual reality, and the company soon attracted an independent VC funding round of $75 million. In 2014, Oculus VR was acquired by Facebook for $2.3 billion.
22
Chapter One
Production start-up. Following a noteworthy failure by General Motors (GM) to introduce an electric automobile, Tesla was launched in 2003 with the intent of developing a commercially viable electric automobile. After year of development efforts, in 2009, Tesla initiated production of the Roadster, the first production automobile to use lithium-ion battery cells and the first electric vehicle with a range greater than 200 miles per charge. Completing a bellwether sale. Snapchat, an online image-sharing application, was launched on a small scale in 2011 and grew to many users over the next few years but with no specific plan for developing a revenue stream. It began to monetize the network in 2014, with a sponsored story from Samsung. By the time of the Snap, Inc. initial public offering (IPO) in 2017, additional key marketers included Starbucks, bareMinerals, Trolli Candy, 20th Century Fox, and Under Armour. First competitive reaction. One indication that a business model is working is when an important competitor decides that the upstart new venture is a significant threat. Amazon.com, as an Internet-based book distributor attempting to compete against brick-and-mortar bookstores, is a particularly good example. At the time of Amazon’s launch, it was not clear whether online book distribution would displace brick-and-mortar, function as a weak substitute, or even complement physical distribution. When Amazon began to significantly draw sales from brick-and-mortar stores, Barnes & Noble, the largest retail book chain, responded with a combination of introducing its own online service and discounting book prices. The competitor’s response is evidence that the potential for success is substantial, and, in the case of Amazon, created opportunities to test additional hypotheses related to repeat purchase potential and the price sensitivity of Amazon users. First pivot, redesign, or redirection. The company now named Groupon was initially launched as The Point in 2007 to implement the idea of its chief executive officer (CEO) and founder, Andrew Mason, of using social media to organize people into collective groups to pursue specific causes or goals. The original idea achieved only modest success. Responding to an opportunity identified by a group of the users, another of the founders, Eric Lefkofsky, encouraged the company to pivot and focus on group buying. The company was renamed Groupon. Groupon’s pivot is an example of how a venture can adjust when its original hypothesized value proposition is rejected but, in the process, a different focus is identified.
Introduction 23
Relationship Between Hypothesis Testing and Real Options When we analyze new ventures in a scientific manner so that key business model hypotheses are formulated and tested, we create opportunities for the entrepreneur to learn and to either validate the assumptions or refute them. Each of the examples of milestones presented earlier offered the entrepreneur a way to learn about the venture and its market and to modify (or pivot) accordingly, or to abandon. Hypothesis-driven entrepreneurship, in effect, creates real options, or rights to undertake decisions in the future about an underlying nonfinancial asset. Because they limit downside risk and increase upside potential, real options are valuable. In Chapters 5 and 6 we discuss more formal modeling of real options and approaches to how they can be valued.
1.4 The Stages of New Venture Development Although there is no typical life cycle for a new venture, firms do go through stages of development: they come into existence; they may undergo episodes of rapid growth, slow growth, or stagnation; and they may fail. But, after coming into existence, a venture can go through stages in any order and can go through one or more stages several times. A firm can even fail more than once. With this caveat, Figure 1.5 depicts the stages of new venture development from inception of the opportunity through its development and eventual harvesting—the stage where the investors and maybe the entrepreneur can successfully unwind their positions. To avoid overgeneralizing, the figure represents a hightech, single-product venture for a product that gains rapid market acceptance after being introduced. The diagram shows the various stages of development, identifies the types of decisions and actions that typically are made at each stage, indicates the types of real options associated with each stage, and in the final row describes the stage in general terms. As the diagram implies, once an opportunity is recognized, some nonfinancial activities can begin immediately. Filing for patents, searching for personnel, assessing the size of the market, and identifying actual and potential competitors are examples of activities that need to be performed regardless of any decisions about development strategy. From a decision-making standpoint, the entrepreneur’s first action is to generate a short list of alternative venture strategies. For some kinds of ventures, only one such strategy may be feasible. For many others, it should be possible
Stages Actions Real Options Description
Opportunity
Development
Start-up
Early growth
Expansion
Exit
Obtain seed financing Assess opportunity Assess strategic alternatives Determine organizational structure Determine organizational form Prepare business plan/model
Obtain R&D financing Build research team Conduct R&D activities, e.g.: Secure patent Develop prototype Build website Test market/market research Assess/update business model
Obtain start-up financing Acquire facilities and equipment Initiate production Build starting inventory Build sales and marketing team Initiate revenue generation Assess/update business model
Obtain early-growth financing Work toward break even revenue Expand team as needed Expand facilities as needed Assess/update business model
Obtain expansion financing Work toward proven viability Expand team as needed Expand facilities as needed Build track record for harvest Assess/update business model
Obtain continuing financing: IPO Acquisition Buyout Early investors harvest Assess/update business model
Continue to next stage Modify concept Abandon
Continue to next stage Extend stage/financing Modify R&D strategy Abandon
Continue to next stage Modify production/financing Modify marketing/financing Abandon
Continue to next stage Extend stage/financing Abandon
Continue to next stage Extend stage/financing
Choose form of exit
All activities through preparation of business plan and before incurring significant expense.
All research and development activity that must be completed before revenue generation can commence.
All activities related to start of production and marketing and initiation of revenue-generating activities.
All activities during the period before the venture reaches a level of sales sufficient for cash-flow breakeven.
All activities during the period after breakeven and before sustainable viability is established.
All activities related to establishing continuing financing and enabling early investors to harvest.
Fi g u r e 1 . 5 Stages of new venture development The figure shows the standard progression of development of a new venture from opportunity identification though stages culminating in exit. At each stage, the figure indicates the kinds of actions that normally are associated with the stage, as well as some of the real options the entrepreneur is likely to have.
Introduction 25
to represent the array of sensible alternatives in terms of a small number of discrete strategic scenarios. At each stage, the entrepreneur needs to identify a limited set of real options for implementation. These real options are choices such as continuing to the next stage; waiting until the potential of the opportunity becomes clearer; abandoning the venture altogether; or modifying the R&D strategy, which typically involves revising the business model and may affect financing as well. This analysis is done for one real option structure at a time, matched with each financing structure that would make sense for that strategy. The real option structure and financing structure are interdependent. Thus, the entrepreneur needs to search for the most valuable financing structure to complement a particular real option structure. Performance evaluation/learning is a continuous process, and is reflected in Figure 1.5 as “assess/update business model.” Reassessment of the venture’s progress occurs throughout its development. If implementation and financing results align generally with expectations, then all that may be necessary is to periodically refine and update the business plan. If not, then the strategy is suspect and the entrepreneur should revisit the strategic plan, considering alternatives ranging from major refocusing to abandonment. As indicated in Figure 1.5, the final stage for a successful venture is exit, which allows investors and the entrepreneur to harvest their investments.
1.5 Financial Performance and Stages of New Venture Development Figure 1.6 shows how financial performance can be used to delineate the stages of new venture development for a typical venture that supplies a physical product and makes a positive investment in net working capital. The figure reflects the same stages shown in Figure 1.5 after the opportunity has been identified and a decision to proceed has been made. The horizontal axis measures time; the vertical axis, dollars. Time 0 represents initiation of sales. The three curves in the figure are revenue, net income, and cash flow available to investors. Revenue and net income are measured in the conventional ways for a firm that uses accrual-based accounting. Cash flow available to investors is defined in the figure as cash flow from operations after tax and before interest expense, less the net new investments in working capital and fixed assets that are needed to support the revenue levels and prepare for future growth. If cash
Chapter One
Dollars
26
0 Revenue Net income Cash flow
0
Development
Start-up
Time
Early Growth
Expansion
Exit
Fi g u r e 1 . 6 Financial performance and stages of new venture development The figure reflects five stages that are typical of new venture development. During the development stage, the venture generates no revenues and both net income and cash flow are negative. Start-up begins when the firm acquires the facilities, equipment, and employees required to produce the product. During early growth, revenue is growing, but both net income and cash flow available to investors are negative. Expansion is the last stage during which external financing is required. During exit, the rate of growth declines to the point where cash flow available to investors is positive.
flow available to investors is negative, the venture must finance the shortfall. If it is positive, the firm can distribute the surplus to investors, including interest payments on debt, debt redemption, dividends, and share repurchase, or it can pursue other investment opportunities.
Development This stage encompasses all activities leading up to the start of revenue generation. During this stage, the entrepreneur develops prototypes, performs beta testing, and invests in any capital equipment needed for product development. The venture generates no revenues during this period. Net income is negative and may become increasingly negative as the number of people and other resources involved in product development increases. Cash flow initially is very negative as the firm invests in capital equipment needed for development. Under existing U.S. accounting conventions, firms must record the depreciation expense of equipment over time. As a result, net income is not as negative as cash flow.
Introduction 27
Start-Up The demarcation between development and start-up is when the firm acquires the facilities and equipment required to produce the product and invests in the employees and marketing that support production. The decline in cash flow preceding revenue generation reflects investment in production equipment, facilities, and net working capital. This pattern would be somewhat different for a venture that does not supply a physical product, such as one that receives payment in advance for software or a subscription-type service or one that could operate with negative net working capital (e.g., if accounts payable is higher than accounts receivable). For such ventures, cash flow could be higher than net income.
Early Growth Revenue is growing (possibly rapidly) in the third stage. Both net income and cash flow available to investors are negative. Cash flow of a conventional venture exceeds net income, essentially because periodic depreciation expenses are larger than the increase of investment in working capital and fixed assets needed to support the growth of revenue. This is referred to as the earlygrowth stage because, although growth may be rapid in percentage terms, the base from which we calculate revenue growth is low. When we consider the forms of financing that are available to ventures at different development stages, it will be useful to think of the early-growth stage as one where, if the venture were to stop growing and to spend and invest accordingly, operating income would still be negative and the venture would not be viable as a going concern.
Expansion There is no clear line of demarcation between the early-growth stage and expansion. Many new ventures experience growth in the early stage but fizzle out before they achieve sales that are sufficient to sustain the business. Hence, entrepreneurs can benefit by being cognizant of signs that the venture has reached the expansion stage. As shown in Figure 1.6, net revenues are increasing but are at a level where, if growth were to stop and the revenue and expenditures were to be adjusted to the no-growth scenario, the venture would have positive operating income and be a viable entity. For a product-based business, rapid sales growth during this stage puts heavy demands on the entrepreneur to locate the financing needed to sustain the corresponding growth of
28
Chapter One
working capital and fixed assets. For subscription businesses and other businesses with negative net working capital, rapid growth can provide positive cash flow during expansion.
Exit The exit stage requires that there be a viable market for the company’s shares either via an acquisition or merger or via an IPO. This implies that investors, either buying in the IPO or making the acquisition offer, believe that insiders (the entrepreneur and existing investors) are representing the company in an accurate way so that outsiders are not skeptical of the reason insiders want to divest (e.g., selling shares because they believe that future performance will be poor). This does not imply that all uncertainty is reduced or even that the firm is profitable (in fact, many high-tech and biotech firms are acquired or go public with negative net income), but it does suggest that the growth rate and track record of the firm have reached a point where future growth can be forecasted. As long as outsiders do not believe they are at a serious information disadvantage, successful exit, even from very risky ventures, is feasible. Generally, as Figure 1.6 indicates, during the exit stage, the venture’s rate of growth declines to the point where cash flow available to investors is positive. The venture can provide returns to debt and equity investors without needing to increase outside financing. This is an obvious time for investors to harvest and realize the returns on their investments.
1.6 The New Venture Business Model Formal written business plans are not as common as they once were. Instead, entrepreneurs are encouraged by investors to think of the business model as a means to identify the value proposition, explain why the founders are the ones who should pursue the venture, clarify the assumptions underlying their business strategy, and lay out expectations of what is achievable if the assumptions are validated. Because of the importance of the test-learn-pivot approach, the modern business model is dynamic and flexible, evolving over time as the results of the experimentation and testing unfold and the venture progresses. It is easier to attract investors with a business model that sets out explicit financial projections and milestones than with one that is vague. By making the model specific and by including documentation and support, the entrepreneur invites oversight and evaluation and provides a mechanism that makes it easier for investors to reevaluate their investments as the venture progresses. Investors
Introduction 29
can use the milestones and financial projections to test the entrepreneur’s beliefs. The model can also facilitate reaching an initial investment agreement if there are contractual terms that tie the entrepreneur’s ownership share or control to attainment of milestones such as growth targets or landing a critical customer. New venture business plans serve multiple purposes. In contrast, for an established enterprise, the business plan may be strictly an internal document. Decisions by outside investors, such as the decision to continue financing an established enterprise, are likely to be based on experience with management, track record of the business, and current financial health. Investors in established ventures are not likely to depend significantly on review of the company’s projections. However, for a start-up, the reverse is true. Consequently, the business plan for a start-up must be prepared with an expectation that it will be scrutinized by outsiders who may not be very familiar with the business. Many things that are unnecessary to include in the plan of an established enterprise are essential in a plan that outside investors expect to see and on which they will rely. The common features of a business plan include identification of the problem to be solved, the “solution” or value proposition, the market potential, the competition, the team, financial metrics, accomplishments to date, a time line, and (if appropriate) the financing “ask” and explanation of how the funds will be used. It sometimes is appropriate to have multiple (but consistent) plans, with each plan containing information that is tailored for the specific information needs of a particular audience. Plans can be circulated to potential customers, investors, partners, and founders.
What Makes a Business Plan Convincing? Some entrepreneurs prepare and circulate business plans in an effort to test investor interest in an idea. In such a case, the entrepreneur operates with the perception that the critical success factor for the venture is the idea, and that investors, recognizing the merits of the idea, will come forth with funding. Obviously, this approach—of looking for investors before committing to the project—reduces risk for the entrepreneur. Unfortunately, it also reduces the potential for obtaining financing. Credible evidence of the entrepreneur’s commitment and beliefs about the validity of projections presented in the business model is critical to securing funding. Credible evidence of commitment. Although a good idea is necessary, investors are looking to the plan for evidence that the entrepreneur (along with key members of the team) is committed to the venture. The most convincing
30
Chapter One
evidence is the investment of effort and capital that the entrepreneur has already made and is continuing to make in the venture and for which a return cannot be realized unless the project goes forward and is successful. There are many ways to demonstrate such investments. The key element for their functioning as credible commitments to the project is that they are sunk, meaning that they are not recoverable or will not generate a return for the entrepreneur unless the entrepreneur continues to commit effort to the project.10 It is not the action per se that matters but the irreversibility of the action. The loss of salary that comes with resignation of current employment is credible as a signal only if the entrepreneur would have difficulty finding new employment of equal value.11 Evidence of reputation and certification. Experienced entrepreneurs sometimes attract investment without the need to make credible demonstrations of commitment to individual projects.12 Later in the book, we explore the connection between past success and reputation, and we will see other ways in which reputation is important in entrepreneurial settings. What about the first-time entrepreneur who has yet to establish a reputation or an entrepreneur whose track record is less than perfect?13 We’ve already seen that the entrepreneur may be able to make a credible commitment to the venture, but without a reputation the commitment alone may not be sufficient to attract investors.14 One solution for the first-time entrepreneur is to rely on the reputations of others.15 Important suppliers or customers who have publicly committed to transact with the venture may provide the level of certification that is important for attracting investment, as can relational partners who become involved in implementing the business plan.
1.7 Summary Having completed this chapter, you should now have a clear sense of the important ways that entrepreneurial finance is difference from finance for a public corporation. You should also recognize what it means to be an entrepreneur and that entrepreneurship exists in a variety of forms, including endeavors ranging from replicative ventures that are intended to remain small to visionary enterprises that are intended to disrupt the status quo and change the ways we do things. You will have seen that the failure rates of new ventures are quite high, but also that even some ventures that last only a few years were
Introduction 31
designed to capitalize on transitory opportunities. While our main focus is on the difficult challenges that face stand-alone, high-tech ventures with the potential for rapid growth, we also note that much entrepreneurial activity occurs within established firms and that some entrepreneurial activity is driven partly by the desire of the entrepreneur to pursue a social agenda. Perhaps the most important lesson from this introductory chapter is that good entrepreneurship has an intrinsically scientific nature that is designed to help steer resources to the places where they can be most valuable. Entrepreneurial finance at the early stages in the life of a venture is a progression of successively setting out and testing hypotheses about the merits of the idea and the ability of the entrepreneur to bring the idea to fruition. Hypothesis-driven entrepreneurship is an approach that can quickly and inexpensively identify ideas that are not worth pursuing and can continue to nurture and refine those that are on a path toward success. This chapter develops some important terminology and concepts that are related to the maturation of successful ventures. As a venture matures, it progresses through a number of developmental stages. At each stage, the kinds of hypotheses that are being explored change, and the kinds of resources that will be available to support the venture change. The chapter concludes with a brief discussion of how business planning is different for entrepreneurial ventures than for established firms.
Review Questions
1. What are the key distinctions between corporate finance and entrepreneurial finance? 2. What, according to this chapter, is an “entrepreneur”? 3. What makes a venture a social venture? 4. How is the distinction between innovative and replicative entrepreneurship related to the distinction between opportunity-based and necessitybased entrepreneurship? 5. What is hypothesis-driven entrepreneurship and how does it relate to real options? What are some examples of real options? 6. What are the main stages of new venture development? What kinds of things go on during the different stages? 7. How are the stages of development related to the financial performance of the venture? How do the stages relate to availability of financing? 8. Explain how milestones are related to development stages and to real options.
32
Chapter One
9. What are some important differences between new venture business plans and business plans of established firms? 10. How can entrepreneurs make their business plans more convincing? Notes
1. A “project” is the fundamental building block of corporate finance. In basic courses it is treated a black box where cash flow forecasts are unbiased and the task for the manager is to assess risk and choose the right discount rate when determining net present value. In more advanced courses, projects are analyzed with techniques such as sensitivity analysis, simulation, and other computer-assisted techniques that look more objectively at the uncertainty associated with cash flows. See, for example, Brealey, Myers, and Allen (2017). 2. Kirzner (1979) discusses early views of entrepreneurship. 3. Drucker (1985), p. 20. Bull and Willard (1994) survey definitions of entrepreneurship and conclude that most are permutations of, or derivative of, Schumpeter’s view. 4. See Headd (2003). 5. For insights on returns to entrepreneurship, see Hall and Woodward (2010), who study returns to an important class of entrepreneurs. They find that the typical venture-capital-backed entrepreneur received an average of $5.8 million in exit cash. Almost three quarters of entrepreneurs receive nothing at exit and a few receive over a billion dollars. See also Moskowitz and VissingJorgenson (2002), who find evidence of a “private equity premium puzzle” in that returns to entrepreneurs who invest in their own ventures would have done just as well if they had invested in public equity instead. However, Kartashova (2014) examines a longer time span and finds that the private equity puzzle occurs in only some time periods. All of the authors point out the riskiness of entrepreneurship and the nondiversified position that entrepreneurs face. 6. GEM is a not-for-profit academic research consortium that makes international data on entrepreneurial activity publicly available. 7. Mycoskie used a portion of his proceeds from the sale to establish the TOMS Social Entrepreneurship Fund, which invests in other social entrepreneurs. http://w ww.reuters.com/article/us-toms-baincapital/exclusive-bain -capital-to-invest-in-shoemaker-toms-sources-idUSKBN0GK1ZZ20140820. 8. https://w ww.inc.com/magazine/201605/ leigh-buchanan/toms-founder -blake-mycoskie-social- entrepreneurship.html. 9. A large number of incubators and applied courses on entrepreneurship use this approach when guiding new venture development. Useful reference books for practitioners include Blank and Dorf (2012) and Ries (2011).
Introduction 33
10. See Williamson (1985) and Ghemawat (1991) for seminal work on the role of sunk investment in strategic settings. 11. Levesque and MacCrimmon (1997) report that one study (of the owners of Inc. magazine’s list of fastest-growing companies) found that, on average, the entrepreneurs kept their existing jobs for four months beyond the founding of the new venture. They offer a theoretical model of the choice of when to resign, based on the tolerance for continuing to work and the marginal productivity of devoting effort to the venture. They indicate that timing is often affected by the entrepreneur’s need for cash to defray living expenses or to help fund the venture. The evidence is generally consistent with a view that resignations tend to occur at a point when the entrepreneur perceives the specific opportunity to be more attractive than continued current employment and when resignation would lend credibility to the entrepreneur’s capital raising efforts. 12. Gompers, Kovner, Lerner, and Scharfstein (2007) show that entrepreneurs with track records of success are more likely to succeed than firsttime entrepreneurs or those who have previously failed. They also find that VCs are more likely to identify and invest in first-time entrepreneurs who are likely to become serial entrepreneurs. 13. Wright, Robbie, and Ennew (1997) study VC investment decisions and find a strong preference for investing in projects of entrepreneurs who have played major roles in previous successful ventures. 14. For an overview on the role of reputation, see Milgrom and Roberts (1992). Klein and Leffler (1981) provide a formal analysis of reputation formation, the role of sunk investment as a bonding mechanism, and the effect of reputational capital on product price. 15. A model of certification, applied to new issue underwriting, is presented in Booth and Smith (1986). Hsu (2004) evaluates the certification benefit that entrepreneurs gain by accepting financing from reputable VCs. The entrepreneurs expect to gain credibility by associating with reputable VCs, in part through their access to networks of relationships that they otherwise would not have. He finds that high-reputation VCs acquire start-up equity at a 10 to 14% discount relative to less reputable counterparts, implying that entrepreneurs see the relationship as value-enhancing. References and Additional Reading
Baumol, W. 2002. The Free-Market Innovation Machine: Analyzing the Growth Miracle of Capitalism. Princeton, NJ: Princeton University Press. Blank, S., and B. Dorf. 2012. The Startup Owner’s Manual: a Step-by-Step Guide for Building a Great Company. Pescaderno, CA: K&S Ranch.
34
Chapter One
Block, Z., and I. C. MacMillan. 1985. “Milestones for Successful Venture Planning.” Harvard Business Review 63 (5): 184–96. Booth, J. R., and R. L. Smith. 1986. “Capital Raising, Underwriting, and the Certification Hypothesis.” Journal of Financial Economics 15: 261–81. Brealey, R. A., S. C. Myers, and F. Allen. 2017. Principles of Corporate Finance, 12th ed. New York: McGraw-Hill. Bull, I., and G. Willard. 1994. “Towards a Theory of Entrepreneurship.” Journal of Business Venturing 9: 183–95. Chemmanur, T. J., and P. Fulghieri. 2014. “Entrepreneurial Finance and Innovation: An Introduction and Agenda for Future Research.” Review of Financial Studies 27: 1–19. Drucker, P. F. 1985. Innovation and Entrepreneurship: Practice and Principles. New York: HarperCollins. Ghemawat, P. 1991. Commitment: The Dynamics of Strategy. New York: Free Press. Headd, B. 2003. “Redefining Business Success: Distinguishing Between Closure and Failure.” Small Business Economics 21: 51–61. Gompers, P., A. Kovner, J. Lerner, and D. Scharfstein. 2007. “Performance Persistence in Entrepreneurship.” Journal of Financial Economics 62: 731–64. Hall, R. E., and S. E. Woodward. 2010. “The Burden of the Nondiversifiable Risk of Entrepreneurship.” American Economic Review 100: 1163–1194. Hsu, D. H. 2004. “What Do Entrepreneurs Pay for Venture Capital Affiliation?” Journal of Finance 59: 1805–44. Kartashova, K. 2014. “Private Equity Premium Puzzle Revisited.” American Economic Review 104: 3297–3334. Kirzner, I. M. 1979. Perception, Opportunity, and Profit. Chicago: University of Chicago Press. Klein, B., and K. B. Leffler. 1981. “The Role of Market Forces in Assuring Contractual Performance.” Journal of Political Economy 89: 615–41. Levesque, M., and K. R. MacCrimmon. 1997. “On the Interaction of Time and Money Invested in New Ventures.” Entrepreneurship Theory and Practice 22: 89–110. Moskowitz, T. J., and A. Vissing-Jorgensen. 2002. “The Returns to Entrepreneurial Investment: A Private Equity Premium Puzzle?” American Economic Review 92: 745-78. Milgrom, P., and J. Roberts, 1992. Economics, Organization and Management. Englewood Cliffs, NJ: Prentice-Hall. Ries, E. 2011. The Lean Startup. New York: Crown.
Introduction 35
Schumpeter, J. A. 1912. Theorie der Wirtschaftlichen Entwicklung. Leipzig: Dunker and Humblot. Translated by Redvers Opie as The Theory of Economic Development. Cambridge, MA: Harvard University Press, 1934. Von Mises, L. 1966. Human Action, 3rd rev. ed. Chicago: Henry Regnery. Williamson, O. 1985. The Economic Institutions of Capitalism. New York: Free Press. Wright, M., K. Robbie, and C. Ennew. 1997. “Venture Capitalists and Serial Entrepreneurs.” Journal of Business Venturing 12: 227–49. Wulf, J., and J. Lerner. 2007. “Innovation and Incentives: Evidence from Corporate R&D.” Review of Economics and Statistics 89: 634–44.
C h a p t e r T wo
N e w Ve ntu r e Fi nan ci n g : Co n s i d e r atio n s an d Ch oice s
T h e p r i n t i n g p r e s s , the automobile, the microchip, satellites, and
the Internet are inventions that have transformed markets and changed the course of history. But how are such great ideas and inventions financed and brought to market? Financing needs vary considerably depending on the underlying technology. Automobile manufacturing, for example, is extremely capital intensive, whereas many mobile apps can be produced at little cost. The stage of development also affects the menu of available financing options. A pre-revenue venture has limited options and cannot, for example, access traditional debt markets, but a later-stage company may well consider debt as an option. The choice of financing method is pivotal to whether an idea or product reaches the market quickly and successfully. Financing decisions during early stages of development also dramatically affect the value the entrepreneur can derive from the endeavor. The choice of exit financing (e.g., acquisition vs. IPO) can also dramatically affect value. In this chapter, we provide an overview of financing sources for new ventures and the regulatory considerations that relate to financing choices. In addition to introducing the institutions and terminology, we examine determinants of the choice. There are four important questions to keep in mind as we survey financing choices: 1. Is the need for financing immediate (urgent) or is there time to shop the alternatives? 2. How large is the near-term financing need? 36
New Venture Financing 37
3. Is the near-term financing need permanent or transitory? 4. How does the near-term financing need relate to the cumulative need? The answers depend on the development stage of the company. A profitable later-stage venture with stable revenue growth has different options than a very early-stage company with high growth and negative cash flow and net income. Capital intensity is important as well. As an illustration, over the last decade, start-ups gained the ability to take advantage of outsourced cloud-based computing instead of making large capital investments in hardware and storage. At the same time, the availability of early-stage financing by VCs increased dramatically. The innovation in cloud computing significantly reduced financing needs for start-ups that rely on these services compared to the financing needs of more asset-intensive ventures like biotech and driverless cars. Ewens, Nanda, and Rhodes-Kropf (2017) attribute the changes in financing choices to the technological shift marked by the introduction of Amazon Web Services (AWS). AWS and competitors have allowed cloud-based start-ups to economize on physical space and development tools and allowed them to scale as needed rather than buying or developing their own hardware and software.
2.1 The Sequence of New Venture Financing The stages of development discussed in Chapter 1 are also reflected in the top row of Figure 2.1. The figure illustrates how the realistic financing options vary as the firm evolves. At an early stage, the entrepreneur looks for ways to finance the venture from bootstrap sources such as personal savings and borrowing sources that do not rely on venture success. In some cases, development efforts are subsidized through grant-based financing from government programs directed to universities and other research organizations. As development progresses, the entrepreneur may look to other sources of financing, including angel investors and VCs. Customary venture financing terminology reflects the correspondence between feasible financing choices and development stages. The earliest external financing is known as seed financing. Seed financing consists of small amounts of money to support exploration of a concept. It may cover such things as the cost of assessing market size and customer validation. For high-technology ventures, seed financing may provide initial funds for R&D. In cases where R&D efforts
38
Chapter Two
Fi g u r e 2 .1 Sources of new venture financing Dark gray shading indicates primary focus of investor type; light gray shading indicates secondary focus, or focus of a subset of investors.
Entrepreneur Friends and family Angel investors Corporate strategic partner Venture capital Government programs Asset-based lender Venture leasing Franchising Trade credit/vendor financing Factoring Commercial bank lending Mezzanine lender Public debt IPO Acquisition, LBO, MBO
Development
Start-up
Early growth
Expansion
Exit
are expensive and protracted, R&D financing could be required beyond what is typically regarded as seed financing. The critical risk exposure in such cases is that development efforts could fail. Start-up financing covers activities around the initiation of sales. Generally, start-up financing is provided when a concept appears to be worth pursuing, key members of the team are in place, and most of the risks related to initial development have been resolved. Start-up financing includes such things as investment in infrastructure related to manufacturing and financing of inventory. For software companies it includes financing for investment in platforms and systems for distribution of products. At this point, the main risk exposure is related to whether a cost-effective approach to manufacturing/distributing technology can be put in place. Later-stage financing is associated with the early-growth and expansion stages of development. VCs sometimes refer to financing rounds by number, such as “round 1 financing,” “round 2 financing,” or by letter, such as “A-round financing,” “B-round financing.” These terms are somewhat arbitrary but do imply sequencing based on meeting a particular milestone and moving to the next. To be consistent with terminology, we refer to later-stage financing as encompassing two general types of financing: 1. Financing provided to a company that is generating revenue but (normally) has not yet achieved profitability. The firm has a marketable product, but substantial uncertainty remains as to achievable sales and profitability. The critical element of risk is whether the venture can reach a level of sales sufficient to attract and compensate investors in an exit. 2. Financing to support the continuing growth of a venture that is operating around the breakeven point of profitability. Many profitable ventures are not yet generating sufficient cash flow to support anticipated expansion. Uncertainty remains about ultimate market potential and profitability.
New Venture Financing 39
In practice, a venture may go through many rounds of growth financing. The limiting factors on the number of times the venture can “go to the well” are practicality and the desire of the investor to maintain a close monitoring relationship. The downside of increasing the number of rounds, particularly if the investments are being made by a VC or someone else with fiduciary responsibility, are that each negotiation can be a distraction from pursuit of the venture and each round requires a valuation. One venture with which we are familiar went through 16 rounds of VC-backed financing, with a valuation at each stage! Referring again to Figure 2.1, the exit stage is where the founders and other investors attempt to harvest their investments. Harvesting techniques include taking the company public with an IPO, arranging for an acquisition or management buyout (MBO), and other approaches discussed later.
2.2 Sources of New Venture Financing Self, Friends, and Family Bootstrap financing does not depend on investor assessment of the merits of the opportunity or whether venture assets are sufficient to secure a loan. Bootstrap financing includes financing from personal resources, family, and friends. Family and friends generally have years of experience with the entrepreneur and probably have a sense of the entrepreneur’s reliability, trustworthiness, and ability to handle adversity. They may be incapable of assessing the merits of the opportunity and are investing because they believe in the entrepreneur or feel compelled by family relationships. The obvious starting point for the entrepreneur is to use personal resources to advance the project to a point where outside financing is feasible. The entrepreneur’s resources include not only personal savings and assets but also debt capacity. Stories of entrepreneurs whose earliest financing was achieved by taking out second mortgages on their homes and “maxing out” credit card borrowing are common.1 A study of nascent entrepreneurs indicates that personal savings is the most important early financing source. The evidence indicates that more than 90% of entrepreneurs rely on personal savings, followed by personal loans and credit cards (28%) and loans from friends and family (7%). Approximately 5% rely on equity investments by friends and family.2 In dollar terms, entrepreneur-backed bank loans, such as home mortgages, represent the primary sources of financing during the first year.3 Bootstrap sources rely on the entrepreneur’s reputation and credit history and not on the
40
Chapter Two
merits of the venture. Accordingly, personal debt capacity gives the entrepreneur access to capital without the need to convince investors that the opportunity is worth pursuing. Many ventures are legendary for their success in spite of getting started with bootstrap financing. Steve Jobs and his partner, Steve Wozniak, sold a Volks wagen and a programmable calculator to raise the $1,350 they used to build the first Apple computer. Bill Gates started his venture with Paul Allen from Gates’s dorm room in 1975 and later relocated to a hotel room in Albuquerque.4 They funded the start-up from savings and built a shoestring operation into Microsoft, a company with more than $110 billion in 2018 sales. But many more ventures begun by bootstrapping have failed miserably. Some undoubtedly were based on good ideas but were underfinanced.
Accelerators and Incubators Accelerators offer aspiring entrepreneurs work space, mentoring, networking opportunities, and relatively small amounts of funding (typically $25,000 to $150,000), in exchange for equity (usually 4–8%). Access to accelerators is competitive (e.g., Y Combinator, DreamIt, and Techstars); accelerators are cohortbased, offer involvement for a fixed amount of time (3–6 months), and usually culminate in a “demo day” where participating entrepreneurs have an opportunity to pitch their ideas to potential investors and others. Most accelerators are attracted to technology-based, early-stage ventures with high-growth potential that do not require large investments. Accelerators have proliferated in recent years, increasing from very low numbers to over 500 worldwide.5 Incubators also offer resources for early-stage entrepreneurs. They, too, have competitive selection, but they work with a broader array of ventures. They offer longer-term assistance (1–5 years) than accelerators and may offer workspace, access to networks, assistance with rounding out the management team, and help with obtaining external equity financing. They are distinct from accelerators in several ways, with the primary distinction being that incubators usually do not invest directly in the company, although they may take an equity stake for their efforts and consulting services.6
Crowdfunding It is important to distinguish between donation-based crowdfunding and equity crowdfunding. Platforms like Kickstarter and GoFundMe are donation based. They are designed to enable people to support projects in which they are interested, most commonly artistic endeavors like music, film, video
New Venture Financing 41
games, or unique product ideas. Financial support is provided in exchange for tickets to prospective performances or merchandise. Donation-based crowdfunding can function as a means of validating the level of interest in a venture that would help attract angel or VC funding at a later stage.7 Crowdfunding has evolved to offer funding and marketing exposure for a broader array of early-stage ventures in exchange for equity. Equity crowdfunding allows a large number of online investors to contribute small amounts of money to a company in exchange for equity. They can do so through online platforms approved by the Securities and Exchange Commission (SEC) such as AngelList and CircleUp. As explained shortly, there are regulatory constraints on the fund-raising and issuing equity through such sites, including investor maximum investment amounts and minimum wealth restrictions.
Initial Coin Offerings (ICOs) An initial coin offering is a new kind of capital-raising approach that is similar to crowdfunding. It is difficult to generalize about ICOs because the approach is evolving rapidly. Currently, ICOs are not generally used for raising equity but rather are used for building a network or ecosystem of users for platformbased business. Such ventures could include those focused on a software utility, such as a blockchain technology or a platform to assign and transfer cloud storage capacity. The ventures seek to use an ICO to build a network of users and can use an ICO to raise funds for working capital to support early development efforts and to provide information about the extent of interest in the idea. They do that by offering utility tokens to be used on the platform. In contrast to an IPO, where the offering document is a prospectus, an offering document of an ICO often is a Simple Agreement for Future Tokens (SAFT). Tokens are normally tradable rights to use the utility the entrepreneur seeks to develop. The agreement is for “future tokens” because at the time of the ICO the utility normally is still in development (and development success is far from certain). In contrast to ownership shares, utility tokens convey usage rights. In many ICOs, purchasers of tokens can pay either currency or cryptocurrency, such as Bitcoin. ICOs in the U.S. are regulated by the SEC but the nature of the regulation is still evolving, as the SEC is catching up with this new approach to funding entrepreneurial activity.
Angel Investors For ventures based on concepts that require lengthy development efforts, the earliest source of outside financing is often from high-net-worth individuals
42
Chapter Two
who invest their own capital either independently or through an organized angel investor group. These angel investors generally are freelancers who are interested in investing relatively small amounts of money ($25,000 to $500,000) in early-stage projects.8 Angel investors often provide seed capital to develop an idea to the point where more formal outside financing becomes feasible. Many are willing to invest over horizons of 5 to 10 years. In contrast, traditional VC funds usually prefer larger investments and somewhat shorter investment horizons. Angels seek to add value by identifying ventures with high potential for success and helping them to progress. They usually seek to realize a return by taking equity in the venture. Most estimates put the amount of angel capital invested in the U.S. in recent years at around $20 billion annually, excluding capital provided by friends and family. According the University of New Hampshire Center for Venture Research, 64,000 entrepreneurial ventures received angel financing in 2016, with an average round size of $330,000. Many angel investors work alone; others are affiliated with angel networks. In the late 1980s, angels began to organize into informal groups that could share deal flow and due diligence. Participants co-invest with each other, enabling the group to make larger commitments and bring more experience to each deal. There are several well-known groups in geographic areas known for high start-up activity: the Band of Angels (Menlo Park), Tech Coast Angels (eight locations in Southern California), and New York Angels (New York City). Angel groups range in size from 40–100 members. There also are global angel groups, such as Keiretsu Forum, which has groups on three continents and 2,500 members. Networking organizations, like the Angel Capital Association (ACA), facilitate connections among angel groups. The ACA has 400 angel groups in its database, and there are many more around the world. Some angel groups, such as Investors’ Circle, focus on social venturing while others, like Golden Seeds, focus on specific types of ventures—namely women-led early-stage ventures. The sizes of angel group investments in start-ups vary widely. The 2016 Halo Report of the Angel Resource Institute reports that the median angel group investment reached $127,000 in 2016. Round sizes are larger because groups sometimes co-invest with other angel groups, VCs, or other investors. The ACA reports that average group size in 2016 was around 42 members, with total annual investments around $2 million in 10 deals per year. While each group has distinct membership and goals, angel groups share common characteristics: they meet regularly to review business proposals, select entrepreneurs to make presentations to the membership, decide individually whether to invest
New Venture Financing 43
in each venture, work together to conduct due diligence to validate the plans, and evaluate the entrepreneurial team. While precise numbers are hard to come by, Panel A of Table 2.1 shows a time series of statistics provided by the Center for Venture Research (CVR). The CVR estimates that U.S. angels invested about $21.3 billion in 2016, in over 64,000 deals; many of these investments were in seed/start-ups (a combined category in the CVR data) or early-stage ventures. The time-series statistics in Panel A show that the total size of the angel investment market was negatively affected by the 2008 financial market crisis, mainly through smaller investments and a reduced focus on seed and start-up investments in more seasoned ventures (defined in the CVR data as post-seed/start-up), as well as a reduced focus on initial funding in favor of follow-on funding in ventures that had already received funding. While the total amount of angel investments has recovered since 2009, and the number of angel deals has increased, average deal size has remained lower than in the precrisis period and the focus on seed and start-up (in favor of early-growth) and initial angel investments has remained low. While angels may jointly investigate and discuss opportunities, U.S. securities laws dictate that each investor must make an independent investment decision and take individual responsibility for her own due diligence. It is common for individuals in these groups to invest $25,000 to $50,000 in a company. Some angel investors are interested strictly in return on investment and do not want to become deeply involved with the companies in which they invest. Others are more active, often motivated by desire to add value to the companies beyond their capital commitments. These angels typically focus on the types of companies with which they have experience. They can be good sources of information concerning financing and strategy. Angel syndicates. In addition to organizing into groups, angel investors sometimes form syndicates to invest collectively. In contrast to the group structure, the syndicate structure is a special-purpose vehicle that is created to invest in a single deal. Typically a syndicate has one lead investor who is experienced in attracting deal flow, conducting due diligence, and providing ongoing moni toring of portfolio companies. In recent years notable angel investors that often act as lead investors in angel syndicates have increasingly relied on website platforms like AngelList to identify individual angels who are interested in joining the lead investor’s syndicates. Angel investors who join syndicates must meet U.S.-accredited investor standards (discussed later) and be accepted by the syndicate leader. Many syndicated deals are led by well-known investors, including partners in top VC firms. Angel syndication is a recent development, and when VC partners lead syndicates, it appears that they may be pursuing opportunities
44
Chapter Two
Tab le 2 .1 US angel and VC investments: 2006–2016
Panel A
US Angel Investments
Angel Investors
Total Investments ($Billion)
Deals Funded
Average Deal Size ($Thousand)
Percent of Deals Seed/Startup
Percent Initial Funding
2006
$25.60
51,000
$502.0
46.0%
63.0%
2007
$26.00
57,120
$455.2
39.0%
63.0%
2008
$19.20
55,480
$346.1
45.0%
63.0%
2009
$17.60
57,225
$307.6
35.0%
47.0%
2010
$20.10
61,900
$324.7
31.0%
41.0%
2011
$22.50
66,230
$339.7
42.0%
52.0%
2012
$22.91
67,030
$341.8
35.0%
50.0%
2013
$24.81
70,730
$350.8
45.0%
50.0%
2014
$24.11
73,400
$328.5
25.0%
50.0%
2015
$24.56
71,110
$345.4
28.0%
44.0%
2016
$21.26
64,380
$330.2
41.0%
43.0%
Percent of Deals Seed/Angel
Percent Initial Funding
Panel B
US Venture Capital Investments
VC Investors
Total Investments ($Billion)
Deals Funded
Average Deal Size ($Thousand)
2006
$29.17
3,320
$8,785.2
13.1%
38.7%
2007
$35.46
4,291
$8,263.3
17.4%
39.0%
2008
$37.25
4,730
$7,876.1
19.1%
37.3%
2009
$26.57
4,456
$5,963.0
26.1%
37.3%
2010
$32.00
5,403
$5,922.5
31.2%
38.5%
2011
$44.66
6,775
$6,591.9
38.4%
41.7%
2012
$40.72
7,986
$5,098.5
44.7%
41.7%
2013
$44.89
9,325
$4,813.4
50.1%
38.5%
2014
$68.66
10,565
$6,498.6
52.2%
35.6%
2015
$79.04
10,483
$7,539.6
54.4%
32.4%
2016
$70.83
8,469
$8,363.0
50.4%
29.3%
source :
Center for Venture Research, University of New Hampshire and National Venture Capital Association, 2017 Yearbook.
that would not be good prospects for their VC funds or may be prospecting for opportunities that could attract VC funding in a subsequent round.9
Venture Capital Venture capital (VC), along with angel investing and other nonpublic financing sources, is part of the private equity (PE) market. VC funds are organized by
New Venture Financing 45
VC firms, where the VC firm serves as the general partner (GP) of each VC fund it organizes. VC funds in the U.S. are limited partnerships in which the limited partners (LPs) provide almost all of the capital and the GP is responsible for managing the fund, including investment selection, working with entrepreneurs, and harvesting the investments. VC funds have finite lives and are normally established to last for about 10 years. While some VC firms establish seed-stage funds for funding start-ups, most focus on funding ventures that are about to embark on the start-up stage with the potential for high growth that will follow. Because VC funds involve small numbers of LP investors (financial institutions and wealthy individuals) that are presumed to be sophisticated enough not to require government oversight, VC funds are exempt from the key federal regulations of the securities industry. As PE has grown in importance and more companies have opted out of the public capital markets, PE has increasingly attracted regulatory scrutiny in the U.S. and elsewhere. VC-backed ventures generally require few tangible assets or other assets that could serve as collateral and provide a basis for secured lending; they also have the potential to achieve high growth with high profitability but with the commensurate high risk. Because both angel investing and VC are key to enabling high-growth entrepreneurship, we devote Chapter 3 to a comprehensive description of the institutions. Here we focus more narrowly on identifying some of the considerations that help entrepreneurs decide when to seek VC financing, as opposed to alternatives. Panel B of Table 2.1 provides statistics on the size of the U.S. VC market since 2006. Total VC investment has increased from being about the same as the level of angel funding in 2006 to over $70 billion—more than 3 times the size of the angel market. Still, there are many more angel-funded deals than VC-funded deals. Offsetting the numbers of deals, the average size of VC deals is many times greater, indicating an important difference in the primary foci of angels and VCs. It is noteworthy, however, that over time the two have grown closer. Seed/angel-stage VC deals are now more than half of all VC deals, but angels are still the primary sources of small early investments; in 2016, for example, 26,400 angel-backed deals were classified as seed/start-up, whereas 4,300 deals were classified by VCs as seed/angel stage. VCs are also less inclined than angels to undertake initial funding of a venture. From the entrepreneur’s perspective, a number of factors help determine whether VC is appropriate and which VC firm to select. First, timing is important. The venture must be developed to a point where the VC firm can expect to add value, not just money, and the firm must expect to earn a return that will compensate for the VC’s human capital investment and where the entrepreneur recognizes the potential value of VC involvement.
46
Chapter Two
Fi g u r e 2 . 2 New VC investments by sector National Venture Capital Association, PitchBook 2018-Q1 Report. source:
100% 90%
Other
80%
Software Pharma & biotech
70%
Media
60%
IT hardware Health care services & systems
50%
Health care devices & supplies
40%
Energy
30%
Consumer goods & recreation Commercial services
20% 10% 0%
2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017
Fi g u r e 2 . 3
50%
Capital invested Investee companies
40%
30%
20%
10%
s te sta
O th er
lo
ra d
o
a Co
ns
yl
va
ni
U ta h Pe n
in oi s Ill
n to ng
s
a hi
W
as
Fl or id
xa Te
ts
sa
ch
us
et
Y or k M
N ew
as
a
0%
Ca lif or ni
source: National Venture Capital Association, 2017 Yearbook, PitchBook data.
60% Percentages of capital invested and investee companies
Top 10 states for VC investing: 2016
Second, the venture must be in an industry sector where the VC has expertise. Figure 2.2 shows the sectors in which VC funds typically have invested. Their focus is on sectors with potential high-growth, high-return opportunities. Third, VCs have tended to concentrate investments in particular geographic areas and to focus on opportunities that are “close to home.” While, historically, New England has been the focus of much VC activity, in more recent years the geographic focus has shifted. The National Venture Capital Association (NVCA) reports that over the 2014 to 2016 period, 57.1% of the value of VC investments
New Venture Financing 47
was on the West Coast, followed by 16.1% in the Mid-Atlantic. Figure 2.3 identifies the top 10 states for VC investment. Fourth, the investment horizon and investment objectives of VC funds make VC better suited for some projects than others. In part, because of their agreements with LPs, VCs are likely to focus on projects with potential for very high returns but low probabilities of success and that can be harvested during the planned life of the VC fund. They are not likely to invest in projects that can generate positive cash flows quickly but have limited growth potential or in those that are likely to be harvested outside a 3-to-8-year window.
Corporate Venturing A number of prominent corporations have, at one time or another, implemented an in-house VC program to aid in developing new business opportunities. Some corporations view continuous innovation as essential to their sustainability and growth. Such firms have found creative ways to supply capital to support ideas that do not “fit” with their traditional product or service lines but may have market potential. Corporate venture funds can help to retain creative employees and provide some assurance that great ideas will not “escape.” A well-known example of an idea that escaped is the establishment of Hewlett-Packard by former employees of Bell Labs who became frustrated with the corporate research process at Bell Labs. Subsequently, two employees of HP left to form Intel. Former Google employees have founded a number of start-ups. Perhaps in response to concerns with losing talent and losing out on opportunities, in 2009 Google’s parent company (Alphabet) formed GV (formerly Google Ventures) as its VC investment arm. GV provides seed- and growth-stage funding to technology companies and operates independently from Google. In contrast, Apple is a technology company where talented employees have left and founded start-ups of their own; however, Apple does not have a VC arm. The primary motivation for corporate venture capital (CVC) is strategic, but CVC can be driven by several motivations. One is defensive—a way for firms to keep track of emerging competitors and innovative technologies. CVC can also be an offensive strategy by investing and potentially acquiring or partnering with ventures that provide complementary services or products. CVC activity can also be a way to keep and incentivize employees who are entrepreneurial and would otherwise leave to start their own firms and rely on private VC for funding. CVC is more likely in firms that depend on innovation to sustain competitive advantage. For 2016, the CBInsights Global CVC Report indicates that there were 752 investments by U.S. corporations and 1412 worldwide.
48
Chapter Two
The average CVC deal size in the U.S., according to the Global CVC Report, was $21.4 million, much larger than the average VC investment. However, there is not a bright line between CVC activity and innovative activity that occurs within firms. Firms like Apple have engaged in numerous acquisitions of innovative firms, and they devote considerable internal resources to cutting-edgetechnology development. Similarly, the R&D investments of pharmaceuticals firms are largely internal because of such considerations as shared resources and the extensive testing required for new drug treatments. There is considerable variation in the way CVC activity is set up within the corporate organization. Drover, Busenitz, Matusik, Townsend, Anglin, and Dushnitsky (2017), in their review, indicate that some units are organized like independent VC firms that are staffed by experienced VCs, receive funding from a dedicated capital pool, and have discretion over investment selection and support. Incentives for performance are similar to those of independent VC firms. Others are in-house units, staffed by salaried corporate employees who most likely are looking for strategic insights or complementarities with the parent and are not as concerned with harvesting per se. The financial performance of CVCs is mixed. Some studies document that strategic alliance formation can positively impact firm value but that quantitative assessment of the financial benefits to the parent firm is difficult. Drover et al. (2017) report that very few CVC units have generated “top tier” returns. While CVC can benefit start-ups, entrepreneurs need to be aware that corporate incentives may not align with those of the entrepreneur, especially in cases where ownership of intellectual property is not clear-cut and appropriation of ideas is a concern.10 If the entrepreneur’s innovation is complementary to the corporation’s product, this is less of a concern. Empirically, based on their review, Drover et al. conclude that start-ups that accept CVC funding appear to perform at least as well as VC-backed peers. The dollar magnitude and number of CVC investment varies widely from year to year. Based on data from the NVCA Yearbook, Figure 2.4 shows total U.S. CVC investment relative to total VC investment. Corporate venturing is pro-cyclical, as a percentage of total VC investment and CVC peaks when total VC investment is highest. The average size of the CVC deals is larger than traditional VC deals. In recent years in the U.S., corporate venture fund investments represent around 10–15% of all VC deals (including CVC), but around 30–45% of all VC deal dollar value.
Venture Debt Debt financing for entrepreneurial firms encompasses a variety of forms. Term lengths can vary, loans may be collateralized or not, and returns may be interest-only or a combination of interest and warrants or convertible to ac-
New Venture Financing 49 Fi g u r e 2 . 4
12,000
Number of independent VC deals and deals with CVC involvement
The table shows, by year, the total number of VC deals, the number of independent VC deals, and the number of deals with CVC involvement, 1999–2017.
8,131 7,473
Total number of deals
source: National Venture Capital Association, PitchBook 2018-Q1 Report.
9,129 9,090
10,000
8,000
5,790
7,282
7,002
5,996
6,000 4,207
4,801 4,008
3,473
4,000
3,624
2,516
2,450
2,505 2,550
3,968
2,766
2,000 2,123 1,301
0
1,005
577
461
567
578
568
680
699
497
577
738
863
1,468 1,351 1,370 1,095 1,338
1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 CVC deals
Independent VC deals
quire equity. Lenders include venture debt funds, sovereign wealth funds, specialized banks, and institutions such as pension funds, and others that make direct loans to ventures. Asset-based lending. Asset-based lenders provide debt capital to ventures that have tangible or intangible assets that can serve as collateral. The lender is not relying on the venture’s cash flow for repayment but on the ability of the venture to liquidate business assets, if necessary, for debt servicing. Loans may be secured by accounts receivable, inventory, equipment, real estate, royalty rights to intellectual property, or other assets with verifiable market/liquidation values. Taking the collateral is not something the lender wants to do. Accordingly, asset-based lenders can also be expected to be concerned about the financial health of the venture. Based on the Commercial Finance Association’s Annual Asset-Based Lending and Factoring Survey, 2016 U.S. commitments from responding lenders totaled $199 billion. The actual number is likely to be much higher because of nonresponses.
Debt financing and business development companies. Venture debt is a form of early-stage financing represented by loan agreements between earlystage ventures and the providers of the loans. These loans generally are not collateralized as they are with asset-based lending. There are four categories of
50
Chapter Two
providers of venture debt: private debt funds, direct lenders, business development companies (BDCs), and enterprise banks like Silicon Valley Bank. The loans are complementary to equity investments in that they sometimes fill gaps in funding between rounds of equity so that the firm has funds to enable it to reach the next milestone. The loans can work like mandatory redeemable preferred stock with a premium since the provider is making a bet that the firm will be able to get future financing or operating cash flow that will enable the debt to be paid off. The interest on the debt can accrue or be paid off each month. The loans can be short- or long-term. Those intended as bridge financing between equity rounds or prior to an IPO or acquisition may have terms as short as a few months. Interest rates on unsecured loans that do not include equity sweeteners can be high—in the 11–14% range or higher. The loans usually include warrants, but these normally represent a tiny fraction of the capitalization of the company. One attractive feature of the financing is that a reputable provider can help the borrower with introductions, networking, and so on. Many such loans can be restructured if the borrower runs into trouble.11 BDCs are active participants in this market. They were authorized by the U.S. Congress in 1980 under a modification to the 1940 Investment Company Act to provide funding to small and medium-sized enterprises, especially during early stages of development. BDCs are set up as closed-end investment funds and many are publicly held firms, with shares traded on major stock exchanges. The investments that BDCs make in emerging firms are a mix of debt (primarily), equity, and hybrid securities like convertible debt.12 BDCs, like VCs, are closely involved with providing advice to their portfolio companies, which are usually private firms. To qualify as a BDC, an investment company must invest at least 70% of its assets in U.S. firms with market values less than $250 million. BDCs are pass-through investment vehicles that are able to qualify as tax exempt as long as they distribute over 90% of their profits to BDC shareholders. BDCs comprise a small fraction of the total market for new venture financing. As of 2017, the Closed-End Fund Association reports that there are 52 publicly traded BDCs with a total market value of approximately $35 billion. Revenue-based financing. For firms that are growing rapidly and have solid income statements, revenue-based financing has emerged to fill the gap between bank debt and VC. Loans for this category are generally in the $4–$10 million range. More traditional forms of venture debt and bank debt generally include liquidity covenants, which may not be appropriate for start-ups or growing firms, both of which will almost certainly have liquidity issues. VC, in
New Venture Financing 51
contrast, involves the investor taking an equity stake in the firm. Recognizing a gap in the financing menu, revenue-based financing emerged as a new vehicle offered by limited partnerships that are organized like VC firms, but where the portfolio companies are growth firms with (relatively) predictable revenue growth and are not the typical high-growth-potential firms that VCs pursue. Revenue-based financing is a structured debt instrument where the investor does not receive equity or warrants.
Venture Leasing Venture leasing is an alternative to asset-based lending. In asset-based lending, the venture owns the asset and the asset serves as collateral on the loan. In venture leasing, the lessor continues to own the asset and the venture leases it. The periodic lease payment is a substitute for debt service. Unlike most leases, leases to ventures are often subject to considerable risk. To compensate for potential loss, the lessor’s return may be tied to the financial performance of the venture, so that if the venture does well, the lessor realizes more than the expected return, and conversely. Venture leasing usually involves assets that are key to the operation of the venture. For example, in past years a number of ventures were launched to provide whole-body computerized tomography (CT) scans. The entrepreneurs in these ventures had to buy or lease a CT scanner, which, if purchased, was a multimillion-dollar investment. With venture leasing, the entrepreneur’s initial investment is reduced, but some of the upside is shared with the lessor. Normally, the providers of leases are the companies that produce and distribute the equipment, as they are the ones who can most easily redeploy them if the venture fails. There can also be tax advantages to leasing as compared to owning, especially for a venture that has yet to reach a profitable level of operation. Depending on its organizational form, if the venture were to purchase, it might not have sufficient income to be able to realize the tax benefit of recording depreciation expense related to the outlay for the equipment. In contrast, a profitable lessor who buys the machine and leases it to the lessee can realize the tax benefit of depreciation. If lessors compete to supply the equipment, some of the tax benefit should make its way to the venture in the form of more favorable lease terms.
Government Programs Many countries have established government-supported programs to provide loans and other financing for start-ups, small firms, and firms with growth potential, and to support R&D. In the U.S., Congress created the Small Business
52
Chapter Two
Administration (SBA) in 1953 to foster innovation and growth of small businesses.13 The SBA is primarily a guarantor of loans made by banks and other financial institutions and does not lend directly to small businesses.14 In addition to the SBA programs described here, government supports innovation in other ways such as grants to support scientific research, research labs, and even incubators. The SBA’s Small Business Investment Companies. Most SBA-guaranteed financing is provided through Small Business Investment Companies (SBICs). SBICs are privately owned and managed, for-profit organizations that are licensed by the SBA. With their own capital (raised from private sources including banks, corporations, individuals, and others) and with borrowed funds, SBICs provide capital to new and established small businesses.15 SBIC financing usually takes the form of interest-bearing loans that are guaranteed by the SBA. Because of government support, the capital that SBICs invest is obtained at below-market rates. An SBIC that is faced with competition to supply funds to a venture may be compelled to pass along some of these savings. More recently, the program has broadened to enable financing that is more like equity, such as loans that include warrants where the total return is tied to venture success. While SBICs have invested in such companies as Apple, Costco, FedEx, and Whole Foods, the impact of SBA loan guarantee programs on entrepreneurial activity is small. In 2016, the SBA Office of Advocacy reports that its loan/guarantee portfolio totaled $124 billion. However, the effect of the loan guarantees is mainly to subsidize the institutions that are the actual lenders by reducing the cost of default risk, with the expectation that at least a portion of the subsidy gives rise to more loans and lower interest rates for borrowers.16 The SBA also offers some grant programs, but the grants generally support not-for-profit organizations, lending institutions, and state and local government entities that provide small business management or assistance to small businesses. Although SBICs are involved in financing at about the same stages as VC, the normal interest-bearing structure of the financing makes it better suited for firms that are less risky than typical VC-backed firms. SBICs prefer firms with more limited growth potential that need financing for working capital and have the ability to achieve profitable operations quickly. The Small Business Innovation Research program. If the venture involves technology, several government agencies sponsor grants under the auspices of the Small Business Innovation Research (SBIR) program. Partici-
New Venture Financing 53
pants include the Departments of Agriculture, Commerce, Defense, Education, Energy, and Health and Human Services, as well as the National Aeronautics and Space Administration (NASA) and the National Science Foundation. SBIR Plan I research grants provide funding defined by milestones; in Phase I, they provide seed capital and funding for up to $150,000 to explore commercial feasibility, while Phase II grants are usually around $1,000,000 for up to two years. The goal is to expand R&D and further assess viability. Phase III involves transitioning from the lab to the market, during which time the firm searches for follow-on private sector financing or other government programs; the SBIR does not fund this phase. Government programs in other countries. Most developed economies provide government-based assistance grants and other means to facilitate creation and growth of new businesses. Prominent examples include Canada’s Small Business Finance Centre and the U.K.’s Department for Business, Energy and Industrial Strategy (BEIS). Among the programs of the BEIS is a multimillion-pound equity finance program, Enterprise Capital Funds. The program is designed to provide up to £5 million of initial funding for ventures that require funding within what some have termed the equity gap—more than what angel investors generally supply but less than what VCs would consider.
Trade Credit Trade credit, or vendor financing, is the largest source of external short-term financing for firms in the U.S. and is even more important in emerging economies, where risk capital can be scarce. Vendor financing arises whenever a business makes a purchase from a supplier that offers trade credit. For example, a venture that buys supplies on terms of net 30 receives the supplies right away but does not need to pay until 30 days later.17 In effect, the supplier is providing the venture a zero-interest-rate loan for 30 days. Trade credit appears on the borrower’s balance sheet as an account payable. The same loan appears on the supplier’s balance sheet as an account receivable. The difference between a firm’s accounts payable and its accounts receivable is its net trade credit. The net trade credit balance defines the firm in terms of whether trade credit functions as a net source or net use of funds. If accounts payable exceeds accounts receivable, trade credit is a net source—that is, borrowing. Not all trade credit is free. When it is not, heavy reliance on trade credit can be very expensive. For example, terms of 2/10 net 30 are common. The buyer is offered a 2% discount if the invoice is paid within 10 days; otherwise the full payment is due in 30 days. If the buyer decides to forgo the discount, there is a
54
Chapter Two
sizable opportunity cost. In effect, the firm borrows the invoiced amount for 20 extra days by paying 2% more. The compounded implicit interest rate on such a loan is 44% per annum.18 Managers of any new venture must consider this cost when deciding whether to pay during the net period or take advantage of the discount. If the need for cash is transitory or bank loans are not available, forgoing the discount could make sense.19 Entrepreneurs may not have much actual control over the decision to offer trade credit to customers. Credit terms and availability are determined by competition and tend to be uniform in a given industry. A venture that is competing against established firms may find that it must offer trade credit to get customers to try the product, whereas the venture’s suppliers may insist on receiving cash until the venture has proven itself.
Factoring When a business offers trade credit, it generates accounts receivable. In lieu of borrowing to finance the receivables, the venture may be able to sell the receivables to a factor. A factor is a specialist who buys receivables and manages the collection activities. Factoring comes in two basic types: with and without recourse. If factoring is without recourse and a customer does not pay its bill, the factor absorbs the loss. Factoring with recourse means that if the customer does not pay, the factor can collect from the venture directly. The three basic parts of a factoring transaction are (1) the advance (the factor advances a percentage of the face value of the receivables to the seller), (2) the reserve (the remainder of the total invoice amount that the factor holds until it receives payment), and (3) the fees associated with the transaction, which are deducted from the reserve. The advance generally ranges from 70% to 90%. The fee is calculated as a percentage of the face value of the receivables. It includes compensation for handling the collection and (for a nonrecourse transaction) compensation for assuming the risk of nonpayment. Normal fees can be 2–6% and can include implicit interest for the period over which the funds are advanced or an additional component for explicit interest over the period. Research indicates that factoring is more common for smaller sellers and is most beneficial for sellers with geographically dispersed buyers and those with a few repeat customers. Although the cost of factoring may seem high, the terms are competitively determined. Factoring may be attractive to a cash-poor venture until it grows to a size that makes integrating the collections function more economical.20 Factoring is a large and international institution. FCI, a trade association of many of the world’s factoring firms, estimates that 2016 total factoring volume
New Venture Financing 55
was €2.375 trillion, concentrated mainly in Europe (67.05%) and Asia-Pacific (23.38%), with North America in fourth place (4.0%). In a number of European countries and a few Asian countries, the World Bank reports that the ratio of factoring volume to gross domestic product (GDP) in 2015 was greater than 10%, whereas in the U.S. it was less than 1% of GDP. While its use in the U.S. has been static, factoring is a growing and significant source of financing worldwide. This trend may reflect differences in institutional constraints on bank lending that make factoring more appealing in some countries, whereas in others, bank loans that are secured by receivables are more common. It may also reflect differences in typical firm size and the composition of industries.
Fintech Bank loans, factoring, and trade credit are long-lived traditional institutions that have been challenged on many fronts by financial technology (fintech) ventures that are looking to disrupt old ways to providing credit to start-ups and small and medium-size enterprises (SMEs). These include peer-to-peer (P2P) lending firms, firms that have adopted business models that streamline SBA loans, online lenders (e.g., Kabbage), innovations in creating markets and auctions for trade credit balances (e.g., C2FO.com), and electronic markets for corporate receivables (e.g., NYSE Euronext). All these innovations are forms of disintermediation and are aimed at more efficient, speedier financing for firms. While most of these innovations in financing have focused on the larger markets in the U.S. and Europe, there is a global movement toward nonbank lending as well.
Franchising and Independent Contracting Franchising is another way to finance growth. In business format franchising, such as a fast-food restaurant, the franchisor establishes a business format and offers franchising opportunities to prospective franchisees. The franchisee invests in a facility in a particular locality. The facility carries the franchisor’s brand and must conform to the general standards of the franchise network. Franchisors provide a range of services that can include site selection, training, product supply, marketing, and assistance in arranging financing. The franchisee normally pays a franchise fee and makes periodic payments that are partly based on revenues. Similarly, companies like Uber and FedEx use independent contractors in at least some components of their enterprises. Uber drivers and FedEx Ground operators acquire their own vehicle and, in some cases, may employ drivers.
56
Chapter Two
Uber and FedEx coordinate the activities of the independent contractors and establish service standards. Reasons for franchising and use of independent contractors are not limited to financing advantages, but the arrangement does allow the franchisor to expand the size and geographic reach of the business rapidly, without having to raise all the capital by itself.21 Among other things, the franchisee may be responsible for identifying the local market opportunity and acquiring the land and facilities, which, if the outlet were company owned, would have been costs borne by the franchisor. Franchising is common in a number of industries such as quick-service restaurants, lodging, car rental, automobile dealerships, and full-service restaurants. About one third of all retail sales in the U.S. are made through franchised outlets (including car dealers and gas stations).22 The practice has also increased the international reach of well-known consumer brand names. McDonald’s, for example, is so pervasive that the “Big Mac Index” is sometimes used as an indicator of purchasing power parity across countries. There are also companies that do not “franchise” in a formal legal sense, but instead enter into licensing agreements for operation (e.g., Shake Shack). Still, the licensing revenue stream provides the parent firm with financing. The legal distinctions between licensing and franchising are nuanced and important to consider before choosing to adopt one approach versus the other.
Mezzanine Capital “Mezzanine financing” usually refers to capital raised after the firm has established a record of positive net income with revenues approaching $10–$20 million or more. Some VC firms and other private equity firms offer this type of financing. The financing generally is a hybrid of senior debt and common equity. A frequently used instrument is subordinated debt with an equity “sweetener” or warrants. Warrants entitle the holder to buy shares of the firm’s common stock at a stated price for a period of time. For example, each $1,000 of borrowing might provide the mezzanine lender with warrants to purchase 20 shares of stock at $5 each.
Private Debt Debt financing may make sense for a rapidly growing venture. There are two primary reasons for using debt to finance growth. First, because interest payments are tax deductible, debt may be less expensive than equity for a firm that has consistent earnings. Second, debt holders usually cannot vote. There-
New Venture Financing 57
fore, equity owners do not lose voting control when debt is issued. However, debt is a contractual obligation, so in the event of bankruptcy, bondholders have priority over equity owners and may effectively gain control. Private debt, such as commercial bank debt, sometimes is appropriate for small businesses, particularly those with steady, verifiable cash flows, where the business is not expected to grow rapidly. This type of financing is more likely for replicative types of entrepreneurial ventures, such as those that employ established business models and are created to serve a growing population—such as independently owned dry cleaners, bakeries, restaurants, and health clubs. The venture should be in a position to make the interest payments and to take advantage of the tax deductibility of interest payments.
Publicly Traded Debt Small and new firms do not have access to the public bond market. Investment banking firms seldom underwrite bond issues smaller than $10 million in gross proceeds. Unless a firm has a substantial asset base and steady cash flows and is in need of a significant amount of capital, it is unlikely to be able to arrange a bond issue. New ventures are more likely to borrow from commercial banks, life insurance companies, or an SBIC or other government-related entity.
Private Placements of Equity and Debt Sometimes a firm would like to raise a specific amount of capital quickly and would prefer to avoid the cost and time required to complete a public offering. This can be accomplished by private placement of debt or equity. The private placement market generally is more attractive than the public market for small equity or debt issues or for debt issues backed by complex security arrangements where flexibility is desired. Angel and VC financing are equity private placements at early stages. The term “private placement,” however, applies more broadly to any sale of equity or debt securities to a selected group of investors by means other than a public offering. In the U.S., public offerings of debt and equity are regulated by the SEC. Similar regulatory bodies oversee public offerings in other countries. One advantage of private placements is that the venture avoids the complex, ongoing reporting that would be required if it were to raise capital via public offering. In addition, an entrepreneur can use private placement to limit the number of people who gain access to strategic information about the venture. While privately placed securities avoid SEC registration and reporting requirements, only certain types of offerings qualify as private. Important factors
58
Chapter Two
that bear on the choice are the number of offerees, the relationship of the offerees to the issuer, the offering size, and the manner of offering. Often more important, however, is the sophistication of the investors and their access to information about the issuer. Stakeholders in the new venture, including employees, distributors, retailers, suppliers, and so on, are potential investors in equity private placements, as are insurance companies, pension funds, and high-net-worth individuals. Institutional investors such as insurance companies and pension funds are candidates for buying privately placed debt. Often, private equity is structured as convertible preferred stock. Preferred stock typically converts to common at an IPO. Privately placed debt may have some advantages as compared to a public issue. As with equity, the costs of a private placement of debt tend to be lower and the placement can be quicker. Also, it is possible to negotiate greater flexibility in the terms than would be possible for a public offering. A significant advantage of a privately placed issue is that it may facilitate monitoring. A public debt issue normally has many investors. The resulting diffuse ownership encourages free riding in terms of investor effort devoted to monitoring the company. In contrast, if a debt or equity issue is placed privately, ownership is concentrated and there are significant incentives to monitor management. The small number of investors also facilitates renegotiating the terms of the investment agreement if circumstances change.
Initial Public Offering In an IPO, the issuing company raises capital by selling registered equity shares to the public via a formal offering process. Since the IPO gives rise to a liquid market for company shares, it can be a convenient exit or harvesting mechanism for VCs and other investors.23 The importance of VC in the IPO market has increased over the past few decades. For the period from 1980 to 2016, 37% of IPOs were VC-backed, but the level rose to an average of 55% from 2010 to 2016.24 Going public provides a way for the venture to raise equity capital and for early investors to realize returns on their investments, achieve liquidity, and diversify.25 An IPO also provides a market-determined valuation that can be used as a basis for negotiating merger and acquisition (M&A) transactions. Where the venture’s track record is clear, large amounts of capital are needed, and potential synergies with other firms are absent, an IPO may bring a higher share price than a private placement. Public ownership also provides a way to create equity incentives for employees. The stock price is a barometer of market expectations and can provide information about how investors expect economic events and managerial
New Venture Financing 59
decisions to affect future earnings. Finally, a publicly traded firm may be able to raise additional capital more quickly and cheaply than if it were private. On the cost side, IPOs are expensive in terms of time and costs of the offering and ongoing compliance with SEC regulations and reporting requirements.26 Already-scarce managerial time must be diverted from the business and devoted to the IPO process and ongoing reporting. The presence of a visible stock price may also induce management to be unduly concerned with short-term stock price fluctuations. In recent years, the private equity market has developed to the point where some of the historical benefits of being public are less apparent.
Direct Public Offering Between an IPO and an equity private placement, U.S. securities regulations provide a variety of safe harbors whereby a firm can issue equity directly to small numbers of investors without going through the formal public offering process. Generally, the safe harbors relate to raising capital from small numbers of investors, employees, and sophisticated investors. SEC safe harbors operate on the premise that the investors’ specialized or inside knowledge will protect them sufficiently even in the absence of SEC oversight. As an example, in 1984 Ben & Jerry’s, a company founded in Vermont, built its first ice cream plant by selling $750,000 of equity directly to Vermont residents in a direct public offering (DPO). The firm later raised $6.5 million in a public offering. This experience suggests that DPOs can make sense for a firm that has a loyal customer and employee base.
Later-Stage Financing Alternatives The list of financing sources could be extended considerably. Other possibilities identified in Figure 2.1 include merger and acquisition (M&A), leveraged buyout (LBO), and management buyout (MBO). We address these alternatives later in the text when we analyze the choice of how early-stage investors harvest their investments. M&A, for example, is a much more likely exit outcome for a successful start-up than is an IPO.
2.3 What’s Different About Financing Not-for-Profit Ventures? Ventures that are organized as not-for-profit entities can compete in the same market as for-profit ventures. The main difference for our purposes is that
60
Chapter Two
they have different access to financing sources. For example, there are specific federal and state grant programs and foundation grant programs that target nonprofits. There also are sponsored competitions that focus on areas of specific social concern (such as improving water quality, education, or health).27 While not-for-profit and for-profit firms have similar access to many financing sources, especially debt financing sources such as trade credit and private and public debt offerings, not-for-profits cannot raise capital by issuing equity. In some cases, not-for-profits have established for-profit subsidiaries to carry out activities that do not directly fit within the not-for-profit mission but can provide resources to support the nonprofit mission. For example, Mozilla Foundation is the nonprofit provider of the well-known browser Firefox. The foundation’s mission, broadly stated, is to provide free access to the Internet and to develop software to advance that mission. Firefox does not provide its own search engine but instead uses a third-party provider as its default search engine. In 2005, Mozilla Foundation established Mozilla Corporation as a for-profit subsidiary to contract for search engine services. Until late 2014, Mozilla Corporation was contracting with and receiving payments from Google and was using Google as its default search engine. In 2012, for example, CNET reported that Mozilla Corporation received $311 million in revenue from Google. In 2014, Mozilla switched to Yahoo. As part of the deal, Yahoo agreed to support Do Not Track technology, which would enable users of Firefox to prevent tracking of their web use for advertising purposes.28 While Mozilla Corporation is wholly owned by Mozilla Foundation, a potential benefit of establishing a for-profit subsidiary is that the subsidiary can raise capital by selling shares to outside investors. Although nonprofits cannot issue equity directly, they can offer debt claims that have payoff structures that are somewhat similar to equity. For example, a nonprofit can raise capital by issuing revenue participation rights or other creative financing instruments that combine the properties of equity and debt. A quasiequity debt security is particularly useful for ventures that are legally structured as nonprofits since they cannot issue equity. Revenue participation rights are technically debt, but with returns that are linked to the organization’s financial performance. The investor in such an instrument does not have a direct claim on the voting control or ownership of the enterprise, but the loan is designed to provide greater returns if the nonprofit venture does well. The Bridges Social Entrepreneurs Fund in the U.K., for example, committed £1 million as a social loan to HCT Group, which uses surpluses from its commercial London buses and related services to provide community transportation for people who are unable to use conventional public transportation. The loan has a quasi equity feature in that the investor receives a percentage of revenues, thereby sharing some of the business risk and gains.29
New Venture Financing 61
2.4 Organizational Form and Financing Choices One of the early decisions an entrepreneur must make concerns the organizational form of the venture. The choice has implications for a variety of factors, including ability to attract financing, tax treatment, liability, succession, and ability to attract employees. The organizational form choice can be addressed most effectively by considering the growth potential of the venture and the availability of financing sources. Answering a few key questions can aid the process: Are not-for-profit status and the attendant tax exemption worthwhile? Should liability be limited, or should losses be passed on to the company’s owners? Is it important to be able to switch corporate forms easily as the company evolves? How important is it to avoid corporate-style taxation (i.e., double taxation)? Who are the best monitors of the firm—owners, investors, or managers? How will the monitors themselves be monitored? In the U.S. the most common organizational forms for for-profit entities are sole proprietorship, general partnership, limited partnership, limited liability company (LLC), S corporation (S corp), and C corporation (C corp). In the U.S., many not-for-profit firms are organized under Section 501(c)(3) of the Internal Revenue Code, which allows the company to avoid paying tax on earnings in excess of expenses. The rationale is that the excess is not distributed to owners but rather is retained to fund the ongoing activities of the enterprise. The organizational form choice can be changed, but not without cost. A venture at an early stage might be organized as an S corp or LLC because operating losses are expected and these forms allow losses to be passed through to owners. Later, when the venture needs to access larger capital sources and early owners are looking to exit, it may be reorganized as a C corp. Even a venture that starts as a 501(c)(3) firm can be reorganized as a taxable entity, but doing so is difficult. While tax exemption may seem like a tremendous advantage, when income is a small fraction of revenue the economic advantage can also be small. A forprofit venture has advantages associated with equity financing and more effective monitoring of managers. These benefits can more than make up for the tax benefits associated with not-for-profit or pass-through status. Assuming the venture under consideration is to be operated for profit, it is important to consider long-run capital needs. C corps and limited partnerships can raise large amounts of money from investors whose involvement with the enterprise may be passive. The other forms all restrict capital-raising ability by limiting the number of investors or limiting investment to parties who are actively involved in the venture. Important distinctions between a C corp and a limited partnership relate to transferability of ownership and tax treatment
62
Chapter Two
of earnings. A C corp, particularly if it is publicly held, can most easily raise capital from diverse groups of investors, and organizing as a C corporation facilitates transfer of ownership. Limited partnership interests are less easily transferable. However, the limited partnership structure avoids taxation of earnings at the entity level and facilitates structures that allocate taxable gains and losses among partners. Both forms offer limited liability to investors who are not actively involved in the venture. Limited liability is essential to large-scale fund-raising from investors who wish to play passive roles in the venture. If raising a large amount of capital is not important, a venture may be organized as a general partnership. Doing so subjects the partners to unlimited liability but enables venture earnings to flow through to the partners untaxed. Sometimes the number of partners can be so large and the activities of the enterprise so diffuse that the pure partnership form is an impediment to growth. In that case, an LLC form can preserve the tax advantages of partnership and still offer some protection against liability. Protection from liability can break down, however, if partners act negligently in monitoring the activities of the venture. The early choice of organizational form can be of strategic importance to a new venture. Sole proprietorships and small partnerships are relatively easy to convert to other forms, but as the venture grows and more parties become involved as partners or owners, ability to transition from one form to another decreases. If the firm is planning to raise venture capital or angel financing, then a C corp makes sense (and may be required by the investors). If there is no plan to raise equity, then an S corp or LLC is a good choice as it has tax benefits relative to a C corp and retains the liability protections.30
2.5 Regulatory Considerations Entrepreneurs and investors are confronted with a myriad of securities laws when considering how to finance a venture and whether to raise capital in the public markets. The laws governing access to the public capital markets, and to operating as a corporation with publicly tradable securities, are complex and fluid. In Table 2.2, we identify the main legislation that affects entrepreneurial ventures in the U.S. While we focus on the U.S., all countries have laws that affect public market access and that require regular financial disclosure. We identify the major components of U.S. federal securities laws; discuss state regulations (known as “blue sky laws”) that affect security issuance; describe the SEC’s securities registration process; and review some exemptions and safe harbors that may affect financing choices.
New Venture Financing 63 Tab le 2 . 2 Key U.S. securities market regulations for entrepreneurs and investors
Regulation
Description
Relevance to Entrepreneurs and Investors
Securities Act of 1933
Regulates the offering and sale of securities, establishes the law related to securities fraud, and sets requirements to register securities formally with the federal government. It delegates to the SEC the authority to enforce and interpret the 1933 act.
Securities and Exchange Act of 1934
Regulates aftermarket trading of public securities. It also establishes the law on insider trading and buying on margin. Regulates the organization of companies that engage primarily in investing and trading securities. The Act includes disclosure requirements including valuations and investment objectives. Reforms aimed at enhancing corporate responsibilities and financial disclosure.
Pertains to the primary offering of securities by a company and the ongoing disclosure of information once the securities are issued. Firms wishing to issue public securities must file a prospectus (Form S-1, or SB-1 for issues less than $10 million). Relates to the exchange of publicly traded securities after they have been issued, including reporting requirements. Requires closed-end mutual funds to report fair value of non-public securities, discouraging funds from investing in non-public firms.
Investment Company Act (ICA) of 1940
Sarbanes-Oxley Act (SOX) of 2002
Jumpstart Our Business Startups (JOBS) Act of 2012
Aims to help emerging growth companies raise funds in public capital markets by streamlining regulatory requirements for certain firms and offering sizes.
Reporting requirements are substantial and greatly increase the cost of being a public firm, which may impact the decision to go public, especially for smaller firms. Reg. A+ allows for public equity offerings for up to $50 million with streamlined regulations and reporting requirements. The “equity crowdfunding” provisions allow companies to sell securities through open platforms (e.g., Kickstarter, Indiegogo). The Act led to Regulation Crowdfunding in 2016.
The business failures and personal bankruptcies of the U.S. Great Depression led to enactment of the two federal laws that form the basis of regulation concerning the offering, sale, and trading of securities. The Securities Act of 1933 is the more relevant to start-ups, as it pertains to the primary public offering of securities by a company and the ongoing disclosure of information once the securities are issued. The Securities and Exchange Act of 1934 relates to exchange of publicly traded securities among investors after the securities have been issued. The 1933 Act seeks to foster transparency at the time of a company’s public offering of securities and afterward. Under it, corporations that desire to issue securities to the general public can do so only by filing a registration statement with the SEC (Form S-1, or, for small businesses seeking to raise $10 million or less, Form SB-1). The Act mandates dissemination of a disclosure document known as a prospectus and creates a complex liability scheme concerning
64
Chapter Two
misinformation in the registration statement or the prospectus or during the distribution of the securities to public investors. The Act also places specific limitations on the timing and content of pre-issuance communications. The 1933 Act has been amended numerous times and has become expansive in its reach and vision. The issuer’s responsibilities do not end with filing a registration statement and circulating a prospectus. After the IPO or public debt offering, the company is subject to ongoing reporting requirements under the 1933 Act and under Sarbanes-Oxley. Public companies must provide timely annual reports that include audited financial statements and are subject to other reporting rules that are intended to help level the playing field for investors. The cost of ongoing reporting can be a significant burden for a small company, a consideration that prevents all but the most promising ventures from going public. In the U.S., the Jumpstart Our Business Startups (JOBS) Act of 2012 was a major policy effort aimed at encouraging funding of small businesses by easing securities regulations. The changes include creating a way for companies to use crowdfunding to issue securities, which was not previously permitted. The JOBS Act includes changes in Regulation A (“Reg A”) by expanding the regulation into two tiers and increasing the threshold for its applicability. Under Reg A+ there are two tiers for securities offerings—tier 1 for offerings up to $20 million and tier 2 for offerings up to $50 million. The JOBS Act streamlined regulations and reporting requirements for firms making Reg A+ offerings. Securities sold and traded within a single state are insulated from federal regulation but still are subject to state law. The state laws that pertain to companies with publicly traded securities are known as “blue sky laws.” The phrase derives from the concern that shady operators would try to sell unsuspecting investors pieces of clear blue sky. As of now, every state in the U.S. has securities laws. Unlike federal securities laws, which constitute a disclosure-based regulatory system, some states use merit-based regulation—the individual states are empowered to decide whether an offering deserves the attention of investors. The state laws do not regulate securities exchanges, interstate trading markets, or disclosure by public companies in connection with shareholder trading or voting.
Exemptions from Registration in the U.S. Registration of securities can be very costly in terms of out-of-pocket expenses for both lawyers and underwriters and because of potential liability. The 1933 Act exempts from registration those transactions for which registration may
New Venture Financing 65
be unduly expensive given investor sophistication or the issuer’s modest financial needs. But even if a security is exempt, securities laws apply. Thus, the SEC can proceed against any issuer if there is misrepresentation. Exempt transactions include purely intrastate offerings, private placements, and certain small public offerings by the issuer. In 1982, the SEC consolidated its exemptions for small and private offerings into a set of rules known as Regulation D (“Reg D”), which were subsequently modified by the JOBS Act. Reg D contains three exemptions from the federal registration requirements, and thereby facilitates private placements of equity. 1. An exemption for small offerings (Rule 504). Companies can offer and sell up to $1 million of their securities to “accredited investors” in any 12-month period without registration. Accredited investors are high-networth individuals who are presumed not to need the investor protections of the 1933 Act. 2. An exemption for offerings of restricted securities to accredited investors or limited numbers of investors (Rule 505). Companies can offer and sell, without registration, up to $5 million of their securities in any 12-month period to an unlimited number of accredited investors and up to 35 other persons. Investors receive “restricted securities,” meaning that they cannot be sold for six months or longer without first being registered. 3. An exemption for private placements (Rule 506). Under the private offering exemption, companies can raise an unlimited amount of capital without registration and can advertise to the general public, provided they sell only to accredited investors and take reasonable steps to verify that all purchasers of an offering are accredited. Securities acquired through private placement cannot be sold for at least a year without being registered. Unlike the second exemption, all investors must be sophisticated— that is, they must have sufficient knowledge of and experience with financial and business matters to make them capable of evaluating the merits and risks of the prospective investment. An additional exemption from registration was created under Regulation Crowdfunding in 2016. This exemption covers offerings up to $1,000,000 (which is adjusted up over time with inflation) and requires that the offering be made through an SEC-registered online platform. Investors in crowdfunded issues are subject to limitations on the fractions of their net worth they can invest. Under this regulation there is a requirement for annual reporting to the SEC, but an audited financial statement is not required. Securities purchased via crowdfunding generally cannot be resold for a year.
66
Chapter Two
2.6 International Differences in Financing Options Resources to fund entrepreneurial ventures vary dramatically around the globe, in terms of both the availability of funds and the types of financing. The reasons for these differences include such infrastructure factors as enforceability of contracts, protection of minority interests, institutionalization of money management, and the economic wealth of the society. Financing sources are also affected by the nature of the entrepreneurial activity that is prevalent in the country and, conversely, source availability affects the nature of entrepreneurial activity. In the U.S., where a significant fraction of the activity is focused on high tech with the potential for very high returns, VC has emerged as an effective funding mechanism, and the availability of VC funding has fostered entrepreneurship. Even so, total funds provided to entrepreneurs by VC are quite small compared to, for example, bank loans. In emerging economies, entrepreneurial activity is more focused on launching and growing traditional enterprises that are relatively safe and can generate positive cash flow quickly. Because the mix of entrepreneurial activity varies so much, it is difficult to generalize about availability of financing sources. Based on survey data compiled by the World Bank to examine global patterns of financing of SMEs, external financing as a fraction of total financing does not depend on country infrastructure, but the mix of external financing does.31 Enterprises in less-developed countries are not as able to obtain external debt and equity financing and, therefore, rely more on operations-based financing such as trade credit. Greater reliance on vendor financing in emerging economies is not surprising since such economies tend to have limited availability of financing from other sources. According to World Bank data, in most low-income countries, loans are the only institutional source of financing, but the amount of loan financing is small in relation to GDP, which, in low-income countries, is also quite small on a per capita basis. For the most part, if entrepreneurs need funding in such countries, they are compelled to rely on non-institutional sources, such as personal wealth, friends and family, and vendors and strategic partners who are willing to extend credit. But from where do the companies that extend credit obtain their funds if bank lending in such countries is low? One important answer is that providers of credit are often foreign and have access to financing sources that local businesses do not. The outlook for entrepreneurs is much better for the high-income countries. Bank lending is high compared to GDP, but in most countries it is dominated by private and public debt issues and equity offerings. Of course, these are also the countries where financing sources such as trade credit, angel investment, and
New Venture Financing 67
VC are relatively high and where bootstrap funding and funding from friends and family are likely to be more significant.
2.7 Recap: Considerations in the Choice of Financing We began this chapter with 4 questions regarding the choice of financing: How urgent is the financing need? How large is it? Is it permanent or transitory? How does it relate to the cumulative need for financing? Now that we’ve reviewed the varied sources of new venture financing, it should be clear why these questions are important. One could rely, for example, on trade credit with expensive terms to finance a small, temporary financing need until the company can obtain external financing, but if the cost of paying late is high, it does not make sense to rely on trade credit for a sustained period. Mezzanine loans prior to an IPO are also a temporary form of financing and are viable only when a company is nearing an exit that will fund repayment of the loan. To add structure to the decision-making process, a useful starting point is to assess the current stage and condition of the venture. The realistic menu of financing alternatives is constrained by these factors. Additional considerations include the value of outside advice, the asset intensity of the venture, track record, the level and stability of earnings and cash flow, and influences of existing financing.
Stage of Development Financing typically begins with bootstrapping. Seed capital normally is external financing, such as angel investment, although as we have seen, angel investment can go beyond seed financing. Start-up financing provides funds to initiate operations. Later-stage financing options are used to finance growth before the venture is profitable or during the early-growth and expansion stages. Mezzanine financing and other forms of debt are best suited for ventures that are generating taxable income and approaching the conditions for harvesting. Exit financing, including buyout financing and public debt or equity issues, is designed to enable the entrepreneur and early-stage investors to realize returns on their investments.
Asset Intensity of the Venture The ability to collateralize assets is an important characteristic that affects the choice of financing. Even a high-risk venture with no track record and an
68
Chapter Two
nproven entrepreneur can attract debt financing if assets are available to seu cure the loan. These advantages to the lender translate into lower borrowing costs, because they commit the entrepreneur to dealing reasonably with future adversity and because they signal confidence. An offsetting concern is that a secured lender has little interest in the success of the venture and is unlikely to provide much useful advice other than ways to ensure that the loan is repaid.
Immediacy of Financing Needs If financial needs are immediate, most government-based financing is precluded, given the length of time to process applications. Similarly most equity funding sources are precluded, unless the entrepreneur has an established relationship with the investor. VC firms and angel investor groups can have lengthy due diligence and approval processes, especially for initial investments. The same is true of corporate venture funds. Similarly, late-stage financing like an IPO or an acquisition would take months of planning to complete. Even more expedient methods of raising equity, such as private placements, are unlikely prospects if needs are urgent. The most realistic sources of immediate financing are those for which little or no negotiation or approval process is required or for which financing is preapproved. For early-stage firms, the realistic alternatives are the personal resources of the entrepreneur, friends and family, loans secured by personal guarantees, and possibly existing shareholders. Early-stage ventures that are producing revenues can approach customers or suppliers for funds, but it is unlikely that a suitable agreement can be reached on short notice since such relationships are not sufficiently strong at that point. Temporary use of trade credit (by forgoing prompt payment discounts) and factoring are feasible if the firm is at a stage where it is producing revenue.
The Influence of Near-Term Financing Needs Two primary considerations influence the venture’s choice of near-term financing. First, the choice of near-term financing depends on the permanency of the need. Transitory needs are generally best financed with short-term sources. Second, the choice depends on the amount the venture is seeking. For small amounts, likely sources include the SBA, crowdfunding, banks, and private placements of equity with a few individuals or a group of angel investors. For larger amounts, private placements of debt with institutional lenders, VC investors, and strategic partners gain importance. If required financing is very large, the entrepreneur could consider a registered public offering of debt or equity.
New Venture Financing 69
The cost of short-term financing can be volatile. If the financing is used to acquire long-term assets, it may increase the uncertainty of cash flows or even threaten venture survival. Equity and long-term debt are the obvious sources for meeting long-term needs of large and established firms. The issue cost percentages clearly favor debt issues. Offsetting the lower cost of issuing debt are two important considerations: first, the issuance costs of debt are recurring whereas equity is a source of permanent financing; and second, compared to equity issues, debt issues limit flexibility.
The Influence of Cumulative Financing Needs If the entrepreneur expects that the venture will soon begin to generate positive cash flow sufficient to fund growth, then he can make near-term financing choices with little regard to their potential effect on the availability or cost of future financing. The important scenarios, however, are the extremes. How should near-term financing be arranged if the entrepreneur’s cumulative needs may be substantially higher or substantially lower than the venture’s ability to generate free cash flow? If cumulative needs are expected to be higher than present needs, current financing must not seriously impair the ability to raise capital later. Investor options to make additional investments at pre-agreed prices, or rights of first refusal for new financing, or financing arrangements that include other options and rights can be problematic for the new venture in the future. If such terms are accepted, their effects may be offset by other provisions, such as call options on the financing or buyout options that enable the venture to restore flexibility, such as by paying off the existing investors or lenders. If the entrepreneur expects future needs to be lower than near-term needs, the arrangements should enable the venture to pay off the financing with little or no penalty as cash becomes available. Call options and buyout rights are useful under these conditions, as well as when the venture is expected to grow rapidly.
Other Considerations That Affect Financing Choices Diversification of risk. Outside investors are generally capable of more fully diversifying risk. In contrast, entrepreneurs tend to be highly underdiversified, as they put much of their human and financial capital into their ventures. Because investors who can diversify do not need to be compensated for bearing idiosyncratic risk, outside investors can justify lower expected returns for a given level of total risk. The implication for choice of financing is compelling. All else equal, greater reliance on external financing can increase the value of
70
Chapter Two
the entrepreneur’s ownership stake. An offsetting consideration is the cost associated with managing relations with the investors, including reporting and access to information and the possible loss of control that can occur. Taxes. Not all organizations generate taxable income, and the applicable tax rate depends in part on organizational form. For example, tax rates and other aspects of taxation differ for corporations versus proprietorships. Also, the tax treatment of interest payments on debt differs from that of returns to equity investors. Debt is not attractive for many new ventures, partly because if they are not generating taxable income they may be unable to benefit from the tax deductibility of interest payments. The value of outside advice. An important consideration in the choice of early-stage financing is whether the venture can benefit from active involvement of an outside investor.32 Angel investors and VCs are active investors. Active involvement may be limited to monitoring the operation and serving on the board. Or it may be as extensive as stepping in as CEO for a period to help get the venture on track, making key strategic and tactical decisions, and helping to build the management team.33 In most cases, the active investor’s return comes mainly through ownership in the venture. The cost of any monitoring services provided by outside investors ultimately must be borne by the entrepreneur in the form of reduced ownership share. There are several reasons why it might be worthwhile for the entrepreneur to incur the cost of subjecting the venture to monitoring. One pragmatic reason is that monitoring is a precondition of getting outside capital from an angel or VC firm and that alternatives that do not involve monitoring are usually unavailable to early-stage, high-risk ventures. Overall, the empirical evidence on the value of monitoring suggests that both angels and VCs can add value beyond the benefits associated with just the financing.
Assessing the Effects of Financing Choices The effects of these financing considerations on the value of the entrepreneur’s claim are complex and sometimes counterintuitive. A sensible approach to making the financing choice is to model the venture, including revenue growth projections and underlying cost structure, and study the effects of alternatives on the value of the entrepreneur’s claim. More often than not, the realistically available choices are limited to a few. But even if the entrepreneur has few op-
New Venture Financing 71
tions, he will want to evaluate such variables as the amount of the funding, the form of the commitment the investor plans to make (e.g., debt versus equity, passive versus active commitment), and the effects of restrictions the investor imposes on future actions (through restrictive covenants or by other means).
2.8 How Financial Distress Affects Financing Choices The financial condition of a venture affects its financing opportunities. A profitable business with steady cash flow can arrange financing more easily than one that is struggling. For many new ventures, however, at some stages of development, struggling to meet financing needs is normal. The term “financial distress” means more than simply that a venture is in need of cash to fund its operations. A distressed firm is one that has fundamentally disappointed investors. If the investors are creditors, a distressed firm may have violated important debt covenants or may have defaulted on its repayment obligations or be about to do so. For equity-financed ventures, financial distress can mean that the venture is so short of cash that it is unable to carry out its business plan. The cash shortage can be symptomatic of missed milestones or other disappointing performance. Such a firm may be compelled to terminate or significantly scale back operations and may be at risk of losing key employees.
Why Turnaround Financing Is Different Financing distressed firms is different from financing high-risk start-ups for two reasons. First, distress means that the entrepreneur has already failed to achieve a level of success consistent with earlier projections. That failure can undermine the entrepreneur’s credibility. Although it is common to search for external causes for underperformance, the evidence suggests that the usual causes are fundamental failures of management. Management issues include lack of critical skills, excessive turnover, and similar deficiencies. Inadequate financial planning and controls can contribute to unanticipated cash shortages, poor pricing decisions, lack of control over costs, and similar problems. Given that the venture has failed in fundamental dimensions, why should investors assume that projections in a revised plan are achievable or that the entrepreneur is capable of managing the operation? In the U.S., an entrepreneur who has encountered trouble can take advantage of a legal presumption that the cause of financial distress, if not completely external, is curable. A Chapter 11 bankruptcy offers a window of protection against creditors’ demands for
72
Chapter Two
repayment of their loans. The underlying concept is that if a venture is fundamentally viable, then creditors, as a group, are better off remaining on board. The statute gives the entrepreneur time to try to achieve a reorganization plan that creditors find acceptable. This leads to the second reason why financing of distressed firms is different. The financial structures of most firms are based on a premise that the venture will be successful. Creditors, including trade creditors and even employees, expect that they will be paid. When a venture gets into trouble, the orientations of creditors change. Secured creditors begin to look more intently at the value of collateral as a source of repayment. They worry that if the entrepreneur is allowed to continue to operate the business, the collateral may depreciate. They may press for liquidation as a means of repayment. Suppliers to distressed firms no longer assume that they will eventually be paid for merchandise they sell on credit; they may demand cash payment. Employees may assume that their jobs are at risk and try to protect themselves by seeking new employment. Even customers may become concerned and stop purchasing. In the U.S., once a company declares bankruptcy, individual creditors can no longer attach the company’s assets. Instead, the assets become part of a common pool, and they are not dispersed unless the creditors reach an understanding about how the various debts will be settled. For example, the creditors as a group might conclude that the assets should be liquidated and the creditors paid out of the proceeds, or they might decide to refinance the failing company. The objective in turnaround financing is to devise a new financial structure such that each party expects to be better off by maintaining or increasing its investment than by collecting the liquidation value of its financial claims. Financial restructuring can involve significant changes in the financial claims held by different investors. Claims of creditors, for example, may be exchanged for common stock or for securities convertible to common. In general, restructuring must achieve three results: it must reduce the cash needs of the venture for servicing existing debt; it must reallocate the going-concern value toward creditors and away from the entrepreneur and other equity investors; and it must provide a means of raising enough cash to restart the venture.34 For some kinds of ventures, the real costs of financial distress are small. Business assets, though they may be specialized to a particular use or geographic location, are readily marketable, and success or failure of the venture does not depend on the special knowledge or skills of employees. In such cases, financial distress can result in little more than a reallocation of ownership claims and may not materially affect the revenue or profit stream of the venture. For other kinds of ventures, even the rumor of financial distress can have important real consequences. Key employees, who would be difficult to replace,
New Venture Financing 73
may resign. Customers may switch to alternative suppliers to avoid the possible consequences of dealing with a firm that might fail. Managers who anticipate failure may engage in excessive risk taking or may depreciate the assets of the venture. When the potential costs of financial distress would be large, earlier financing choices should be based, in part, on the value of avoiding financial distress. Among other things, this could imply raising a higher level of capital at the outset and using a financial structure that limits the risk of defaulting or violating debt covenants.
2.9 Summary Investing in new ventures is very different from investing in the shares of a public company. Because of the high degree of uncertainty about the potential for survival and success, financing commitments tend to be made in stages that are related to development of the venture. Customarily, development is described in terms of hypothesis testing and milestones that correspond to resolutions of significant uncertainty. Related to this approach to entrepreneurship, maximum value for the entrepreneur will not be achieved by raising all of the anticipated financing needs at one time. As business model hypotheses are tested and validated, the riskiness of successful ventures declines, and staging can produce higher expected value for the entrepreneur. Appropriate forms of financing depend on the stage of new venture development. Financing typically begins with bootstrapping (tapping the resources of the entrepreneur and friends and family). Seed capital normally is external financing, such as angel investment, although angel investment can go beyond seed financing. High-tech ventures with long development stages may require specific R&D financing. Start-up financing provides funds to initiate operations. Financing options, such as VC, are used to finance growth before the venture is profitable or during the early growth and expansion stages. Mezzanine financing and other forms of debt are best suited for ventures that are generating taxable income. Exit financing, including buyout financing, acquisition, and public debt or equity issues, is designed to enable early-stage investors to realize the returns on their investments. This chapter describes some of the main financing choices available to the entrepreneur, defines terms and concepts that are used frequently in the industry, and identifies considerations that bear on the choice, such as securities regulations, and institutional differences in capital markets across countries. We also provide an approach for assessing financing needs and the timing of those needs.
74
Chapter Two
Review Questions
1. Identify types of financing that are associated with each of the following stages of new venture development: research and development, start-up, early growth, rapid growth, and exit. 2. Why are the following important factors when choosing among financing sources: immediacy of financing need, the size of the need, the permanency of the need, and the cumulative need? 3. What, besides providing financial capital, are the advantages to a new venture of attracting angel group financing? Venture capital financing? 4. What factors should an entrepreneur consider when differentiating between CVC and VC investors? 5. When is vendor-provided credit (trade credit) an important form of financing for new ventures and how do the credit terms affect the cost of using trade credit? 6. How do government programs encourage new venture development? Give examples of government programs that facilitate financing of new ventures. 7. Why do small firms, especially new ones, have limited access to the public debt market? Under what conditions are ventures likely to be able to access the bond market? 8. What are the differences in financing choices for for-profit firms and not-for-profit firms? 9. Identify the basic U.S. securities laws that affect new venture financing options. How has the JOBS Act (2012) affected new venture financing? 10. What are accredited investors and why are they important to new ventures seeking registration of their equity securities? 11. How is turnaround financing different from financing associated with a healthy growing firm? Notes
1. Chatterji and Seamans (2012) illustrate the importance of credit cards to entrepreneurial activity. They study the impact of credit card deregulation and find that deregulation increases the probability of entrepreneurial entry, with a particularly strong effect for black entrepreneurs. 2. See Stouder (2002), who analyzes data on 74 entrepreneurs. 3. Reported by Robb and Robinson (2014), who base their work on the large-scale Kauffman Foundation Firm Survey.
New Venture Financing 75
4. Gates (1995). 5. Drover, Busenitz, Matusik, Townsend, Anglin, and Dushnitsky (2017). 6. For comparison of accelerators and incubators, see Atkins (2011). Universities and colleges also offer incubating services, which allows them to leverage access to faculty and student research. The scale of support and terms of investment vary considerably across programs. As examples of prominent incubators, see the University of Texas, https://ati.utexas.edu/, and the University of Chicago, https://polsky.uchicago.edu/programs- events/polsky -incubator/. 7. Strauz (2017) models the trade-off between the benefits that crowdfunding provides entrepreneurs by reducing demand uncertainty and allowing them to screen more effectively for valuable projects, and to gauge the expected costs associated with the entrepreneur’s incentives for moral hazard. 8. A few angels may commit to significantly larger amounts ($500,000 to $3 million). Angels in this category include people like Mitch Kapor, founder of Lotus; Paul Allen, cofounder of Microsoft; and Ron Conway, a Silicon Valley angel investor who was an early-stage investor in Google, Ask Jeeves, and PayPal. 9. See Itenberg and Smith (2017) for further discussion. 10. See Hellmann (2002) concerning the effects of conflicts of interest arising in corporate venturing. He suggests that VC firms are less prone to such risks. 11. Kwan and Carleton (2010) report that in a sample of privately placed loans by corporations, a large fraction were renegotiated. In fact, one reason for a firm to place a large loan privately is that, compared to public debt, privately placed debt adds the flexibility of being able to renegotiate covenants. See also Boot, Gopalan, and Thakor (2006). 12. Cornelli and Yosha (2003) study the role of convertible securities in new ventures with staged financing. They suggest that with straight debt, the entrepreneur may have an incentive to “window dress” to attract the next round of financing. Financing with convertible securities reduces the entrepreneur’s incentive to engage in such actions because the lender shares in equity returns. 13. Research by Brown and Earle (2017) shows that each million dollars of loans supplies an increase of 3–3.5 jobs. They estimate that taxpayer cost per job created is in the range of $21,000–$15,000 per job. 14. In addition to the SBA, the Farmers Home Administration, the Department of Commerce, the Department of Energy, the Department of Housing and Urban Development, and the Department of the Interior all offer loans or loan-guarantee programs. Other state and federal government
76
Chapter Two
nancing sources include state business and industrial development corporafi tions, the Export-Import Bank of the United States, and the Small Business Innovation Research (SBIR) program. 15. SBICs generally invest their own money, along with borrowed funds covered by SBA loan guarantees. Investments that can generate steady streams of cash flow are most appropriate for these sources. 16. Craig, Jackson, and Thomson (2009) find evidence of a small positive impact of the SBA’s guaranteed lending programs on the financial performance of companies that receive guaranteed loans. 17. See Petersen and Rajan (1997). Ng, Smith, and Smith (1999) analyze trade credit terms across industries. 18. The effective annual rate is calculated as: (1 + discount percent)(365/credit period) − 1. 19. Huyghebaert, Van de Gucht, and Van Hulle (2007) find that entrepreneurs prefer trade credit over bank debt, even if it is more expensive. They argue that suppliers are more willing to negotiate with borrowers in the event of default, rather than compelling them to liquidate. 20. Mian and Smith (1992) and Smith and Schnucker (1994) examine firms’ choices to factor their receivables. Factors are more likely to be used by firms when information costs about buyers and monitoring costs are high. 21. Some states have classified Uber drivers and FedEx Ground operators as employees for specific purposes such as requiring the employer to provide benefits. The legal determination of employment status, however, does not affect the fundamental economic nature of the relationship where the venture is funding its activities partly through resources provided by the operators. 22. Brickley and Dark (1987). 23. See Gompers (1995); Barry, Muscarella, Peavy, and Vetsuypens (1990); and Lin and Smith (1998). 24. See https://site.warrington.ufl.edu/r itter/ipo- data/. 25. Early-stage investors are often precluded from selling shares they own in or during the IPO. For the most part, they harvest by selling in the public market after the shares have been trading for several months. 26. The cost of a public offering has three components: the spread between the offer price and net proceeds to the issuer (which constitutes the underwriter’s fee), issue costs borne directly by the issuing firm, and underpricing. For equity issues, in addition to the fee and expenses, it is standard practice for the underwriter to set the offering price to around 15% below what the market price is expected to be at the time of the offering. If the underpricing is considered to be part of the issue cost, then the total cost of an
New Venture Financing 77
IPO, as a percentage of net proceeds to the issuer, could be in the 15–30% range, depending on actual underpricing, issue size, and other factors. 27. Many online sources provide listings of nonprofit funding opportunities, including corporate funding; see http:// w ww .fundraiserhelp .com/ corporate-grants-source-list.htm. 28. For links to relevant references see, https:// en .wikipedia .org/ w iki/ Mozilla_Foundation. 29. See Bugg-Levine, Kogut, and Kulatilaka (2012). 30. In terms of absolute numbers of business entities, small proprietorships dominate, especially those with annual receipts less than $1 million. But C corps become more prevalent as total business receipts increase and account for the majority of total business receipts. S corps, LLCs, and partnerships, the flow-through entities, are more prevalent in the midsize receipt classes. See the IRS’s Statistics of Income Bulletin (annual). 31. See Beck, Demirgüç-Kunt, and Maksimovic (2008). 32. Bettignies and Brander (2007) model the entrepreneur’s choice between bank and VC financing. Their model implies that VCs cannot survive purely as financial intermediaries and that VC funding will be most attractive to entrepreneurs who value the skill set or industry knowledge provided by the VC. Ueda (2004) compares bank and VC investment decision making and monitoring. See also Winton and Yerramilli (2008). 33. Inderst and Mueller (2009) model the impact of active investor involvement on new venture growth. They observe that involvement by active investors can create a competitive advantage over rivals. The advantage comes through better and faster information conveyance, quicker shutdown of less promising ventures, and faster growth of ventures with better prospects. 34. Pindado, Rodrigues, and de la Torre (2006) find that for nondistressed firms, the amount of long-term debt is related to both tax shields and insolvency costs. These relationships do not hold for distressed firms. They conclude that equity issues and renegotiation with creditors are used by distressed firms to reduce leverage.
References and Additional Reading
Atkins, D. 2011. “What Are the New Seed or Venture Accelerators?” http:// www.nbia.org/resource_ library/review_ archive/0611_01.php. Barry, C., C. Muscarella, J. Peavy, and M. Vetsuypens. 1990. “The Role of Venture Capital in the Creation of Public Companies: Evidence from the Going Public Process.” Journal of Financial Economics 27: 447–71.
78
Chapter Two
Beck, T., A. Demirgüç-Kunt, and V. Maksimovic. 2008. “Financing Patterns Around the World: The Role of Institutions.” Journal of Financial Economics 89: 467–87. Bettignies, J., and J. Brander. 2007. “Financing Entrepreneurship: Bank Finance Versus Venture Capital.” Journal of Business Venturing 22: 808–32. Boot, A. W. A., R. Gopalan, and A. Thakor. 2006. “The Entrepreneur’s Choice between Private and Public Ownership.” Journal of Finance 61: 803-836. Brickley, J., and F. Dark. 1987. “The Choice of Organizational Form: The Case of Franchising.” Journal of Financial Economics 18: 401–20. Brown, J. D. and J. Earle. 2017. “Finance and Growth at the Firm Level: Evidence from SBA Loans.” Journal of Finance 72: 1039–80. Bugg-Levine, A., B. Kogut, and N. Kulatilaka. 2012. “A New Approach to Funding Social Enterprises.” Harvard Business Review (January–February): 118–123. Chatterji, A., and R. C. Seamans. 2012. “Entrepreneurial Finance, Credit Cards, and Race.” Journal of Financial Economics 106: 182–95. Cornelli, F., and O. Yosha. 2003. “Stage Financing and the Role of Convertible Securities.” Review of Economic Studies 70: 1–32. Coyle, J. F., and J. M. Green. 2014. “Contractual Innovation in Venture Capital.” Hastings Law Journal 66: 133–83. Craig, B. R., W. E. Jackson III, and J. B. Thomson. 2009. “The Economic Impact of the Small Business Administration’s Intervention in the Small Firm Credit Market: A Review of the Research Literature.” Journal of Small Business Management 47: 221–31. Dempwolf, C., J. Auer, and M. D’Ippolito. 2014. Innovation Accelerators: Defining Characteristics Among Startup Assistance Organizations. Report issued by the SBA Office of Advocacy. http://w ww.sba.gov/advocacy. Drover, W., L. Busenitz, S. Matusik, D. Townsend, A. Anglin, and G. Dushnitsky. 2017. “A Review and Road Map of Entrepreneurial Equity Financing Research: Venture Capital, Corporate Venture Capital, Angel Investment, Crowdfunding and Accelerators.” Journal of Management 43: 1820–53. Ewens, M., R. Nanda, and M. Rhodes-Kropf. 2017. “Cost of Experimentation and the Evolution of Venture Capital.” Harvard Business School working paper 15-070. Journal of Financial Economics, forthcoming. Gates, B. 1995. The Road Ahead. New York: Viking. Gompers, P. 1995. “Optimal Investment, Monitoring, and the Staging of Venture Capital.” Journal of Finance 50: 1461–89. Gompers, P., and J. Lerner. 2001. “The Venture Capital Revolution.” Journal of Economic Perspectives 15: 145–68.
New Venture Financing 79
Hellmann, T. A. 2002. “A Theory of Strategic Venture Investing.” Journal of Financial Economics 64: 285–314. Huyghebaert, N., L. Van de Gucht, and C. Van Hulle. 2007. “The Choice Between Bank Debt and Trade Credit in Business Start-Ups.” Small Business Economics 29: 435–52. Inderst, R., and H. M. Mueller. 2009. “Early-Stage Financing and Firm Growth in New Industries.” Journal of Financial Economics 93: 276–91. Itenberg, O., and E. Smith. 2017. “Syndicated Equity Crowdfunding: The Trade-off Between Deal Access and Conflicts of Interest.” Simon Business School working paper FR 17-06. https://papers.ssrn.com/sol3/papers.cfm ?abstract_id=2933822. Kerr, W. R., R. Nanda, and M. Rhodes-Kropf. 2014. “Entrepreneurship as Experimentation.” Journal of Economic Perspectives 28: 25–48. Kerr, W. R., and R. Nanda. 2015. “Financing Innovation.” Annual Review of Financial Economics 7: 445–62. Kwan, S. H., and W. T. Carleton. 2010. “Financial Contracting and the Choice Between Private Placement and Publicly Offered Bonds.” Journal of Money Credit and Banking 42: 907–29. Lerner, J., A. Schoar, S. Sokolinski, and K. Wilson. 2018. “The Globalization of Angel Investments: Evidence Across Countries.” Journal of Financial Economics 127: 1–20. Lin, T. H., and R. L. Smith. 1998. “Insider Reputation and Selling Decisions: The Unwinding of Venture Capital Investments During Equity IPOs.” Journal of Corporate Finance 4: 241–63. Mian, S. L., and C. W. Smith Jr. 1992. “Accounts Receivable Management Policy: Theory and Evidence.” Journal of Finance 47: 169–200. Ng, C., J. K. Smith, and R. L. Smith. 1999. “Evidence on the Determinants of Credit Terms Used in Interfirm Trade.” Journal of Finance 54: 1109–29. Petersen, M. A., and R. G. Rajan. 1997. “Trade Credit: Theories and Evidence.” Review of Financial Studies 10: 661–91. Pindado, J., L. Rodrigues, and C. de la Torre. 2006. “How Does Financial Distress Affect Small Firms’ Financial Structure?” Small Business Economics 26: 377–91. Robb, A., and D. Robinson. 2014. “The Capital Structure Decisions of New Firms.” Review of Financial Studies 27: 153–79. Smith, J. K., and C. Schnucker. 1994. “An Empirical Examination of Organizational Structure: The Economics of the Factoring Decision.” Journal of Corporate Finance 1: 119–38. Stouder, M. D. 2002. “The Capital Structure Decisions of Nascent Entrepreneurs.” PhD dissertation, Rutgers University.
80
Chapter Two
Strauz, R. 2017. “A Theory of Crowdfunding: A Mechanism Design Approach with Demand Uncertainty and Moral Hazard.” American Economic Review 107: 1430–76. Ueda, M. 2004. “Banks Versus Venture Capital: Project Evaluation, Screening, and Expropriation.” Journal of Finance 59: 601–21. Winton, A., and V. Yerramilli. 2008. “Entrepreneurial Finance: Banks Versus Venture Capital.” Journal of Financial Economics 88: 51–79.
C h a p t e r Th r ee
Ve ntu r e Capital an d An g e l I nve sti n g
T h e t e r m “ v e n t u r e capital” is sometimes used in a generic sense to mean
capital invested in new ventures. It more accurately refers to capital supplied by a specific type of financial institution, the venture capital (VC) firm. VC firms have a unique organizational structure and focus on a specific market niche of high-risk ventures with potential for rapid and significant growth. Most new ventures are not suitable for VC investing, and accordingly, most capital supplied to new ventures does not come from VC firms. On the other hand, VC has been an early source of financing for many successful companies and is an important driver of economic growth. As discussed in Chapter 2, angel investing historically has focused on seed-stage companies to help finance their early development, and VC investing generally follows as firms reach a rapid-growth stage and require larger amounts of financing to sustain growth. In recent years, however, the lines have blurred, as some VC firms have funds that focus on seed-stage financing and some angel investors, through organized groups and syndicates, have increased their presence in seed-stage investing, but also delve into growthstage investments. Accordingly, in this chapter, we discuss both angel and VC investing. We review the evidence on the structure of the markets in which VCs and angels participate, their business practices, their financial performance, and the impact of their involvement with young firms. Chapter 4 provides an economic analysis of contractual terms used in the venture “deals” between entrepreneurs and VCs and angels, and the types of financing instruments and provisions they use. Understanding the natures of VC firms and angel organizations is important for any entrepreneur who plans to undertake a venture that has the potential 83
84
Chapter Three
for rapid growth and for anyone considering making an angel investment or an investment in a VC fund. The entrepreneur’s concerns are where to search most effectively for financing and what to expect from the relationship. Entrepreneurs whose ventures are not well suited for VC financing, for example, can easily waste time and effort seeking funding from firms that, because of their orientation, are unlikely to find the entrepreneur’s enterprise attractive. VC firms, angel groups, and syndicates are of interest for other reasons. First, the contract structures they use, both when they invest and when they raise capital from investors, illustrate market-based solutions to many of the information and incentive problems that are endemic to new venture financing. Second, new venture financing structures and terms are evolving rapidly. Being aware of the markets for early-stage investing can provide a useful frame of reference for examining and evaluating new organizational forms and practices and for evaluating opportunities for innovation in entrepreneurial finance. VC is a particular type of private equity. Private equity refers to any investment in equity where the financial claims cannot be readily traded on an organized exchange. This includes VC-style investments in early-stage companies but can also include investments in unregistered shares of public companies that have freely tradable registered shares. Firms that are engaged as general partners in VC funds may also be engaged in other kinds of private equity funds such as later-stage growth funds. Private equity funds that are focused on buyouts may also allocate a portion of their investment capital to VC deals. Investments in private equity can take the form of leveraged buyouts, VC, distressed debt investments, mezzanine capital, and so forth. With a VC investment, the firm typically invests in young start-ups with growth potential and generally does not obtain majority control. In contrast, when a private equity firm undertakes a leveraged buyout (LBO), the firm seeks to acquire majority control of an existing firm. Given that we are most interested in financing arrangements for new ventures, our focus is on VC as a subcategory of private equity.
3.1 Development of the Venture Capital Market The first modern VC fund in the U.S. was organized in 1946 by American Research and Development (ARD). ARD’s fund was organized as a closed-end mutual fund. In contrast to the VC funds of today, it was open to investment by any investor. In contrast to open-end funds, where new investors can join at any time by purchasing new shares of the fund, the total capital investment in
Venture Capital and Angel Investing 85
a closed-end fund is fixed. After initial sale of the shares by the fund, a person who wishes to invest must buy existing shares from someone else who owns them and wants to sell. ARD established the practice, which persists today, of searching for high-risk deals with the potential for big wins. The fund’s early-stage investment in just one of its ventures, Digital Equipment Company, accounted for roughly half of the entire return to fund investors over a period of more than two decades.1 Several other funds that were launched during the 1950s and 1960s imitated the structure and orientation of ARD. As documented later in this chapter, early growth of the VC industry in the U.S. was slow, restricted by regulation, tax policy, case law, and professional investment management practices.
The Investment Company Act and the SEC In the U.S., mutual funds and other investment companies are regulated under provisions of the Investment Company Act of 1940 (ICA), enforced by the SEC. Based on ARD’s early success, it might appear that mutual funds would be a viable vehicle for VC investment. However, because of SEC interpretations of some provisions of the ICA in the late 1960s, the ARD style of closed-end mutual fund was (and still is) largely precluded for VC investment. Specifically, the SEC required that the “fair values” of fund investments be determined “in good faith by the fund’s board of directors.” Thus, reliance on consultants to determine fair value would not absolve the board of potential liability for reporting valuations later determined to be inconsistent with fair value. Particularly detrimental to VC, the SEC equated fair value with liquidation value, even with regard to assets where the fund has no near-term intent to liquidate, and the SEC prohibited reliance on formulaic approaches to estimating fair value. The SEC’s emphasis on liquidation value and good-faith liability of fund directors precipitated a withdrawal of closed-end funds from investing in private equity. Over most of the following decade, the VC market languished. Closedend funds effectively were foreclosed as vehicles for investing in VC. Consistent with the impact of the SEC’s interpretations of the ICA fair value standard, Figure 3.1 shows that new capital commitments to VC generally declined from 1969 through 1977.2 Capital commitments are reputation-enforced promises to invest up to a specified amount in the fund. The decline over this period is 77%, from $171 million in 1969 to $39 million in 1977. To show percentage volatility more clearly, the figure is presented in log scale in Panel A. Panel B is in non-log scale.
86
Chapter Three Panel A $1,000,000
Millions of dollars (log scale)
$100,000
$10,000
$1,000
$100
1995
1997
1999
2001
2003
2005
2007
2009
1997
1999
2001
2003
2005
2007
2009
2017
1993
1995
2015
1991
1993
2013
1989
1991
2011
1987
1989
1985
1987
1983
1981
1979
1977
1975
1973
1971
$1
1969
$10
Panel B $120,000
Millions of dollars
$100,000 $80,000 $60,000 $40,000
2017
2015
2013
2011
1985
1983
1981
1979
1977
1975
1973
1971
$0
1969
$20,000
Fi g u r e 3 .1 New commitments of VC in the United States: 1969–2017 Data through 2005 are drawn from the following sources: U.S. Census Bureau, Statistical Abstract of the United States (various issues); Thomson Reuters Venture Capital data (http://thomsonreuters.com); Sahlman (1990); National Venture Capital Association (http:// www.nvca.org). Data from 2006 on are from the 2018 PitchBook-NVCA Venture Monitor.
sources:
Panel A is presented in log scale to show percentage volatility more clearly. Panel B is in levels.
VCs could avoid being subject to SEC interpretations of the ICA by taking advantage of safe harbor provisions in the SEC Acts (discussed in Chapter 2). The safe harbor provisions drove the firms that were interested in raising capital for investment in nonpublic companies to organize as limited partnerships and
Venture Capital and Angel Investing 87
to only raise capital from investors who are deemed by the SEC to be sophisticated and not requiring the protections of the SEC Acts. The VC limited partnership structure that is common in the U.S. today is designed to fit within the safe harbors. To be able to invest in a VC fund and other alternative private equity funds, an investor either must be a qualified institutional buyer such as a pension fund, endowment, or insurance company or must be an accredited individual investor. To be considered an accredited investor, an individual must qualify under either an income test or a wealth test that only about 7% of the U.S. population can pass. Thus, overwhelmingly, individual investors are precluded from investing in VC. Recognizing that many individuals would like to allocate a portion of their wealth to VC and private equity, in recent years, several prominent private equity firms (e.g., KKR and Blackstone) have completed IPOs by selling a partial ownership interest in their management companies. Investing in the management company provides somewhat of a workaround of the ICA and allows individual investors to access alternative asset classes. By investing in the management company, the investor shares some of the management fees and carried interest associated with the firms’ private equity funds.
The Prudent Investor Standard and the Employee Retirement Income Security Act (ERISA) An additional problem impeded the early growth of formal VC limited partnerships. To achieve their intended purpose, VC funds need access to patient capital from investors who are able to (1) accept illiquidity and high risk in assets with long holding periods and (2) reliably commit to providing investment capital when called upon to do so. The natural sources of such funds are wealthy individuals and large institutional investors such as pension funds and endowments. Historically, however, professional asset managers with fiduciary liability viewed themselves as being largely precluded from investing in high-risk or nonmarket assets. The constraints were derived from an array of old-fashioned state and federal regulations, legal precedents, and investment policies. One specific constraint was the traditional “prudent person” standard of fiduciary responsibility. Under this standard, responsibility to clients was based on the risk of each investment separately, without regard to portfolio diversification. Using this standard, courts would attempt to assess whether each investment was appropriate for the manager’s client (e.g., an employee participant in a retirement plan). Investment managers who were held to such a standard generally avoided investing in anything but safe securities such as “blue chip” stocks.
88
Chapter Three
Significant changes in the VC market in the U.S. began to occur in the late 1970s. First, in 1979 the Department of Labor reinterpreted the prudence standards governing investments by pension funds. The reinterpretation enabled managers to view the risk of an individual investment in the context of its contribution to the overall risk of the pension fund portfolio. Under the new prudent investor standard, pension funds could invest in securities that traditionally had been viewed as too risky (imprudent). The result was a rapid increase in the level of pension fund investments in VC funds. Other asset managers with fiduciary responsibilities similar to those of pension fund managers, such as managers of endowment funds, have adopted the new prudent investor standard. Second, reduction of the capital gains tax rate in the Revenue Act of 1978 spurred an increase in entrepreneurial activity. Workers had more incentive to leave salaried employment to pursue new ventures in which the principal form of compensation would be a capital gain on sale of the venture’s stock. The rate reduction also increased incentives to invest in financial instruments that produce capital gains.3 Corresponding to the capital gains tax rate reduction, annual new commitments to VC funds increased from $39 million in 1977 to $600 million in 1978. After 1978, as shown in Figure 3.1, the level increased dramatically, reaching $106.1 billion in 2000 before dropping precipitously to $3.9 billion in 2002. New commitments grew again during the mid-2000s, reaching $35.5 billion by 2007 before again starting to decline. In 2016, the industry had its best fund-raising year in more than a decade, raising $40.6 billion across 277 funds, still well below the high water mark in 2000. The level of new VC commitments each year is related to what is happening in the public equity markets. The early 1980s and much of the 1990s were periods of rising stock market values, high levels of new investment in the capital markets, and high levels of IPO activity. The late 1980s was a period of lower economic growth and a generally less active capital market. The decline beginning in 2000 corresponds to the collapse of the dot-com industry. The growth of new commitments during the 2000s and the decline in 2008 parallel stock market performance during the period, as does the recent uptick in venture commitments. Despite its rapid growth, the VC market in the U.S. remains small relative to public equity and debt markets. Investment-grade bonds issued in the U.S. in 2016, for example, were $1.27 trillion. Moreover, VC commitments in 2016 were comparable in size to SBA loan guarantees to small businesses, which in the 2017 U.S. budget totaled $46 billion. VC funding can dry up quickly when the equity capital market declines. The data underlying Figure 3.1 shows that new commitments dropped precipitously
Venture Capital and Angel Investing 89 $120
12,000
10,000
$100 Dollars invested (billions)
8,000
$60
6,000
$40
4,000
$20
2,000
Number of companies
Dollars invested (billions)
Number of companies $80
0 2017
2016
2015
2014
2013
2012
2011
2010
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1995
1990
1985
1980
1970
$0
Fi g u r e 3 . 2 VC investments in the United States: 1970–2017 Data through 2005 are drawn from the following sources: National Venture Capital Association (2009); PricewaterhouseCoopers/ National Venture Capital Association MoneyTree Report (https://w ww.pwcmoneytree.com), based on data from Thomson Reuters. Data from 2006 on are from the 2018 PitchBook-NVCA Venture Monitor.
sources:
The figure shows the total dollars invested by VCs and the number of companies invested in over time.
in 2009, down by 60% from 2008. However, the impact on the flow of investment capital from new commitments to new investments can be spread over several years. Figure 3.2 shows the decline in new investments from 2008 to 2009. This 28% decline is much less dramatic than the decline in new commitments.
Worldwide Growth of Venture Capital Institutional factors similar to those that slowed the early growth of organized VC in the U.S. have restrained growth in other parts of the world. The pension funds that invest in VC in the U.S., for example, are defined-benefit plans in which asset allocation choices are made by a board that acts as the agent of plan participants. In many countries, defined-benefit retirement plans do not exist or account for only a small fraction of retirement savings. Some countries rely primarily on redistribution systems similar to the U.S. Social Security system, which is unfunded; instead, current beneficiaries receive payments from the contributions paid in by current workers. In some countries, retirement plans have formally imposed restrictions on investment choices or are required to invest heavily in the sovereign debt of the country or securities selected for political reasons. Similar restrictions sometimes limit the investment choices available to insurance companies, another important source of VC funding in the U.S.
90
Chapter Three
In general, the factors that limit availability of funding to VC are more onerous in other countries than in the U.S. One response has been the introduction of VC funds that raise capital from U.S. investors for the purpose of investing in ventures in developing economies. Another has been that VC firms outside the U.S. tend to raise funds from traditional financial institutions, such as commercial banks. When they do, the management practices of the VC firms tend to be more conservative and based partly on the collateral or personal guarantee of the entrepreneur. Venture capital in Europe. Many factors influence the level of VC activity in a country or region. Generally, the level of formal activity is higher in developed economies than in emerging economies. Europe ranks second in overall VC activity. As shown in Figure 3.3, VC new commitments in 2008 reached over €9 billion and declined by more than half during the financial market collapse. Similarly to the U.S. experience, the 2016 level of new commitments to VC in Europe, at almost $10 billion, modestly exceeds the level in 2008. As a comparison, in 2016 the level of commitments in Europe is about one fourth of the level in the U.S.4 In general, the level of VC activity is positively related to the economic development and financial health of a country. Figure 3.4 shows investments by VCs in portfolio companies that are located in Europe, expressed as percentages of
€10 €9 €8
Billions of euros
€7 €6 €5 €4 €3 Fi g u r e 3 . 3 European new VC commitments: 2008–2017 PitchBook 2017 European Venture Annual Report. source:
€2 €1 €0
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
Venture Capital and Angel Investing 91
0.07% 0.06% 0.05% 0.04% 0.03% 0.02%
Greece
Ukraine
Romania
Luxembourg
Czech Republic
Italy
Poland
Portugal
Austria
Bulgaria
Netherlands
Baltic Countries
Norway
Hungary
Belgium
Germany
United Kingdom
Spain
Denmark
0.00%
France
0.01% Sweden
The data are compiled by the portfolio company and not by the location of the VC fund.
0.08%
Switzerland
source: Invest Europe: https://w ww.investeurope .eu/media/651727/i nvesteurope -2016 - european -private - equity-activity -final.pdf.
0.09%
Ireland
New VC investments in Europe as percentages of country 2016 GDP
Finland
Fi g u r e 3 . 4
country GDP. Investors can be local or international. Standardizing by GDP allows us to evaluate the importance of new venture activity to each country. While the UK, Germany, and the Netherlands, as examples, have much larger absolute VC investments, Ireland has outpaced other countries in VC investments as a percentage of GDP. Ireland’s climate for new companies is very favorable, as reflected in its corporate tax rate, which is one of the lowest in Europe, and in other policies aimed at attracting and retaining new ventures. The primary areas of investment are life science and tech, including fintech, traveltech, and edtech.5
The Changing Role of Venture Capital In Chapter 2 (Figure 2.2), we presented evidence that the sector focus of VC investment has changed over time in response to fluctuations in technological innovation and changes in the need for investment capital for high-risk ventures. The stage focus has also changed. In the U.S. in the 1980s and early 1990s, as shown in Figure 3.5, around 40% of investment dollars went predominantly to seed/start-up and early-stage ventures. In more recent years, after the 2000 tech rally ended, VC funds have focused their resources more on late-stage investments. For the period after 2000, the proportion of resources devoted to seed/ start-up and early-stage investments has averaged around 35%. Although the percentage allocation of investment dollars has remained small in recent years, the number of seed and early-stage investments by VC funds has increased. In 2008, VC funds invested $1.5 billion in seed/start-up ventures,
Chapter Three 100%
Fi g u r e 3 . 5
80% Late-stage
70%
Early-stage
60%
Seed/start-up
50% 40% 30% 20% 10%
2016
2014
2012
2010
2008
2006
2004
2002
2000
1998
1996
1994
1992
1990
1988
1986
1984
0%
1980
Data through 2005 are drawn from National Venture Capital Association, Yearbook (various years); PricewaterhouseCoopers/National Venture Capital Association MoneyTree Report (https://w ww.pwcmoney tree.com). Data from 2006 on are from the 2018-Q1 PitchBook-NVCA Venture Monitor.
sources:
90%
Percentage of new commitment dollars
U.S. VC investments by stage of development: 1980–2017
1982
92
compared to $9 billion invested by angels. More recently, the number of seedstage investments has increased. This reflects changes in technology and the emphasis on experimentation at the development stage. Ewens, Nanda, and Rhodes-Kropf (2017) describe the change in investment strategy as “spray and pray,” where VCs and other early-stage investors provide small amounts of funding and limited oversight to a large number of start-ups, most of which they abandon when the experiment fails, but a few of which are very successful. They associate this change in strategy with the advent of inexpensive cloud computing. This innovation has allowed start-ups that leverage the cloud to shift their larger capital requirements to later stages, when some of the early-stage uncertainty has been resolved. Another notable change over time is the increasing presence of corporate venture capital in the VC market. Figure 3.6 provides information on CVC involvement over the 12 years ending in 2017. While the percentage of VC deals with CVC involvement has remained in the 10–15% range, CVC has increasingly focused on the larger deals so that the number of deals with CVC involvement had risen from 24% of total VC investment dollars in 2009 to 46% in 2016.
Who Invests in Venture Capital Funds? VC requires investors who are able to commit substantial amounts of capital over long periods and who do not regard the illiquidity of their VC investments as an important cost. As of 2003, the latest available year, the primary sources of funding for VC in the U.S. were pension funds (43%), endowments
Venture Capital and Angel Investing 93 Fi g u r e 3 . 6
50%
Corporate VC involvement in U.S. VC deals: 2007–2017
45%
source: PitchBook-NVCA Venture Monitor, 2018-Q1.
The figure shows the percentage of VC deals that have CVC involvement and the percentage of total deal value that includes CVC involvement. Total CVC investment dollars is not reported.
VC deals with CVC involvement (% of all VC deal value) Number of deals with CVC involvement (% of all VC deals)
40% 35% 30% 25% 20% 15% 10% 5% 0%
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
(25%), and insurance companies (20%). Other investors in VC include highnet-worth (accredited) individuals, family offices, and corporations.6 Earlier, we indicated that VC in countries other than the U.S. has been restrained by limited access to the kinds of institutions that are the primary sources of VC in the U.S. Europe is second to the U.S. in terms of the size of the VC industry. Yet even there, we can see the impact of limited access to institutional sources of patient risk capital. In sharp contrast to the sources of U.S. funding, the most important sources for European VCs in order of importance (in 2016) are government agencies (25%), corporations (15%), private individuals (12%), funds-of-funds (10%), and family offices (8%). Endowments and foundations, insurance companies, and pension funds collectively account for only 15%, and “other” accounts for 15%.7 A relatively new source of VC funding is country-sponsored Sovereign Wealth Funds (SWFs). These funds may invest directly in new ventures or indirectly as limited partners of VC funds. In recent years there has been an increase in SWF interest in investing in high-profile new ventures, including “unicorns,” defined as VC-backed companies valued at $1 billion or more. SWFs invested in 6 of the 10 most highly valued start-ups in 2017. Temasek, a SWF in Singapore, for example, has participated in 88 rounds of 74 start-ups since 1999.8
The Impact of Venture Capital on the Economy Early-stage VC investment is small compared to corporate and government R&D investments. For example, during 2015, VC firms received $78.6 billion
94
Chapter Three
in new VC commitments, which was a ten-year high. This represents less than a small fraction of the total $462 billion in R&D expenditures in the U.S. in 2015.9 Though VC is a small industry, it has a large economic impact. Many of today’s largest and most research-active corporations received early-stage funding from VC firms. Gornall and Strebulaev (2015) study how firms that were VC-financed in early stages have profoundly changed the U.S. economy. Public companies with VC backing employ 4 million people and account for one fifth of the market capitalization and 44% of the R&D spending of U.S. public companies. A longer-term perspective shows that of all U.S. public companies that were founded after 1974, 42% were VC-backed. These firms now represent 63% of U.S. market capitalization and 85% of total R&D. Five of the largest companies in the U.S. in terms of market cap are Apple, Alphabet (Google), Microsoft, Amazon, and Facebook, all of which were VC-backed.10 Sedlacek and Sterk (2017) show that employment created by start-ups is volatile and pro-cyclical, meaning that successful large firms tend to be born during periods of booming consumer demand, when it is relatively easy for firms to acquire new customers. Interestingly, the positive impact of start-ups on employment persists over time (rather than being mean-reverting), as highly scalable businesses need time to reach their full potential. Why is the impact of VC so great? The answer lies in the specialized market niche that VC serves. VC investors seek young companies that have the potential for rapid and substantial growth and that will need significant capital investments to finance that growth. VC thereby fills a niche between very early-stage private investment by entrepreneurs, their friends, and angel investors, and the market for public capital or private corporate acquisition. Firms with limited product markets and slower growth potential generally are not candidates for VC. Such firms tend to have more stable capital needs and may be better suited for debt financing.
3.2 The Organization of Venture Capital Firms Successful VC investing depends on finding solutions to an array of problems. The fund manager must sort through a plethora of business plans, each describing a venture with a negligible operating history. In addition to identifying the few ventures that have some potential for success, the investor must be able to add enough value to the deals to cover the extra costs of managing the fund’s investment portfolio, including the return on the VC’s commitment of effort (human capital). Beyond addressing these problems, the manager must
Venture Capital and Angel Investing 95
be able to commit the fund’s capital for long periods with little hard evidence that value is being created for investors. The VC fund manager must be able to expand, contract, and refocus efforts in response to changes in opportunities and new information about prospects for success, and must be able to redeploy the VC firm’s human capital when its ability to add value to a venture wanes.
The Limited Partnership Structure To address these challenges in the context of the regulatory environment described earlier, the VC limited partnership has become the dominant organizational form for VC investing in the U.S. Figure 3.7 is a schematic representation of the limited partnership form of a VC fund. Unlike a corporation or
Fi g u r e 3 .7 Organizational structure of VC funds VC funds are usually organized as limited partnerships. The GP manages the fund, while LPs provide most of the investment capital. The fund invests in a portfolio of new ventures.
• • • • •
Effort and 1% of capital
General Partner
Generate deal flow Screen opportunities Negotiate deals Monitor and advise Harvest investments
Annual management feeFee 2%–3% 2-
Carried interest 20%–30% of gain 20 -
Investment capital and effort
Venture Capital Fund
Capital appreciation 70%–80% of gain
99% of investment capital
• • • • •
Financial claims
Limited Partners
Pension plans Life insurance companies Endowments Corporations Individuals
Portfolio Companies • Value creation
96
Chapter Three
open-end mutual fund, a VC limited partnership has a finite life span, typically 7–10 years. The general partner (GP), usually a VC firm that is itself organized as a partnership, is the fund organizer and is responsible for raising investment capital from the limited partners (LPs) and deploying the capital by investing in portfolio companies. On the capital deployment side, the GP screens opportunities based on quality and compatibility with the GP’s capabilities and with timing of the fund’s capital flows. When an attractive investment prospect is identified, the GP negotiates the terms for investing. Committing the fund’s financial capital to a venture also commits some of the GP’s human capital to ongoing involvement in monitoring and advising. The intensity of these efforts varies across funds and across portfolio companies. Philosophies differ, and some ventures warrant more active investor involvement than others. Finally, the GP is responsible for harvesting the investment. Harvesting enables the fund to provide a return to the LPs and allows the GP to redeploy human capital to other investments. Raising the capital for a VC fund is a costly endeavor. The GP commits a substantial amount of time to marketing the fund to prospective investors and to managing relations with existing investors. Most institutional investors require periodic (at least annual) reporting, including valuations and status reports on the fund’s portfolio companies. The GP’s primary contribution to the fund is in the form of effort. In addition, though specifics vary, the GP normally commits 1% of the fund’s capital; LPs provide the other 99%. A small fraction of invested capital is used each year to cover costs related to managing the operation of the fund. The balance is invested in portfolio companies, in exchange for financial claims on the companies. If and when these investments are harvested, the returns are distributed to LPs, first to repay their initial investments, with the balance (the capital gain) being shared between the LPs and the GP. Normally, the LPs receive 70–80% of the gain and the GP receives the balance as “carried interest.”
The Process of Investing in VC Figure 3.8 illustrates the VC investment process and its relation to fund maturity. After the fund concept is developed, the GP makes concurrent efforts to secure commitments from investors and to generate deal flow. The fund closing and first capital call mark the point when the GP begins to build a portfolio of investments. Some VC performance statistics are reported based on funds grouped by “vintage year.” The vintage year is the year of the first capital call or first investment, even though some activities commence months earlier. Most funds are intended to last about 10 years.
Venture Capital and Angel Investing 97 Development of Fund Concept
Fi g u r e 3 . 8 The VC investment process
Secure commitments from investors
Closing of Fund First capital call
Year 0
2–3 years
Generate deal flow
Screen business plans
Evaluate and conduct due diligence
Negotiate deals and staging
Additional capital calls
Invest funds
Value Creation and Monitoring 4–5 years
2–3 years or more
Board service Performance evaluation and review Recruit management Assist with external relationships • Help arrange additional financing
• • • •
Harvesting Investment • IPO • Acquisition • LBO • Liquidation
Distributing Proceeds • Cash • Public shares • Other
10 years plus extensions
The role of the GP changes over the life of the fund, from activities related to investing capital to those related to managing and monitoring investments and finally to those related to harvesting. Normally, it takes 2 to 3 years before a fund is fully invested. During this period, the GP is busy screening plans, conducting due diligence on prospective investments, and negotiating deals. Corresponding to these efforts, the GP makes additional capital calls and seeks to place the entire commitment with new ventures. While the subscription agreement makes reference to a “call-down schedule,” in practice the timing of capital calls can be accelerated or extended depending on how quickly appropriate deals are identified. In addition to capital calls made to existing committed investors, a fund can also have more than one “closing.” A fund closes when sufficient commitments of capital have been made and sufficient investment opportunities have been found to warrant going forward. A closing is a legal process in which the commitments
98
Chapter Three
are used to define an ownership group. During the fund-raising stage, the GP may bring in new investors and oversee a second or third closing. A closing defines a group of investors who are treated identically as the fund progresses. Managing a fund with multiple closings gives rise to potential opportunism as it enables investors to get into the fund at different times. Existing investors may be concerned with the potential for opportunism by new investors who might try to buy into existing successes at valuations that are too low. Conversely, new investors may be concerned that existing investors will try to exploit information advantages or get better deals for themselves. Thus, having a second closing with a different investor group elevates the importance of accurate valuations of portfolio companies and fund investments in them. Fund managers can try to address potential opportunism by segregating existing and new investments into separate funds. Of course, an investor can also avoid opportunism by acquiring the right, ex ante, to participate in each closing at the same level (e.g., 5% of each closing). However, this may increase the investor’s financial commitment beyond what was desired. Following investment in a portfolio company, the GP’s responsibility shifts to value creation and monitoring. Service on boards of portfolio firms is routine, as is continual performance evaluation. In addition, the GP often is involved in recruiting the management team, building relations with trading partners, and helping to arrange subsequent financing. Most GPs hope to harvest their investments about 5 years after the investments are made. Ideally, the GP continues to work with a portfolio company as long as the GP is able to add value and until a point when a liquidity event is possible. Harvesting enables investors to realize the gains on their involvement, through receipt of either cash or publicly tradable shares. Distributions are natural milestones for the GP in its efforts to establish another fund. The harvesting phase is typically 2 to 3 years. The long window allows the GP to time the exits in light of market conditions and company-specific factors. Because portfolio companies progress at different rates, the periods of investment, value creation, and harvesting within a single VC fund overlap. Typically, the LPs can agree to extend the life of the fund for several years to permit orderly liquidation of investments. The finite life of the fund limits and controls the GP’s behavior. A GP that is successful in adding value will have little trouble generating commitments to a new fund, whereas one that has not been successful is unlikely to attract new investors. In practice, VC firms usually try to launch new funds more frequently than every 10 years but with enough time in between so that investors can assess whether the fund seems to be on track for success. Here again, long-term relationships and reputation are important. Unlike with the public equity mar-
Venture Capital and Angel Investing 99
ket, prospective investors are few and can easily communicate with each other. Concern with reputational damage disciplines the GP and protects investors from short-run opportunism.
Terms in a VC Limited Partnership Private Placement Memorandum The VC private placement memorandum (PPM) is similar to the prospectus of a public company. It sets out conditions for investing, requirements for closing, distribution requirements, and the other terms such as industry or stage focus for investments in portfolio companies. Table 3.1 contains a summary of some of the more important terms in the PPM.
Tab le 3 .1 Summary of terms in a VC limited partnership private placement memorandum (PPM)
Purpose
General Partner
Limited Partnership Interests General Partner’s Investment Minimum for Closing
Payment of Subscriptions
Term Allocation of Profit and Loss Distributions
Management Fee
The “purpose” identifies the fund’s focus and investment objective, such as “to earn superior returns from early-stage investments in e-commerce ventures.” Provisions may limit the amount invested in any one venture and/or restrict the types of investment vehicles the fund can use. Normally the general partner is a partnership of individual general partners. Investors in the fund may seek access to the contract between the limited partnership and the individual general partners to assess their incentives and perhaps seek revisions and limitations to the agreement. The provision specifies the total size of the fund that is being raised, such as $250 million, and the minimum per investor, such as $10 million. The usual contribution is 1%, and sometimes may be in the form of a promissory note. The provision specifies the dollar commitment, such as $100 million, that must be received before an initial closing. If there is a subsequent closing, it may be required to occur within a fixed period, such as one year after the initial closing. Investors must provide a specified amount of capital at the time of closing (such as 30% of their total commitment). Subsequent calls often are made on short notice to afford maximum flexibility for investing the funds. Penalties for late or missed calls will be enumerated. The term of the agreement usually is 10 years, with options to extend under certain conditions, which permits orderly liquidation of investments. The provision specifies the allocation and distribution of returns from investment. The most common allocation is 80% to limited partners and 20% to the general partner. The provisions ensure prompt distribution of invested capital and gains as investments in portfolio companies are harvested. Normally, distributions can be in the form of cash and public securities. Terms may specify that the limited partner investors receive a full return of their invested capital before the general partner receives any distributions. An alternative is to distribute 20% of the gains to the general partner, but if there are losses on some investments, then the agreement obligates the general partner to make up for the overcompensation through a clawback provision. Normally, the management fee is 2% of contributed and committed capital for established VCs managing large funds, and up to 3% for smaller funds with newer VCs. The fee may be linked to expenses, may provide for specific offsets to be charged, and may be set in light of other funds under management.
100
Chapter Three
In contrast to public capital markets, VC is a market where the LP’s reputation is as important as the LP’s money. When a new fund is created, the GP seeks capital commitments from investors at the same time that it assesses investment opportunities and negotiates deals. Each LP’s commitment is formalized in a subscription agreement. Actual investments in new ventures are not made until the fund’s closing. Each investor’s commitment is conditional on the fund generating sufficient commitments from other investors to reach the minimum total for closing. When the closing occurs, the GP can make an initial “capital call” on the investors. Following a capital call, the investors have a short time (such as 30 days) in which to deliver the funds. To achieve the best performance, the GP makes capital calls only when there are immediately attractive portfolio investment opportunities. The investors may receive several capital calls during the first few years of the fund’s life. Because each investor is expected to deliver capital when called upon, the investors’ reputations are important to fund operation. Failure to respond to a capital call could cause the fund to miss investment opportunities and otherwise disrupt fund operation. Because of this, penalties for missed capital calls are substantial. Use of closing dates and capital calls enables a VC fund to operate with very little internal liquidity. The need for liquidity does not disappear, however; it is shifted instead to the investors. Consequently, the typical investors are those whose ordinary levels of liquidity are sufficient to enable them to respond to unpredictable capital calls. The upshot is that GPs seek investors who can reliably commit their capital for the entire life of the fund. What better place to look for such investors than institutions such as pension funds, endowments, and life insurers, all of which have predictable needs for liquidity that are small relative to the overall sizes of their investment portfolios? By extension, the GP screens for individuals who can reliably commit to maintaining their capital investments.
Why Limited Partnership? The LP structure solves many of the managerial problems that arise in new venture investing. Because investors can participate in multiple funds and can diversify across a broad array of other investments, the GP is free to concentrate on the kinds of opportunities in which the partner’s expertise adds the most value. The finite life of a VC fund subjects the GP to ongoing market discipline. The LP structure also enables the pool of VC funds to expand or contract depending on the opportunities perceived by the GP. Finally, the structure makes efficient use of the liquidity that naturally accrues to large institutional investors. The downside, compared to the closed-end fund struc-
Venture Capital and Angel Investing 101
ture, is that most individual investors in the U.S. are foreclosed by SEC regulations from investing directly in VC partnerships.
Venture Capital Contracts with Investors Traditional corporate control mechanisms are not available to LPs. For example, there is no board of directors to represent the LPs’ interests and no market for corporate control to discipline the GP. Thus, other mechanisms must align GP and LP incentives. A VC limited partnership agreement (LPA) formalizes the PPM and addresses concerns related to shirking and opportunistic behavior by the GP. Table 3.2 describes some of the more common fund management covenants associated with LP agreements. These are in addition to terms described in the PPM (Table 3.1). With respect to overall management, investors may be concerned with the GP’s incentives to take on excessive risk, to favor existing funds over new one, and to increase management fees by delaying disbursements. To address the incentive to take excessive risk, the agreement may limit the investment in any single venture and restrict the ability of the GP to add leverage. Another concern is that the GP may make investment decisions that favor its existing funds. Consider a VC firm that manages several funds. Suppose an older fund has an investment in a portfolio company that is seeking second-round financing. The GP may take advantage of LPs in the new fund by using their capital to make a follow-on investment in the venture on terms that benefit the older fund. This is more of a concern if the VC is not well established and wants to use the early-fund performance to attract capital. To obviate this concern, a fund contract may give investors the right to review the GP’s decisions to make follow-on investments in ventures held by its other funds or specify that it may do so only as a co-investor along with another first-time investor in the venture. To avoid fee manipulation, the agreement may restrict the GP’s ability to reinvest fund assets. This limits the ability of the GP to increase its management fees (which usually are calculated as a percentage of the value of assets under management) by postponing distributions. In addition, as a means of controlling potential shirking while collecting management fees, the agreement may limit the investment the GP can make in financial instruments that offer low expected returns and do not require significant commitments of VC effort. Additional concerns regarding the activities of GPs relate to self-dealing and dilution of effort. To address self-dealing, GPs’ partners and employees may be precluded from buying or selling investments in a venture in advance of transactions on behalf of the fund. To control shirking, the agreement may limit the amount of outside investment a GP can make. It also may restrict the
102
Chapter Three
Tab le 3 . 2 Key covenants in a VC limited partnership agreements (LPA)
Overall Fund Management Key Person Provision
Investment in a Single Firm Use of Debt
Co-Investment by Venture Capitalist’s Other Funds Reinvestment of Profits
The provision addresses what occurs if there are important changes in core investment personnel. Limited partners may have the right to suspend investment or terminate the fund. As a way to restrict risk-taking by the venture capitalist, the covenant limits the amount of the fund’s investment in a single firm. The covenant limits borrowing by the fund and possibly debt maturity. It also may limit borrowing by the general partner (venture capitalist) or guarantees of loans by the venture capitalist. The covenant addresses concerns with leveraging the fund and creating conflicts of interest. To limit opportunism and avoid incentive conflicts, the covenant provides for monitoring of decisions to invest in any portfolio company that is in one of the venture capitalist’s other funds. The covenant may require co-investment with other funds. The covenant restricts or prohibits reinvestment of profits from a fund and concerns possible conflicts of interest regarding the venture capitalist’s management fees.
Activities of General Partner
Personal Investing in Portfolio Companies
Sale of Interest by General Partner Fund-Raising
Outside Activities
Addition of General Partners
The covenant limits the ability of the members of the general partnership to invest alongside of, or in advance of, the fund. The provision serves several purposes, including controlling incentives of the venture capitalist to overinvest efforts in those funds in which it has a personal investment, and to avoid conflicts about when to harvest. Limits on the timing of investment may prevent front running by the venture capitalist. The general partner may not be permitted to sell its interest in a portfolio company separately from selling the limited partners’ interest. This provision protects the limited partner from self-dealing. The covenant may limit ability of the general partner to raise new funds until existing funds are substantially committed (to avoid dilution of effort and to prevent efforts to try to increase fees). The provision ensures that the general partner devotes sufficient effort to portfolio companies and limits its outside activities or involvement with outside companies, especially during the portfolio companies’ early stages. The covenant limits additions of partners to ensure that the quality of effort to the fund remains high.
Types of Investments Restrictions on Asset Classes
The provision limits the ability of the general partner to invest in certain classes of assets, notably those that require less effort (such as public securities), other venture funds, or portfolio companies that compete in industries where partners lack expertise. The provision limits opportunistic behavior by the venture capitalist.
ability of the VC to raise capital for new funds. Further, it may limit the outside activities of the partners, as well as the addition of new partners, and may set expectations for intensity of monitoring. As might be expected, the most commonly negotiated terms in a LPA are management fees, the percentage split of the carried interest, clawbacks, and key person provisions. The LPs have ultimate control of the fund through their ability to withdraw from the partnership or terminate the agreement. As a
Venture Capital and Angel Investing 103
practical matter, however, withdrawal is unlikely because substantial sanctions usually are contractually imposed for withdrawal and because early termination usually is not in the interest of the partners.
3.3 Investment Returns and Compensation As noted earlier, the gain on investment in a portfolio company is shared between the LPs and the GP according to the terms specified in the limited partnership agreement. While this sounds simple, implementation is challenging. Beyond this, since the GP is itself usually a partnership with employees, returns to the GP must be allocated to the individuals who are involved.
Returns to Limited Partners in VC Funds The returns to LPs are commonly reported either as internal rates of return (IRRs) or “net multiples.” The IRR is computed over the period from when the money is actually received by the fund from the LPs until it is returned to them. A net multiple is essentially just the ratio of cash paid out to cash invested. So a return of $3 for each $1 invested is a net multiple of 3—the investors got their money back plus $2 more for each $1 invested. Both measures provide useful information and help to discipline the GP to act in the interest of the LPs. Since holding on to large amounts of cash before it is needed for good projects will reduce the IRR, the measure discourages the GP from making capital calls before they are actually needed and encourages them to pay out cash quickly once it is no longer needed. The net multiple does somewhat the opposite. It discourages the GP from harvesting too soon just to be able to report a high IRR. The two measures work best together. The IRR is sometimes maximized by harvesting good projects quickly because holding longer would lower the return, even if the project would continue to earn a rate higher than its cost of capital. The net multiple, conversely, encourages holding too long, provided that the project is still generating cash, even at a low rate. The actual returns to VC funds are subject to considerable volatility over time. Figure 3.9 shows quarterly net returns to LPs. The GP’s carried interest generally is equivalent to about one fourth of the return to the LPs. In many periods in the figure the LPs’ returns are small or negative; in some years, however, the return is very high. As can be seen in the figure, LP returns from VC are pro-cyclical. They were quite high during the market run-up associated with the tech rally, but plunged and became negative after mid-2000 when the high-tech market collapsed.
104
Chapter Three
80%
Quarterly percentage return
60%
40%
20%
2016 Q1
2015 Q1
2014 Q1
2013 Q1
2012 Q1
2011 Q1
2010 Q1
2009 Q1
2008 Q1
2007 Q1
2006 Q1
2005 Q1
2004 Q1
2003 Q1
2002 Q1
2001 Q1
2000 Q1
1999 Q1
1998 Q1
1997 Q1
1996 Q1
1995 Q1
1994 Q1
1993 Q1
1992 Q1
1991 Q1
1990 Q1
1989 Q1
1988 Q1
1987 Q1
1986 Q1
1985 Q1
1984 Q1
1983 Q1
1981 Q1
(20%)
1982 Q1
0%
Fi g u r e 3 . 9 Net returns by quarter to limited partners of VC funds: 1981–2016 sources:
Returns through Q1 2009 are from the National Venture Capital Association. Returns after that date are from Preqin.
oreover, the very high returns of the late 1990s were generated over unusuM ally short holding periods, so they can overstate the investment return potential. A similar but less dramatic episode surrounds the 2008 financial market collapse. In Figure 3.9, the simple quarterly average of the returns to VC investing over the entire time period is 3.3%. However, the simple average overstates the long-run rate of return an LP would have realized by investing at the beginning and holding to the end. Converting to continuously compounded returns, the quarterly average is 2.9%, and the compound annual IRR is 12.0%. We defer, until later, the question of whether this level of return is sufficient to cover the opportunity cost of taking money away from other investment opportunities.
Returns to General Partners Waterfalls. Returns to LPs and the GP follow a progression that is sometimes referred to as a “waterfall.” The progression is depicted in Figure 3.10.
Venture Capital and Angel Investing 105
Investment Return Management fee LP return of principal LP preferred return GP carry on preferred return LP 80% of residual capital gain
GP 20% carry on residual capital gain
Fi g u r e 3 .1 0 Depiction of a VC waterfall provision In this illustration, the distribution of returns follows a pattern: payment of management fees, return of the LP’s initial capital investment, payment of LP’s preferred return (if any), a catch-up payment of the GP’s carry on any preferred return to the LPs, distribution of 80% of the residual capital gain to the LP and 20% of the residual capital gain to the GP.
To illustrate, consider a small fund with a $100 million commitment. The GP comprises five partners and a number of other employees. For simplicity, the fund pays a 2.5% fee on committed capital each year for the operating budget and there is no preferred return to the LPs. The GP is entitled to a carried interest of 20%. Suppose the fund has a 10-year life span and a 5-year average holding period of investments, with 20% annual appreciation after fees. For simplicity, suppose the fund invests $20 million per year in Years 1 through 5 and realizes uniform returns of $49.8 million per year in Years 6 through 10 (a 20% per year return on each investment). The management fee is $2.5 million per year, or $500,000 per partner of the GP, but is used to pay for fund operations including base salaries. The GP’s gross return due to capital appreciation is the 20% carried interest [0.2 × ($49.8 million − $20 million)], or about $5.95 million annually, equivalent to about $1.2 million per partner of the GP in each of years 6 through 10. These returns, of course, are before partnership expenses. The LPs invested $20 million in each year, 1 through 5, and receive a $20 million return of principal in each year 6 through 10. They also receive 80% of the $29.8 (i.e., $23.84) million gain in each of the last 5 years. Clawbacks. In the preceding illustration, the GP receives a portion of the carried interest each time a venture is harvested. This works in the example
106
Chapter Three
because each investment has a positive payout that is more than the amount invested. The reality is more likely to be that good investments are harvested relatively quickly and bad investments stay on the books until the end of the fund life. Challenges arise because if payouts on good investments are made early, the GP may receive more than 20% of the gain. In a strict application of the waterfall, the GP would not realize any carried interest return until all of the assets of the fund had been liquidated through sale, acquisition, or failure. Normally, the GP does not want to wait 10 years or more until all of the returns are realized. Accordingly, the waterfall distributions are often handled asset by asset, as in the preceding example. Thus, if the fund invests in a venture that has a successful IPO early in the life of the fund, the carried interest fraction of the proceeds may be distributed to the GP at the time. But what happens if the later ventures do not do well, so that the LPs do not realize the overall return to which they were entitled? In such cases, the investment agreement may include a “clawback” provision that requires the GP to return “excess distributions” to the partnership. A clawback represents the GP’s promise that it will not receive a greater share of the distributions than was bargained for. Consider a simple example of a fund that makes only two investments of $10 each. The first investment is later harvested for $60. The LPs get $50 (the $10 investment plus 80% of the gain). The GP gets $10 (20% of the gain). Now, the second investment turns out to be a failure and is written off. In this case, the fund invested $20 and the total payoff was $60, so the total gain is $40. The GP is entitled to 20% of the gain, or $8. Since the GP has already received $10, $2 would be clawed back. While it seems risky for LPs to allow the GP to be rewarded before the entire fund is liquidated, the normal exposure is not large. The amount that is clawed back in this simple example is small relative to the total the GP had received.
VC Firm Internal Structure and Compensation Because VC firms are small and private, data are sparse regarding their internal organization and compensation. Some data are available from Gompers, Gornall, Kaplan, and Strebulaev (2016), who surveyed a large sample of VCs and find the average firm employs 14 people, 5 of whom are senior partners. Most VC firms have a few junior deal-making personnel as well. Others who work at VC firms include entrepreneurs-in-residence, analysts (at larger firms), and staff. VC firms that focus on early-stage investment tend to be smaller, and those with a later-stage focus require more personnel for due diligence ef-
Venture Capital and Angel Investing 107
forts. Gompers et al. also find that in 60% of the funds, partners will specialize in different activities, including fund-raising, deal making, deal sourcing, and networking. The typical respondent in their sample (a senior partner) held five board seats, and the data indicate that the single largest amount of time spent by VCs is in working with portfolio companies. Based on the survey, most VC firms compensate partners depending on their individual success, although, in recognition of the value added by the organization, larger and more well-established VC firms are less likely to allocate compensation heavily based on individual success.11 The compensation structure must create incentives for individual effort and for cooperation. The optimal mix is likely to vary with the size, stage, and industry focus of the funds that are being managed by the VC firm. The typical compensation package for partners of a VC firm includes a salary, a bonus, and a share of the carried interest. Not all VC firm partners receive equal shares of the carry, particularly in early-stage VC firms. The compensation will vary with the size of the fund. An estimate by Ramsinghani (2014) is that for larger well-known VC firms, the compensation for a managing GP of one of its VC funds could consist of salary ($700,000), bonus ($350,000), and carry ($101,000) for a total of $1.151 million. An analyst would receive a salary and bonus of about $110,000 and would not participate in the carried interest. An individual partner of the VC firm may manage assets for more than one VC fund. Note, however, that these estimates are for successful funds, so there is a selection bias in the numbers; if a fund performs poorly, the partner’s career as a VC may be brief.
3.4 Impact of Compensation on Investment Selection The compensation arrangement between the GP and LPs is an important determinant of investment attractiveness. The financial claim of the GP has the characteristics of a call option: a limited downside and a significant upside. If the fund does well, the GP realizes a significant return through its carried interest. Conversely, if the investments perform poorly, most of the loss accrues to the LPs, who contribute the bulk of the financial capital; the GP continues to collect a management fee. This structure can affect the kinds of ventures that attract the GP’s interest, as well as the terms of financial contracts between the fund and its portfolio companies. A venture with an attractive expected return but limited upside potential is unlikely to be a prospect for VC, even if downside risk is small. Other
108
Chapter Three
things equal, such a venture does not offer the potential for the GP to benefit significantly by sharing in the gain. The attractiveness of an investment to the GP also depends on the terms of the financial contract. A high-risk venture that is structured to create financial claims that offer safe but reasonable returns to investors is unlikely to attract VC funds. On the other hand, a low-risk venture can receive funding if the financial claims are structured to offer substantial upside potential by shifting downside risk to the investor. The financial claims most commonly sought by VC investors are those that align the interests of the GP with those of the LPs. Interests are aligned when both types of partners participate in new venture success but are protected against losses if the venture performs poorly. Convertible preferred stock, for example, is often used in VC financing. Convertible preferred offers some downside protection (ahead of common stock) and still preserves the potential for significant gain. Structures involving staging of investment or put options for the investor can have similar effects. We discuss the economics underlying VC contracts with portfolio companies in Chapter 4.
3.5 Aspects of the VC Industry Structure Geographic Clustering A notable feature of the VC industry is that entrepreneurial firms and VC firms tend to cluster in geographic areas (Figure 2.3). California, Massachusetts, and New York attract the most VC investments, with these three states accounting for 75% of U.S. VC dollars invested in 2016 and 52% of all investee companies. Geographic concentration suggests the presence of agglomeration economies. Such economies arise when firms’ costs are reduced as a result of being proximate to their rivals and other market participants, such as suppliers and specialized labor. These characteristics fit the VC industry. Value is added through close working relationships between VCs and portfolio firms. For portfolio firms, proximity to rivals is likely to keep them competitive and better informed. When new venture activity is concentrated, VC firms can more easily monitor industry trends. In addition, VC firms and portfolio companies have specific human capital requirements. It is no accident that the two largest pockets of VC activity in the U.S. are located near large and talented university populations. Consistent with this, a VC that decides to fund a firm outside its geographic area may predicate funding on relocation of the venture.
Venture Capital and Angel Investing 109
Tian (2011) documents a relationship between the use of VC staging and geographic proximity of the venture and the VC. The author finds that VC investors located farther away from an entrepreneurial firm tend to finance the firm with a larger number of rounds of shorter duration between successive rounds, and smaller amounts per round. Thus, when firms are not close to their investors and monitoring costs are higher, staging and deal structure can, in part, substitute for direct monitoring.
Syndication Syndication occurs when VC funds co-invest in ventures. Typically, for any given venture, one fund is the lead investor and the others are co-investors. The lead normally is the fund with the highest level of direct involvement with the venture and is the one most likely to serve on the venture’s board of directors. Syndication is a reciprocal, ongoing, informal relationship, in which VC funds tend to collaborate by distributing the responsibility of serving as lead investors or coinvestors. The practice enables VC firms to pool their human capital resources, to invest more economically in geographically dispersed markets, and to spread the investments of their funds over larger and more diverse portfolios.12 Because the LP investors in VC are normally well diversified over many investments, diversification within a fund is normally of little direct value to them. Fund-level diversification is mainly of direct value to those involved in the VC firm. Fund-level diversification can indirectly benefit VC LPs because the managers whose interests benefit from diversification can agree to lower levels of compensation and can limit the effects of their personal concerns about fund-level underdiversification on investment selection.13 In countries where VC investing is well established, syndication among reputable VCs is common. In emerging economies, the lack of both attractive investments and established VC firms is an important impediment to syndication and industry growth. There can be different reasons for later-round syndication than for earlyround. Based on a study of biotech investment rounds, Lerner (1994b) finds that in first-round investments, established VC firms tend to syndicate with each other and avoid less experienced firms. Later rounds are more likely to involve syndicated investments by less well-established VC firms. Lerner interprets the evidence as being consistent with “window dressing” in the syndication of laterround investments by less well-established VCs. Also, in cases where a less wellestablished VC happens upon an good early-stage investment that does well, better established VCs may join the syndicate later and provide additional funding, greater access to harvesting opportunities, and additional expertise and contacts.
110
Chapter Three
Admati and Pfleiderer (1994) develop a rationale for the allocation of ownership in later-round syndication based on informational asymmetries. Because the first (lead) investor may have an informational advantage over subsequent investors, opportunism in later rounds can be avoided if the lead maintains a constant share of the firm’s equity. If the venture needs to raise a large amount of capital in a later round, this can imply that additional investors must provide a portion of the later-round financing. The lead investor sometimes is paid for assisting in the syndication. The typical fee is 2% of the money raised from co-investors. This fee is sometimes shared with the other members of the initial syndicate.
3.6 How Venture Capitalists Can Add Value GPs charge substantial fees for fund management and share significantly in the success of their investments. Although GP returns may seem excessive, VC funds attract most of their financial resources from sophisticated investors. The investors, either directly or through gatekeepers, continuously monitor the actions and decisions of the GP. The sophistication of VC investors supports the inference that the compensation structures are justified. If so, the fund managers must contribute significantly to performance; sophisticated investors do not reinvest with GPs in whom they do not have confidence.14 VC firms, if they are to survive, must be able to add value sufficient to cover their compensation. Beyond the efforts to create value for an individual fund, the GP seeks to create value for the VC firm. VC firms expect to realize the benefits of information and other scope economies by operating multiple funds. They also expect to maximize the value of the firm’s human capital by deploying experienced partners to work with multiple portfolio companies. When one fund is being liquidated, another often is being formed, so that the firm’s human capital is used efficiently over the course of the investment/monitoring/harvesting cycle. Managing a VC fund is complicated because the fund supplies a joint product consisting of investment capital and consulting/monitoring services. The gross return to the VC fund must cover both the opportunity cost of the LPs’ investment capital and the human capital services of the GP. VC firms seek to accomplish this in a variety of ways.
Selecting Investments and Negotiating Deals The ideal ventures to include in a portfolio are those in which the GP can add significant value without an excessive commitment of time. Other things being
Venture Capital and Angel Investing 111
equal, an investment is more attractive if it appears to be well managed and unlikely to require much assistance. Such a company, however, should also be able to negotiate a favorable deal with the GP. Beyond this, a venture is more attractive if its needs meld with the specific capabilities of the GP. Simple metrics of fit include industry focus, stage of development, and location. A prospective venture is more attractive if the expected time commitment for VC involvement corresponds to a period over which the market can come to recognize the value of the venture so that exit is possible. If the market takes too long to recognize value, the GP may be compelled to continue to devote time to the venture. This involvement can extend beyond the point where the GP is adding value, especially if it causes the GP to forgo work on other ventures. Because LPs are concerned with earning returns that exceed what they can earn from other investments, the GP does not want to raise capital before it is needed. Similarly, the GP does not want to hold a portfolio company after the market recognizes its value. Market conditions affect the decision calculus. Ideally, the manager seeks to create the fund during a period when opportunities to invest in new ventures are abundant and seeks to harvest when the market is receptive to public offerings. There is some evidence that VCs can time distributions to correspond with periods when market values are high. Lerner (1994c) finds that in choices between IPOs and private rounds, VC-backed biotech companies tend to select IPO after run-ups in the market and before market declines. Ball, Chiu, and Smith (2011) examine VC-backed exits by IPO or merger over a longer period and more sectors. Consistent with Lerner, they find that IPOs are selected after market or sector run-ups. However, they find no evidence that VC-backed firms can spot opportunities to issue in anticipation of market declines.
Selecting Entrepreneurs VC firms evaluate thousands of ventures each year. A primary component of the evaluation is assessing the entrepreneur’s knowledge and skills. Gompers et al. (2016) survey the methods VC firms use to evaluate entrepreneurial firms and the management team. On average, VC respondents report that they spend 118 hours on due diligence per firm, but more on late-stage than earlystage firms. Respondents rank the team as the most important factor in selection, followed by business model and product. Kaplan, Sensoy, and Stromberg (2009) find that in a sample of VC-backed firms, the business plans of the firms are “remarkably stable” over time while management turnover is “substantial.” Based on observed behavior, they infer that investors in start-ups place more weight on the business model than on the management team.
112
Chapter Three
Changing the Management Team A study of Silicon Valley portfolio companies by Hellman and Puri (2002) indicates that professional managers replace more than half of founding entrepreneurs.15 In about 40% of these cases, the founder retains a position at the company; in the remainder, the founder has no discernible tie with the company after being replaced. In another study, based on survey data, Fiet, Busenitz, Moesel, and Barney (1997) provide perspective on the types of “mistakes” that can lead to dismissal and the contractual covenants that are effective in aligning managers’ incentives with the interests of the VC fund. They find that likelihood of dismissal is reduced when manager salaries are low, contracts include earn-out provisions, revenue per employee is growing, and VC control of the board is low. These results do not establish causation. A VC that is concerned about the capabilities of the entrepreneur is more likely to negotiate performance-based covenants and require board representation. An entrepreneur who wants to signal confidence may be willing to accept unfavorable board composition in exchange for more funding or favorable terms.
Monitoring and Advising Portfolio Companies VCs can seek to add value by selecting and monitoring portfolio companies and by providing services to the companies, including introductions that facilitate the hiring of key employees and facilitate exit via an acquisition or IPO. There is evidence that portfolio companies value association with reputable VCs and that VC backing is associated with job creation, higher levels of patent awards, and more citations to patents.16 VC backing also appears to play a certification role in the IPO process, as VC backing is associated with less severe IPO underpricing, lower total cost of going public, and going public sooner.17 How much of the return to VC is attributable to successful screening of ventures (as opposed to post-investment involvement)? Bernstein, Giroud, and Townsend (2016) try to sort this out by adopting an identification strategy to hold company selection fixed. They conclude that on-site involvement can increase both innovation and the likelihood of a successful exit.
3.7 Luck Versus Skill: What Accounts for Venture Capital Success? Kaplan and Schoar (2005) find that the returns to VC persist over funds of the same VC firm. That is, high LP returns for one VC fund imply that LP returns
Venture Capital and Angel Investing 113
of subsequent funds offered by the same VC firm will also be high. There is significant evidence that more reputable VCs have better results in terms of returns to investors in the fund and better exit results for portfolio companies.18 These results raise the interesting question of whether persistence of fund performance and portfolio outcomes is due to the skill of the VC firm or luck. Conceivably, a VC firm that is lucky with an early fund can attract better deal flow and raise capital from investors with less effort than a first-time fund or one with a track record of poor performance. To the extent that search costs for investors and ventures are reduced by prior success, even if the success was due to luck, some of this gain can be expected to be shared with investors. Scholars have tried to distinguish between luck as a plausible explanation with the competing hypotheses that top VC firms have a unique skill that is difficult for new entrants to develop. Sorensen (2007) finds evidence that luck plays an important role but that the direct influences of the VC firm also matter. Smith, Pedace, and Sathe (2011) find that both style persistence and sector experience are positively related to fund performance, but they also find that agility, the ability to move quickly into a new sector, contributes positively to returns. Consistent with VC skill playing a role in venture success, Ewens and RhodesKropf (2015) find evidence of both skill and exit style differences among VC firms and even among individual partners at the same VC firm. Their estimates suggest that, in explaining performance, an individual partner’s human capital is 2 to 5 times as important as the VC firm’s organizational capital. Also, evidence from Hsu (2004) suggests that VCs add value in that entrepreneurs “pay extra” to affiliate with VCs that are highly reputable. Hsu finds that reputable VCs acquire start-up equity at a 14% discount. Entrepreneurs apparently believe that VCs provide more than financial capital and that the extra services are valuable to them.
3.8 The Role of Reputation in the Venture Capital Market Reputation plays a role in the functioning of most markets; as we have seen, the market for VC is no exception. Reputation is an important enforcement mechanism for explicit contract terms, such as the LPs’ commitments to respond quickly to capital calls. The alternative is to insist that investors place committed funds in escrow accounts until the GP needs them. But this alternative impedes the LPs’ abilities to invest the funds efficiently until they are needed. Accordingly, there can be significant advantages to working with incomplete, flexible contracts, where reputation substitutes for explicit contract terms. If the GP can depend on the LPs not to make opportunistic decisions, then elaborate provisions to limit the choices of the LPs can be avoided. If the
114
Chapter Three
VC can trust the entrepreneur to act in the interest of the venture, then specific provisions designed to shift risk to the entrepreneur are less important. Contractual relationships that are based partly on the reputations of the parties can generate higher returns for both parties. Evidence of this implication is manifested in several ways. First, the preference of VCs to raise funds from institutions, whose reputations are more easily assessed, demonstrates that the cost of dealing with individual investors is expected to be higher. Second, wellestablished VC firms can raise larger amounts of capital more quickly, reducing the overall costs of fund-raising. Third, well-established VCs can command higher fees and a larger carried interest, but normally do not. The most common terms are a 2% fee and 20% carried interest. Finally, the ability of established entrepreneurs to raise capital more easily than first-time entrepreneurs indicates that investors rely on the entrepreneur’s experience and demonstrated commitment as an element of the negotiation. Several studies support the view that reputation is important for understanding the functioning of the market for VC. Some of these studies find that IPOs with VC backing are less underpriced than those without such backing.19 These findings suggest that VC investors perform an important monitoring function and a certification role. One economic rationale for a VC firm’s investing in reputation is that, because the firm’s human capital is specific to new ventures, the firm benefits by developing a reputation for not selling overpriced shares in IPOs. Thus, VC reputation elicits more interest from IPO investors and also leads to lower underwriter fees (Megginson and Weiss, 1991). Evidence indicates that more established VC firms are better able to bring portfolio companies public at early stages of development than are other VCs. This VC reputation becomes highly sought after by investment banks, who sometimes even offer GPs personal incentives to do business together (Loughran and Ritter, 2004). The data indicate that VCs with established reputations seek to maintain their reputations by selling shares in IPOs only if they expect the IPOs not to be overpriced. Established VCs are also more likely to forgo selling shares when IPO shares are overpriced or are fully priced. In effect, they are more inclined than less well-established VCs to sacrifice immediate return for long-run gain.20
3.9 Angel Investing Angel investors are affluent individuals who provide capital for start-ups in exchange for either equity (convertible preferred stock) or convertible debt. The latter is a type of loan that converts to equity at a trigger point, which
Venture Capital and Angel Investing 115
typically is a subsequent financing event where there is a valuation (such as a VC investment). One of the advantages of convertible debt, as analyzed in Chapter 4, is that it is simple because it allows the investor to wait for a thirdparty valuation, presumably at a point when some major sources of uncertainty about the early-stage venture have been resolved. In addition to providing capital, angels may also add value by providing expertise, introductions, monitoring, and so on. Angels may invest individually or in groups. While individual angel investors have a long history, angel groups are a more recent phenomenon beginning in the mid-1990s (Lamoreaux, Levenstein, and Sokoloff, 2004). Traditionally angel groups have been local organizations comprised of around 10 to 150 members (median size 45). The groups enable the angels to share research costs and due diligence efforts and to pool their investment capital. The number of angel groups has been increasing both worldwide and in the U.S. Most estimates put the number of angel groups worldwide at around 500. The Halo report estimates that there are about 300 angel groups in the U.S. A recent survey of angel groups shows that while all major cities in the U.S. have angel groups, California attracts the most investment. California is also the location of 30% of the companies that are involved in angel-backed deals.21 The same report shows that the median funding round size of an angel group is $950,000, which can include multiple angel group and VC investments. The median angel group investment in these deals is $127,000.22 In terms of market size, recent estimates suggest that U.S. angels pump around $20–$25 billion per year into start-ups.23 While the focus of angel investments is usually early-stage, some angels invest in expansion and late-stage as well.24 The formation of angel groups reflects an economic rationale for more efficient investment. The angel organization allows members to pool their capital so that investments can be larger than they would be if made individually. The members can also make smaller investments in more opportunities, thereby diversifying. They can also economize on due diligence costs. Start-ups benefit from the experience and expertise of the various group members who invest in the firm. Angel group business practices. Each angel group has its own set of procedures, policies, and ethos. Common practices include regular meetings to review business proposals and listen to pitches. Angels work together to conduct due diligence. Many angel groups co-invest with other groups, individual angels, and VCs who make early-stage investments. Like VCs, angel groups invest in a range of industries. Among the most common are software, medical devices, telecom, and manufacturing. Most groups follow a procedure
116
Chapter Three
of screening business plan submissions to eliminate applications that are not appropriate. If selected, the founder is invited to an “investment meeting” to make a presentation/pitch, followed by Q&A.25 If a sufficient number of angels indicate they are interested in pursuing a venture, then a team of members will conduct due diligence, and if the new venture survives due diligence and a final screen, then the group or a subgroup will negotiate a term sheet and draft an investment agreement.26 Evidence of angel impact on firm performance. Because angel investing is a private industry, data on financial performance is not readily available, and when available, it is subject to a concern with selection bias. That is, investments that generate low or zero returns are unlikely to be reported. Instead of relying on self-reported financial performance estimates, researchers look at objective measures of success such as venture survival, growth, successful exit, and measures of innovation (patents) or impact of innovation (citations to patents). Several studies suggest that angels are beneficial to the growth, performance, and survival of start-ups.27 As with VCs, an important question is whether angels add value primarily through superior deal selection or go beyond it and add value through providing technical and managerial expertise and monitoring and facilitating subsequent funding. Kerr, Lerner, and Schoar (2014) attempt to distinguish between the two hypotheses by comparing ventures that receive angel group funding to ventures that sought funding but were narrowly rejected. The results show that while the selection metrics were similar, ventures that received the funding performed better than those that did not. Hence, it appears based on their sample that improved performance is attributable to the types of skills and expertise that angels can bring to the table. Lerner, Schoar, Sokolinski, and Wilson (2018) use a similar methodology (regression discontinuity) and generate similar results using data from angel groups across multiple countries. Frequently, for ventures that hold promise, an angel investment will be followed by VC. In other cases, ventures receive only angel funding and in others, only VC funding. As noted, some VCs compete directly with angels in seed-stage financing. Dutta and Folta (2016) try to exploit the variation to explore how VCs and angel groups may be associated with different outcomes for ventures. Using a longitudinal sample of tech ventures, the authors find that VCs and angel groups contribute equally to the innovation rates of ventures they finance, but that VC-backed ventures produce more impactful innovation (measured by patent citations) than angel-backed ventures and also faster commercialization. VCs are also associated with more successful exit. They do not find marginal benefits in rates of innovation to receiving VC funding if the venture has already
Venture Capital and Angel Investing 117
received angel group funding, which suggests that the impact of VC funding may have been overstated in previous research.28 Angel Syndicates. In recent years, angel syndicate platforms such as AngelList and OurCrowd have altered the structure of the industry, which historically has been fragmented by geography. The online investment platforms were created to enable diffuse groups of angel investors to aggregate investment capital on a deal-by-deal basis. Through these platforms, accredited investors can invest beyond their local angel groups. This enables them to gain some geographic diversification and may give exposure to more opportunities. An angel syndicate is a fund created to make a single investment and is led by an individual who is an experienced technology investor or is another entity such as a VC firm. The lead identifies an investment opportunity and elects to lead the syndicate as a way to raise funds to pursue the opportunity. Once the syndicate is established, accredited investors may seek to join, but their participation must be approved by the lead. As compensation for the lead’s efforts to find the opportunity, conduct due diligence, and monitor the venture after the investment is made, participants in the syndicate agree to pay the lead’s carried interest. AngelList provides an example of how this works:29 Suppose a notable angel investor decides to lead a syndicate to invest in a promising new venture. The lead agrees to invest $250K in the venture. To do so she makes a personal commitment to invest $50,000, establishes a syndicate as an LLC, and offers the remaining opportunity to prospective investors for the remainder of the deal ($200,000). The investors who join as members of the syndicate agree collectively to invest $200,000 in the deal and to pay the lead 15% of any realized gain as a carry. If the investment has a successful exit, the syndicate investors first receive their $200,000, after which every dollar of the syndicate’s profit is split 80% to the syndicate, 15% to the lead, and 5% to AngelList advisors. There also will be out-of-pocket costs for each deal, such as fees paid to regulatory agencies and accountants, which syndicate investors typically agree to pay. Much like a VC fund, the syndicate investors do not invest directly in a venture. Rather, the syndicate’s fund will invest by purchasing preferred shares, convertible debt, or other instruments issued by the company. The syndicate members invest in a special-purpose fund that is specifically created to invest in companies selected by the syndicate lead, and the fund is formed as an LLC. Both individuals and VC funds can form syndicates. By current regulation, each syndicate can only accept up to 99 investors. In contrast to the LPs of VC funds, syndicate investors can choose which start-ups they want to invest in and can stop investing at any time, and they
118
Chapter Three
also have much lower minimum investments. Syndicates generally do not charge management fees and use the term deal carry (not fund carry). Also, in general, the syndicate lead will put a larger portion of the total investment into the syndicate than the GP of a VC fund typically puts into a fund.
3.10 Summary The modern VC firm has its roots in the U.S. in the midtwentieth century. Limited partnership is the most common structure for VC investment. The limited partners provide financial capital, while the general partner locates and cultivates the ventures and arranges for harvesting the investments. Harvesting usually takes the form of arranging for the venture to go public via an IPO, a private sale to another company, or a management buyout. LPs are precluded from active participation in fund management. The aggregate amount of capital invested in U.S. VC funds increased dramatically through 2000, largely as a result of institutional investors having more flexibility in their investment decisions. As commitments of capital increased, so did competition for deals. Intensified competition has heightened incentives to improve valuation methods and to develop more sophisticated approaches to generating and structuring deals, marketing them to investors, and designing exit strategies. Following an investment in a portfolio company, the GP’s responsibility shifts to creating value in the portfolio company. This means that a partner of a VC firm may sit on the board of directors of the venture. The VC firm is responsible for monitoring operations, recruiting management team members, arranging additional financing, and generally providing expertise and industry contacts. The VC fund has a limited life span and at some point, usually after about 5 years, begins to harvest its investments. The chapter reinforces the role of reputation in the VC market. We considered the contracts that characterize the industry: contracts between the LPs and the GP. The terms and provisions of partnership agreements reflect issues regarding risk bearing and incentive alignment. LPs are passive investors who relinquish control of their investment to a GP that does not bear the full cost of its decisions. Among the issues that arise in this setting are sorting (selecting the portfolio companies from the deal flow) and controlling agency costs and operating costs. Contractual compensation reflects the respective functions of the two types of partners. If the fund does well, the GP realizes a significant return through its carried interest; if the fund does poorly, almost the entire loss accrues to the LPs. Hence, GP compensation is linked to value creation in an asymmetric way.
Venture Capital and Angel Investing 119
The final section of the chapter deals with angel investing. Individual angels can make individual investments in start-ups or they can invest through an organized angel group or through an angel syndicate, which is a new and evolving investment platform. We document an increase in angel activity worldwide and the business practices that angel groups and syndicates use when making investments. Because the industry is highly fragmented, data on investment performance is sparse and relies on self-reported numbers. Recent studies suggest a positive role for angels as measured by successful exits and innovative activity. Review Questions
1. Referring to Figure 3.1, review the major developments that have affected the growth in the VC market in the U.S. Explain, for example, the impact of the Investment Company Act, ERISA’s prudent investor rule, the growth of institutional investing, and general stock market activity. 2. What types of deals are most appropriate for VC and why? 3. What are some possible explanations for the variation across countries in the development of VC markets? 4. Describe the primary features of the LP structure of VC funds, including the roles of the GP and LPs, the length of the partnership, and the compensation mechanisms for the partners. 5. How do VCs add value to new ventures? When can syndication add value and why? 6. Describe the steps in the VC investment process (Figure 3. 8). 7. Explain how VCs distribute capital gains associated with a successful exit of a portfolio company (the waterfall). How is a clawback provision related to the overall distribution of returns to LPs and GPs? 8. Explain the economic rationale for key provisions in a limited partnership agreement, such as limitations on additions of general partners and key person provisions. 9. Why is reputation so important in the VC market? 10. How do angel investors add value? What does the evidence show? 11. How do angel syndicates work? Notes
1. See Gompers (1994) for more information about ARD. 2. Smith, Smith, and Williams (2001) provide a more detailed analysis of the SEC’s interpretation of the ICA fair value standard.
120
Chapter Three
3. Gompers and Lerner (1998b). 4. Based on Pitchbook, Q2 2017 European Venture Report. Based on funds raised by European VC firms, Invest Europe, 2016 European Private Equity Activity, reports a lower total of €6.4 billion ($7.5 billion). 5. Invest Europe, 2016 European Private Equity Activity: Statistics on Fundraising Investments & Divestments, https://w ww.investeurope.eu/research/ activity-data/annual-activity-statistics/. The City of Dublin hosts the European headquarters of a number of prominent firms, including Google, Facebook, and eBay. Irish Venture Capital Association publications, http://w ww.ivca.ie. 6. Data are from US Statistical Abstract, 2004. The government no longer collects this information. 7. Invest Europe, 2016 European Private Equity Activity: Statistics on Fundraising Investments & Divestments, https://w ww.investeurope.eu/research/ activity - data/ annual -activity -statistics/. Government is by far the largest source of funds for European VCs; the next largest is corporations. A fund-offunds is an investment strategy in which a fund invests in other types of funds such as VC funds, stock funds, and bond funds instead of investing directly in new ventures, bonds, stocks, and other types of securities. 8. Preqin Special Report, Sovereign Wealth Funds, August 2017. 9. R&D data are from the OECD Science and Technology and Industry Outlook 2015, https://data.oecd.org/rd/g ross- domestic-spending-on-r- d .htmwww.iriweb.org/sites/default/fi les/2016GlobalR%26DFundingForecast. Note that the definitions of R&D measured by Compustat and OECD are different—for example, Compustat includes foreign expenditures of R&D by U.S. multinationals, while OECD does not. See Gornall and Strebulaev (2015). 10. Gornall and Strebulaev (2015). Also see the National Venture Capital Association Yearbook (2017), https://nvca.org/blog/nvca-2017-yearbook-go -resource-venture- ecosystem/. 11. Ewens and Rhodes-Kropf (2015) find evidence of individual style differences within a VC firm. 12. In their survey of VCs, Gompers et al. (2016) support these ideas, finding that the important factors cited by VCs that lead to syndication are complementary expertise, capital constraints (most important), risk sharing, and future deals (least important). 13. See Ewens, Jones, and Rhodes-Kropf (2013). Chahine, Arthurs, Filatotchev, and Hoskisson (2012) examine the effects of syndicate diversity on earnings management. Cumming, Grilli, and Murtinu (2017) find that European syndicates combining independent private firms and governmental VC investors achieve better exit performance than do government-backed ventures.
Venture Capital and Angel Investing 121
14. Puri and Zarutskie (2012) find that cumulative failure rates of VCbacked firms are lower than those of non-VC-backed firms, and that most of the difference arises shortly after VC involvement begins. Chemmanur, Krishnan, and Nandy (2011) report that the overall efficiency of VC-backed firms is higher than that of non-VC-backed firms and that total factor product increases after VC involvement begins. Guo and Jiang (2013) find that VC-backed firms in China outperform non-VC-backed firms and that performance improves after VC involvement begins. Bernstein, Giroud, and Townsend (2016) find that on-site involvement by VCs leads to increases in innovation and the probability of successful exit. 15. Hellman and Puri (2002). 16. Hsu (2004); Kortum and Lerner (2000). 17. Megginson and Weiss (1991). 18. Studies include Sorensen (2006), who finds that companies funded by more experienced (a proxy for reputation) VCs are more likely to go public, and Krishnan, Ivanov, Masulis, and Singh (2011), who find that VC firms with high shares of prior IPO activity tend to have higher percentages of investments that result in an IPO and better performance after the IPO. Nahata (2008) finds that new ventures backed by VCs with superior reputations are more likely to go public and tend to go public earlier. Smith, Pedace, and Sathe (2011) find that fund reputation is positively related to fund IRR and the net multiple return to investors in the fund. 19. See, for example, Barry, Muscarella, Peavy, and Vetsuypens (1990), Megginson and Weiss (1991), and Brav and Gompers (1997). 20. Lin and Smith (1998) report that the initial returns average 11.6% when VCs with established reputations are among the sellers, compared to only 5.1% when they are not. That is, they appear to protect their reputations by forgoing selling in IPOs that are more likely to be overpriced. 21. Halo report (Wilmington, NC: Angel Resource Institute, 2017). 22. Halo report, 2017. 23. See also Shane (2012). 24. Data from J. Sohl, “The Angel Investor Market,” reports from various years, Center for Venture Research, University of New Hampshire. 25. Kerr, Lerner, and Schoar (2014) describe the procedures followed by two major U.S. angel groups: Tech Coast Angels and Common Angels. Sudek, Mitteness, and Baucus (2008) include a template illustration of the TechCoast Angels screening process. 26. Angel Capital Association, www.angelcapitalassociation.org. Other sources include the Center for Venture Research and the Halo report. 27. See, for example, Lerner, Schoar, Sokolinski, and Wilson (2018).
122
Chapter Three
28. See also Hellmann and Thiele (2015). 29. https://angel.co/help/dyndicates/how-syndicates-work.
References and Additional Reading
Admati, A. R., and P. Pfleiderer. 1994. “Robust Financial Contracting and the Role of the Venture Capitalist.” Journal of Finance 49: 371–402. Ball, E. R., H. Chiu, and R. L. Smith. 2011. “Can VCs Time the Market? An Analysis of Exit Choice for Venture-Backed Firms.” Review of Financial Studies 24: 3105–38. Barry, C. B., C. J. Muscarella, J. W. Peavy III, and M. R. Vetsuypens. 1990. “The Role of Venture Capital in the Creation of Public Companies.” Journal of Financial Economics 27: 447–71. Bernstein, S., X. Giroud, and R. R. Townsend. 2016. “The Impact of Venture Capital Monitoring.” Journal of Finance 71: 1591–1622. Bernstein, S., A. Korteweg, and K. Laws. 2017. “Attracting Early-Stage Investors: Evidence from a Randomized Field Experiment.” Journal of Finance 72: 509–38. Brander, J. A., R. Amit, and W. Antweiler. 2002. “Venture Capital Syndication: Improved Venture Selection Versus the Value-Added Hypothesis.” Journal of Economics and Management Strategy 11: 423–52. Brav, A., and P. Gompers. 1997. “Myth or Reality? The Long-Run Performance of Initial Public Offerings: Evidence from Venture and Non-Venture Capital-baked Companies.” Journal of Finance 52: 1791–821. Chahine, S., J. Arthurs, I. Filatotchev, and R. Hoskisson. 2012. “The Effects of Venture Capital Syndicate Diversity on Earnings Management and Performance of IPOs in the US and UK: An institutional Perspective.” Journal of Corporate Finance 18: 179–92. Chemmanur, T., K. Krishnan, and D. Nandy. 2011. “How Does Venture Capital Financing Improve Efficiency in Private Firms? A Look Beneath the Surface.” Review of Financial Studies 24: 4037–90. Chemmanur, T., E. Loutskina, and X. Tian. 2014. “Corporate Venture Capital, Value Creation, and Innovation.” Review of Financial Studies 27: 2434–66. Chen, H., P. Gompers, A. Kovner, and J. Lerner. 2010. “Buy Local? The Geography of Successful and Unsuccessful Venture Capital Expansion.” Journal of Urban Economics 57: 90–102. Cumming, D., L. Grilli, and S. Murtinu. 2017. “Governmental and Independent Venture Capital Investments in Europe: A Firm-Level Performance Analysis.” Journal of Corporate Finance 42: 439–59.
Venture Capital and Angel Investing 123
Dutta, S., and T. Folta. 2016. “A Comparison of the Effects of Angels and Venture Capitalists on Innovation and Value Creation.” Journal of Business Venturing 33: 39–54. Ewens, M., C. Jones, and M. Rhodes-Kropf. 2013. “The Price of Diversifiable Risk in Venture Capital and Private Equity.” Review of Financial Studies 26: 1854–89. Ewens, M., R. Nanda, and M. Rhodes-Kropf. 2017. “Cost of Experimentation and the Evolution of Venture Capital.” Harvard Business School working paper 15-070. Journal of Financial Economics, forthcoming. Ewens, M., and M. Rhodes-Kropf. 2015. “Is a VC Partnership Greater than the Sum of Its Partners?” Journal of Finance 70: 1081–113. Fiet, J. O., L. W. Busenitz, D. D. Moesel, and J. B. Barney. 1997. “Complementary Theoretical Perspectives on the Dismissal of New Venture Team Members.” Journal of Business Venturing 12: 347–66. Gilson, R. J., and D. Schizer. 2003. “Venture Capital Structure: A Tax Explanation for Convertible Preferred Stock.” Harvard Law Review 116: 875–916. Gompers, P. 1994. “The Rise of Venture Capital.” Business and Economic History 23: 1–24. Gompers, P., W. Gornall, S. N. Kaplan, and I. A. Strebulaev. 2016. “How Do Venture Capitalists Make Decisions?” NBER working paper 2287. Gompers, P. A. 1995. “Optimal Investment, Monitoring, and the Staging of Venture Capital.” Journal of Finance 50: 1461–89. Gompers, P. A., and J. Lerner. 1996. “The Use of Covenants: An Empirical Analysis of Venture Partnership Agreements.” Journal of Law and Economics 39: 463–98. ———. 1998a. “Venture Capital Distributions: Short-Run and Long-Run Reactions.” Journal of Finance 53: 2161–84. ———. 1998b. “What Drives Venture Capital Fundraising?” Brookings Papers on Economic Activity: Macroeconomics, 149–204. Washington, DC: Brookings Institution. ———. 1999. “An Analysis of Compensation in the U.S. Venture Capital Partnership.” Journal of Financial Economics 51: 3–44. Gornall, W., and I. Strebulaev. 2015. “The Economic Impact of Venture Capital: Evidence from Public Companies.” Stanford University Graduate School of Business research paper 15-55. http://dx.doi.org/10.2139/ssrn .2681841. Guo, D., and K. Jiang. 2013. “Venture Capital Investment and the Performance of Entrepreneurial Firms: Evidence from China.” Journal of Corporate Finance 22: 375–95.
124
Chapter Three
Hellmann, T. 1998. “The Allocation of Control Rights in Venture Capital Contracts.” RAND Journal of Economics 29: 57–76. Hellmann, T., and M. Puri. 2002. “Venture Capital and the Professionalization of Start-Up Firms: Empirical Evidence.” Journal of Finance 57: 169–97. Hellmann, T., and V. Thiele. 2015. “Friends or Foes? The Interrelationship Between Angel and Venture Capital Markets.” Journal of Financial Economics 115: 639–53. Hochberg, Y., A. Ljungqvist, and Y. Lu. 2010. “Networking as a Barrier to Entry and the Competitive Supply of Venture Capital.” Journal of Finance 65: 829–59. Hsu, D. H. 2004. “What Do Entrepreneurs Pay for Venture Capital Affiliation?” Journal of Finance 59: 1805–44. Kaplan, S. N., and A. Schoar. 2005. “Private Equity Performance: Returns, Persistence and Capital Flows.” Journal of Finance 60: 1791–823. Kaplan, S., B. Sensoy, and P. Stromberg. 2009. “Should Investors Bet on the Jockey or the Horse? Evidence from the Evolution of Firms from Early Business Plans to Public Companies.” Journal of Finance 64: 75–115. Kaplan, S. N., and P. Stromberg. 2003. “Financial Contracting Theory Meets the Real World: An Empirical Analysis of Venture Capital Contracts.” Review of Economic Studies 70: 281–316. Kerr, W., J. Lerner, and A. Schoar. 2014. “The Consequences of Entrepreneurial Finance: Evidence from Angel Financings.” Review of Financial Studies 27: 1–19. Kortum, S., and J. Lerner. 2000. “Assessing the Contribution of Venture Capital to Innovation.” RAND Journal of Economics 31: 674–92. Krishnan, C. N. V., V. Ivanov, R. W. Masulis, and A. K. Singh. 2011. “Venture Capital Reputation, Post-IPO Performance, and Corporate Governance.” Journal of Financial and Quantitative Analysis 46: 1295–333. Lamoreaux, N., M. Levenstein, and K. Sokoloff. 2004. “Financing Invention During the Second Industrial Revolution: Cleveland, Ohio 1870–1920.” NBER Working Paper series 10923, National Bureau of Economic Research. Lerner, J. 1994a. “A Note on Private Equity Partnership Agreements.” Harvard Business School Case 294-084. Cambridge, MA: Harvard Business Publishing. ———. 1994b. “The Syndication of Venture Capital Investments.” Financial Management 23: 16–27. ———. 1994c. “Venture Capitalists and the Decision to Go Public.” Journal of Financial Economics 35: 293–316.
Venture Capital and Angel Investing 125
———. 1995. “Venture Capitalists and the Oversight of Private Firms.” Journal of Finance 50: 301–18. Lerner, J., A. Schoar, S. Sokolinski, and K. Wilson. 2018. “The Globalization of Angel Investments: Evidence Across Countries.” Journal of Financial Economics 127: 1–20. Lin, T. H., and R. L. Smith. 1998. “Insider Reputation and Selling Decisions: The Unwinding of Venture Capital Investments During Equity IPOs.” Journal of Corporate Finance 4: 241–63. Loughran, T., and J. R. Ritter. 2004. “Why Has IPO Underpricing Changed over Time?” Financial Management (Autumn): 5–37. Megginson, W. L., and K. A. Weiss. 1991. “Venture Capitalist Certification in Initial Public Offerings.” Journal of Finance 46: 879–903. Nahata, R. 2008. “Venture Capital Reputation and Investment Performance.” Journal of Financial Economics 90: 127–51. Puri, M. and R. Zarutskie. 2012. “On the Life Cycle Dynamics of VentureCapital- and Non-Venture-Capital-Financed Firms.” Journal of Finance 67: 2247–93. Ramsinghani, M. 2014. The Business of Venture Capital, 2nd ed. Hoboken, NJ: Wiley. Sahlman, W. A. 1990. “The Structure and Governance of Venture Capital Organizations.” Journal of Financial Economics 27: 473–521. Sedlacek, P. and V. Sterk. 2017. “The Growth Potential of Startups over the Business Cycle.” American Economic Review 107: 3182–210. Shane, S. 2012. “The Importance of Angel Investing in Financing the Growth of Entrepreneurial Ventures.” Quarterly Journal of Finance 2: 1250009-54. Smith, J. K., R. L. Smith, and K. Williams. 2001. “The SEC’s ‘Fair Value’ Standard for Mutual Fund Investment in Restricted and Other Illiquid Securities.” Fordham Journal of Corporate and Financial Law 6: 421–74. Smith, R. L., R. Pedace, and V. Sathe. 2011. “Venture Capital Fund Performance: The Effects, of Exits, Abandonment, Persistence, Experience and Reputation.” Financial Management (Winter): 1029–65. Sorensen, M. 2007. “How Smart Is Smart Money? A Two-Sided Matching Model of Venture Capital.” Journal of Finance 62: 2725–62. Sudek, R. C., C. Mitteness, and M. Baucus. 2008. “Betting on the Horse or the Jockey? Impact of Expertise on Angel Investing.” Academy of Management Proceedings. Tian, X. 2011. “The Causes and Consequences of Venture Capital Stage Financing.” Journal of Financial Economics 101: 132–59.
C h a p t e r Fo u r
Ve ntu r e D e al s
I n t h i s c h a p t e r we turn to the contractual relationship between the en-
trepreneur and the investor and explore how financial contracting can be used to benefit both parties. We emphasize the considerations that bear on the choice of contract terms, namely differences between the entrepreneur and the investor in diversification, in the information they possess, in their expectations, and in incentives. Although dealing with investors can be stressful, time consuming, and inconvenient for the entrepreneur, investors can also be beneficial. Potential benefits derive from three sources: • Outside investment enables the entrepreneur to invest less and increase diversification. • Because well-diversified investors have lower required rates of return, increasing outside investment can increase the present value of the venture. • An investor may contribute expertise, advice, and information that enhance value. Of course, there are also potential problems—disagreements over direction, more time devoted to maintaining the relationship, and conflicts of interest, among others. While we recognize the potential problems, we focus on how contract design can be used to address these concerns and how, on net, investors can add value. In financial terms, contracts with investors have three effects: they allocate risk, they allocate expected returns, and they change risk and expected returns. Debt financing, for example, allocates most of the risk to the equity investor/ 126
Venture Deals 127
entrepreneur, allocates expected returns between the equity holder and creditors, and changes overall returns due to the incentive and tax effects.
4.1 The Economic Framework for Financial Contracting Economics provides a useful framework for thinking about risk and return with the objective of understanding, analyzing, and designing the basic elements of venture deals.
Information and Incentive Problems and Financial Contracting Information and incentive problems are at the core of negotiations between entrepreneurs and investors. New venture markets are highly uncertain and subject to being influenced by unanticipated events. The parties involved may have different expectations about venture success and may have difficulty communicating their expectations. Outside parties often cannot know whether venture performance is due to luck or to managerial capability and effort. We refer to such problems as information problems. Information problems affect start-ups in many ways. For example, a prospective entrepreneur may have an idea with significant commercial value but be unable to protect it by patent or copyright. The entrepreneur cannot disclose the idea to prospective investors without risking its appropriation. How do entrepreneurs deal with such dilemmas? How, conversely, can investors determine the entrepreneur’s true expectations? In addition to information problems, incentive conflicts can arise between the investors and the entrepreneur and among different investors. An entrepreneur with a limited equity stake may not always act in the best interest of other shareholders. If outside financing is in the form of debt, the entrepreneur may want to take on more risk than creditors would like. How can incentive problems such as these be overcome?
A Taxonomy of Information and Incentive Problems A useful distinction exists between costs that arise before a contract is entered and those that arise afterward. Before the agreement, the fundamental problem is information; each party is unsure about what the other knows. After an agreement is reached, the problem is incentives for performance. Consider the notion of a perfect contract. A perfect contract anticipates and provides for every contingency. Its terms bind the parties, so that neither can try
128
Chapter Four
to take advantage of the other. The contract must contain sufficient detail so that a third party can fully enforce its provisions. A perfect contract is efficient in the sense that (1) each risk is allocated to the party who can bear it at least cost, and (2) collectively, the contract terms exhaust the possibilities for mutual gain. New venture financial contracts are far from perfect. They often include extensive lists of representations and warranties and complex contingency structures. Such provisions are designed to deal with information and incentive problems. They tend to be complex but incomplete in that they do not address every contingency. Contractual complexity arises from efforts to address information and incentive problems through detailed contract terms.
Precontractual Information Costs Information costs arise before any sunk (nonsalvageable) investment is made. They arise because the future is complex and uncertain, because the parties may have different information and expectations, and because each party has an incentive to distort their true information or beliefs. Such problems are known as “adverse selection” because they can cause good opportunities to be forgone. The choice to buy health insurance is the classic example of adverse selection. If all people are offered insurance on the same terms, those who expect to be healthy and not to use the insurance will opt not to take it, whereas the unhealthy “adverse selectors” will enroll. Adverse selection is driven by three major types of precontractual information costs: bounded rationality, information asymmetry, and impacted information.1 Bounded rationality. The idea that individuals have limited capacity to deal with complexity is referred to as bounded rationality.2 The parties to a venture cannot anticipate every contingency the venture might face. Even if delineating the contingencies were possible, negotiating provisions to deal with each remote contingency would be too expensive to justify. Decision makers weigh the costs and benefits and rationally stop short of explicitly contracting over all contingencies. To illustrate, consider an entrepreneur who is seeking funding. Product development time may be short or long. Quality of the resulting product and consumer demand may be high or low. Rivals may be successful or unsuccessful in their efforts to compete. Catastrophic events (death of the entrepreneur, war, etc.) may intervene. The funding needs of the venture depend on all of these factors. In some scenarios the investor will be eager to provide funding but not in others. In some, the entrepreneur would prefer to abandon the venture. If contracting costs were low, the parties could develop an elaborate list of contingencies and
Venture Deals 129
design a contract that would specify the response to each. Bounded rationality explains why entrepreneurs and investors do not try to do this. Information asymmetry. Information asymmetry means one party has information that the other lacks and cannot easily acquire. Asymmetric information can prevent a bargain from being struck or cause risk and return to be allocated less efficiently than if information were shared. Information asymmetry can be a problem for entrepreneurial ventures. An entrepreneur usually cannot know why a prospective investor is interested in the venture. The investor may only be seeking to assess the venture as a competitive threat. Entrepreneurs claim that investors sometimes only get involved in a venture to keep it from reaching the market. This could happen if the investor was involved with another venture that targeted the same market. Conversely, the investor may not be able to accurately assess the expertise and commitment of the entrepreneur. Impacted information. The problem is not just that the information is held asymmetrically but also that each party has an incentive to distort what they know. Consequently, each fears exploitation by the other and is reluctant to commit to the venture. Information is impacted when one party is uncertain about what the other knows (information asymmetry) and the parties cannot easily communicate what they know to each other. Impacted information raises the cost of market exchange and contracting because prospective trading partners fear being taken advantage of. To complete the exchange, one or both parties must expend resources to overcome real or perceived information disadvantages. For new ventures, the expenditures often take the form of due diligence investigations in advance of contracting or negotiations of contractual contingencies that mitigate the value of information advantages.
Postcontractual Incentive Problems Once a contract has been entered or a sunk investment has been made, incentives change. The parties may act in ways that are not consistent with their original intentions. Incentive problems arise when contracts are incomplete or when parties cannot monitor performance perfectly. Such problems are known as “moral hazard” because behavior changes once the contract is entered. The classic example again comes from health insurance.3 A person who is fully responsible for his or her medical bills makes value-maximizing tradeoffs in deciding whether to visit a doctor. Because insured individuals are not required to pay the full cost of their medical treatments they will tend to overuse medical services in the sense that the benefit of a specific service to the
130
Chapter Four
patient is less than the cost of providing it, a factor that drives up the cost of insurance for everyone. Anticipating moral hazard is a first step toward finding ways to use deal structure to minimize the costs associated with moral hazard. Financial contracts also can give rise to moral hazard problems. For example, once an investor commits to a venture and shares in the benefits of its success, the entrepreneur’s ownership share is reduced and the entrepreneur may devote less effort to achieving that success. It makes sense that the entrepreneur would devote less effort than if he were the exclusive beneficiary of his efforts. Replacing outside equity with debt financing changes the entrepreneur’s incentives, but if the debt is risky, it does not eliminate the moral hazard.4 Moral hazard is driven by three major types of precontractual information costs: bounded rationality, specific investment, and small numbers bargaining. Bounded rationality. Just as with adverse selection, if complete contingent claims contracting were costless, the parties could remove the potential for moral hazard by completely specifying the contingencies. Costless complete contracting requires both the ability to specify the full range of contingencies and the ability to costlessly monitor the performance of the entrepreneur, neither of which is even remotely feasible. Specific investment. Specific investments are investments that support a given activity or relationship but have little value in alternative use. Once a specific investment is made, it is, in effect, sunk. An example is an entrepreneur’s investment of effort and resources in developing a working prototype of a circuit board that is customized to the needs of a particular customer.5 Suppose, given the planned rate of production, that the amortized quarterly economic cost of effort and resources invested in developing the prototype is $4,000 and the quarterly variable cost of circuit board production is $2,000. The circuit board sales contract between the entrepreneur and the customer calls for the customer to pay $7,000 quarterly. This means that the economic gain to the entrepreneur is $1,000 per quarter. This gain is called an “economic rent.” Economic rent is an ex ante concept and is measured as the excess of expected return over the opportunity cost of resources committed. In deciding whether to make the investment to produce the circuit boards, the entrepreneur commits the resources only if the economic rent is positive. Thus, the economic rent is what makes the NPV of the investment positive. Now consider a scenario in which the resource commitment to manufacture the circuit boards has already been made. Economic rent is no longer relevant;
Venture Deals 131
instead, the concern is whether the return is sufficient to cover the alternative use value of the resources (the variable cost). Suppose these resources have a salvage value of $2,500 per quarter. The $7,000 quarterly revenue is more than sufficient to justify continuing to make circuit boards. The $4,500 difference between revenue and salvage value is called a “quasi rent.” Quasi rent is an ex post concept. It measures the difference between revenue received and minimum revenue necessary to justify keeping the equipment in its current use. Quasi rents exist because of sunk investments. Moral hazard problems are related to the existence of quasi rents. Because the entrepreneur will continue to supply circuit boards as long as the quarterly return is at least $2,500, the customer who is paying $7,000 has an incentive to threaten not to purchase unless the entrepreneur reduces the price. Such an attempt to appropriate the quasi rent is an example of moral hazard that arises after a specific investment is made. This example shows that the parties to a new venture must be sensitive to the presence of specific investments. If the investment is not specific, competition can prevent appropriation. If it is, then the party investing in the specific asset faces the risk of appropriation, and the contract structure may need to address the risk. Small numbers bargaining. Market exchange works best when large numbers of buyers and sellers compete. The market for initial outside financing can be such a market. It is easy to see how the presence of large numbers of potential investors facilitates exchange. Any investment the entrepreneur makes at an early stage is likely to be specific to the venture and nonsalvageable. With a large number of investors competing, the entrepreneur does not need to be concerned about making value-enhancing investments before agreeing to a contract. If the entrepreneur is right, competition among investors will result in terms that more than compensate for the sunk investment. The difficulty arises when the number of prospective investors is small; then the entrepreneur cannot rely on competition to give him an opportunity to recover sunk investment. In fact, sunk investment invites appropriation by an investor. Having large numbers of competitors vying in the initial stage does not prevent opportunism in later-stage negotiations. Many ventures evolve from large numbers of potential investors to small numbers as the venture progresses. The investor in the first round may gain a first-mover advantage that would make it difficult for rival investors to compete in later-stage negotiations. It is not useful to characterize this practice of aggressive bargaining by the investor in the second round as opportunistic, particularly if the pattern is characteristic of the market so that the ability to bargain more aggressively in later rounds is reflected in the terms for the initial investment.
132
Chapter Four
Still, in new venture financing the potential for opportunism is large. To obtain initial financing, the entrepreneur may be asked to provide an abandonment option and possibly other rights, such as the right to acquire control or to terminate the entrepreneur. Why do venture financing contracts give so much control to investors, when it would appear that the potential for opportunism could be limited by using a more complex contract that might give similar rights but only under certain well-specified conditions? Apparently, entrepreneurs in such arrangements do not fear opportunism by investors—but why not? The answer is threefold: First, if the venture is successful, the entrepreneur can pursue other financing, even if the investor withdraws. Second, the investor needs to preserve a good working relationship with the entrepreneur. Third, any investor who wishes to be involved repeatedly in new ventures must avoid gaining a reputation for behaving opportunistically.
4.2 Essentials of Contract Design The term “contract” spans a spectrum of agreements or understandings between parties. At one end, “discrete contracts” are those that contain very explicit provisions. The transactions are impersonal, self-contained exchanges, and performance is easily observable and measurable. The contracts are of short duration, and any contingencies are anticipated and provided for explicitly. At the other end of the spectrum are “relational contracts.” These are highly flexible, implicit contracts based on ongoing relationships between the parties. A relational contract is like a constitution that describes what the parties are trying to accomplish and how, in general terms, they aspire to share the benefits.
Discrete Contracting An example of a discrete financial contract is one between an entrepreneur and a bank. If the loan is collateralized, the contract is discrete. The bank does not need to accept the entrepreneur’s sales projections or verify the venture’s financial performance. The entrepreneur is not concerned that the lender will act opportunistically as long as it is clear that the venture can meet its debt obligations. If the lender does not renew the loan in spite of the entrepreneur’s adequate collateral, the entrepreneur can turn to another lender. Discrete contracting works best for simple exchanges, where the parties have similar expectations, objectives are easy to specify, and performance is easy to verify. If the parties have materially different expectations, it may be possible
Venture Deals 133
to design a more complex contract that is still discrete but circumvents the need to have similar expectations. If expectations are held symmetrically and incentives are not a concern, then the better-diversified party should bear most of the risk of a new venture. For entrepreneurial ventures, this implies that the investor (who is likely to be well diversified) would bear most of the risk and the entrepreneur would have a relatively low-risk claim on the venture’s cash flows. However, both information asymmetry and incentive considerations are central to contracts involving investment in entrepreneurial ventures. The contractual solution will be different if beliefs are asymmetrically held; in such cases, deal terms may shift risk toward the better-informed or more optimistic party. Suppose, for example, that the entrepreneur is confident he can achieve $200,000 in monthly sales within 18 months. The investor is less optimistic and projects monthly sales of $100,000. A deal that provides for equity sharing can be based on the investor’s projections but include a provision that if the entrepreneur’s revenue expectation is met, he will receive a specified percentage increase in equity. Even though the diversified investor is better able to bear risk, the entrepreneur now accepts some of the risk to address the parties’ different expectations. Discrete contracting can also address incentive problems. Suppose an entrepreneur is concerned that a first-round investor may attempt to hold up the entrepreneur when second-round funding is needed. In the first round, the entrepreneur can rely on competition among investors, but afterward the selected investor may gain an information advantage over rivals and could exploit that position. Consistent with earlier discussion, the incentive problem arises because the entrepreneur will have made a specific investment in the venture that is appropriable by the investor. The concern can be addressed by discrete contracting: the parties can agree in advance on the terms and conditions under which the investor will provide second-round financing. Discrete contracting, however, can also create incentive problems. It is usually difficult to state explicit conditions that do not distort incentives. If, for example, the entrepreneur’s equity share is tied to sales in month 18, then the entrepreneur may devote too much attention to sales and not enough to profitability. On balance, it is sometimes better to leave certain provisions general or unstated and to rely more on reputation and the relationship.
Relational Contracting and Flexibility Relational contracts rely on implicit mechanisms such as damage to reputation for enforcement of an understanding.6 The virtue of a relational contract is flexibility. Consider the alternatives of seeking financing from either a
134
Chapter Four
assive investor who has limited familiarity with the market or from a VC who p is involved in the industry and would be an active participant in the venture’s management. The passive investor is more likely to seek a discrete contract. Because of bounded rationality, the contract is likely to be designed around crude proxies for states of the world, such as revenue or profit targets that are only rough indicators of the success of the venture. Conditions for later-stage investment, allocation of ownership, and other factors may hinge on whether the venture meets those targets. Many unforeseen conditions could result in failure to meet the targets even if the venture is worth pursuing. The problem is exacerbated if the investor is passive and has no firsthand knowledge of why a performance target was or was not met. VC financing is likely to be more flexible. Essentially, the VC offers a relational contract that generally does not tie funding to specific performance but rather provides continued support (and possibly funding) as long as doing so is in its interest. Deal terms enable the VC to monitor the entrepreneur and even to restrict the entrepreneur’s ability to act without VC consent. The net effect is that if additional funds are needed, the VC is in a good position to evaluate whether to advance the funds. The entrepreneur typically does not have previous experience or reputation for a managing business. One of the potential benefits for associating with a reputable VC is that they have extensive networks they can rely on to provide expertise when needed. The VC cannot know with certainty how much direct involvement will be required. The result is that the contracts are not explicit about the amount of oversight or monitoring the VC will provide.7 Such contracts expose entrepreneurs to the risk of opportunism, but that risk is mitigated by the reputation of the VC. Discrete contracts and flexible relational contracts are different responses to bounded rationality. Which response is better in any particular setting depends on the identities of the parties, the costs and benefits of flexibility, and the ability of the parties to agree on explicit terms.
Incomplete Contracts and Mechanisms for Resolving Information and Incentive Problems Relational contracts do not “solve” information and incentive problems. Because the contracts are incomplete, other mechanisms that address informational asymmetry and opportunism often accompany relational contracts. The mechanisms that are adopted to complement relational contracts involve signaling, screening, bonding, and monitoring. Technically, the terms “signaling” and “screening” apply to mechanisms that address adverse selection
Venture Deals 135
(informational asymmetry), whereas “bonding” and “monitoring” apply to mechanisms that address moral hazard (opportunism). Signaling. It can be difficult for an investor to distinguish high-quality ventures from the “lemons.”8 Part of the problem is that the entrepreneur’s claims of success potential are costly or impossible to verify. Sometimes an entrepreneur can convey positive private information simply by showing the information to investors and leaving them to draw their own conclusions. But this is risky if an investor could decide to appropriate the opportunity. Sometimes, providing the information is not feasible for the entrepreneur. How, for example, without elaborate documentation, can an entrepreneur establish that efforts to develop new voice-recognition software have progressed much more rapidly than expected? Conversely, how can a VC firm show that it is not involved with or considering any competing products that could influence its attitude toward the entrepreneur’s product? The challenge is to do so in a way that is convincing, yet still preserves the value of the information. One solution is for the holder of the information to use a signal. A signal is a credible demonstration that obviates the need to convey the information. Sometimes, deal terms can serve as signals. Suppose an investor is concerned that an entrepreneur’s financial projections are overly optimistic. By proposing a contract that ties his return to performance targets, the entrepreneur “signals” his positive beliefs. Consider the following terms for a contract between an entrepreneur, Miles Stone, and a VC firm, Limited Deals, Ltd., concerning financing for a virtual amusement park venture. Upon signing the agreement, Limited is committed to finance the venture up to $4.0 million for one year, by which time specific benchmarks are to be met. If the benchmarks are met, Limited will invest an additional $2.0 million; if the venture fails to meet one or more benchmarks, Limited can decide not to invest in the second round. If Limited does not invest, then Miles has 60 days to find a new investor; if he is unable to do so, then Limited has the right to force liquidation. Miles retains ownership of the amusement park idea and the related copyrights and trademarks. If the other assets are sold, the proceeds are to be distributed in proportion to shares of invested capital. Finally, if Limited invests the second-round $2.0 million even though the venture does not meet the benchmarks, then Limited gains managerial control. Willingness to abide by these terms signals several things about Miles. First, the benchmarks are based on his projections. By structuring the deal around them, Miles signals confidence that the projections are reasonable. Second, his willingness to relinquish control if performance expectations are not met
136
Chapter Four
s ignals confidence in his managerial skills. The arrangement also affects Miles’s incentives to devote effort to the venture. By working harder, he helps assure that the benchmarks are met and that he maintains control. Thus, signaling also addresses moral hazard. Screening. Screening is much like signaling, except that the party without the private information offers a menu of alternative terms so that the party with information reveals the information through the act of choosing. In a new venture setting, the investors can offer entrepreneurs alternative types of financing. Given the choice, entrepreneurs with low expectations, for example, are likely to prefer financing that does not include liquidation preference for the entrepreneur or terms that reduce the entrepreneur’s own risk exposure. In contrast, entrepreneurs who are confident in their expectations are more likely to accept favorable liquidation preferences and more of the risk exposure, believing that the venture will succeed and everyone will do well at exit. The screening contract in effect elicits information about the entrepreneur’s beliefs and expectations. Bonding. In contract terms, posting a bond is one way to give credibility to a commitment not to engage in opportunistic behavior. A bond is a penalty that will be paid by the party who makes a promise in the event that the promise is broken. The fundamental element of a bond is that the party who posts it is made worse off if the commitment is violated. Thus, a bond provides an incentive to fulfill a commitment. The commitment can be explicit and specific (like agreeing to resign if the investor is not satisfied with the entrepreneur’s effort) or implicit (like tying compensation to the attained level of sales). Investors may be concerned that an entrepreneur will not focus on the right activities and problems. For example, the entrepreneur may be more interested in working on technical problems than in managing a rapidly growing organization, or he may not recognize failure or may not be open to refocusing the venture. A bond can be provided in several different ways. Sometimes a set of specific contract provisions can function as a bond, such as giving up equity if certain conditions are not met: for example, signing a sales contract with a major customer that the entrepreneur promised he can deliver. Other means of implementing a performance bond would be to give the investor the right to terminate the entrepreneur as CEO if the targets are not met, with the caveat that reliance on specific conditions is not always ideal because random aspects of new venture development may be beyond the control of the entrepreneur and unrelated to
Venture Deals 137
his abilities. Moreover, explicit conditions are subject to manipulation in ways that are not in the parties’ interests. It should be easy to see how the reputation of a VC, or of an entrepreneur who has previous successes, can perform the same function as a bond. In the economics literature, a party can be “trusted” when refraining from opportunistic behavior is the party’s higher-valued course of action. The literature refers to trust in terms of reputation and defines “reputational capital” as a nonsalvageable, intangible asset that is most valuable in its current use. Because of its nonsalvageability, reputational capital functions as a bond. The owners of a reputable VC firm do not want to lose their sunk investments in reputational capital, as could result if the firm were to cheat by reneging on an implicit longterm contract. Uncertainty is greater concerning the behavior of a new entrant to the VC industry or a first-time entrepreneur. If the VC firm or the entrepreneur has not made much of an investment in industry-specific nonsalvageable capital, then others cannot rely on reputation to enforce the contract. Thus, entrepreneurs and investors with significant industry-specific reputations can rely more heavily on promises and trust. What happens in relationships where only one party has reputational capital? A VC with an established reputation, for example, might require that the contract contain explicit provisions regarding the entrepreneur’s performance, whereas the first-time entrepreneur dealing with an established VC might settle for more generally worded promises without explicit enforcement mechanisms. There are indirect means of bonding performance as well. Certification by a third party can fill the gap when reputation of the contracting parties is insufficient.9 For example, reliance on a sales forecast can be supported by tangible evidence that a significant customer is interested in buying the product or has placed a large order. In effect, the entrepreneur is “borrowing” the reputation of one of its customers and using it to enhance the credibility of projections. Similarly, a firm’s affiliation with a reputable VC can provide certification for an untested entrepreneur. Monitoring. The other way opportunistic behavior is controlled is by monitoring. In lieu of bonding, the parties can rely on observation. Monitoring can be direct, as in the case of an investor who serves on the venture’s board, or indirect, as in the case of requiring financial statements to be periodically audited by a reputable accounting firm. The relationships between VC firms and their portfolio companies involve elements of both bonding and monitoring. The deal structures frequently involve
138
Chapter Four
bonding arrangements such as termination rights or warrants, but the investors also closely monitor performance. Some monitoring involves formal triggers; for example, hypothesis testing and staging give the investor specific, periodic opportunities to evaluate whether to continue investing. For ventures where borrowing is possible, the use of debt covenants is an example of monitoring that employs a trigger. Debt covenants are contractual terms used by lenders to place restrictions on borrowers. Debt covenants may restrict the borrower from taking on additional debt or making capital investments beyond what originally was contemplated. They may also require the borrower to adhere to certain constraints with respect to various financial ratios and may limit the ability of the borrower to make distributions to other investors. Use of convertible debt can be a way to allay the lender’s concern that the entrepreneur may take on risks that do not benefit the lender. The debt is convertible at the lender’s option (or triggers specified in the contract) to a predetermined number of shares of common stock. In that way the lender is positioned to benefit from the firm’s performance if it is especially strong and the entrepreneur or firm manager has less incentive to take unwarranted risk. The conversion option thus aligns the interests of the entrepreneur more closely with those of the lender. It is common for high-risk new ventures to borrow using convertible debt or to issue convertible preferred stock.10 In the following sections, we discuss both convertible debt and convertible preferred stock as two main types of securities used in financing high-risk ventures.
4.3 Elements of VC Deal Structure A central aspect of an entrepreneur’s attempt to raise capital is negotiating “the deal” with potential investors.11 The most important and negotiable aspects of the deal between an entrepreneur and a VC or angel investor are stated in the “term sheet.” The term sheet defines the allocation of risk and return as well as the rights and obligations of the entrepreneur and the investor. Because every new venture is different, arrangements between the entrepreneur and investors defy generalization. Nonetheless, certain elements and documents are common. First, it is common for the parties to reflect their mutual understanding in the term sheet. In particular, a typical term sheet will include a statement of covenants and undertakings that indicates what the parties are agreeing to seek to accomplish and what they agree not to do. The statement serves to focus the efforts of the entrepreneur on activities that are intended to advance the
Venture Deals 139
agreed-upon objective and constrains the entrepreneur from devoting resources in other directions. In addition, it is common for a term sheet to include a statement of representations and warranties. The statement may, for example, indicate that the investor has no conflict of interest that would be contrary to pursuing the success of the venture, that the entrepreneur’s educational background is as indicated in a résumé, and that the entrepreneur owns the intellectual property that is critical to venture success. The term sheet typically reflects an agreed-upon valuation and sets out the amount of investment that is to be made, as well as the ownership claims the investor will receive. In addition, the term sheet may identify some of the options, rights, and responsibilities of each party. For example, the investor may have the right to make appointments to the board of directors and, under some conditions, may have the right to withdraw from the project or to terminate the entrepreneur. The entrepreneur may have the right to call on the investor for additional funds in the event that certain milestones are achieved. Often, holders of convertible securities have the right to vote their ownership claims on an as-converted basis and may have additional voting rights in certain situations. Table 4.1 summarizes the basic provisions of a VC term sheet, where the investment is in the form of convertible preferred stock. The term sheet provides the essential information about ownership and risk allocation that, along with other provisions, is a step on the path to a formalized legally enforceable investment agreement. A well-structured deal that takes account of information and incentive problems and of differences in expectations and tolerance for risk can create value for both the entrepreneur and the investor. In the following discussion we consider the basic categories of terms and provisions contained in both the term sheet and the investment agreement.
Investment Form By far the most common type of security that VCs use to make the investment is convertible preferred stock. In contrast, the entrepreneur holds straight equity and is the residual claimant on the cash flows of the firm. To attract investment, the entrepreneur can signal confidence by agreeing to compensate the investor with preferred stock instead of common. Holders of convertible securities generally have the right to convert at any time, but there are also conditions when conversion is automatic, such as when a company goes public. Preferred stock investors are often entitled to receive dividends, and debt investors are entitled to receive interest. However, because
140
Chapter Four
Tab le 4 .1 Standard provisions of a VC term sheet for convertible preferred investment
Terms and Provisions
Description and Purpose
Amount Raised, Price per Share, Pre-Money Valuation, Capitalization
Describes the general parameters of the financing round. Pre-money valuation is the value implied by the price per share before including the proceeds of the round. Capitalization describes the VC structure, including preferred shares used in the funding round. Describes how dividends will accrue to preferred shareholders and timing and conditions under which dividends are paid. Describes the order of payment of proceeds from liquidation, such as: first return the proceeds invested by preferred shareholder; then pay preferred dividends; then pay common dividends; then, if investor shares are participating, share the balance either pro rata or according to terms of the agreement. Requires the company to provide for an employee pool of common shares to aid in recruitment and retention of employees. Provides the right to avoid dilution of ownership stake in a subsequent funding round that results in lower value per share; weighted average ratchets are most common. Requires preferred investors to participate in down rounds or convert to common or lose some preferred rights. Describes the conditions of the right to convert preferred shares to common. Mandatory conversion specifies conditions under which a public offering of the venture would force conversion of the preferred shares. Provides a first right to purchase shares offered by the company and right to sell when other owners sell. Describes any special voting rights (on an as-converted basis) and the right to vote separately to elect a specified number of directors. Describes whether preferred shareholder gets a seat and vote on the board. Describes rights of preferred investors, as a group, to demand redemption of investment from available funds. Describes conditions for registration of shares issued to preferred investors, and the lockup provisions on trading. When there is a public offering of shares, this provision provides for registration of investor shares along with the shares of the company. Provides rights to deal with decisions to sell or take the venture public when there are majority and minority shareholders. For example, a majority who want to sell can drag along the minority, compelling them to sell their shares. Conversely, in special cases where a potential buyer only cares about a majority of shares, then tagalong rights allow for the minority to sell shares on the same terms as the majority. These may include noncompete clauses, employment contracts, key person insurance, information rights, board meeting requirements, and stock and stock option vesting terms.
Dividends for Preferred Shares Liquidation Preference and Participation (if any)
Establishment of Employee Pool Anti-Dilution
Pay-to-Play Conversion Rights
Rights of First Refusal and Co-Sale Preferred Stock Voting Rights Board of Directors Redemption and Registration
Piggyback Registration Drag-along, Tag-along
Other Rights
entrepreneurial ventures are cash-constrained, dividends and interest payments are normally accrued rather than being paid out. Accrued dividends and interest are reflected as increases in the amount invested. In addition, investments in the venture may be “sweetened” with warrants that would enable the investor to acquire additional shares at a prespecified price. Also, the entrepreneur and employees may be compensated partly in the form of stock options.
Venture Deals 141
Price and Valuation In any investment round, the term sheet specifies the amount invested. In investment rounds that include common or preferred shares, the term sheet also specifies the number of shares to be acquired in exchange for the investment. The terms for investing implicitly or explicitly establish a price per share for the investment round. This price per common or preferred share effectively establishes a new valuation for the venture. It is customary and useful to state the valuation on a “fully diluted” basis, where full dilution assumes the conversion of all preferred shares or other convertible securities to common and the exercise of all outstanding stock warrants and options (including any that are issued in conjunction with the financing round). For the purpose of valuation, the normal approach is to ignore any sweeteners of preferred shares that make them more valuable than common shares. Term sheets commonly specify the establishment of an employee pool of shares and stock options. Often, employees in new ventures receive material fractions of their compensation in the form of stock options or shares of stock that may not vest until after several years of employment in the venture. The larger the employee pool, the larger will be the fully diluted number of shares and, therefore, other things being the same, the lower will be the value per share to which the investor will agree. Pre-money and post-money valuation. Two concepts of valuation are frequently used in new venture deals—pre-money and post-money valuation. Post-money value refers to the value of the venture that is implied by the negotiated share price multiplied by the total fully diluted number of shares that would be outstanding after the investment round (i.e., post-money). Pre-money value can be inferred either by subtracting the amount being invested from the postmoney value, or directly by multiplying the negotiated share price by the number of fully diluted shares before the round. Both pre- and post-money values for a given investment round are based on the negotiated share price in that round. It is easy to confuse pre-money value with the post-money value from the prior round. These are not the same. In fact, the difference between the current pre-money value and the prior post-money value is a measure of the increase or decrease in the total value of the existing shares that has transpired since the prior round. If this change in value is positive, the current round is an “upround” in that the value per share has increased. If the change is negative, the current round is a “down-round,” suggesting that the venture is not performing up to the expectations that were reflected in the prior round.
142
Chapter Four
Consider a simple deal in which an investor contributes $1 million in exchange for 400,000 shares of common stock and the entrepreneur retains 600,000 shares. Effectively, the investor is acquiring the shares for a price of $2.50 per share. The term post-money value refers to the total value of the venture that is implied by multiplying $2.50 by the entire 1 million shares outstanding after the new investment. In this case, the post-money value (sometimes referred to as capitalization) is $2.5 million. Pre-money value is the implied value of the venture before the investment. In this case, the pre-money value is $1.5 million (i.e., the post-money value less the $1 million investment). The post-money value is a measure of the value of the venture to the outside investor. The entrepreneur may have different beliefs about value, and those beliefs may affect the negotiation, but they do not appear in the term sheet. If, in the next round, an investor agrees to invest $2 million in exchange for 500,000 shares, the new implied value per share is $4.00. The post-money value in the round is $6 million ($4 time 1.5 million shares), and the pre-money value is $4 million, a $1.5 million increase over the prior post-money value. A caveat on valuation. As noted previously, the post-money value does not reflect the effects of any sweeteners that might have caused the investor to agree to a higher price per share. Accordingly, when the investor’s claim is preferred equity, post-money value is an upwardly biased measure of what the venture is worth. Suppose, in the preceding example, the investor pays $2.50 per share, but the shares are convertible preferred stock (the most common form of VC investment). Because the owner of preferred shares has a prior claim to common stockholders (i.e., the entrepreneur) in a liquidation event, post-money value as calculated above overstates value relative to what the investor would have been willing to pay for common shares. The more sweeteners the entrepreneur is willing to surrender to the investor, the more the investor should be willing to pay for the shares. The result is a higher valuation, both post- and pre-money, but in reality value is not necessarily increased. Rather, it is transferred from the entrepreneur to the investor. The ultimate concern of the entrepreneur is not the post-money valuation but the true value of the entrepreneur’s ownership interest. Many entrepreneurs make the mistake of focusing on post-money value and ignore the value of the “sweeteners” they have promised to the investor. Of course, an entrepreneur who is optimistic about the future of the venture may not perceive it as very costly to give a liquidation preference to a less optimistic investor. In fact, the entrepreneur’s willingness to agree to such a sweetener is one way the entrepreneur can signal confidence in the venture.
Venture Deals 143
Liquidation Preference and Participation Normally, preferred stock has a liquidation preference, where a liquidation event is defined broadly to include such occurrences as bankruptcy, acquisition, sale of voting control to an outside party, or sale of substantially all of the company assets. Under this definition, an IPO is not a liquidation event but rather is another funding event. The IPO generally triggers the immediate conversion of preferred shares to common stock.12 When a liquidation event occurs, the preferred stock investors will have preference to common equity and, hence, will receive distributions first. A typical liquidation preference will specify that the preferred investor first gets a return of some multiple of her investment—such as 1.5 times her investment or 3.0 times her investment. (She may also receive accrued dividends that are paid out as common shares.) If there is an IPO at a sufficient price per share, the convertible preferred equity immediately converts to common equity. Conversion removes the liquidation preference and any other preferences and thus facilitates the IPO. The preferred stock liquidation preference creates a contingent payoff structure for the investor. If the company goes public, conversion is automatic and the investor’s return comes from the increased value per share relative to the value when the preferred stock was purchased. If a liquidation event occurs, the preferred investor can opt to accept the liquidation preference in lieu of converting to common. The investor is not required to accept the liquidation preference. If there is a high-value acquisition, for example, the preferred investor might do better by converting to common than by accepting the liquidation preference. In addition to the liquidation preference, the preferred shares may also be entitled to participate in the proceeds from a liquidation event. These shares are referred to as convertible participating preferred shares. Participation usually is pro rata (proportional to ownership). In many cases it is common for participation to be capped. For example, participating preferred equity with a 1.0 times liquidation preference and a 3.0 times cap on participation would return a total maximum of $3 on each $1 invested.
Antidilution Rights An antidilution right is another sweetener that works to the benefit of preferred stock investors. Antidilution rights are designed to protect the investor from loss of value and ownership share in the event that the next investment round is a down-round. The most common protections are commonly referred to as “ratchet” provisions. Such protections can be full or partial. Under a full ratchet, if the valuation per share declines from what the investor paid, the e ntrepreneurial
144
Chapter Four
firm will issue the investor enough new shares for free (or for a nominal price) so that the investor’s average cost per share is the same as the cost to a new investor. Since a ratchet provision either fully or partially protects the investor from loss of value, an investor who has a ratchet can be expected to demand a smaller fraction of ownership in exchange for a given level of investment. Inclusion of a ratchet, however, can also make raising subsequent financing more difficult. The most common antidilution protection is a “weighted average” ratchet. This provision adjusts the investor’s average price per common share based on (1) the amount of money raised and price per share in the prior round and (2) the amount of money being raised and price per share in the current round. Following a down-round, which is the only time a ratchet provision is activated, this weighted average price is always lower than the original purchase price. The weighted average formula for two financing rounds can be simply stated as (Round 1 investment + Round 2 investment) (Round 1 shares + Round 2 shares) So, for example, if 1 million shares are issued in round 1 at a price of $2.00, and 3 million shares are issued in round 2 at a price of $1.00, the total investment is $5 million (i.e., $2 million in round 1 and $3 million in round 2), and a total of 4 million shares are issued. The weighted average formula will lower the first-round investor’s average price to $1.25 (i.e., $5 million/4 million shares). To lower the first-round investor’s average cost from $2 to $1.25, the investor will need to receive an additional 600,000 shares for free. It is important to recognize that the existence of an antidilution provision from a prior round complicates the valuation in the next round. In the preceding example, the second-round investor has to take account of how the additional free shares going to the first-round investor will affect total shares outstanding. Had there been no such provision, there would be 600,000 fewer shares outstanding and the second-round investor would end up with a larger fraction of the total shares. Accordingly, the ratchet must reduce the amount the second-round investor would be willing to pay per share. In the preceding example, the $1.00 price in the second round would have been higher if there had been no ratchet.
Pay-to-Play Provisions Closely related to antidilution rights are pay-to-play provisions. Such a provision makes an investor’s ability to benefit from an antidilution right contingent on the preferred stock investor’s full participation in the subsequent downround. That is, an investor who does not participate fully in the next round does not receive any antidilution protection. In a more extreme form, the pay-
Venture Deals 145
to-play provision can specify that nonparticipation triggers conversion of the preferred shares to common and the loss of all preferred stock preferences. Pay-to-play provisions are important because it is difficult to attract new investors to a down-round if existing investors are not participating. The provision is often waived in up-rounds, when attracting funding from new investors may not be difficult. The exact definition of “full participation” can vary from the expectation that investors in the prior round fully fund the next round to a more limited case where they fully fund a percentage of the next round, which percentage is determined by the board of directors.
Rights of First Refusal/Preemptive (Pro rata) Rights Somewhat the converse of a pay-to-play provision is a right of first refusal or preemptive right. Such rights give prior investors the right to maintain their pro rata ownership percentages by participating in subsequent financing rounds or the right to prevent sale to an outside party as long as the investors are willing to match the negotiated price. Among other things, preemptive rights enable existing investors to maintain their percentages of voting control, and rights of first refusal enable existing investors to prevent bringing in new investors who could be problematic to deal with later on. Rights such as these are sometimes referred to as non-price-based antidilution provisions since they enable the investor to avoid dilution of voting rights. One potential problem with a right of first refusal is that it can impair the entrepreneur’s ability to shop for better terms from a new investor. Deciding whether to invest in a venture and on what terms is a costly undertaking for a potential investor. If an existing investor has a right of first refusal, the potential investor could be unwilling to undertake the evaluation since any deal the potential investor proposed could be preempted by existing investors via the exercise of their right of first refusal. The net effect is that the round negotiation over investment terms becomes more bilateral, weakening the entrepreneur’s ability to rely on competition with potential investors to produce a good valuation. As a precursor to effective negotiations with potential new investors, it can be helpful for the entrepreneur to seek a waiver of the right of first refusal from existing investors. On the other hand, partial participation in the round by existing investors can signal to new investors that the terms for investing are attractive.
Board Representation, Protective Provisions, and Information Rights New venture term sheets generally include a provision that specifies the size of the board of directors and gives the investor the right to select a given number
146
Chapter Four
of the directors. They also can include protective provisions that require the venture to gain approval of the investor before undertaking certain actions. For example, investor approval may be required before the venture can increase the number of shares, issue shares or other claims to which the existing preferred shares would be subordinate, declare a dividend, borrow money, or make major purchases. Additionally, the term sheet normally indicates information rights the investor may have to budget projections, audited financial statements, and other similar information.
Rights Related to Harvesting and Exit A key concern of investors is that they have a clear path to harvesting their investment. Since investors commonly do not have majority control, if they do not provide a mechanism that can be used to force redemption of their investment, they can become locked into the venture and the entrepreneur may be satisfied with simply continuing to operate the venture and collecting a salary. A number of provisions can work together to assure the investor with a means of harvesting. Redemption rights. Investors may have the right to demand of the entrepreneur that their shares be redeemed based on prespecified terms, such as the initial investment plus accrued but unpaid dividends. This right can be valuable if the venture is successful as a going concern but is not a good candidate for IPO or high-valued acquisition, and particularly where the likely arm’s-length value of the shares would be less than the price indicated by the formula. Demand registration rights. Redemption rights are unlikely to be found in the absence of registration rights. Demand registration rights give the investor the ability to demand that company stock be registered even if the company does not need to raise capital and the entrepreneur does not want to take the company public in an IPO. The main value of demand registration rights is that they can force the entrepreneur to consider either taking the company public or negotiating to buy out the investor. If the public market value is expected to be high, an entrepreneur who wants to keep the company private may need to buy out the investor at a value that is even higher than the public market value. If the company is not doing well, the main benefit of demand registration rights is to force the entrepreneur to negotiate a buyout. Redemption rights can function as a floor on the buyout price. Piggyback registration rights. These rights generally are triggered when a company elects to go public and specify that the investor’s shares are to be
Venture Deals 147
registered along with the shares of the company. Normally, the full cost of registration is borne by the company. The investor may also agree that even though her shares are registered, she will refrain from selling in the public market until (normally) 180 days after the IPO. The main purpose of piggyback registration rights is to enable the investor to gain a public market for her shares without incurring the cost of registration. Drag-along and tag-along rights. It can be difficult to negotiate the sale of a venture unless all shareholders agree to the sale. This gives rise to a potential holdup where a minority of shareholders refuses to sell unless they get a disproportionately large fraction of the sale proceeds. Also, an entrepreneur who holds a minority interest may prefer to continue managing the venture, rather than selling out to an acquirer. Drag-along rights give the majority that wants to sell the right to drag the minority along by selling their shares on the same terms as the majority. In contrast, tag-along rights give the minority the right to sell on the same terms as the majority. A tag-along right can be valuable in the case where a potential buyer cares only about acquiring majority control so that the minority potentially can be frozen out.
Capitalization Table A term sheet commonly will include a capitalization table (“cap” table) that describes who owns the company both before and after the financing round. The table is a spreadsheet or table that shows capitalization, or ownership stakes, in a company, including equity shares, preferred shares and options, and the prices paid by stakeholders for these securities. Table 4.2 is an example of how a simple cap table could display the ownership interests and changes over the first two financing rounds after the founder’s original investment. The founders collectively invest $100,000 and do so at a valuation of $0.01 per share. Based on the investment amount and price per share, the founders receive a total of 10 million shares of common stock. At that point, since no other investor is involved, the founders’ combined ownership share is 100%. In the table, as shown, two things happen. First, a group of angels invests a total of $500,000 in exchange for 2 million shares, implying a valuation of $0.25 per share. In addition, at this point, the investors collectively decide to establish an employee pool that can be used as partial compensation to employees. Initially, this pool can be shares, stock options, or a combination. The pool will be allocated to employees over time, so at any time there may be some shares or options that are allocated to specific employees and others that are not. The important point for the cap table is that at the time the pool is established, no
Tab le 4 . 2 Simple capitalization table
Founders Angel Investors Employee Pool VC Series A Preferred Total
Initial Capitalization
First Round Investment
Investment ($000)
$ per Share
Shares (000)
Percent Ownership
$100
$0.01
10,000
100.0%
Investment ($000)
$500
$ per Share
$0.25
Second Round Investment
Shares (000)
Percent Ownership
10,000
Investment ($000)
Shares (000)
Percent Ownership
66.7%
10,000
58.8%
2,000
13.3%
2,000
11.8%
3,000
20.0%
3,000
17.6%
2,000
11.8%
17,000
100.0%
$4,000 10,000
100.0%
15,000
100.0%
$ per Share
$2.00
A capitalization table describes who owns the company both before and after financing rounds. The founders begin with 100% of the company but after the angels and VCs invest, the percentage ownership is 58.8%. Caveat: percentages shown do not take into account the nature of the equity investment (e.g., convertible preferred versus common) or sweeteners such as liquidity preferences. Those distinctions are clearly documented in a full cap table.
Venture Deals 149
investment is made by employees. In the figure, the employee pool is established to reserve 20% of outstanding shares for employees, or 3 million shares. Based on the total shares outstanding at this point, and the $0.25 value per share from the angel investment, assuming that the angels invest at fair value (a zero net present value), the overall value of the venture is implied to be $3.75 million. Notice that the cost of the employee pool is borne entirely by the founders. If there were no employee pool, the venture would still be worth $3.75 million, of which $500,000 would have been provided by the angels. The remaining $3.25 million would be the value of the founders’ shares, or $0.325 per share. Including the employee pool reduces the per-share value to $0.25 and the resulting value of the founders’ shares to $2.5 million. The $750,000 difference in the founders’ position reflects the implied value of the employee pool. Through the angel investment and the establishment of the employee pool, the founders’ ownership percentage is reduced to 66.7%. In the second outside round, VC investors contribute $4 million in exchange for 2 million shares, an implied value per common share of $2.00. In the cap table the preferred shares are reported on an “as-converted” basis, and no adjustment is made for the value of any other sweeteners that could have affected per-share value. With 17 million shares outstanding at this point, the implied value of the venture is $34 million. The value of the VC shares is the negotiated $4 million (i.e., a zero net present value investment). The angel shares also have an apparent value of $4 million, and the apparent value of the employee pool is $6 million. The residual value of the founder’s shares is implied to be $20 million and the founders’ ownership percentage is reduced to 58.8%. Because the VC investors received preferred shares, the implied valuations of the holdings of the other investors are likely to be somewhat overstated.
4.4 Analysis of Key Term Sheet Provisions In negotiating the key provisions of a term sheet, it is important for both the entrepreneur and the investor to consider a range of possible future valuations and to understand how wealth allocation would be affected by the provisions.
Liquidation Preference To begin, consider an example where a VC investor invests $1 million in convertible preferred stock that has a 3X liquidation preference or is convertible to 20% of the common shares. Based on the investment of $1 million for 20% of common shares, the implied post-money value of the venture at the time of
150
Chapter Four
18.0 16.0
Value of ownership share
14.0 12.0
Investor conversion value Investor preference value Founder residual value Founder share value
10.0 8.0 6.0 4.0 2.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0 10.5 11.0 11.5 12.0 12.5 13.0 13.5 14.0 14.5 15.0 15.5 16.0 16.5 17.0 17.5 18.0 18.5 19.0 19.5 20.0
0.0
Venture value Fi g u r e 4 .1 Liquidation preference: $1 MM investment with a 3X preference or convertible to 20% of common equity The solid lines in the figure compares the preferred stock liquidation value with a 3X liquidation preference against the value of converting immediately to 20% of common equity. The solid black line shows liquidation value with a 3X preference. The solid gray line shows the value of converting to 20% of common. As shown, when venture value exceeds $15 million, the holder of the convertible preferred equity would convert to common equity and forgo the liquidation preference. Dashed lines show the values of the founders’ shares depending on the choice made by the preferred stock investor.
investment is $5 million. At the time of a liquidation event, the investor must choose either the liquidation preference or conversion to common. Figure 4.1 shows the liquidation values of the investor’s alternative claims (the 3X value or the conversion value to common equity) as solid lines. The respective values of the founders’ claims are shown as dashed lines. Note that if the liquidation value is less than $3 million, the venture cannot fully meet the liquidation preference. Instead, the investor receives the entire value of the venture and the entrepreneur gets zero. As shown by the solid lines in the figure, the investor’s liquidation preference is more valuable than the conversion option as long as the exit valuation is below $15 million. If the exit is above $15 million, it is better for the investor to convert to common and forgo the liquidation preference. The investor can be expected to select the alternative that results in highest value. For the founders, the effect is the converse. Since the investor can be expected to choose the greater of the liquidation preference or conversion, the entrepreneur (founder) will get the lesser of the dashed line payoffs in the figure.
Negotiating Investment Terms Figure 4.1 shows only the payoffs that would result from an agreed 3X liquidation preference that is convertible to 20% common stock. Suppose, however,
Venture Deals 151
6.0
Value of ownership share
5.0
Max of 3X or 20% common 25% common
4.0
3.0
2.0
0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0 10.5 11.0 11.5 12.0 12.5 13.0 13.5 14.0 14.5 15.0 15.5 16.0 16.5 17.0 17.5 18.0 18.5 19.0 19.5 20.0
1.0
Venture value Fi g u r e 4 . 2 Liquidation preference: $1MM investment with a 3X preference or convertible to 20% of common stock, versus an alternative contract of 25% of common stock The figure shows the liquidation values of two different investor claims as the value of the venture increases. The comparison is between the maximum value of a 3X preference convertible to 20% of common stock (the black line) versus 25% common (the gray line).
that this is only one of two alternative terms the investor is considering. The other is that instead of the convertible preferred stock with liquidation preference, the investor could receive common stock representing a 25% interest in the venture. Figure 4.2 compares the investor’s payoffs under these alternative provisions. The black line in the figure shows the maximum values from Figure 4.1 of either taking the liquidation preference or converting to 20% common. The gray line shows the alternative value of 25% common stock (with no preference). As you can see, the preferred stock offers a higher payoff for the investor if liquidation value is below $12 million, but a lower payoff if liquidation value is above $12 million. The founders, in contrast, would do better by agreeing to the liquidation preference if the liquidation value turned out to be above $12 million. It should be apparent now how the parties can use variations in investment terms to work around their different perceptions of value. If the investor expects a low liquidation value but the entrepreneur expects a high one, the parties would probably both want the preferred stock with liquidation preference. The investor sees the preferred payoff as attractive, whereas the more optimistic entrepreneur does not see offering the preference as very costly (i.e., he expects a liquidation value above $12 million). It should also be apparent that negotiating over terms
152
Chapter Four
such as these is a way for a party to “put their money where their mouth is.” If the investor is truly pessimistic, she should be attracted by the preference, whereas if the entrepreneur is truly optimistic he should want to offer the preference rather than giving up a larger percentage of the common stock.
Convertible Participating Preferred In addition to liquidation preference, preferred stock sometimes also participates in high-valued liquidations. Such participation may be unlimited or capped at a certain level, usually specified as a multiple of the preferred investment. Figure 4.3 compares four alternative preference and participation contracts from the perspective of the investor. The two solid lines in the figure are the same as in the prior figure—solid black is the maximum value of a 3X liquidation preference or immediate conversion to 20% of common equity. The solid gray line is an alternative provision where the investor would receive 25% of equity in exchange for the $1 million investment. As noted, between these two, an investor who expects a liquidation value below $12 million would prefer the 3X liquidation preference to a larger fraction of common stock ownership.
6.0 Max of 3X or 20% common
5.0
1X preference with full participation
Liquidation value
2X preference with 4X cap
4.0
25% common
3.0 2.0
0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0 10.5 11.0 11.5 12.0 12.5 13.0 13.5 14.0 14.5 15.0 15.5 16.0 16.5 17.0 17.5 18.0 18.5 19.0 19.5 20.0
1.0
Exit valuation Fi g u r e 4 . 3 Comparison of four alternative terms for $1MM investment: 3X liquidation preference or convertible to 20% of common stock; 1X preference and full participation with 20% ownership; 2X liquidation preference with 20% participation up to a 4X cap; and 25% of common stock The figure shows the liquidation values of four different investor claims as the value of the venture increases. The comparison is between the two former contracts shown in Figure 4.2 and two participating convertible contracts: 1X preference with full participation and 2X preference with a 4X cap on participation.
Venture Deals 153
The black dashed line in the figure shows an alternative contract term where the investor has a 1X liquidation preference and participates fully (with 20% ownership) in any excess of the exit value over the $1 million preference. The gray dashed line shows the value of preferred stock with a 2X preference and a cap on the full 20% participation of 4X. The cap is reached at an exit value of $12 million. Given these four alternative contracts, the most conservative/ cautious investor should prefer the 3X preference with no participation, and a somewhat less conservative investor would prefer the 2X preference with capped participation at 4X. This contract is the best alternative for exit values between $7 and $16 million. An optimistic investor who expected an exit value above $16 million would do best by forgoing preferred stock and selecting 25% common stock. If all four alternatives are available, there is no valuation where the 1X preference with full (20%) participation does best. But in pairwise comparisons, it is better than 25% common equity at low valuations and better than both 3X and 2X preferences at high exit valuations. To be clear, we do not expect the entrepreneur to be offered such a full array of options. Rather, the entrepreneur and investor should understand the basic features of the feasible options and be able to evaluate them and then focus negotiation on the structure each finds most appealing given his or her own expectations for the performance of the venture.
Antidilution Antidilution provisions—either full or partial ratchets—are complicated to evaluate. Once a ratchet is in place, if there is a subsequent down-round, the investor in the next round will need to recognize that her decision to invest will cause the earlier investor to acquire additional shares, essentially at a zero price. Since the total number of shares will increase because of the ratchet, the amount the investor should be willing to pay per share should go down. Many presentations of the effects of ratchet provisions ignore the dilutive effect of the provision on the price the new investor is willing to pay. They simply take the new round price as a given and use it to determine the relative holdings. This is clearly misleading as there is simultaneity at work: the price the new investor is willing to pay and the adjustment in the number of shares that will be issued must be determined simultaneously. To see the effect of a full ratchet provision, consider the following example. An entrepreneur holds 4 million shares of the venture. Subsequently, in an A round, a VC invests $4 million in exchange for 2 million shares that include a full-ratchet antidilution provision. Based on the $2 per-share value in the A round, the post-money value of the venture is $12 million. Subtracting the new
154
Chapter Four
money, the entrepreneur’s shares have an apparent value of $8 million (ignoring the fact that the ratchet would have positively affected the price paid by the VC in the A round). Now, suppose that the venture has lost half of its value and needs another $2 million. As shown in the first panel of Table 4.3, had there been no ratchet, the B round investor would have been willing to pay $1 per share—a downround. If there were no ratchet provision, there would then be 8 million shares outstanding at a value of $1, indicating a post-money value of $8 million. With no ratchet, the entrepreneur’s shares would appear to be worth $4 million and the A round VC’s shares would be worth $2 million. From the A round to the B round, both would have lost 50% of their value. Now, consider the effect of the ratchet. First, we can assume that the total value of the company is still the same $8 million as computed above. We also know that the Series A investor’s full ratchet will provide enough additional shares to maintain the value at $4 million and that the new investor will acquire shares worth $2 million. This leaves only $2 million as the residual value for the entrepreneur. We also know that the entrepreneur has 4 million shares. To achieve the $2 million valuation for the entrepreneur, the shares must be worth only $0.50 each. Since the Series A investor has a $4 million position, she must have a total of 8 million shares (6 million additional shares over the original 2 million). Likewise, for the Series B investor to invest $2 million on a zero-NPV basis, the investor needs to receive 4 million shares. So after exercise of the ratchet, the second panel of Table 4.3 shows that there will be a total of 16 million shares and the entrepreneur’s fractional ownership will have fallen from 66.7% after the A round to 25% after the B round—potentially a loss of control for the entrepreneur. There are two important points to make about the mechanics of ratchet provisions. First, the new shares to fulfill the ratchet obligation come from the venture and not from the entrepreneur. The entrepreneur’s shares remain constant relative to the prior round. Second, the entrepreneur bears the full cost of the ratchet provision. That is, the increase in shares that go to the Series A investor to maintain the value gives rise to a per-share value reduction whereby the cost of maintaining the Series A value reduces the value of the entrepreneur’s shares by an equivalent amount. If post-money value is known, evaluating the effects of a full ratchet is straightforward. Simply subtract the amount of the new investment and the amount of the prior investment that is protected by the ratchet, and divide the residual value by the entrepreneur’s total share holdings. This gives the round price per share and from there it is easy to determine how many shares will be held by each party.
Tab le 4 . 3 Comparison of anti-dilution provisions: full ratchet and weighted-average ratchet
Entrepreneur Shares Entrepreneur Value Percentage Ownership A Round Shares A Round Value Percentage Ownership B Round Shares B Round Value Percentage Ownership Total Shares Value per Share Total Value
No Ratchet Provision
Full Ratchet Provision
Weighted Ratchet Provision
Founders
A Round
B Round
Founders
A Round
B Round
Founders
A Round
B Round
4,000,000 NA 100.0%
4,000,000 $8,000,000 66.7%
4,000,000 $4,000,000 50.0%
4,000,000 NA 100.0%
4,000,000 $8,000,000 66.7%
4,000,000 $2,000,000 25.0%
4,000,000 NA 100.0%
4,000,000 $8,000,000 66.7%
4,000,000 $2,666,667 33.3%
2,000,000 $4,000,000 33.3%
2,000,000 $2,000,000 25.0%
2,000,000 $4,000,000 33.3%
8,000,000 $4,000,000 50.0%
2,000,000 $4,000,000 33.3%
5,000,000 $3,333,333 41.7%
2,000,000 $2,000,000 25.0% 6,000,000 $2.00 $12,000,000
8,000,000 $1.00 $8,000,000
4,000,000 $2,000,000 25.0% 6,000,000 $2.00 $12,000,000
16,000,000 $0.50 $8,000,000
3,000,000 $2,000,000 25.0% 6,000,000 $2.00 $12,000,000
The figure shows the impact of two types of anti-dilution provisions when new rounds of equity are raised in a down-round: full ratchet and a weighted ratchet.
12,000,000 $0.667 $8,000,000
156
Chapter Four
Evaluating the effects of a weighted average ratchet is more difficult. The standard equation for the weighting of a ratchet that was shown earlier in the subsection on antidilution calculates the dilution price as a value-weighted average of the sum of the prior and new investments divided by the sum of the prior and new shares. But unless the number of new shares is given, this equation cannot be used to determine the dilution price for the protected investor. In the third panel of Table 4.3, we use the information we do know to determine the round price. We know that that the Series A investment was $4 million and that the Series B investment is $2 million. We use these two amounts to compute the weighted value of the Series A that takes account of the antidilution provision (i.e., 2/3 × $4 million + 1/3 × $2 million = $3.33 million). We also know that the post-money value is $8 million, that the Series B investor will have a 25% interest, and that the entrepreneur has four million shares. We can find the per-share value for the entrepreneur by subtracting the new investment and the diluted value of the A round investment from $8 million. The indicated value of the entrepreneur’s shares is $2.67 million, resulting in a value per share of $0.667. This is the value used to determine the number of shares in the B round—3.0 million shares. Then the number of shares for the A round investor is the diluted value divided by the round price of $0.667, or 5.0 million shares. This is one example of a weighting approach. Actual approaches can vary. IPO ratchet. Ratchets can be negotiated in any round of investment, including a round right before an IPO. In recent years some notable late-stage rounds of convertible preferred investments have included IPO ratchets. These provisions protect the investor, and can provide a promised minimum return, by specifying that if the IPO is priced below a stated price per share (perhaps a price that would be insufficient to generate an agreed-upon return on investment), then the IPO conversion of those shares to common is adjusted so that the investor receives enough additional shares such that the predetermined minimum return on the investment is assured. In effect, it is IPO price protection. Square’s Series E preferred stock investors, who invested right before the IPO, for example, negotiated such protection.13
4.5 Deal Structures of Angel Investments The preceding discussion is related primarily to investments by VC funds. Angel investors often enter into investment agreements that are less complex. They may be as simple as providing a brief description of the venture
Venture Deals 157
and the parties’ aspirations for it, then specifying an amount of investment in exchange for a percentage ownership interest. The agreement might call for periodic reporting, board participation, or other access to information. Angel deals often do not involve formal staging commitments, preferential forms of investment, antidilution protection, or provisions designed to force a liquidity event. It may seem, given that uncertainty is probably greatest at the stages suitable for angel investment, that the deal structures would be similarly complex. Angel investors tend to rely on simple structures for several reasons: (1) the preexistence of complex deal structures that would need to be unwound before VC investment is possible may discourage VC investment; (2) in contrast to VCs, angels rely on informal rather than contractual methods of screening and monitoring; and (3) it is not cost effective to prepare elaborate contracts for small investments where time until the next investment round is needed is likely to be short. With the rise of organized angel investor groups, angel investment agreements have become somewhat more structured, moving more in the direction of VC deals. Ibrahim (2008) ascribes the change to the greater professionalism of angel groups, the higher transactions costs of group investing, and the larger amounts that the groups are able to invest.14
Nonpriced Rounds Often, angel investments are made at very early stages in the life of a venture. In such cases, there is little basis upon which to establish a valuation. Moreover, the amount of investment may not justify spending effort to agree on a valuation. Recently, in the face of these difficulties, some early-stage investors have been using convertible debt instruments to circumvent the need for explicit valuation. Even more recently, investors have begun to use a similar, but non-interest bearing instrument known as a SAFE (Simple Agreement for Future Equity). In fact, some angels now invest only with these types of financial instrument. Financing rounds where such instruments are used are referred to as “nonpriced rounds.” The logic of this type of financing is that in the event of a good outcome, the loan will convert to stock (usually convertible preferred stock) but will do so on terms that are directly tied to the valuation in the subsequent “priced round,” where the subsequent investors put a dollar valuation on the company. In the event of a bad outcome, the convertible debt holder has a prior claim on the value of the venture’s assets up to the value of the loan plus any accrued but unpaid interest. Usually, when such an instrument is used, the loan agreement will specify either a discounted conversion price relative to the price in the next
158
Chapter Four
round, a cap on the conversion price that does not depend on the next round price, or both. The following are examples. Convertible note with discount and cap. Initially, an entrepreneur has 10 million shares of common stock. In a seed round, the entrepreneur raises $1 million in a convertible note from an angel investor. Conversion is triggered if there is a priced equity financing round of at least $2 million. The note has a 30% discount and a $12.5 million cap on the pre-money value. The note will convert at whichever is more favorable to the note holder. Pre-money at the time of the angel investment refers only to the shares held by the entrepreneur. A few months later, the entrepreneur receives a Series A term sheet from a VC firm offering $5 million for 20% of the company. Because of this triggering event, the note holder will convert the loan into preferred equity based on the implied valuation of $25.0 million ($5 million/20%). Because the note has a discount and a cap, the angel investor can choose whichever provision is more valuable. Since the cap terms are prespecified, we can easily determine that the maximum share price under the cap is $1.25 (the $12.5 million pre-money value divided by the entrepreneur’s 10 million shares). On this basis, the $1 million note would convert to 800,000 shares. In total, the entrepreneur and the angel would hold 10.8 million shares, which would represent 80% of the shares. Accordingly, the total number of shares would be 13.5 million and the VC investor would get 2.7 million shares. Given the $5 million investment, the implied value would be $1.8518 per share. Under the cap, with the $25 million valuation, the angel would own 5.93% of the outstanding equity and the entrepreneur would own 74.07%. In either case, the cap or the discount, the VC would own 20%. Pricing under the discount is more complicated since the dilutive effect of the discount depends on the price that is negotiated in the VC round. That is, the round price and total number of shares are determined simultaneously. We do know that the VC is seeking a 20% ownership interest in exchange for a $5 million investment. We also know that the entrepreneur has 10 million shares, that the angel has invested $1 million convertible to shares at a 30% discount, and that together the entrepreneur and the angel will hold 80% of the shares. Algebraically, $5MM/Price = 0.2 × Shares and 10MM + $1MM/(0.7 × Price) = 0.8 × Shares Solving both equations for the number of shares gives $25MM/Price = Shares and 12.5MM + $1.785714MM/Price = Shares
Venture Deals 159
Equating these and solving for Price gives Price = $1.87714 Then substitution this value for Price in one of the original equations gives Shares = 13.46156MM We can then find the conversion price for the note as Price = 0.7 × $1.87714 = $1.30 and the allocations of ownership are as follow: Entrepreneur:
10,000,000 shares
74.29%
Angel investor:
769,231 shares
5.71%
VC investor
2,692,328 shares
20.00%
In this case, with a valuation of $25 million, the cap is more favorable to the angel investor than is the discount. The capped value is $1.25 per share and the discounted value is $1.30. Under the cap, the angel investor would have 5.93% of the equity and under the discount the ownership would be 5.71%. The additional ownership fraction under the cap comes from the entrepreneur. At a post-money valuation of $25 million, the advantage of the cap is slight. At higher valuations the cap would be even more valuable. At lower valuations, the discount would gain value relative to the cap. One final point should be made about the use of caps and discounts. While a discount is harder to evaluate than is a cap, a cap may be harder to negotiate than a discount. While both are referred to as “nonpriced,” the cap implies that there is at least a rough understanding about the range of possible valuations in the next round. If the parties disagree too much about the possible values, it may not be possible to agree on a cap. That is, the entrepreneur may believe that a cap proposed by an angel (or VC) in the first round is too low and will give too much ownership to the angel, whereas an angel may believe that one proposed by the entrepreneur is too high and may actually be higher than the next round valuation, making the cap nonbinding. Possibly for this reason, some angels and VCs negotiate discount terms but avoid using caps.
4.6 Summary The chapter develops a conceptual framework for evaluating contractual features of new venture financing. Differences in attitudes toward risk and
160
Chapter Four
ifferences in expectations about value and potential success give rise to a d significant opportunity for an entrepreneur to contract with outside investors in ways that both parties perceive to be beneficial. The chapter demonstrates that thoughtful design of contracts between entrepreneurs and outside investors can enhance the value of projects and turn unacceptable projects into attractive ones. Adverse selection arises when the parties enter a negotiation with asymmetric information and expectations. Informational asymmetry is an impediment to effective contracting between entrepreneurs and prospective investors, since each may be concerned that the other knows more and is trying to take advantage. Solutions to adverse selection include signaling and screening. A signal is a mechanism that a party with positive information can use to distinguish herself from those with negative information. A screen is a mechanism that a party without information can use to compel the party with information to reveal whether the information is positive or negative. Screening and signaling are similar in that parties with negative information find that it is not economical to imitate those with positive information. Moral hazard arises after an agreement is entered. If a sunk investment has been made and is specific to a relationship, a party to the agreement can attempt to appropriate the value of the investment. Techniques for controlling moral hazard problems include bonding and monitoring. Bonding involves making a commitment or investment such that the party who has the opportunity to appropriate would be worse off by exploiting the sunk investment than by not doing so. Monitoring can be used by one party to limit the ability of the other party to act opportunistically. Given the venture and the entrepreneurial team, the financial contract emerges as a potentially important determinant of success or failure. A welldesigned contract can contribute dramatically to the value of an idea and help allay the concerns of investors about the capabilities and commitment of an untested entrepreneur. A poorly designed contract, however, can just as easily prevent a good idea or product from reaching the market. Information and incentive problems arising from adverse selection and moral hazard are important determinants of financial contract terms and organizational structures. The chapter shows how the provisions in venture capital contracts with entrepreneurs are tied to concerns with the many information asymmetries and incentive issues that characterize start-ups. Use of convertible equity securities (along with liquidation preferences and participation rights) shifts more of the risk of failure to the entrepreneurs. Covenants also address investor concerns with dilution of ownership that may arise in subsequent financing rounds; they define governance issues that address monitoring concerns; and they address
Venture Deals 161
issues related to exit through provisions such as defining “qualified IPOs” and redemption and registration rights. The last part of the chapter focuses on a quantitative analysis of several key features of VC contracts, namely liquidation preferences, participation rights, and antidilution provisions (ratchets). You should now be able to recognize the differences in types of convertible preferred securities and analyze the costs and benefits of provisions from the perspectives of both the entrepreneur and the investor. The chapter concludes with an analysis of convertible note contracts that are commonly used by angel investors and some VCs in seed-stage financing. Firms at this stage are subject to so much uncertainty that investors may not want to put much effort into valuation and hence prefer to invest by way of convertible debt, which converts to equity at a later stage when the firm has received an equity round (and hence a valuation). You should now be able to evaluate the impact for the entrepreneur and the investor of the basic features of convertible notes, in particular the discount and the valuation cap, and how they interact. Review Questions
1. Explain the difference between moral hazard and adverse selection. Why do these two concepts pose contracting problems for new ventures? 2. What is the difference between an information problem and an incentive problem? Give an example of each in an entrepreneurial setting. How do these two problems relate to moral hazard and adverse selection? 3. What is the difference between a screen and a signal? How can screens and signals help to mitigate contracting problems? 4. What are the meanings of bonding and monitoring in the context of new venture finance? What, from an economic standpoint, makes a bond effective? 5. Why is convertible preferred equity the most common instrument used by venture capitalists when investing in new ventures? What incentive and/or information problem does this form of equity address compared to straight equity? 6. Liquidation preferences and participation are both common features of term sheets. How do these provisions work when there is a liquidation event like a winding up of the company or an acquisition? 7. What problem does a ratchet (antidilution provision) try to address? Why would a full ratchet be less appealing to the entrepreneur than is a weighted average ratchet?
162
Chapter Four
8. Why is a convertible note an attractive type of financing instrument for seed-stage ventures? 9. What are the roles of a cap and a discount in a convertible note instrument? 10. Identify some common provisions of a term sheet. What information and incentive problems do these provisions anticipate, and how do they help to mitigate the problems?
Notes
1 Williamson (1975, 1985) refers to these problems in his work on organizational choice and form. 2 Williamson’s (1975) discussion of bounded rationality in a contracting context draws on work by fellow Nobel laureate Herbert Simon. Simon first introduced the concept of bounded rationality in the context of employer-employee relations. See Simon (1957, 1961). 3 The example was developed by Akerlof (1970) 4 For the effects of moral hazard problems on firm value and financing choices, see Jensen and Meckling (1976), Myers (1984), Darrough and Stoughton (1986), and Harris and Raviv (1991). 5 See Klein, Crawford, and Alchian (1978) for seminal work on the risk of appropriation of sunk investments and the choice of contractual arrangement and organizational form. 6 For discussion and illustration of the distinction between discrete and relational contracting, see Joskow (1985, 1987). 7 Gorman and Sahlman (1989) report that VCs regularly monitor their portfolio firms but are not normally involved in day-to-day management. Also see Tian (2011). 8 A well-known example of signaling is Michael Spence’s job market signaling model (1973, 2002) where potential employees send a signal about their ability level to the employer by acquiring (costly) education credentials. The informational value arises from the employer’s belief that there is a positive correlation between education and greater ability given that it is more costly for a lower-ability employee to acquire the education, and they must bear a greater risk of losing the job when true ability is revealed. Leland and Pyle (1977) are among the first to analyze the role of signals in a finance context, specifically the IPO process. They show that “good companies” can send clear signals to the market when going public by keeping control of a significant percentage of the company. The signal must be too costly for “bad companies” to imitate. If no signal is sent, the result can be adverse selection in the IPO market.
Venture Deals 163
9 For evidence on certification, see James and Wier (1990). Thakor (1982) provides a model of third-party certification, and Millon and Thakor (1985) discuss the role of information-gathering agencies. 10 See Kaplan and Stromberg (2002) and Gompers, Gornall, Kaplan, and Strebulaev (2016), who provide evidence on the use of convertible securities in term sheets. 11 See Feld and Mendelson (2016) for elaboration on deal structure and negotiation of terms. 12 This conversion may require the consent of a majority of the preferred stockholders because conversion means that shareholders relinquish their preferences. The term sheet may also identify the size requirement for a “qualified IPO,” so that an IPO is not used opportunistically as a means of removing liquidation preferences. 13 TechCrunch, “Square’s S-1: Of Ratchets and Unicorn Valuations” https:// t echcrunch . com/ 2 015/ 11/ 10/ s quares - s -1 - of - ratchets - and - unicorn -valuations/. 14 Ibrahim (2008); also see DeGennaro and Dobson (2017), who discuss the interactions between VC and angel investing.
References and Additional Reading
Akerlof, G. 1970. “The Market for Lemons: Quality Uncertainty and the Market Mechanism.” Quarterly Journal of Economics 84: 488–500. Alchian, A., and H. Demsetz. 1972. “Production, Information Costs and Economic Organization.” American Economic Review 62: 777–95. Arrow, K. 1963. “Uncertainty and the Welfare Economics of Medical Care.” American Economic Review 53: 941–73. Beatty, R. P., H. Bunsis, and J. R. M. Hand. 1998. “Indirect Economic Penalties in SEC Investigations of Underwriters.” Journal of Financial Economics 50: 151–86. Beatty, R. P., and J. R. Ritter. 1986. “Investment Banking, Reputation, and the Underpricing of Initial Public Offerings.” Journal of Financial Economics 15: 213–32. Bernardo, A., and I. Welch. 2001. “On the Evolution of Overconfidence and Entrepreneurs.” Journal of Economics and Management Strategy 10: 301–30. Bhide, A., and H. H. Stevenson. 1990. “Why Be Honest If Honesty Doesn’t Pay.” Harvard Business Review 68 (September–October): 121–29. Bitler, M., T. Moskowitz, and A. Vissing-Jorgensen. 2005. “Testing Agency Theory with Entrepreneur Effort and Wealth.” Journal of Finance 60: 539–76.
164
Chapter Four
Booth, J. R., and R. L. Smith. 1986. “Capital Raising, Underwriting, and the Certification Hypothesis.” Journal of Financial Economics 15: 261–81. Brennan, M., and A. Kraus. 1987. “Efficient Financing Under Asymmetric Information.” Journal of Finance 42: 1225–43. Coase, R. H. 1937. “The Nature of the Firm.” Economica 4: 386–405. ———. 1960. “The Problem of Social Cost.” Journal of Law and Economics 3: 1–44. Coyle, J. F., and J. M. Green. 2014. “Contractual Innovation in Venture Capital.” Hastings Law Journal 66: 133–83. Cumming, D. 2005. “Capital Structure in Venture Finance.” Journal of Corporate Finance 11: 550–85. ———. 2008. “Contracts and Exits in Venture Capital Finance.” Review of Financial Studies 21: 1947–82. Cumming, D., and S. Johan. 2008. “Information Asymmetries, Agency Costs, and Venture Capital Exit Outcomes.” Venture Capital: An International Journal of Entrepreneurial Finance 10: 197–231. Darrough, M. N., and N. M. Stoughton. 1986. “Moral Hazard and Adverse Selection: The Question of Financial Structure.” Journal of Finance 41: 501–13. DeGennaro, R., and E. Dobson. 2017. “The Future of Angel Finance.” In The World Scientific Reference on Entrepreneurship, vol. 2, ed. D. Siegel and N. Dai. Hackensack, NJ: World Scientific. Dessein, W. 2005. “Information and Control in Ventures and Alliances.” Journal of Finance 60: 2513–49. Feld, B., and J. Mendelson. 2016. Venture Deals, 3rd ed. Hoboken, NJ: Wiley. Gompers, P., W. Gornall, S. N. Kaplan, and I. A. Strebulaev. 2016. “How Do Venture Capitalists Make Decisions?” NBER working paper 2287. Gompers, P. A. 1995. “Optimal Investment, Monitoring, and the Staging of Venture Capital.” Journal of Finance 50: 1461–89. Gompers, P. A., and J. Lerner. 1999. “Conflict of Interest in the Issuance of Public Securities: Evidence from Venture Capital.” Journal of Law and Economics 42: 1–28. Gorman, M., and W. A. Sahlman. 1989. “What Do Venture Capitalists Do?” Journal of Business Venturing 4: 231–48. Grossman, S., and O. Hart. 1986. “The Costs and Benefits of Ownership: A Theory of Vertical and Lateral Integration.” Journal of Political Economy 94: 691–719. Harris, M., and A. Raviv. 1991. “The Theory of Capital Structure.” Journal of Finance 46: 297–355. Hart, O., and J. Moore. 1988. “Incomplete Contracts and Renegotiation.” Econometrics 56: 755–86.
Venture Deals 165
Hellmann, T. A. 2002. “Theory of Strategic Venture Investing.” Journal of Financial Economics 64: 285–314. Ibrahim, D. M. 2008. “The (Not So) Puzzling Behavior of Angel Investors.” Vanderbilt Law Review 61: 1405. James, C., and P. Wier. 1990. “Borrowing Relationships, Intermediation, and the Cost of Issuing Public Securities.” Journal of Financial Economics 28: 149–71. Jensen, M. C., and W. H. Meckling. 1976. “Theory of the Firm: Managerial Behavior, Agency Costs, and Ownership Structure.” Journal of Financial Economics 3: 305–60. Joskow, P. L. 1985. “Vertical Integration and Long-Term Contracts.” Journal of Law, Economics, and Organization 1: 33–80. ———. 1987. “Contract Duration and Transactions Specific Investment: Empirical Evidence from the Coal Market.” American Economic Review 77: 168–83. Kaplan, S. N., and P. Stromberg. 2000. “Venture Capitalists as Principals: Contracting, Screening, and Monitoring.” American Economic Review Papers and Proceedings 91: 426–30. ———. 2002. “Characteristics, Contracts, and Actions: Evidence from Venture Capitalist Analyses.” Journal of Finance 59: 2177–210. ———. 2003. “Financial Contracting Theory Meets the Real World: An Empirical Analysis of Venture Capital Contracts.” Review of Economic Studies 70: 281–316. Kerr, W. R., and R. Nanda. 2015. “Financing Innovation.” Annual Review of Financial Economics 7: 445–62. Klein, B., R. G. Crawford, and A. A. Alchian. 1978. “Vertical Integration, Appropriable Rents, and the Competitive Contracting Process.” Journal of Law and Economics 21: 297–326. Klein, B., and K. B. Leffler. 1981. “The Role of Market Forces in Assuring Contractual Performance.” Journal of Political Economy 89: 615–41. Krishnan, C. N. V., V. Ivanov, R. W. Masulis, and A. K. Singh. 2011. “Venture Capital Reputation, Post-IPO Performance and Corporate Governance.” Journal of Financial and Quantitative Analysis 46: 1295–333. Landier, A., and D. Thesmar. 2009. “Financial Contracting with Optimistic Entrepreneurs.” Review of Financial Studies 22: 117–50. Leland, H. E., and D. H. Pyle. 1977. “Informational Asymmetries, Financial Structure, and Financial Intermediation.” Journal of Finance 32: 371–87. Millon, M. H., and A. V. Thakor. 1985. “Moral Hazard and Information Sharing: A Model of Financial Information Gathering Agencies.” Journal of Finance 40: 1403–22.
166
Chapter Four
Myers, S. C. 1984. “The Capital Structure Puzzle.” Journal of Finance 39: 575–92. Myers, S. C., and N. S. Majluf. 1984. “Corporate Financing and Investment Decisions When Firms Have Information That Investors Do Not Have.” Journal of Financial Economics 13: 187–221. Simon, H. 1957. Models of Man. New York: Wiley. ———. 1961. Administrative Behavior, 2nd ed. New York: Macmillan. Spence, A. M. 1973. Market Signaling: Information Transfer in Hiring and Related Processes. Cambridge, MA: Harvard University Press. ———. 2002. “Signaling in Retrospect and the Informational Structure of Markets.” American Economic Review 92: 434–59. (Also available as his Nobel Prize lecture: http://nobelprize.org/economics/laureates/2001/ spence-lecture.pdf.) Thakor, A. V. 1982. “An Exploration of Competitive Signaling Equilibria with Third Party Information Production: The Case of Debt Insurance.” Journal of Finance 37: 717–39. Tian, X. 2011.”The Causes and Consequences of Venture Capital Stage Financing.” Journal of Financial Economics 101: 132–59. Williamson, O. 1975. Markets and Hierarchies. New York: Free Press. ———. 1983. “Credible Commitments: Using Hostages to Support Exchange.” American Economic Review 73: 519–40. ———. 1985. The Economic Institutions of Capitalism. New York: Free Press.
C h a p t e r Five
N e w Ve ntu r e Str ategy an d R e al O p tio n s
Yo u m ay wo n d e r why, in a book on entrepreneurial finance, we would
devote this chapter and the next to strategic planning. There are three key reasons. First, strategic planning is about choosing a course of action that is designed to achieve a particular objective. In business settings, the overarching objective is financial return. Even in not-for-profit settings where the primary objective may be philanthropic, adequacy of financial return must be an intermediate goal. Second, almost any strategic plan affords opportunities to change course after the initial direction has been selected. In financial terms, these choices are described as real options. Finance provides a means of valuing real options and taking account of those values in the initial strategic choice. Third, for new ventures, the ability even to pursue a particular strategy can depend on whether financing can be found. This chapter establishes the basics of strategic planning in an entrepreneurial setting and develops a framework for evaluating alternative strategies.1 The framework describes real options as decision trees and uses investment valuation to evaluate alternative strategies. The framework begins with identifying the objective and the strategic alternatives for achieving it. In general, we assume that the objective is maximum NPV for the entrepreneur. The alternatives are structures of real options that can be described and evaluated as the branches of a decision tree. We begin by highlighting the interconnected nature of productmarket, financial, and organizational choices.
5.1 Product-Market, Financial, and Organizational Strategy Especially for new ventures, financial strategic decisions must be determined jointly with product-market and organizational strategic decisions. The 169
170
Chapter Five
otential for success is greatest when the decisions are in harmony. Financial p strategy defines the type and timing of financing. Strategic financial decisions include the amount of outside financing, target capital structure, staging of cash infusions, and so on. Product-market strategy involves the targeted sales growth rate, product price, product quality, product differentiation, whether to produce multiple products, and the like. Organizational strategy concerns the horizontal and vertical boundaries of the firm and in whom decision-making authority resides. As an illustration, suppose an entrepreneur were to conclude that rapid sales growth was the preferred strategy in the product market. The strategic choice to grow rapidly would commit the firm to a limited menu of financing options. Rapid growth usually requires external capital. The firm could choose to operate with high financial leverage, in which case it might sacrifice product-market and organizational flexibility and the entrepreneur would realize a residual return after debt service. Or, it might turn to equity capital, in which case the entrepreneur might have to secure agreement of other shareholders before acting and would have to share the total return with the outside investors. It does not make sense to settle on a product-market strategy without considering how that choice restricts financing options. It also makes little sense to place product-market strategy ahead of financing in the decision hierarchy. While it may be possible to settle on a financing option that makes the choice of product-market strategy viable, sequencing the two decisions can lead to second-best outcomes. Even though rapid growth may appear to be attractive, the entrepreneur might be better off if the firm were to grow more slowly. Similarly, there are interactions between financial and organizational strategy and between product-market and organizational strategy. Compared to new ventures, the interdependencies among product-market, organizational, and financial decisions can be less significant in large, wellestablished organizations. If a large company contemplates a strategy of rapid growth in a market that accounts for a small fraction of its total activity, that decision does not necessarily commit the company to a highly leveraged capital structure or the need to raise equity capital by selling stock. Thus, investors may not be harmed if the investment decision and financing decision are treated separately. In fact, this principle of separation is the hallmark of large public corporations and an important difference from entrepreneurial firms. Our point is that for new ventures, product-market, organizational, and financial decisions need to be viewed simultaneously. Simultaneous consideration helps guarantee that the first-best overall strategy is identified and selected. Thinking about product-market, organizational, and financial strategies as simultaneous rather than sequential choices takes us beyond the limits of
New Venture Strategy and Real Options 171
intuitive decision making. Later in the chapter, we use decision trees as devices to capture the interplay between simultaneous and sequential decisions and to identify and evaluate alternative strategies.
5.2 The Interdependence of Strategic Choices: An Example To illustrate the interdependencies of strategic choices, consider the case of C2FO.com, one of several companies that have been established to arbitrage well-known inefficiencies in the management of working capital. In many industries, trade credit terms are established as industry norms to which suppliers must normally adhere. For example, net 30 is the standard term in some industries and means that customers have up to 30 days to pay for products they have purchased and received. Traditionally, as discussed in Chapter 2, a supplier that was short of cash but had a significant balance of uncollected accounts receivable might be able to use the receivables as collateral for a bank loan or might be able to sell the receivables to a factor. At the same time, customers that were offered net 30 terms might be holding substantial cash balances that were invested at low rates of return. This traditional structure could often result in markets with cash-poor suppliers and cash-rich customers, suppliers resorting to expensive means of borrowing, and customers settling for low returns on cash holdings. The rigidities of the trade credit, bank lending, and factoring markets adversely affected the profitability of both customers and suppliers. C2FO is one of the firms that saw this inefficiency as an opportunity that could be addressed with new technology that would match cash-rich customers with cash-poor suppliers and enable them to more efficiently negotiate with each other to transfer liquidity from customers to suppliers. Costco, for example, deals with multiple suppliers and has been a customer on the C2FO network since 2011. Using the C2FO marketplace, Costco can post an available cash balance. Suppliers to Costco, such as those that had sold on terms of net 30 and are in need of cash, can post offers of discounts on their receivables from Costco in order to collect their receivables more quickly. Costco can then decide which offers to accept. The C2FO example illustrates how product-market, organizational, and financing strategies all work together. We can think of cash-short suppliers as potential borrowers that are willing to pay for cash if the loan terms are right. Rather than funding loans to these companies directly and raising its own capital to do so, C2FO effectively intermediates a lending operation (an organizational choice for C2FO that facilitated a new financing strategy for its customers). The
172
Chapter Five
cash comes from purchasers of products supplied by the purchaser’s vendors. C2FO earns a return by intermediating these exchanges in a way that is more efficient and more adaptable to specific needs than what a bank or factor might do. C2FO captures an intermediation spread related to transaction volume.2
5.3 What Makes a Plan or Decision Strategic? It is useful to distinguish strategic decisions from other decisions along three dimensions. First, strategic decisions are consequential. Unlike the decision of which way to drive home from work, strategic decisions involve substantial commitments of time and resources. They are of sufficient magnitude and importance that they are rare, and little precedent exists on which to base them. Second, strategic decisions are both active and reactive. The decision is made in a competitive setting, with regard to the possible actions and reactions of others who have competing or complementary objectives. In selecting among alternatives, the decision maker must consider the choices that may already have been made by others whose objectives overlap and must recognize that others may react to the decision. The decision of when, where, and how to locate a retail store is strategic in this sense. Third, strategic decisions are costly to reverse. If a decision maker selects a wrong course of action, she cannot simply retract it. Investments made to pursue the first course of action are, to some extent, sunk. Sunk investments limit flexibility because the full cost of changing direction must be compared to only the incremental cost of continuing in the same direction. Consequently, an initial wrong strategic choice is one from which the decision maker may never fully recover.
5.4 Financial Strategy Financial strategic choices have the same three elements. A financing choice can limit future financing options in a variety of ways. For example, contractual provisions of a debt agreement may restrict the firm’s ability to redeem the debt and replace it with equity. Existing debt financing may limit the financing sources available for new projects. And debt service requirements may limit the firm’s ability to undertake new projects that would generate negative cash flows in the short run. Competitive interdependencies are also present. Financing choices that are costly to reverse or change involve sunk investments in arranging the financing. In such cases, a firm’s financing choices may credibly commit the firm to a
New Venture Strategy and Real Options 173
particular course of action that is observable by competitors, suppliers, customers, employees, and stockholders. Such credible commitments can influence the actions of these groups. For example, securing project financing can discourage others from moving ahead with plans of their own. Conversely, announcement of intent to develop a new product, if not backed by a credible commitment, could touch off a scramble among rivals to be first in the market. For example, when Airbus announced in 2000 that it was proceeding with the production of its super jumbo jet, the A380, Boeing immediately said that it too would introduce a new super jumbo jet to replace the aging 747. It was not until Airbus had secured financing and began investing in A380 production capacity that Boeing abandoned its plans for a 747 replacement and pursued the 787 Dreamliner. The scope of financial strategy is quite broad. It goes beyond the simple debt-versus-equity financing decision and includes such considerations as the connections between financing choices and growth, flexibility, and control. In addition, financial strategy includes such choices as the use of financial contracts to address or overcome informational asymmetries between entrepreneurs and investors and to better align the incentives of entrepreneurs and employees with the interests of investors.
5.5 Deciding on the Objective For many entrepreneurs, the decisions of whether to launch a new venture or continue in current employment and how to develop the venture are made by intuition. Yet these are not decisions where intuition is likely to lead to the best outcome. Also, some entrepreneurs place weight on qualitative considerations. They may, for example, take pride in having created something new or they may value self-employment.3 While subjective considerations can be important, our objective is to lay out an analytical framework that can help entrepreneurs and investors make better decisions. Here, we propose a two-step process: first, select the strategy that yields the highest estimated NPV; second, make qualitative adjustments to the NPV by assigning subjective values to the considerations that are important to the prospective entrepreneur. Our emphasis is on the first step, maximizing NPV. Generally, we examine NPV from the perspective of the entrepreneur by structuring the terms for outside investment to yield an NPV of zero and keeping the full residual for the entrepreneur. More formally, the first step in developing the analytical framework for strategic planning is to clearly specify the objective. Making rational choices is a
174
Chapter Five
forward-looking concept. Rationality does not mean that the choice is always right, ex post, only that it is expected to be right given the information available at the time of the decision. A prospective entrepreneur considers a variety of alternatives and selects the action that she expects will result in the highest level of satisfaction. A number of qualitative factors bear on the choice. A risk-tolerant person is more willing to abandon secure employment to pursue a venture than is a person who places a high value on security. A person who enjoys work-related challenges is more willing to start a venture than is one who values leisure time. These are the entrepreneur’s personal choices, and we can do little in a formal way to assess such qualitative trade-offs. But we can provide an analytical approach to help the entrepreneur evaluate the trade-offs. Strategic planning reduces the importance of intuition in the decision process so that entrepreneurs and investors can make better-informed decisions. To achieve this, we must begin with an objective that can be measured. We assume that the entrepreneur’s objective is to maximize the value of the venture to herself. Thus, in purely financial terms, 40% ownership of a $5 million business is more valuable to the entrepreneur than 15% of a $10 million business. The value of the entrepreneur’s interest is what should influence the strategic choice.4 This does not mean that the entrepreneur should substitute value maximization for utility maximization as the ultimate determinant of the choice. Once the quantitative value of the entrepreneur’s interest is determined, it is easier for the entrepreneur to assess the qualitative trade-offs related to such factors as control and security. To illustrate, suppose, as an entrepreneur, you are offered two choices by a prospective investor. Under one you would retain control, and under the other you would not. In exchange for relinquishing control, however, you would receive additional compensation having a present value of $100,000. You can then assess whether you would be willing to relinquish control in exchange for a $100,000 increase in the expected value of your interest.5 The entrepreneur is likely to be concerned with the financial return, and risk is also an important consideration. Beyond this, entrepreneurial decisions are likely to be influenced by, among other things, liquidity, diversification, and flexibility; by transferability of ownership; and by the effects of the choice on the entrepreneur’s control and accountability to others. At a more fundamental level, the primary determinant of value is the tradeoff between risk and expected return. Factors such as liquidity, diversification, transferability, and flexibility are taken into consideration through their effects on expected return and risk. The value of control is an additional qualitative factor that goes beyond the objective assessment of present value. As a final note, value maximization does not mean that an entrepreneur is greedy or that a venture is exploitative of either consumers or employees. A vi-
New Venture Strategy and Real Options 175
able venture must offer a product or service that is attractive enough to draw consumers away from alternative purchases. Compensation packages offered to employees must be sufficient to attract them away from other positions.
5.6 Strategic Planning for New Ventures New ventures are different from established businesses because their plans are unconstrained by previous decisions. Questions of financing, organizational design, and product-market strategy are all open. The planning process needs to reflect simultaneous consideration of all three components of the strategy. The distinction between simultaneous and sequential consideration of strategic choices is fundamental. Suppose that on weekends while working in your current position you have perfected a technology for making calorie-free ice cream. You are trying to decide whether to resign and start a venture that will employ the technology. Should you decide to proceed, you must make a number of other decisions. To keep matters simple, in the product market you must choose either a high-margin, slow-growth approach or a low-margin, highgrowth approach. With respect to organizational design, you must choose either to enter only at the manufacturing level and contract for distribution or to enter into both manufacturing and distribution. Let’s suppose that the nonintegrated option involves selling through grocery stores and the integrated approach involves creating a network of branded ice cream stores. Figure 5.1 illustrates the implications of the various productmarket and organizational choices for the venture’s financing needs. If you enter only manufacturing and pursue a slow-growth strategy, your own resources are sufficient to fund the initial investment. The growth rate is determined by the operating cash flows of the venture. With one-level entry and a plan of rapid growth, you must rely on outside financing to supplement the financing available through operating cash flow. If you decide to enter both manufacturing and distribution, even with slow growth the initial investment is too large for you alone, and additional start-up financing is required. With both vertical integration and rapid growth, outside financing is needed both for start-up and to sustain growth. Figure 5.1 displays the two product-market alternatives, along with the two organizational structure alternatives. These choices have direct implications for the available financing options. Each cell in the table represents a combination of a product-market and organizational choices and describes the implied financing choice. The amount shown in the cell along with the financing description is the entrepreneur’s expected NPV associated with that combination of choices. The
176
Chapter Five
Product Market Choice Slow growth Rapid growth
One-level entry Organizational Choice Integrated entry
Initially financed by entrepreneur, growth financed with operating cash flows
Initially financed by entrepreneur, growth financed with operating cash flows and outside financing
NPV = $40
NPV = $120
Initial financing includes outside equity, growth financed with operating cash flows
Initial financing includes outside equity, growth financed with operating cash flows and outside financing
NPV = ($20)
NPV = $70
Fi g u r e 5 .1 Financial implications of product-market and organizational strategic choices Product-market and organizational strategic choices are interdependent with financing choices. One-level entry combined with slow growth minimizes immediate and ongoing needs for external financing. Integrated entry and rapid growth normally require higher levels of immediate and ongoing external financing. NPV reflects the expected value to the entrepreneur.
most valuable is a strategy of entering only at the manufacturing level (relying on contracting for distribution), pricing aggressively to foster rapid growth, and supporting the enterprise initially with your own resources followed by operating cash flows and outside financing. This choice provides the entrepreneur with an expected NPV of $120. However, suppose you do not think through all of the alternatives and settle first on the integrated organizational strategy. In that case, no matter what the product-market strategy, you cannot expect to do better than $70. Moreover, if you choose a high-price/slow-growth strategy, then the expected NPV is negative. Perhaps you can go back to the investors and try to change the deal. But even if the change is possible, it will require renegotiation, and your expected value can never be as high as if you had studied the alternatives first and planned accordingly. Even in this simple illustration, the product-market and organizational decisions limit the financing options. If entry is integrated, available outside financing is likely to be equity. You may have to sacrifice voting control or give other control rights to providers of financing. With rapid growth, financing needs to occur after the venture is established. In the early-growth stages, equity-like claims may still be required, but the fraction of ownership that must be relinquished per dollar raised is likely to be less than if funding is required at start-up. As the venture grows and begins to generate taxable income, debt financing becomes a more realistic and attractive possibility. In addition, alternatives
New Venture Strategy and Real Options 177
such as franchising of distribution outlets and equipment leasing may have advantages over either equity or debt financing. A well-thought-out strategic plan takes such alternatives into consideration. Of course, any commitment of resources to a particular strategy limits flexibility. If, for example, the entrepreneur builds the ice cream factory in California, it would be costly to change and decide that New York would be a better location. But here we are concerned with something more—a loss of flexibility that arises from reliance on outside financing. Almost any outside financing that the entrepreneur can raise comes with limitations on how it can be used and even on how the other resources of the venture can be used. Commonly, such limitations are tied to the strategy that was described in the business plan and was presented to the investors. Even if it were clear to the entrepreneur that a change of plans would be good for the venture, the change often cannot be made unless the investors are also convinced. Convincing them is likely to be difficult, particularly if they have not been heavily involved in the operation or if their interests conflict with those of the entrepreneur. The surest way to limit this problem is to methodically plan the venture before presenting it to investors.
5.7 Recognizing Real Options Strategic planning is not a one-shot exercise. With the passage of time, original targets will be exceeded or missed, and new developments will render the initial plan obsolete. Rather than planning a single immutable course of action, it is more useful to select the strategy that offers the highest expected value in light of the flexibility that the strategy affords for dealing with surprises. Opportunities to abandon a venture, expand it, or change direction are real options. You can think of the focus of strategic planning as deciding which real options to acquire, retain, and abandon at key decision points.6
Option Basics An option is the right to make a decision in the future. In the stock market, for example, a call option is a right to buy a share of stock at some future date for a price that is negotiated today. The right to buy a share of Spacely.com common stock anytime during the next 3 months at a price of $110 is a call option with an exercise price of $110, where the underlying asset is a share of Spacely common stock. The value of the option depends on several factors. Most basically, a call option gains value the more uncertainty there is about the eventual value of
178
Chapter Five
the underlying asset. It gains value if the market price of the underlying asset rises, and it loses value if the price of the asset falls. Suppose Spacely is selling at expiration for $118 per share. The $110 call option would yield an $8 payoff ($118 − $110). If the price of Spacely goes to $125, the value of the call option increases—in this example the payoff almost doubles to $15 ($125 − $110). It follows that call options with low exercise prices are more valuable than those with high exercise prices. The right to buy Spacely for $100 per share obviously is more valuable than the right to buy for $110. Panel A of Figure 5.2 shows the payoffs for this call option with a $110 exercise price. The figure shows clearly that options limit downside risk. Since option risk is one-sided, the more volatile the underlying asset, the higher the value of an option on the asset. Suppose, over the next three months, Spacely is equally likely to increase in value to $140 or to decline in value to $100. If the price of Spacely stock increases to $140, a call option at $110 will be in the money and Call Option Value at Expiration
Panel A
(Exercise price = $110)
$120
Option value at expiration
Expiration date values of call and put options on Spacely.com stock
$100 $80 $60 $40 $20 $0
$0
$50
$100
$150
$200
$250
$200
$250
Stock price at option expiration
Put Option Value at Expiration
Panel B
(Exercise price = $110)
$120
Option value at expiration
Fi g u r e 5 . 2
$100 $80 $60 $40 $20 $0
$0
$50
$100
$150
Stock price at option expiration
New Venture Strategy and Real Options 179
can be exercised to acquire the stock for a saving of $30 per share. If the price of Spacely falls to $100, the call option is out of the money and will not be exercised. Thus, buying a call option limits the downside risk of investing in Spacely but preserves the potential for gain. Because the exposure to risk is one-sided, an option is more valuable the higher the risk of the underlying asset. Because volatility increases with time to expiration, long-term options are more valuable. Similarly, because new ventures are very risky and can take years to harvest, real options on new ventures can be very valuable. The time value of money affects option values because buying an option works like borrowing. If you buy a call, you do not have to come up with the money to exercise the option and purchase the shares until later, if and when you exercise the option. This is different from buying the stock outright, where you must pay the full price today. Accordingly, buying a call option is like borrowing the exercise price without having to pay interest. Since the value of not having to pay interest is greater the higher the interest rate for borrowing, call option value increases with increases in the cost of money. A put option provides the right to sell an underlying asset during a specified period at a specified exercise price. If the owner of a put decides to exercise, he receives the exercise price. In contrast to calls, puts gain value when underlying asset values are low and when exercise prices are high. As can be seen in Panel B of Figure 5.2, a put on Spacely at $110 is more valuable if Spacely is selling for $80 than $90. Like call options, put options are more valuable when the underlying asset is riskier. Because the owner of a put is effectively lending the exercise price without charging interest, a put is less valuable if the cost of money is high (i.e., more interest income is forgone). Calls and puts can be bought or sold. The price of an option contract is called the premium. The seller (or writer) of a call option is obligated to deliver the underlying asset in exchange for the exercise price if the call is exercised. The writer of a put is obligated to buy the underlying asset at the exercise price if the put is exercised. For agreeing to this obligation, option sellers collect the option premium. Puts and calls can be used to allocate the risk of investing in the underlying asset. For example, an investor in Amazon common stock who buys a put option has reallocated the downside risk to the writer of the option. In this case, the stockholder is hedged against the downside risk, and the put writer is acting as an insurer.
Comparisons Between Real and Financial Options Puts and calls on shares of stock are financial options. The underlying asset is a financial asset, and exercising or not exercising the option does not affect the
180
Chapter Five
value of the underlying asset. In several respects, real options are similar. As with financial options, the values of real options increase with the riskiness of the underlying asset and time to expiration. The values of real options are also affected similarly by the difference between the exercise price and the underlying asset value.7 Yet real options differ from financial options in important ways. First, the markets for financial options are often complete, meaning that calls and puts on a stock are available with the same exercise price and expiration date, the underlying asset is freely traded, and riskless borrowing is possible. If the market is complete, option value can be determined by appealing to the ability of investors to arbitrage risklessly any pricing disparities. In contrast, real option markets are not complete and simple arbitrages are not generally feasible. Second, real options are often interdependent in ways that make the application of formal option pricing models inappropriate. Financial options can be bought, sold, and exercised separately, and the value of a portfolio of financial options is simply the sum of the values of the individual options. Real options, in contrast, are often interdependent, and the decision to exercise one may have implications for the values of others. Consequently, the value of a portfolio of real options often cannot be determined by simply adding up the values of the individual options. The term “real options” is used to stress their similarity to financial options while preserving the notion that there are important differences. Returning to the calorie-free ice cream example, once you have developed the technology to manufacture the ice cream, you have options to exploit the technology in many ways. Some are reflected in Table 5.1. You also have the option to do none of those things and instead continue working in your existing position. In addition, you may have the option to delay and, for example, start the venture next year (though in a changed environment). Once you begin the venture, depending on contractual financing arrangements, you or the investor may have the option to abandon it if things do not work out as well as you hoped. We can also intentionally incorporate real options into strategic choices to create additional flexibility. For example, the ice cream venture might acquire options to lease facilities in both California and New York. Effective strategic planning takes account of the values of important real options. Because of the importance of real options to most new ventures, it is useful to incorporate formal decision analysis techniques into the strategic planning process.
The Real Option Premium The cost of acquiring an option is called the option premium. Generally, financial options must be purchased from counterparties who consider the op-
New Venture Strategy and Real Options 181 Tab le 5 .1 Examples of real options
Category of Option
Description
Examples
Option to wait
If it is possible to postpone an action, the decision maker has an option to wait. Decisions involve comparing the value of acting now with the expected value of waiting. Learning options are similar to waiting options except that the focus is on the resolution of uncertainty. The option holder can either make the best choice in light of the uncertain future or wait until some important uncertainty is resolved. An entrepreneur who has the ability to increase the scale of a venture has an option to expand. One who has the ability to downsize has an option to contract.
The owner of a forest can cut the trees today or wait until next year when the trees will have grown larger and lumber prices may have changed.
Option to learn
Options to expand or contract
Options to switch inputs or outputs
An entrepreneur has an option to switch inputs or outputs when she can alter the mix of a production process in response to market prices.
Option to abandon
An abandonment option is a right to discontinue an activity and redeploy the real assets to some other use.
A person who hopes to receive job offers from two different employers either could choose to minimize the expected commute today by buying a house located between the two or could wait for an offer and then locate near the ultimate employer. An entrepreneur might decide to acquire an expansion option by purchasing a facility that is larger than the anticipated need. An entrepreneur might acquire an option to contract by closing down a production line if demand is less than anticipated. The option can be acquired by building a flexible production process. An entrepreneur who designs a facility to operate on either electricity or natural gas has the real option to switch between these two inputs. A refiner who can switch between producing heating oil and gasoline has the real option to switch outputs. Abandonment options include the options to discontinue a research project, close a store, or resign from current employment.
tion premium to be sufficient to compensate them for bearing the risk that the option will end up being in the money. One important difference between real and financial options is that the real option premium bears no necessary relation to the value of the option. The value of an abandonment option, for example, depends on the highest alternative-use value of the assets. Automobiles, for example, have high option value since there is a well-functioning market for used cars. If the original purchaser no longer has a use for the car, it can be easily resold. Although the car has high option value, the new-car market has many competitors, so that the price of the new car is driven mainly by production and distribution costs. Because of the competitive new-car market and well-functioning used-car market, and because the new-car seller is not the counterparty in the used-car market, the new-car seller is not able to demand a price that reflects the option value to the buyer. Thus, the cost of acquiring a real option may have little to do with the option value.
182
Chapter Five
It follows that the value of a strategy can be dramatically improved by making good choices about what real options to acquire and by using them effectively. The terms of a new-car lease contract, in contrast, do reflect an important financial option premium. When a car is leased, the lessee acquires a financial option. At the end of the lease term, if the market value of the car is high, the lessee can purchase the car by paying a residual value that was negotiated at the time of the original lease transaction. Alternatively, if the market value is low, the lessee can return the car to the lessor so that the lessor bears the downside risk about the future value of the car. Although the underlying asset is real, the car lease contract is a financial option. It is effectively a bet about the future market value of the car, and not an instrument that redeploys the actual (real) use of the asset.
5.8 Strategic Planning and Decision Trees A decision tree is a useful way to conceptualize strategic alternatives that involve real options. Constructing a decision tree imposes discipline on the evaluation process and helps the entrepreneur identify relevant real options and the points at which critical decisions must be made. It also enables the entrepreneur to assess, in a structured way, the connections between decisions made today and the value of the venture in the future.
Building and Pruning Decision Trees A decision tree incorporates both decisions and uncertain events that affect the value of the outcome. It identifies a sequence of decisions in which the range of available choices is limited by previous decisions, and the best decision depends on which state of the world is realized. The decision maker is uncertain about which state will be realized but knows or estimates the probabilities of the different states. In addition, the decision maker estimates the NPV of a choice conditional on which future state attains. A few simple techniques can ensure that a decision tree accurately reflects the important choices and help to value them correctly. • Focus on the most important choices. Because the number of branches expands geometrically, decision tree analyses can become intractable. Focusing on a few critical decisions and a few discrete choices is usually all that is needed. • Reason forward to construct the tree. Sequencing is chronological. The tree should illustrate how each choice limits the options for subsequent
New Venture Strategy and Real Options 183
decisions. You can build the tree by first determining the important choices available today. For each of these, determine the choices that would be available at the next key decision point (node), and so on. • Keep track of what is known and unknown at each node. If the decision on optimal scale of entry depends on the level of future demand (a state of the world that is uncertain today), you can only base the decision you make today on expected future demand. • Evaluate choices recursively. Start with the last decision point (the terminal node) and compare the values of the alternatives that emanate from that node as if you had followed the branches up to that point. These are the payoffs conditional on having made the choices that are represented at the prior nodes. • Prune the tree. Select the branch with the highest expected value, conditional on the earlier choices, and eliminate, or “prune,” the inferior choices. For example, if conditional on high demand, you would choose to expand, then you can eliminate the other choices. Move backward to the next earlier decision point and evaluate the choices considering only the highest-valued branch from the subsequent node. • Select the branch of the tree with the highest expected value. This process of backward induction, working from the best future decision conditional on choices made previously, leads to a set of valuations that reflect the values of the embedded options in the decision process.
An Illustration Suppose an entrepreneur is considering investing in a retail venture such as a restaurant or a seller of Internet-of-Things (IoT) products. For simplicity, she assumes that demand for the venture’s products could turn out to be any one of three states of nature: high, moderate, or low. She is considering building a Large Facility, building a Small Facility, or not entering the business at this time. The cost of the Large Facility is $750,000. The cost of the small one is $600,000. The entrepreneur has $400,000 to invest and plans to bring in an investor for the balance. The investor requires 1% of the equity for each $10,000 invested in the venture, resulting in a 35% interest in the Large Facility or a 20% interest in the small one. The entrepreneur retains the balance of ownership. At this point, our focus is on strategic planning and our purpose is to demonstrate the use of decision trees for making strategic choices. To keep things simple, we assume (for now) that the entrepreneur has already estimated the present values (PVs) of the different facility sizes conditional on the different
184
Chapter Five
states of nature and has already estimated the probabilities of the different states. We will relax these assumptions in the next chapter. Naturally, the Large Facility has the potential to generate more revenue, but it also necessitates a larger investment and greater fixed operating expenses. The entrepreneur’s estimates of the PVs in different states of the world reflect these economies. If the high-demand state occurs, the limited capacity of the Small Facility constrains its value. In the high-demand state, we assume that the entrepreneur expects the PV of the Large Facility to be $1.5 million and the PV of the small one to be $800,000. If the moderate-demand state occurs, the PV of future cash flows of both the Large and the Small Facility is expected to be $800,000. In the low-demand state, the PV of the Large Facility is $300,000 and that of the Small Facility is $400,000. The difference is due to the higher fixed cost of operating the Large Facility.8
Evaluating the Venture as an Accept/Reject Decision To establish a baseline for evaluating the real options, consider the project as a simple accept/reject decision with mutually exclusive alternatives—the conventional way of evaluating simple capital investment projects. Figure 5.3 represents the choices in the form of a decision tree. The square in the figure represents the one decision point. The circles represent uncertain outcomes that are beyond the entrepreneur’s control; they reflect states of the world with respect to demand. You can think of “nature” as choosing a state of the world. Each state has an associated probability. In Figure 5.3, the entrepreneur has simplified the decision problem by thinking of it in terms of three states of demand and has estimated the probability of each. The probability of high demand is 30%, the probability of moderate demand is 50%, and the probability of low demand is 20%. In the accept/reject decision, the entrepreneur must decide which, if any, facility to open before the uncertainty about demand is resolved. Her investment is $400,000 regardless of which facility she chooses. Because we are examining the choice from the entrepreneur’s perspective, the difference between the Large and the Small Facilities is shown as the fraction of the value that accrues to the entrepreneur. The triangles in the figure are terminal nodes. Next to each is the NPV the entrepreneur expects if she selects the corresponding branch and the specified level of demand is realized. For example, if the Large Facility is built and the high-demand state is realized, the entrepreneur will invest $400,000 in return for 65% of a facility with a total PV of $1.5 million. The figure shows that $975,000 is the entrepreneur’s 65% share of the PV, so that the resulting NPV of the entrepreneur’s investment is $575,000. In our simple example, the entrepre-
New Venture Strategy and Real Options 185
neur decides how to invest by multiplying each state-contingent payoff by the probability of that state’s occurrence and her fractional ownership interest (.65 or .80) and then subtracting her investment. The result of the calculation is the NPV of that choice to the entrepreneur. Thus, if the Large Facility is selected, her expected payoff is as follows: Expected Payoff = .65 (.3 × $1,500,000 + .5 × $800,000 + .2 × $300,000) = $591,500 As this amount exceeds the entrepreneur’s $400,000 investment by $191,500, it has a positive NPV, which means that it is better than not investing at all. The figure shows $191,500 as the NPV of selecting the Large Facility. By a similar calculation, the entrepreneur’s payoff from investing in the Small Facility is $576,000, an NPV of $176,000. Because both the Large and the Small
30.0%
High Demand
$975,000
= –$400,000 + 0.65 × $1,500,000
Chance
Large Facility –$400,000
$191,500 50.0%
Moderate Demand
$520,000 20.0%
Low Demand
$195,000 30.0%
High Demand
$640,000
50.0% $120,000 20.0% -$205,000 0.0% $240,000
= –$400,000 + 0.65 × $800,000 = –$400,000 + 0.65 × $300,000 = –$400,000 + 0.80 × $800,000
Chance
Small Facility –$400,000
$176,000 50.0%
Moderate Demand
$640,000 20.0%
Low Demand
$320,000 Accept-Reject
30.0% $575,000
0.0% $240,000 0.0% -$80,000
= –$400,000 + 0.80 × $800,000 = –$400,000 + 0.80 × $400,000
Decision $191,500 High Demand
30.0% $0
0.0% $0
Chance
Do Not Enter $0
$0 Moderate Demand
50.0% $0
Low Demand
20.0% $0
0.0% $0 0.0% $0
Fi g u r e 5 . 3 Decision tree for accept/reject decision to invest in the venture Prepared using PrecisionTree™, Palisade Corporation. With a one-time accept/reject decision, the entrepreneur cannot anticipate the level of demand that will be realized. The investment decision and choice of level of investment are made in light of existing uncertainty by maximizing expected NPV.
186
Chapter Five
Facility have positive NPVs, either one is better than not investing. Because the Large Facility offers the entrepreneur a higher NPV, that choice is better than investing in the Small Facility. Thus, the entrepreneur would want to prune the alternatives of investing in the Small Facility or not investing. Ex post, it may turn out that selecting the Large Facility was the wrong decision. If demand turns out to be intermediate, the Small Facility would have been a better choice. If demand turns out to be low, not entering would have been better than either of the other choices. At the time of the decision, however, the entrepreneur cannot know which state of the world will occur. Given what is known and the estimated probabilities of the different states, the Large Facility is the best alternative. In Figure 5.3, the decision facing the entrepreneur is a one-shot choice. There are no real options reflected in the figure. This is the “base case,” with an NPV of $191,500. We now want to expand the range of possibilities by considering three types of real options: (1) the option to wait and learn more about market demand, (2) the option to expand if the Small Facility is selected initially, and (3) the option to abandon.
The Learning Option The ability to wait and learn more before acting is a call option that is open to most prospective entrepreneurs. Rarely does an investment opportunity involve a now-or-never choice. Waiting can add value because uncertainty is reduced, or technology advances, or expenditures of resources are deferred until they are more immediately needed. The offsetting cost is that waiting to invest may encourage others to enter or market conditions may change.9 Returning to our example, suppose that by not investing immediately in the venture, the entrepreneur can learn more about which state of the world is likely to occur, but waiting increases the likelihood of entry by competitors, so that the expected payoffs in the various states are reduced. For simplicity, assume that by waiting the entrepreneur can learn market demand with certainty. Accordingly, in each case she will build the facility that maximizes her NPV. If the good state of nature attains, the entrepreneur can respond by investing in a facility that is optimally sized for high demand. As with any call option, the value of the learning option is higher if the level of uncertainty is high and if delay will materially reduce the uncertainty. Because of the delay and likelihood of competitive entry, we assume that the PV of the Large Facility declines (relative to the base case in Figure 5.3) to $1.3 million and the PV of the Small Facility declines to $700,000. To examine the value of the learning option, we compare the NPV with delayed investment to the base case NPV. The difference is the value of the
New Venture Strategy and Real Options 187
learning option. Figure 5.4 modifies Figure 5.3 to reflect the learning option. Because waiting only affects the “Do Not Enter” branch of Figure 5.3, we have collapsed the two branches that involve investing today. For them, we show only the entrepreneur’s NPV. Under the learning option, conditional on the high-demand state, the entrepreneur’s best course of action is to invest in the Large Facility. Her NPV of waiting and then investing in the Large Facility is $445,000 (net of the $400,000 investment). Because, in the event of high demand, the “Small Facility” and “Do Not Enter” choices have lower NPVs ($160,000 and $0, respectively), we will be able to prune those branches. Figure 5.4 shows that conditional on high demand, the entrepreneur chooses the Large Facility, and the resulting NPV is $445,000.
Large facility
FALSE –$400,000
Small facility
FALSE –$400,000
Learning option
Chance
+
$191,500
+
$176,000
Chance
Decision $213,500 Large facility
TRUE $445,000
Small facility
FALSE $160,000
High demand
30.0% $0
TRUE
Chance
$0
$213,500
FALSE
0.0%
$0
$0
FALSE $55,000
Small facility
TRUE $160,000
50.0% $0
FALSE
FALSE Large facility –$205,000 Small facility
FALSE –$80,000
$0
0.0% $55,000
= –$400,000 + 0.65 × $700,000
50.0% $160,000
= –$400,000 + 0.80 × $700,000
$160,000 $0
20.0%
= –$400,000 + 0.80 × $700,000
Decision
Do not enter
Low demand
$160,000
$445,000
Large facility
Moderate demand
= –$400,000 + 0.65 × $1,300,000
0.0%
Decision
Do not enter Wait
30.0% $445,000
0.0% $0 0.0% –$205,000
= –$400,000 + 0.65 × $300,000
0.0% –$80,000
= –$400,000 + 0.80 × $400,000
Decision $0 Do not enter
TRUE
20.0%
$0
$0
Fi g u r e 5 . 4 Decision tree for investing in a venture, with the option to delay investing until uncertainty about market demand is resolved Prepared using PrecisionTree™, Palisade Corporation. Not investing today may preserve an option to wait until more information is known about the true state of demand.
188
Chapter Five
Similarly, the Small Facility is the best choice for the moderate-demand state. The NPV of investing in the Small Facility if the moderate-demand state occurs is $160,000. By waiting, the entrepreneur also avoids the potential mistake of investing only to learn that the low-demand state has been realized. If low-demand is realized, not entering is the best choice. Because, at present, we do not know which of the three states will be realized, we can only compare the expected value that would result from learning against the most favorable alternative available by choosing not to wait. The immediate question facing the entrepreneur is whether to build now and face the uncertainty about demand or to wait for more information about demand. For this, we need to compare the NPV of investing now (and, as already established, building the Large Facility) against the NPV of waiting. The expected NPV conditional on using the learning option is simply the probabilityweighted average of the three possible outcomes (.3 × $445,000 + .5 × $160,000 + .2 × $0), or $213,500. Because this value is greater than the $191,500 base case expected NPV of the best act-now choice, it is better to wait. The learning option adds $22,000 to the value of the project; this is the value of the real option. The value comes from two sources: the upside gain from building the Large Facility when we know demand is high, and avoidance of the negative NPV outcome that occurs if we were to build any facility in the lowdemand state. These two benefits outweigh the costs associated with waiting.
The Expansion Option A second kind of option is the option to expand the venture after the initial investment has been made. Suppose that after an initial investment of $600,000 in the Small Facility (including $400,000 invested by the entrepreneur), and after learning market demand, the facility can be expanded to the large size by investing an additional $200,000. Assume that the $200,000 (if needed) comes from the outside investor, but on more favorable terms. The terms are better because the second investment is less risky given that uncertainty about demand is resolved. Specifically, the second $200,000 is raised in exchange for a 10% ownership share (1% of the equity for each $20,000 invested), bringing the investor’s total to 30% in the event of expansion (rather than 35% in the nowaiting, Large-Facility scenario). Because the initial investment is sufficient to establish a market presence, we assume that the PV of the Large Facility is $1.4 million, higher than if no immediate investment were made but lower than if the Large Facility were built today. What is the value of the expansion option? Figure 5.5 modifies Figure 5.3 to reflect this option. Here, we have ignored the learning option and have col-
New Venture Strategy and Real Options 189
lapsed the branches where the expansion option does not apply. As the problem is structured, if the Large Facility is chosen initially, the expansion option is implicitly forgone. If the Small Facility is built, the entrepreneur can either expand or not. Expanding, if the high-demand state is realized, increases the entrepreneur’s NPV to $580,000 ($980,000 minus the entrepreneur’s original $400,000 investment). Not expanding leaves the NPV at $240,000, as in Figure 5.3. As this is below $580,000, we can prune the “High Demand/Do Not Expand” branch. Figure 5.5 also compares the values of expanding to not expanding in the moderate- and low-demand states of the world. Expansion is not a good idea with moderate demand because investing reduces the entrepreneur’s ownership share without increasing the value of the venture. So we can ignore the “Moderate Demand/Expand” branch. Not surprisingly, if demand is low, the best choice is “Do Not Expand.” Clearly, expansion is better if the high-demand state obtains. But is it a good idea to invest in the Small Facility first and wait to see what happens to demand?
Large facility
FALSE
–$400,000
+
Chance $191,500 Expand
High demand
30.0% $0
TRUE $980,000
$580,000 FALSE $640,000
0.0% $240,000
= –$400,000 + 0.80 × $800,000
Chance
TRUE
–$400,000
$278,000 Expand
FALSE $560,000
Moderate demand
50.0% $0
Expand
TRUE
FALSE $210,000
20.0%
= –$400,000 + 0.70 × $800,000
$240,000 $640,000
Low demand
0.0% $160,000
Decision
Do not expand
50.0% $240,000
= –$400,000 + 0.80 × $800,000
0.0%
–$190,000
= –$400,000 + 0.70 × $300,000
Decision
–$80,000 Do not expand
TRUE $320,000
Initial choice
= –$400,000 + 0.70 × $1,400,000
Decision
Do not expand Small facility
30.0% $580,000
20.0%
–$80,000
= –$400,000 + 0.80 × $400,000
Decision $278,000 Do not enter
FALSE $0
+
Chance $0
Fi g u r e 5 . 5 Decision tree for investing in a facility, with the option to expand the initial investment Prepared using PrecisionTree™, Palisade Corporation. The option to expand if demand turns out to be high is an example of the flexibility associated with real options.
190
Chapter Five
The answer in this case is yes. As shown in Figure 5.5, the expected NPV from investing initially in the Small Facility and expanding only if the high-demand state is realized is as follows: NPV = .3 × $580,000 + .5 × $240,000 + .2 × ($80,000) = $278,000 where the parentheses indicate a negative value. Comparing this to the base case $191,500 expected NPV makes the expansion option worth $86,500. This is the value of the expansion option compared to the simple accept/reject decision. In this case, we can compare the expansion option strategy that has a NPV of $278,000 with the $213,500 NPV of the learning option (Figure 5.4). According to these calculations, the NPV from investing immediately in a Small Facility (preserving the option to expand) is higher than from investing in a Large Facility ($278,000 > $191,500) and higher than from waiting to invest ($278,000 > $213,500). The incremental value of the expansion option compared to waiting is $64,500. The key point of Figure 5.5 is that ignoring the option to expand would have led the entrepreneur to select a less valuable strategy. In some cases, a project may be passed up entirely because of the failure to recognize that initial investments sometimes create valuable options that can be exercised in the future if the environment is right. Here we see a difference between real and financial options that sometimes is important: the waiting option and the expansion option are mutually exclusive ways to use the real asset. The entrepreneur cannot exercise both. Normally, financial options are independent of each other—they can be acquired separately and decisions to exercise them can be made independently. But because real options relate to how an underlying real asset is used, they sometimes are mutually exclusive. In this case, if the entrepreneur wants to acquire the expansion option, the investment must be made today, thus forgoing the learning option. That is, unlike financial options, where exercise choices can be made independently, sometimes the holder of a real option has to choose from available options—exercising one can preclude exercising others.
The Abandonment Option The final type of option we consider is the option to abandon the venture if things do not work out as well as expected. Suppose the facility has an alternative use as office space. If converted to office space (for a negligible net expenditure), the PV of the Large Facility would be $600,000 and the value of the Small Facility would be $300,000.
New Venture Strategy and Real Options 191
You can see from the numbers that the option to abandon the Small Facility is worthless. This is because a Small Facility, even in the low-demand state, has a PV of $400,000, which is more than its PV as office space. But the option does have value for the Large Facility, because $600,000 is more than the $300,000 present value as a facility in the low-demand state. The best strategy up to this point is to build small and then expand if demand turns out to be high. Is the value of the option to abandon the Large Facility enough to tip the balance in favor of building the Large Facility immediately? The “Large Facility/Low Demand/Abandon” path would mean a $390,000 payoff to the entrepreneur (.65 × $600,000). The entrepreneur’s resulting NPV from investing immediately in the Large Facility with the option to abandon is as follows: NPV = −$400,000 + .3 × $975,000 + .5 × $520,000 + .2 × $390,000 = $230,500 This is less than the $278,000 NPV of initially investing in the Small Facility with the option to expand but higher than any other alternative. If the option to expand did not exist, then investing in the Large Facility with the option to abandon would be the preferred strategy. It would exceed the NPV of the next best alternative—waiting until the state of the world is realized—by $17,000 ($230,500 − $213,500). The option to abandon a venture that does not work out as well as expected can be critical to value and thus to the decision to launch the venture in the first place. It can also impact the willingness of investors to fund the venture. Test pilots of new aircraft normally take along parachutes in case something goes wrong; otherwise, they would be much less interested in testing new designs.
5.9 Decision Trees and Contract Negotiation In the preceding example, we have assumed that the entrepreneur has found an investor who will agree to invest on the indicated terms (i.e., 1% of equity for each $10,000 invested in most cases). More realistically, the entrepreneur will need to negotiate and the terms will need to be mutually beneficial. In this illustration, we do not know whether the discount rate that would be used by the investor would be the same as that used by the entrepreneur. We also do not know whether the investor agrees about the probabilities of the different outcomes or the expected cash flows conditional on an outcome or real option choice. It could also be that the investor’s equity claims have a preference relative to the entrepreneur’s shares.
192
Chapter Five
If we assume that the investor has the same expectations as the entrepreneur and uses the same discount rate, we can use the decision tree information to determine that the best option for the entrepreneur, build the Small Facility and expand if demand is high, would not be attractive to the investor. The investor’s NPV from this strategy would be $38,000. NPVInvestor = [0.3 × (–400 + 0.3 × 1,400) + 0.5 × (–200,000 + 0.2 × 800) + 0.2 × (–200 + 0.2 × 400)] × $1,000 NPVInvestor = $6,000 – $20,000 – $24,000 = –$38,000 The NPV is positive if demand is high but negative otherwise. The willingness of the investor to accept the terms in the example indicates that the investor has a different perspective on the venture, such as a lower discount rate, lower perceived risk, or a higher probability that expansion would be warranted. For example, as discussed in Chapter 1, the entrepreneur’s valuations could be based on a high discount rate due to necessary underdiversification. A welldiversified investor could appropriately use a lower discount rate. If doing so were to double the present values of the cash flows under the expansion option strategy, the investor’s NPV would be positive $184,000 and would be positive in both the high- and moderate-demand outcomes. While the details of discounting cash flows to present value is beyond the scope of this chapter, the point here is that the entrepreneur needs to think about the venture terms from both perspectives and seek to develop a financing strategy that would appeal to both parties. Examining the decision tree from both perspectives can enable the entrepreneur to assess this.
5.10 Rival Reactions and Game Trees Decision trees do not explicitly incorporate the reactions of rivals or counterparties. For example, consider a venture that offers an innovative line of 3D-printed athletic shoes. Its entry strategy—of price, advertising expenditures, geographic scope, and so on—may depend critically on how it expects incumbent firms to react. The firm may decide to enter with an aggressive pricing strategy, assuming that incumbents will keep their prices constant. But if incumbents react by pricing aggressively, then the effectiveness of the new entrant’s pricing strategy will be reduced and perhaps a toehold entry strategy would discourage rival reaction and be better for the entrepreneur in the long run. Similarly, the terms of an agreement that is negotiated between a VC and an entrepreneur can have game-theoretic implications. For example, a VC might
New Venture Strategy and Real Options 193
be concerned that committing too much capital at an early stage might lead the entrepreneur to work less diligently on achieving the next important milestone, whereas investing a smaller amount might elevate the entrepreneur’s concern about the difficulty of negotiating a fair valuation in the next round. Rival (and counterparty) reactions are not an issue in perfectly competitive markets because other firms will not react specifically to the entry of a new rival and competition will mitigate the potential for holdup by an early investor (counterparty). For a small venture entering a large market or an entrepreneur who can deal with many potential investors, it may make sense to think of the market as nearly perfectly competitive. Furthermore, rival reactions are not an issue for a venture that offers a unique product and anticipates no entry, and therefore has market power. But planning for rival reactions can be important in settings where there are only a few actual or potential competitors. In such situations, the decisions of the competing firms and counterparties are highly interdependent. One way to evaluate decision choices is to assign probabilities to rival reactions and use a decision tree to evaluate the choices. This approach is not likely to yield reliable results, however, because the probabilities effectively yield a weighted average reaction from among the various possibilities, whereas the actual reaction will be one choice or another. There is a better way to think about the decision when rival reactions are important. The alternative is to explicitly assess what reaction would be in the best interest of each rival. The underlying reasoning is that a firm’s actions and its rivals’ are interdependent. Each firm is assumed to behave rationally by trying to select the strategy that maximizes value given what it believes its rival will do. Strategic interactions can be modeled by relying on the contributions of game theory. Game theory is concerned with analysis of optimal decision making when decision makers are aware that their actions affect each other’s behavior and take these interactions into account. To illustrate the uses of game theory in formulating strategy, we need to introduce some new terminology. A game consists of (1) a set of players, (2) an order of play, (3) the information set available to the players, (4) the set of actions available to each player, and (5) the payoff schedule that results from the actions. The players are simply the decision makers, such as an entrepreneur, a firm manager, a VC, or a rival. The players make decisions at various points in a game (decision nodes). In a decision tree, the decisions are based on alternative states of the world, which are not strategic. In a game-theoretic setting, however, player actions are strategic and driven by rationality. The sequence in which decisions are made is the order of play. If all players make their decisions one at a time in a sequence, then the
194
Chapter Five
game is a sequential-move game. If all decision makers act at the same time, then the game is a simultaneous-move game. For an entrepreneur, it usually makes sense to think of the game as sequential, with the entrepreneur as the first mover. A sequential-move game can be analyzed with a game tree.10 A game tree is a joint decision tree for the players and is composed of nodes and branches like a decision tree. Each node represents a decision point for one of the players. Each branch represents a possible action for a player at that point.
An Illustration Consider a common problem of a prospective entrepreneur. Ethan would like to quit his bartending job and open his own establishment, Ethan’s Bar. His options are (1) to enter the market with a large bar, (2) to enter with a small bar, and (3) to wait to see if the town’s economy will support another bar. He has estimated the various costs and revenues associated with the two establishment sizes. His biggest concern is a rumor that a national chain, Naomi’s Pub, is considering opening a bar in the same town. Naomi has two options: (1) enter the market or (2) do not enter. Ethan’s decision depends on how Naomi might respond to Ethan’s decision. The decisions are illustrated as a game tree in Figure 5.6. The tree represents Ethan’s perspective; both players’ payoffs, shown in the two columns on the right, are expressed in terms of NPV. By acting quickly, Ethan can make the first move. Thus, the first decision node (working left to right) belongs to Ethan. The middle 3 decision nodes belong to Naomi. If Ethan chooses largescale entry, then Naomi’s Pub would make a loss of $100,000 if Naomi elected to enter and $0 if Naomi elected to stay out. In this case, the rational choice for Naomi is to stay out. As with decision trees, we can ignore (or prune) the branches that represent inferior choices for the decision makers. Thus, we should assume that if Ethan opens a large bar, Naomi will choose not to enter. If Ethan chooses small-scale entry, then Naomi’s Pub would earn a $200,000 NPV if Naomi were to enter and $0 if she decided to stay out. The rational decision for Naomi in this case is to enter, and we can prune the decision tree accordingly. Finally, if Ethan decides to wait for the market to develop further, Naomi must decide between entering and staying out without knowing what Ethan is going to do. If Naomi enters, she would earn $100,000, $210,000, or $300,000, depending on Ethan’s decision. If Ethan has decided to wait, then clearly Naomi would prefer to enter because all the choices associated with an entry strategy generate positive payoffs for Naomi’s Pub. Hence, Ethan can prune the bottom branch of the game tree, “Naomi stays out.” If Ethan decides to wait, he knows
New Venture Strategy and Real Options 195
Ethan’s Payoff Naomi enters Large bar
Naomi’s Payoff
$380,000
$380,000
$380,000 –$100,000
$425,000
$425,000
$425,000
$0
$250,000
$250,000
$250,000
$200,000
$400,000
$0
$300,000
$300,000
$100,000
$190,000
$190,000
$210,000
$0
$300,000
$370,000
$370,000
$0
$350,000
$350,000
$0
$0
$0
Logic $425,000 Naomi stays out
Ethan’s bar
Decision $425,000 Naomi enters Small bar
Logic $250,000 Naomi stays out $400,000
$400,000 Ethan enters with large
TRUE $300,000
Ethan enters with small
FALSE $190,000
Naomi enters
TRUE $0
Decision $300,000 Ethan stays out
FALSE $0
Wait
$0
Logic $300,000 Ethan enters with large
TRUE $370,000
Ethan enters with small
FALSE $350,000
Naomi stays out
FALSE $0
Decision $370,000 Ethan stays out
FALSE
0
$0
0
Fi g u r e 5 . 6 Entry decision game tree Prepared using PrecisionTree™, Palisade Corporation. In this sequential-move game, Ethan is the first mover. Ethan assumes Naomi will react rationally to Ethan’s investment decision. Ethan can select the choice that maximizes value for him in light of Naomi’s expected reaction.
Naomi will enter. Given that, the figure shows that Ethan’s best response is to open the large bar, which gives Ethan a $300,000 NPV and leaves Naomi with $100,000. We can ignore the other branches, “Ethan enters with small” and “Ethan stays out,” of this subtree. Examining the remaining branches, we can adopt the same technique we used with the decision tree: backward induction. Start at the terminal decision nodes that display the NPVs of the various alternatives for the two parties. Beginning at the top, if Ethan enters with a large bar, Naomi will stay out (avoiding the $100,000 loss) and Ethan’s payoff is $425,000. Moving down, if Ethan enters with a small bar, Naomi will also enter, making Ethan’s payoff $250,000. Finally, if Ethan
196
Chapter Five
waits, we know Naomi will enter. Ethan will then respond with a large bar, giving himself a $300,000 payoff. Since Ethan has the first move, and because he knows how Naomi will react, his optimal strategy is to enter immediately with a large bar.
Nash Equilibrium The preceding example is a noncooperative game. In a noncooperative game, the players cannot enter into binding, enforceable agreements with each other. Any solution of the game must be a “Nash equilibrium.”11 A Nash equilibrium is a set of strategies such that each player’s strategy is optimal given the strategy of the other player(s). The Nash equilibrium in our example is the pair of strategies such that (1) Ethan’s strategy maximizes his payoff, given the strategy of Naomi’s Pub, and (2) Naomi’s strategy maximizes her payoff, given Ethan’s strategy. In the entry-decision game, the Nash equilibrium is (Ethan: “Enter with large”; Naomi: “Stay out”). That is, Ethan does not wait to see whether Naomi’s Pub opens; instead, he enters first at a scale that makes it unprofitable for Naomi to open. Both parties maximize their profits given the other player’s action. If each expects the other to choose its Nash equilibrium strategy, then both will. In other words, in equilibrium, expected and actual behaviors converge.
Games Entrepreneurs Play Strategic “games” include a large class of activities in which a decision maker takes into account the actions and reactions of others. Strategic games commonly played by entrepreneurs include the following: • The business plan. An entrepreneur must decide how much optimism to build into the projections that are included in the plan. Overoptimism can be dangerous. The investor may counter with a proposed deal that ties the entrepreneur’s return to achieving the projections. • Strategic partnering. An entrepreneur must decide whether to bring in a vertically integrated company as a distributor and strategic partner or risk the possibility that, if not invited to partner, the corporation will independently develop a competing product. • Control. An entrepreneur must decide how much control to forsake in exchange for securing funding. If the entrepreneur is unwilling to give up control, a prospective investor might decide to forgo the opportunity. • Contract negotiation. New venture contracting with investors is not a onetime event. Rather, it can involve several rounds with varying levels
New Venture Strategy and Real Options 197
of information and changing values. The dynamic nature of contract negotiation makes the process game theoretic. • Information disclosure. A new venture’s management must decide whether to patent an idea now and risk copycat entry of rivals or maintain the idea as a trade secret.
Strategic Flexibility Versus Strategic Commitment Game theory forces the entrepreneur to think about a venture from the perspective of competitors, customers, suppliers, and investors. Decision trees and game trees are useful for assessing trade-offs between the value of maintaining flexibility (real options) and the value of committing to a more limited course of action. As in Figure 5.6, even though the waiting option has value, committing first to a large-scale bar may have an even greater NPV. In the figure, strategic commitment is the best choice, whereas in Figure 5.5, strategic flexibility (preserving the expansion option) was the best choice. It is important for the founder of any new venture (and the investor) to systematically consider the values of the important embedded real options that it may have. However, early commitment to a course of action can preempt or limit rival reactions in the best interest of the new venture.
5.11 Real Options with Continuous Distributions In the main example of this chapter, a retail facility, we limited uncertainty about demand to only three possible states of the world—high, moderate, and low demand. This, of course is artificial. In reality there are an infinite number of possibilities and the actual range is much broader than is reflected in the three that are examined. Artificially limiting the possibilities to a few discrete outcomes enabled us to use decision trees in a simple way to evaluate the strategic alternatives. As long as the strategic choices are limited to a few discrete alternatives, we can still use decision trees to evaluate real options that involve continuous distributions of random outcomes (such as about demand). However, doing so requires more comprehensive descriptions of the uncertainty and the use of simulation software to evaluate the strategic choices. That is, in order to evaluate the strategic alternatives, we need to simulate the payoffs over the full range of possible outcomes. The use of simulation to evaluate real options is the focus of Chapter 6. In this chapter, we use a few discrete scenarios to approximate the value we would
198
Chapter Five
find by simulating a continuous distribution of scenarios. Doing so in a reliable way requires careful consideration of the strategic choices, including the use of simulation to identify the boundary conditions under which an option should be exercised. To illustrate the use of discrete scenarios to evaluate real options involving continuous distributions, suppose we are considering a venture similar to the one described in Figure 5.5 and would like to evaluate the inclusion of expansion and abandonment options. At this point, suppose that based on an examination of similar ventures, instead of describing uncertainty about first-year demand as a single discrete outcome, we have determined that demand uncertainty can be approximated by a lognormal distribution with a mean of 1,000 units and standard deviation of 600 units. Based on the cost of expansion, we have also determined that if first-year demand turns out to be more than 1,500 units, it would be beneficial to expand. Conversely, if first-year demand is below 500 units, it would be beneficial to abandon the venture and liquidate the assets. Based on the parameters of the distribution, by simulating demand, we find that there is about a 15.7% probability that expansion will be warranted and a 16.5% probability that it would be best to abandon. This leaves 67.8% as the probability of continuing the venture without expanding. Now that we have estimates of the probabilities of the three choices, we can use conditional expected values to evaluate the project including accounting for the real options. To do this, we will need to estimate the expected demand associated each of the three strategies. Simulation is a simple way to estimate the average demand associated with each. Using the simulation results, we estimate that the average demand from trials resulting in demand greater than 1,500 units is 2,073.4 units, the average for those with first-year demand below 500 units is 382.9, and the average over the trials where the venture would be continued without expansion is 901.6 units. If we would like to value the project using a simple decision tree similar to the one in Figure 5.5, we would assign a 15.7% probability to expansion and an expected first-year unit sales volume of 2,073.4 units under expansion. Similarly, we would assign 16.5% as the probability of abandonment and an average firstyear unit sales volume of 382.9. The full analysis would require additional information, such as the contribution margin on each unit of sales, a model of how first-year sales relate to ultimate sales volume over time, and the level of fixed cost that is associated with each strategy. The main point here is that condensing the simulation data in this way makes it possible to evaluate the venture using a small number of discrete scenarios.
New Venture Strategy and Real Options 199
5.12 Summary Strategic decisions involve major commitments that limit the range of future actions. Comprehensive strategic planning involves product-market and organizational choices that are highly interrelated with financing choices. By considering all 3 simultaneously, the entrepreneur can be assured of identifying the alternative that yields the highest expected value. A new venture can be thought of as a portfolio of real options, including, among others, the options to delay investing, to expand the size of the investment, and to abandon the investment. Strategic planning is a process of identifying these real options and comparing the values of alternative combinations of real options. Decision tree analysis provides a framework for identifying and describing strategic alternatives and for identifying and managing the real options that are embedded in any new venture. The different branches of the tree describe the interplay between alternative strategic choices and uncertain states of the world. By starting at the ends of the branches, determining the highest-valued choice at each stage, and eliminating the other branches from further consideration, it is possible to identify the strategic alternative today that is expected to result in the highest overall expected value of the project. Game tree analysis is a useful extension of decision trees when the reactions of other parties to specific strategic choices are important. Game theory forces the entrepreneur to think about a decision from the perspectives of others who will be affected by it. In the context of a competitive game, it is important to consider the benefits of strategic flexibility relative to the benefits of strategic commitment. Review Questions
1. How are product-market, organizational, and financial strategies interdependent? Why, for new ventures, is it important to consider them simultaneously rather than sequentially? Give some examples. 2. What are the three aspects of a decision that make it strategic? 3. Why is it important for strategic planning to begin with a clear sense of the objective? 4. How might the objective for an entrepreneurial venture be different from that of a project pursued by a publicly held corporation? 5. How do embedded real options affect the values of investment opportunities? Why, when assessing investment alternatives, is it important to consider the embedded real options?
200
Chapter Five
6. What are three important differences between real options and financial options? 7. Describe an investment opportunity that includes at least one real option. On what does the value of the option depend? 8. Explain how you would construct a decision tree that includes the real option(s) and uncertainty. How would you use the tree to decide on a course of action? 9. How are game trees different from decision trees? For what kinds of decisions would you want to use game trees instead of decision trees? 10. How is the business plan related to strategic planning?
Notes
1. Background on the topic of strategy is provided by Porter (1998). 2. A more conventional example of these strategic interdependencies is the early growth of Ford Motor Company. Ford’s product-market strategy emphasized high-volume, low-product-price marketing. Its complementary organizational strategy was to use assembly lines to build automobiles almost entirely from parts supplied by others, and to sell through non-integrated franchisee dealers. By focusing on final assembly, Ford reduced its need for early-stage capital. It could take advantage of vendor financing (trade credit) for the parts it purchased. By selling to dealers for cash, it shifted the burden of carrying finished-goods inventory to the dealers. It is unlikely that Ford’s product-market strategy could have been achieved without the complementary financial and organizational strategies. 3. Moskowitz and Vissing-Jorgensen (2002) estimate returns to the private equity held by entrepreneurs and find a large public equity premium, and pose the following puzzle: why would entrepreneurs willingly invest in a single privately held firm with seemingly far worse returns than investing in the stock market? As noted in the text, qualitative considerations such as valuing being one’s own boss could explain this. In an update to their study, Kartashova (2014) finds that the public equity premium puzzle does not survive in some subsets of the data. The difference between private and public equity returns is positive between 1999 and 2007, whereas in the 2008–2010 period, the returns are very similar. 4. Hammond, Keeney, and Raiffa (1998) offer a survey of the pitfalls of decision making that is guided by intuition or misapplication of more systematic approaches. 5. Bhide (1994) notes that a variety of factors in addition to NPV can influence the suitability of an investment.
New Venture Strategy and Real Options 201
6. Amram and Kulatilaka (1999) illustrate the application of real options approaches to new venture investing and other decisions. See Chen, Kensinger, and Conover (1998); Childs, Ott, and Triantis (1998); and Loch and Bode-Greuel (2001) for specific applications. Dixit and Pindyck (1995) provide perspective on valuing investment decisions as options. Luehrman (1998a) discusses strategic decision making in terms of real options. 7. Luehrman (1998b) describes how simple investment opportunities can be valued as real options using financial option valuation methods. 8. These values are from the perspective of the entrepreneur and are not necessarily the same as values to an investor. Because the focus in this chapter is on learning to recognize and compare real options, we are abstracting from other aspects of the strategic analysis. 9. As an aspect of a theory of entrepreneurship, Baumol (1993) studies the optimal rate of innovation using the trade-off between delaying introduction of new products as a means of improving their quality and the risk that a competitor will enter first. 10. Simultaneous-move games are usually analyzed using payoff matrices that display the players’ outcomes for the simultaneous choices. 11. The Nash equilibrium concept is named in honor of John Nash, a Princeton mathematician who was a pioneer of game theory. References and Additional Reading
Amram, M., and N. Kulatilaka. 1999. Real Options: Managing Strategic Investment in an Uncertain World. Boston: Harvard Business School Press. Baumol, W. J. 1993. “Formal Entrepreneurship Theory in Economics: Existence and Bounds.” Journal of Business Venturing 8: 197–210. Besanko, D., D. Dranove, M. Shanley, and S. Schaefer. 2016. Economics of Strategy, 7th ed. New York: Wiley. Bhide, A. 1994. “How Entrepreneurs Craft Strategies That Work.” Harvard Business Review 72: 150–61. Chen, A. H., J. W. Kensinger, and J. A. Conover. 1998. “Valuing Flexible Manufacturing Facilities as Options.” Quarterly Review of Economics and Finance 38: 651–74. Childs, P. D., S. H. Ott, and A. J. Triantis. 1998. “Capital Budgeting for Interrelated Projects: A Real Options Approach.” Journal of Financial and Quantitative Analysis 33: 305–34. Dixit, A., and B. Nalebuff. 1993. Thinking Strategically. New York: Norton. Dixit, A., and R. Pindyck. 1995. “The Options Approach to Capital Investment.” Harvard Business Review 28 (4): 105–15.
202
Chapter Five
Dixit A., S. Skeath, and D. Reiley. 2015. Games of Strategy, 4th ed. New York: Norton. Ghemawat, P. E. 2009. Strategy and the Business Landscape, 3rd ed. Englewood Cliffs, NJ: Prentice-Hall. Hammond, J. S., R. L. Keeney, and H. Raiffa. 1998. “The Hidden Traps in Decision Making.” Harvard Business Review 76 (5): 47–58. Kartashova, Katya. 2014. “Private Equity Premium Puzzle Revisited.” American Economic Review 104: 3297–334. Kim, Y. J., and G. L. Sanders. 2002. “Strategic Actions in Information Technology Investment Based on Real Option Theory.” Decision Support Systems 33 (1): 1–11. Klein, B., and K. B. Leffler. 1981. “The Role of Market Forces in Assuring Contractual Performance.” Journal of Political Economy 89: 615–41. Kreps, D. M., and R. Wilson. 1982. “Reputation and Imperfect Information.” Journal of Economic Theory 27: 253–79. Loch, C. H., and K. Bode-Greuel. 2001. “Evaluating Growth Options as Sources of Value from Pharmaceutical Research Projects.” R&D Management 31 (2): 231–48. Luehrman, T. A. 1998a. “Investment Opportunities as Real Options: Getting Started with the Numbers.” Harvard Business Review 76 (4): 51–67. ———. 1998b. “Strategy as a Portfolio of Real Options.” Harvard Business Review 76 (5): 89–99. Magee, J. 1964. “How to Use Decision Trees in Capital Investment.” Harvard Business Review 42 (September–October): 79–96. Milgrom, P., and J. Roberts. 1992. Economics, Organization and Management. Englewood Cliffs, NJ: Prentice-Hall. Mintzberg, H. 1994. “The Fall and Rise of Strategic Planning.” Harvard Business Review 72 (1): 107–14. Moskowitz, T. J., and A. Vissing-Jorgensen. 2002. “The Returns to Entrepreneurial Investment: A Private Equity Premium Puzzle?” American Economic Review 92: 745–78. Norton, S. 1997. “Information and Competitive Advantage: The Rise of General Motors.” Journal of Law and Economics 40 (1): 245–60. Porter, M. 1998. Competitive Strategy. New York: Free Press. Rappaport, A. 1991. “Selecting Strategies That Create Shareholder Value.” In Strategy, Seeking and Securing Competitive Advantage, ed. C. A. Montgomery and M. E. Porter, 379–99. Cambridge, MA: Harvard University Press. Spulber, D. 1992. “Economic Analysis and Management Strategy: A Survey.” Journal of Economics and Management Strategy 1 (3): 535–74.
New Venture Strategy and Real Options 203
———. 1994. “Economic Analysis and Management Strategy: A Survey Continued.” Journal of Economics and Management Strategy 3 (2): 355–406. Trigeorgis, L. 1996. Real Options: Managerial Flexibility and Strategy in Resource Allocation. Cambridge, MA: MIT Press.
C h a p t e r S ix
D e ve lo pi n g Ve ntu r e Str ategy Us i n g Si m u l atio n
D ec i s i o n t r e e s a r e used to identify strategic alternatives and to exam-
ine the sensitivity of expected value to discrete changes in individual variables, one at a time. In Chapter 5, we simplified the analysis of strategic choice by assuming there were only a few possible states of the world. However, decision tree analysis based on discrete possibilities has limitations because the future is being described by a few discrete outcomes and related probabilities. To deal with the limitations of discrete scenarios, we now introduce simulation and demonstrate how it can be used to evaluate strategic choices that include real options. We begin with a few simple examples and then return to the main example from Chapter 5 and use simulation to conduct a more robust analysis. In later chapters, we illustrate the broad use of simulation to evaluate staging options for investments in new ventures and for valuation. A simulation model is a representation of the behavior of a complex system through the use of another system. For our purposes, the complex system is the future performance of a new venture. Simulation can take into consideration uncertainty about the environment, the venture itself, and possibly even the reactions of rivals. The normal way to represent uncertainty in a simulation model is to describe each key element of uncertainty as a statistical distribution, for example, a normal distribution with a given mean and standard deviation or a uniform distribution over a given range. To simulate the future of a venture, we must first build a spreadsheet model of the venture and identify the key decision variables. We can then introduce mathematical expressions that describe the important uncertainties that bear on value, cash needs, or other important factors. By using the computer to simulate the model, we can produce thousands 204
Developing Venture Strategy Using Simulation 205
of possible trials (outcomes), where each trial is based on making a random draw from each statistical distribution and computing the combined effect for the venture. For example, you might wish to project the level of net income for a venture at the end of 5 years, given four sources of uncertainty: the economy, the market’s reaction to the product, cost of production, and development lead time. If you simulate the future of the venture one time, the result is a prediction of net income in 5 years, given a specific outcome for each source of uncertainty. The prediction from a single trial is not likely to be very helpful. However, with a computer it is possible to run thousands of iterations of the model and to aggregate the data from the iterations. Not only does averaging the results improve the accuracy of the prediction, but the dispersion of outcomes serves as a measure of the aggregate effect of all the various sources of uncertainty that are built into the model. Moreover, the trial data can be analyzed for the purpose of refining key decisions such as how much cash to invest at the outset or to estimate the value of a particular real option. As a management tool, simulation has been around for a long time. David B. Hertz first advocated using simulation for investment decision making in 1968. The technique was slow to catch on for a variety of reasons. Among the early impediments were confusion about the correct way to apply simulation to investment decisions, lack of low-cost (fast) computational capacity, lack of data useful for calibrating uncertainty, and lack of user-friendly software. In spite of the difficulties, companies like Merck have been using simulation to analyze investment decisions for many years.1
6.1 Use of Simulation in Business Planning: An Example In the U.S., a new chemical with potential therapeutic value must undergo several stages of expensive clinical testing that can require many years to complete. Because of the nature and expense of the drug development process, simulation can be a valuable aid to decision making. Given this, it is not surprising that the pharmaceutical industry was one of the first commercial applications of simulation. The industry is complex and subject to uncertainty arising from many sources, including health care reform, the generic drug market, insurance practices, and tort litigation. Such uncertainties are particularly acute in developing, testing, and marketing new drugs. Investing in the development of new drugs has much in common with entrepreneurial investment in new ventures: uncertainty is very high, a new drug goes through several identifiable stages of development, and the firm faces many
206
Chapter Six
opportunities to abandon or modify its development efforts as it learns more about the drug over time. Even during the 1990s, rather than relying on single-point estimates of the future for allocating its R&D budget, Merck was developing simulation models that incorporated probability distributions for key variables. Under the direction of then CFO Judy Lewent, Merck developed its Research Planning Model, an approach that integrated principles of economics, finance, statistics, and computer science to produce quantitative analyses of the specific strategic decisions that Merck faced. Using simulation, the model synthesized probability distributions for key variables such as revenues, cash flow, and NPV. Much like the approach we will use, the output of Merck’s model was a frequency distribution that showed the probability that a project’s NPV would exceed a certain level. Merck could use the distribution information to compute summary statistics, such as standard deviation, which it could then use in other analyses, such as to price an option to delay investing. Merck’s Research Planning Model would simulate risk and return project-by-project (prior to commitment of funds). Then, by allocating the research budget across the projects, Merck could simulate the contribution of R&D to the financial performance of the entire corporation. Consider how the model can be used to evaluate a drug research project. Lewent describes it this way: We may know at the beginning of a project that there is a market for a specific treatment that includes many thousands of people, and once we reach a certain point in the process, we may know that a certain compound may be effective. But we still aren’t 100% certain that the compound will prove so safe and effective that it can be turned into a drug. So we have to ask ourselves, “Do we continue to invest?” Those are the kinds of decisions we face every day. And these aren’t investments that easily lend themselves to traditional financial analysis. Remember that we need to make huge investments now and may not see a profit for 10 to 15 years. In that kind of situation, a traditional analysis that factors in the time value of money may not fully capture the strategic value of an investment in research, because the positive cash flows are severely discounted when they are analyzed over a very long time frame. As a result, the volatility or risk isn’t properly valued. Option analysis, like the kind used to value stock options, provides a more flexible approach to valuation of our research investments . . . because it allows us to evaluate those investments at successive stages of a project.2 Merck would take a systematic approach to modeling the risks associated with R&D, manufacturing, and marketing. For example, in considering a new
Developing Venture Strategy Using Simulation 207
drug’s market potential, one of the constraints on development would be the expected time required for FDA approval. Merck could model the range of possible time frames for approval. By drawing from the range and simulating cash flows based on alternative assumptions, Merck could synthesize probability distributions for output variables. The output variables of interest are, of course, the standard measures of financial performance: revenue, cash flow, and NPV.3
6.2 Who Relies on Simulation? Each of the early impediments to using simulation to evaluate important strategic decisions either has been or is being removed. Appropriate ways to use simulation for making investment decisions have been developed, computational capacity is inexpensive and fast, appropriate software is inexpensive and user friendly, and data are increasingly available. Furthermore, the growing recognition of the value of viewing investments as portfolios of options points to greater reliance on simulation over time. Most major companies use at least one of the commonly available simulation packages in some parts of their business. Moreover, many universities have site licenses to standard simulation software. Several different software packages are available for running simulations on personal computers. Some are freestanding decision analysis programs. Others function as add-ins to Excel. Some employ random sampling to generate the trial. Others use Monte Carlo techniques in an effort to generate more accurate predictions of expected outcomes and distributions of outcomes using fewer iterations. As computational speed has increased, the cost- and time-saving rationales for Monte Carlo simulation have diminished in importance. Whenever we use simulation in this book, our modeling will use the @RISK® simulation software provided by Palisade. The software is similar to other Excelbased programs such as Crystal Ball®. When we show you the Excel syntax for a simulated cell in a spreadsheet, we use the @RISK syntax. If you are a user of the other most commonly used commercial package, Crystal Ball®, you can study the parallel syntax by opening the appropriate files on this book’s website.4
6.3 Simulation in New Venture Finance Nowhere is the case for using simulation more compelling than for decision making about new ventures. Consider the following examples of where simulation can lead to better decisions:
208
Chapter Six
• Strategy formulation. An entrepreneur is considering a risky opportunity to develop an enterprise and knows that building in options to abandon or change the nature of the venture can reduce risk and make the project more valuable. Simulation can be used to study the effects of different option structures on risk and value. • Deal structures. An entrepreneur and an investor are negotiating investment terms. The investor is willing to accept common stock but wants a large share of total equity in exchange. The entrepreneur would like to add sweeteners to the investor’s claims so that the investor’s ownership fraction can be reduced. Simulation can be used to evaluate the effects of alternative deal structures on the values of both parties’ claims. • Risk allocation. An entrepreneur and an investor have different tolerances for bearing the risk of a new venture to build snow shovels equipped with cardiac monitors. The entrepreneur is more risk averse than the investor. Simulation can be used to design a deal structure that shifts more to the investor, raising the overall value of the opportunity. • Contingent claims. An investor is not convinced by the financial projections of an entrepreneur who wants to produce and market golf balls equipped with location sensors. The entrepreneur is willing to accept financial claims that adjust the entrepreneur’s ownership share contingent on success. Simulation can be used to design a deal structure that is attractive to both parties. • Cash needs. An entrepreneur is trying to determine the total amount of financing that is needed for a prospective new venture to sell a subscription-based service. The service will upload a daily podcast to your phone that contains morning news content that is customized to your interests and commuting time. If the venture performs worse than expected, the need for financing will be greater. Simulation can be used to examine the relation between attained performance and total cash needs. • Staging investments. A VC wishes to invest in a project but would like to stage the investment so that progress can be evaluated at critical milestones. There is uncertainty about when the next milestone can be achieved and about the cost of achieving it. Simulation can help the VC decide on an amount to invest such that the probability of reaching the milestone is reasonable and the potential for overinvesting in a project that will never succeed is limited. • Valuation. An investor is trying to value an opportunity to participate in a venture and knows that the value of the investment depends not just on the expected return but also on the riskiness of the return. Simulation
Developing Venture Strategy Using Simulation 209
can be used to determine the expected return, the riskiness of the return, and the value of the investment. In this chapter, we apply simulation to the problem of designing a new venture strategy. Later in the book we apply simulation to cash needs assessment, valuation, and contract design.
6.4 Simulation: An Illustration Suppose, to capitalize on demand created by e-commerce, you are considering “hit-and-run” entry into a local home delivery service. That is, you expect that the venture would eventually be displaced by a much larger national or multinational networked delivery organization. Before proceeding with the venture, you have decided to do a simple NPV analysis, based on your projections of the required investment, expected revenue, fixed cost, and variable cost percentage. Using information you have been able to find online and your own judgment, you expect that the required investment in the venture will be $120,000. For the first year of operation, you expect that revenue will be $90,000, which you expect will grow at a rate of 5% per year, and that the venture will have a 50% variable cost percentage. Your projections are in real terms and you project that annual fixed cost will be $35,000. Revenue and all expenses are in cash so that net income and net cash flow are the same. Further, you estimate that your opportunity cost of capital is 10% per year. You expect that the venture could serve a niche market for 10 years, after which the venture would have no liquidation value. To assess the merits of the opportunity, you have constructed an Excel spreadsheet based on your expected projections for each item. The analysis is shown in Figure 6.1. To your dismay, although the computed cash flow is positive every year, the positive amounts, when present valued, are not enough to cover the expected investment. It appears that the NPV of the venture would be slightly negative.5 Before deciding to abandon the idea, it is worth recognizing that if you were to take account of the uncertainties surrounding the different expected projections you might reach a different conclusion. In the “Parameters of distributions” panel of Figure 6.2, we have added distribution information that reflects estimates of uncertainty about the key variables. The parameters appear in the columns on the right. The parameter values are estimates you have been able to develop from your research into the opportunity and your judgment. You have determined that investment cost, which has an expected value of $120,000 (from Figure 6.1) could be as low as $90,000 and as high as $150,000 and that
Known inputs Discount rate
Uncertain inputs Investment cost Year 1 revenue Annual fixed cost Annual revenue growth rate Annual variable cost percentage Financial projections Year Investment cost Revenue Fixed cost Variable cost Cash flow DCF calculations Outputs Discounted cash flows NPV
10% Expected $120,000 $90,000 $35,000 5.0% 50.0%
0 ($120,000)
1
2
3
4
$ 9 0 ,0 0 0 $ 3 5 ,0 0 0 $45,000 ($120,000) $10,000
$94,500 $35,000 $47,250 $12,250
$99,225 $104,186 $109,396 $114,865 $120,609 $126,639 $132,971 $139,620 $35,000 $35,000 $35,000 $35,000 $35,000 $35,000 $35,000 $35,000 $49,613 $52,093 $54,698 $57,433 $60,304 $63,320 $66,485 $69,810 $14,613 $17,093 $19,698 $22,433 $25,304 $28,320 $31,485 $34,810
($120,000) ($268)
$9,091
$10,124
$10,979
$11,675
5
$12,231
6
$12,663
7
$12,985
8
$13,211
9
$13,353
10
$13,421
Fi g u r e 6 .1 Venture NPV estimated based on the expected value of each input The figure shows the computed NPV of the venture that is determined by using the expected value of each item (investment, revenue, fixed cost, variable cost percentage, and growth rate) to project and value future cash flows.
Developing Venture Strategy Using Simulation 211
the most likely amount would be toward the lower end of the range. Because actual investment is projected to fall within this range and the distribution is not symmetrical, you have decided to model investment cost as a triangular distribution with the stated minimum and maximum and a most likely value of $100,000. Note that while the value in Figure 6.1 ($120,000) is halfway between the high and low, the mean (i.e., expected value in the statistical sense) is lower because of the skewness of the distribution. The first numerical column in the panel is the simulated outcome of one random draw from each distribution. The cell formulas in this column of cells contain the @RISK functions that generate the simulation results. If you launch @RISK and open the Excel file for Figure 6.2, you can see how the formulas are linked to the parameters, and by pressing the F9 (recalculate) key on your computer, you can generate new draws from the distribution. When you run the simulation, @RISK will make hundreds or thousands of draws from these distributions and will keep track of the results that you specify.6 Returning to the substance of the model, you also have concluded that the uncertainty of the Year 1 revenue projection is best modeled as a triangular distribution with a mode of $95,000 (somewhat above the mean in Figure 6.1), a minimum of $75,000, and a maximum of $105,000. Here again, the mean of this distribution is below the $90,000 value in Figure 6.1 because of skewness. You have concluded that annual fixed cost is best modeled as a symmetrical triangular distribution, with the mode (and mean) equal to the expected value of $35,000 as in Figure 6.1 and a range of $32,000 to $38,000. In @RISK, the general Excel formula for modeling a triangular distribution is RiskTriang(minimum, mode, maximum). So, for example, the distribution of investment cost can be coded as “=RiskTriang(90,000,100000,150000).” As in Excel, generally, the parameters in this function can be replaced with cell references that contain the values or with expressions that use numbers, cell references, or a combination. The best way to understand this is to launch @RISK and then open the Figure 6.2 spreadsheet in Excel.7 So far, we have only modeled revenue in Year 1. We have yet to address revenue growth and the variable cost percentage. In Figure 6.1, the assumed growth rate was constant at 5%. There was no uncertainty about the growth rate and no provision for the growth rate to be different in different years. Figure 6.2 deals with the first of these limitations by modeling uncertainty about the growth rate as a normal distribution with a mean of 5% and a standard deviation of 8%. Note that with such a large standard deviation compared to the mean, there is a pretty good chance that the realized growth rate in a trial will be negative. In modeling revenue growth, it is important to think about the underlying nature of the process. In Figure 6.2, in any given trial, the growth rate is
Known inputs Discount rate
10%
Uncertain inputs Investment cost Year 1 revenue Annual fixed cost Annual revenue growth rate Annual variable cost percentage Financial projections Year Investment cost Revenue Fixed cost Variable cost Cash flow
$118,543 $100,427 $33,195 1.5% 49.9%
0 ($118,543)
($118,543)
DCF calculations Outputs Discounted cash flow NPV
($118,543) $86,896
4.0
Values × 10–6
Parameters of distributions Distribution Parameter 1 Parameter 2 Parameter 3 Triangular 90000 100000 150000 Triangular 75000 95000 105000 Triangular 32000 35000 38000 Normal 5% 8% Normal 50% 2%
1
2
3
4
5
6
7
8
9
10
$100,427 $33,195 $50,132 $17,100
$101,896 $33,195 $50,865 $17,836
$103,386 $33,195 $51,609 $18,583
$104,899 $33,195 $52,364 $19,340
$106,433 $33,195 $53,130 $20,109
$107,990 $33,195 $53,907 $20,888
$109,570 $33,195 $54,696 $21,680
$111,173 $33,195 $55,496 $22,482
$112,799 $33,195 $56,308 $23,297
$114,450 $33,195 $57,132 $24,123
$17,100
$17,836
$18,583
$19,340
$20,109
$22,482
$23,297
$24,123
$20,888 $21,680
Summary Statistics from Simulation (10,000 trials)
NPV –0.125 0.265 5.0% 90.0%
5.0%
Statistics Minimum
($213,118)
Percentile 1.0%
($162,385)
Maximum
$1,048,138
2.5%
($142,617)
$34,207
5.0%
($124,502)
$124,912
10.0%
($100,446)
3.5
Mean
3.0
Std Dev
2.5
Variance
15603125230
20.0%
($67,364)
Skewness
1.277764548
25.0%
($54,037)
Kurtosis
6.134157776
50.0%
$12,170
$12,170
75.0%
$98,451
Mode
($40,561)
80.0%
$123,221
Left X
($124,502)
90.0%
$197,229
Left P
5%
95.0%
$265,279
Right X
$265,279
97.5%
$341,065
Right P
95%
99.0%
$426,457
2.0 1.5
Median
1.0 0.5
Values in millions ($)
1.20
1.00
0.80
0.60
0.40
0.20
–0.00
–0.20
–0.40
0.0
Fi g u r e 6 . 2 Venture NPV estimated by simulation using the distribution of each input The figure shows the parameters of each statistical distribution and the result of one random draw from each distribution. The financial model is constructed and the NPV is computed based on the random draws for this trial. The histogram at the bottom of the figure shows the distribution of NPV results from a simulation run of 10,000 trials. The panel to the right of the histogram shows summary statistics for the sampling distribution of NPVs.
Developing Venture Strategy Using Simulation 213
rojected to be constant over time. The growth rate can be higher or lower than p the expected rate of 5%, but in this model the rate will not change over time. For new ventures, it is worthwhile to think about how you might model a growth rate that starts out with a random draw, as this one does, but declines over time as the venture matures. Some durable goods have high initial growth rates, but as the market becomes saturated, the growth rate can decline and may become negative. Some products might have growth rates that are meanreverting, so, for example, the expected growth rate is 5% but each year there is a new draw from a distribution with a mean of 5%. These are just a few examples, but they can yield much different results. So it is important to think carefully about the nature of the process and how to represent it in a statistical distribution function. The final element of uncertainty in Figure 6.2 is the variable cost percentage. Here, again, we assume a normal distribution. We might get a low-cost trial as a percentage of revenue, or a high-cost trial, but as the model is constructed, there is no year-to-year variation in the variable cost margin. The “Financial projections” panel of Figure 6.2 has the same structure as in Figure 6.1. The only difference is that the randomly generated trials values are used instead of the fixed values from Figure 6.1. Figure 6.2 shows one randomly generated trial. The DCF calculations are also generated in the same way as in Figure 6.1. The reported NPV is the result from the random trial that is shown in the panels above. To assess the NPV of the opportunity, we generated 10,000 trials of the model and retained the NPV result from each trial. The histogram at the bottom of Figure 6.2 shows the results of the 10,000 trials graphically, and the chart to the right shows summary statistics from the sampling distribution. The first thing to note about the distribution is that it is positively skewed, meaning that it has a long right tail with some very high NPV results, whereas the negative side is more compact. We can see this graphically in the histogram, and the summary statistics indicate that the venture has a potential maximum NPV greater than $1 million, whereas the worst outcome is only about negative $200,000. More importantly, the sample mean is $34,207. This is quite a bit higher than the negative $268 that we found when we used the entrepreneur’s projected values of each input instead of the distribution information. Based on the simulation, the project has a positive NPV and should be pursued. So one benefit of simulation is that it helps us consider how uncertainty about the assumptions can affect the decision to invest. You might wonder if we should modify the NPV in light of the skewness and great uncertainty of the outcomes. After all, there seems to be a lot of risk; the most likely outcome is quite negative (the sample mode is negative $40,561 and
214
Chapter Six
the sample median of $12,170 is quite a bit lower than the mean). The answer, if you are confident that the discount rate is correct, is that the NPV you have calculated already takes account of these considerations so that the accept/ reject decision should be based on the statistical mean result. In this case, it is positive so the investment should be made.
6.5 Simulating the Value of a Financial Option Because real options are important to the value of a new venture and the values of the financial claims that make up the deal structure, it is instructive to first see how simulation can be used to value financial options, and then to consider under what conditions the approach is applicable to real options. To begin, consider how options can be used in financial markets. Suppose a share of stock currently sells for $118 and that calls and puts that expire in one year are available with an exercise price of $125. To keep the illustration simple, suppose that securities trade monthly. We assume, initially, that the expected return for investing in the stock is the risk-free rate of interest of 0.3% per month (3.66% per year) but that in any given month there is a 0.3 probability that the return is 4% lower and a 0.3 probability that it is 4% higher. Figure 6.3 shows the simulation model. The share price begins at $118 in cell B3. Then an @RISK function samples from the discrete probability-weighted returns and combines the resultant draw with the base return of 0.3% per month. The callout in the figure shows the syntax for cell C3, which generates the first monthly return and calculates the new share price. Similar calculations are made for the remaining 11 months, each looking back at the result for the prior month. In this illustration of a single trial, the stock price does better than expected, reaching $126.80 by year end for an annual return of 7.5%. The resulting ending value of a call option with an exercise price of $125 is $1.80 ($126.80 − $125). Because the stock price ends above $125, the put option is out of the money and is worth zero at expiration. Figure 6.4 shows the results of simulating this 12-month stock price model 20 times, starting from an initial price of $118. Each line in the figure is a random draw from the possible price paths of the share of stock.8 We used the simulation model to estimate the expected value of the stock at the end of 12 months and, more importantly, the standard deviation of annual returns for investing in the stock. Given that the risk-free rate is 3.66% per year, the true (theoretically correct) expected (mean) value of the stock at the end of one year is $122.32. When we simulated the value with 10,000 iterations of the model, the estimate of expected value (the sample average) turned out to be
Developing Venture Strategy Using Simulation 215 Fi g u r e 6 . 3 Simulation model of stock price and put and call option expiration values The figure shows the key inputs, assumptions, and cell formulas, along with the outcome of one random trial for a 12-period model of stock price performance. The resulting expiration-date values of puts and calls with an exercise price of $125 are shown, as well as their Month 0 PVs, which are found by discounting the payoff back at the riskfree rate.
Fi g u r e 6 . 4
$160
Simulation of 12 months of stock price performance
$150 $140 Price per share
The figure shows the results of 20 random sequences of returns for a stock that trades monthly. The Excel spreadsheet that generated these results also generates the values of puts and calls on the underlying stock and can be used to simulate the effects of different assumptions.
$130 $120 $110 $100 $90 $80
1
2
3
4
5
6
7
8
9
10
11
12
13
Month
$122.37, very close to the true expected value. The estimated standard deviation of ending stock prices from the simulation was $13.00, or 11.02% of the initial stock price, also very close to the theoretically correct value. A call option has value at expiration if the price of the underlying asset is above the exercise price. The value of the call, at expiration, is the excess of the stock price over the exercise price. In the simulation, the call was in the money 42.3% of the time. Over all 10,000 iterations, the average ending value of the call was $4.06. Discounting this by the risk-free rate of interest yields $3.91 as the estimated PV of the call option. Based on the simulation, this is the expected
216
Chapter Six
market value (the expected option premium at time 0) of the 12-month call with an exercise price of $125. A put option has value at expiration if the price of the underlying asset is below the exercise price. The value of the put, at expiration, is the excess of the exercise price over the price of the stock. In the simulation, the put was in the money 57.7% of the time, with an average ending value of $6.68 and a PV of $6.45. How do these simulated values compare to the values that can be derived from option theory using the Black-Scholes Option Pricing Model (OPM)? Using the true expected return and the estimated standard deviation in the OPM, the call is worth about $3.95 and the put is worth about $6.45. The simulated values are close to the theoretical values. The main reason the differences are more than trivial is that the distribution of possible stock prices in the simulation is not quite a normal distribution, which is the assumed distribution of the OPM. The normal distribution has a slightly lower standard deviation than the simulated distribution. If the OPM and simulation yield such similar values, why do we bother with simulation? Why not just use the OPM for valuing real options? In part, we do so because the decision tree approach that we use provides the information required for using a conditional discounted cash flow approach. In part, the answer is that it is often not realistically possible to identify an underlying asset (like a publicly traded stock) that is key to use of the OPM. Beyond that, the other assumptions of the OPM generally are not satisfied when real options are being valued. Simulation has advantages over theoretical option pricing because financial markets related to new ventures are incomplete and because many real options are interdependent. We examine these concerns more fully in Chapter 10.
6.6 Describing Risk We have already illustrated two simple ways to describe risk. In the discounted cash flow (DCF) example, we assumed that the probability distributions of some inputs were triangular and others were normal. In the stock option example, we modeled risk as a sequence over time of discrete outcomes where the probabilities of different outcomes vary. Clearly, this is not very realistic, and there are better ways to describe the risk of a one-month investment in a share of stock. Perhaps a better example would be VC. Based on historical averages, about 10% of a VC fund’s investments go public in an IPO, about 20% are successfully sold to an acquirer, and the rest are total losses or close to total losses.9 With a
Developing Venture Strategy Using Simulation 217
discrete distribution we could use these percentages and the typical number of investments made by a VC fund (approximately 20) to estimate the distribution of fund outcomes in terms of the percentages of IPOs, acquisitions, and failed investments. Figure 6.5 illustrates several other distributions that are often used in simulation models. Perhaps the most familiar way to characterize risk is as a normal distribution. Many kinds of uncertainty can be reasonably described as normal distributions. For example, a normal distribution could have been a better way to describe the evolution of stock prices over time. The discrete distribution information in Figure 6.3 yields a monthly standard deviation of 3.3% (not shown in the figure). Recall that we assumed a mean risk-free return of 0.3% monthly. We could have simulated the first month’s stock price by multiplying the initial price of $118 by a draw from a normal distribution with a mean of 1.003 (i.e., 0.3% expected return) and a standard deviation of 0.033. The @RISK syntax for this normal distribution is = RiskNormal(1.003, 0.033). Sometimes normal distributions can be problematic. In our model and in reality, a share price can never fall below zero. Similarly, sales revenue in a period can never be less than zero. Because normal distributions are unbounded, it is possible for a simulation of the share price or a revenue forecast to produce a negative value, especially after several months of repeated draws and if the periodic standard deviation is large. You can avoid this by adding a constraint that prevents the simulated share price from falling below zero; however, this may result in unintended biases in your results. A better alternative, for some processes, could be to choose a different distribution that does not yield negative prices or to simulate the returns from one period to the next, as we do in Figure 6.3 (doing so will not result in negative prices). Another solution to the problem of negative values is to use a lognormal distribution and apply it to a starting value that is positive. The second panel of Figure 6.5 shows the result of simulating a lognormal distribution with a mean of 0.1 and a standard deviation of 0.2. The mean is an expected growth rate, so a mean of 0.1 indicates that the price is expected to increase by 10%; the standard deviation of 0.2, or 20%, describes the uncertainty of the growth rate [entered in Excel as =RiskLognorm(0.1, 0.2)]. As is evident from the second panel of Figure 6.5, the lognormal distribution never yields a negative value. Thus, a lognormal distribution could work well for simulating a stock price and for many other risky processes where values can never be negative. A triangular distribution is a convenient way to prevent the occurrence of extreme outliers. By choosing the minimum, most likely, and maximum values to approximate what you believe is the true distribution, you can describe risk distributions that are more likely to yield low values [such as the one shown in
Chapter Six
Normal Distribution (10,000 trials, mean = 0, standard deviation = 1)
Lognormal (10,000 trials, mean = 0.1, standard deviation = 0.2)
700
800
600
700 600 500
400
Frequency
300 200
400 300 200
100
100
8.81 22.55
2.25
8.45 21.65
2.17
2.02
8.08 20.76
2.10
1.95
7.71 19.86
1.87
7.35 18.96
1.80
1.72
6.98
1.65
1.57
1.50
1.42
Exponential Distribution (10,000 trials, mean = 1, standard deviation = 1)
450
2000
400
1800
350
1600 1400
300 250
Frequency
200 150
1200 1000 800 600
100
400
50
200
Value
1400
1200
1200
1000
1000
Value
6.61
6.24
5.88
5.51
17.16
16.27
15.37
14.47
13.57
12.67
11.78
9.98
10.88
9.08
8.18
7.29
6.39
5.49
4.59
25.49
24.47
23.45
22.43
21.41
20.39
19.37
18.35
17.33
16.31
15.29
14.27
13.24
12.22
11.20
10.18
9.16
8.14
0
7.12
200
0
6.10
200
5.08
400
3.69
600
400
4.06
5.14
800
2.80
600
1.90
800
1.00
Frequency
1600
1400
3.04
4.77
Binomial Distribution (10,000 trials, 100 draws, success probability = 10%)
1600
2.02
4.41
Value
Poisson Distribution (10,000 trials, lambda (arrival rate) = 10)
1.00
4.04
3.67
3.31
2.94
2.57
2.20
1.84
1.47
1.10
0.00
0.73
0
19.41
18.60
17.80
16.99
16.19
15.38
14.58
13.77
12.97
12.16
11.36
10.55
9.75
8.94
8.13
7.33
6.52
5.72
4.91
4.11
3.30
2.50
1.69
0.89
0.08
0
0.37
Frequency
1.35
Value
Triangular Distribution (10,000 trials, min = 0, most likely = 5, max = 20)
Frequency
18.06
Value
1.27
1.20
1.12
1.05
0.97
0.90
0.82
0.75
0.45
0.60
0
3.38
3.09
2.79
2.49
2.20
1.90
1.60
1.31
1.01
0.71
0.41
0.12
–0.18
–0.48
–0.77
–1.07
–1.37
–1.66
–1.96
–2.26
–2.55
–2.85
–3.15
–3.45
–3.74
0
0.53
Frequency
500
0.67
218
Value
Fi g u r e 6 . 5 Illustrations of @Risk statistical distributions Examples of the results of simulating some of the statistical distributions available from the @Risk software.
the third panel of Figure 6.5, entered as =RiskTriang(0,5,20)] or high values. You can prevent negative draws with a triangular distribution by selecting the minimum to be zero or a positive value. An exponential distribution is appropriate when outcomes at one end of the distribution are very high but unlikely to occur and those at the other end are
Developing Venture Strategy Using Simulation 219
bounded (such as not to be less than zero) and are more likely to occur. Often, time-related uncertainty is described well as an exponential distribution. For example, the length of time a product might last before breaking can be described as an exponential distribution. The example in the fourth panel of Figure 6.5 has a mean of 1.0, such as an expected life of one year [entered as =RiskExpon(1)]. The last two distributions in the figure look similar to each other. The fifth panel of Figure 6.5 is a Poisson distribution with an expected arrival rate of 10 [entered as =RiskPoisson(10)] and the sixth panel is a binomial distribution of 100 draws with a success rate of 10% [entered as =RiskBinomial(100, 0.1)]. For large numbers of draws, the binomial distribution is approximately the same as the Poisson. In both cases, the outcomes are discrete nonnegative numbers. Poisson distributions are often used to simulate such things as how much inventory you need to have on hand if, on average, a certain number of customers per unit of time (such as 10 per day) each want to buy a unit of the product. You can see from the figure that, if your beginning inventory is 20 units, you will almost always have enough to fully supply the demand. Binomial distributions are convenient for yes/no kinds of processes. For example, if you know that the probability of innovating during a given period of time is 10%, you can simulate how long it might take to achieve a successful innovation. The menu of distribution functions is different in different packages, but with a little creativity, those we have discussed are sufficient to describe most risks that an entrepreneur may face.
6.7 Using Simulation to Evaluate a Strategy Now that we have seen some simple illustrations of simulation and some examples of distribution functions, let’s reexamine the retailing venture from Chapter 5. Our purpose is to see how simulation can improve the evaluation of strategic alternatives that include real options. Simulation involves six distinct steps: 1. Identify the strategies to be evaluated. 2. Establish the criteria for evaluating the alternatives. 3. Model the strategies to which simulation is applied. 4. Specify the assumptions and uncertainties that influence value. 5. Run the simulation. 6. Analyze the results.
220
Chapter Six
Step 1: Identifying Strategic Alternatives When simulation is used to evaluate strategic alternatives, the normal practice is to compare simulated results that are generated from different models of the venture, where each model incorporates a particular set of strategic choices. We might, for example, develop separate models of the Large Facility and the Small Facility and compare the simulated values of the two strategies to see which is better under what circumstances. Limiting the choices to a few discrete possibilities (e.g., a Large Facility or a small one) maintains the tractability of the analysis. Although a retail facility can be any size, it should be possible to answer the important questions by looking at only a few possibilities, such as large or small. The effects of including real options (like waiting, abandoning, and expanding) can be studied by making minor modifications to the basic models. Here, as we did in Chapter 5, we would like to consider: • The Large Facility, without and with an abandonment option • The Small Facility, without and with either the abandonment or the expansion option • Waiting to see market demand, then deciding on the response (a learning option)
Step 2: Choosing Evaluation Criteria The choice of evaluation criteria depends on the nature of the business and the focus of the simulation. For a public corporation, maximizing shareholder value is the most sensible overall objective. A marketing department that does not have responsibility for pricing might focus more narrowly on market share. For an entrepreneur (or outside investor), it makes sense to focus on maximizing the NPV of the entrepreneur’s (or investor’s) interest in the venture. A simulation model must be designed to produce information relevant to the evaluation criteria. Accordingly, for the entrepreneur’s decisions, the simulation model must generate information about the NPVs of the various alternatives so the entrepreneur can use the results to make the best decision. Later in the book, we use simulation to generate information about the cash flows the entrepreneur or investor will receive and the related risk. We use this information to compute the PV of the cash flows in a separate step that does not require simulation. In this chapter, we set aside the extra level of complexity and focus on NPV directly.
Developing Venture Strategy Using Simulation 221
Step 3: Modeling the Problem We use the Large Facility to illustrate the design and use of a new venture simulation model. As before, we assume that the appropriate discount rate is already determined and reflected in the results so that the values of the different outcomes are PVs. For now, we also assume that prices and costs are expressed in such a way that it is unnecessary to deal explicitly with time value. The model must specify mathematically how the entrepreneur’s decisions contribute to the PV of cash flows. We first determine the PV of the facility as if it were owned entirely by the entrepreneur. We then adjust that value downward to reflect the fractional ownership interest retained by the entrepreneur. The PV of the facility can be stated in terms of present valued streams of cash flows: PV(Cash Flow) = PV[(Rev − Cash Exp – Depr) × (1 − Tax Rate)] + PV(Depr) Because the business will be privately held, the entrepreneur should be able to avoid most corporate taxes. Hence, we apply a corporate tax rate of zero to simplify the preceding expression to PV(Cash Flow) = PV(Rev) − PV(Cash Exp) To model the retailing venture, we need to specify the underlying determinants of revenues and cash expenses. Revenue is a function of price and unit sales. Unit sales are the lesser of the quantity demanded or facility capacity. The demand side of unit sales can be described as market size multiplied by the facility’s market share. On the supply side, we model unit sales to be limited by a capacity constraint based on facility size. If demand exceeds capacity, then the constrained quantity is what is sold. Otherwise, unit sales volume depends on market demand. Thus, PV(Rev) = PV(Unit Price × Unit Sales) Unit Sales = Lesser of Demand Quant or Capacity Demand Quant = Mkt Size × Potential Mkt Share Capacity = An assumed maximum value For simplicity, we aggregate market size and market share over the expected life of the venture. Unit sales is calculated based on market size and market share over the life of the venture. We model the PV of cash expenses in a similar fashion but include both a fixed and a variable component, where the variable component depends on unit variable cost and unit sales. Thus,
222
Chapter Six
PV(Cash Exp) = PV(Unit Cost × Unit Sales) + PV(Fixed Costs) The preceding structure determines the PV of the venture as if it were owned entirely by the entrepreneur. But the entrepreneur is willing to commit only part of the required capital; the balance must be raised from an outside source. To determine value to the entrepreneur, we need to know what fractional share of ownership the entrepreneur retains. This depends on how much the investor contributes and how much equity the investor receives for the contribution. The PV of the entrepreneur’s interest can be specified as the residual: PV(Entrep Interest) = PV(Cash Flow) − PV(Investor Interest) where all of the PVs are expressed from the perspective of the entrepreneur. The value of the investor’s interest can be expressed as PV(Investor Interest) = PV(Cash Flow) × (Total Invest − Entrep Invstmt) × Pct Equity per Dollar Invested Finally, the NPV of the entrepreneur’s investment is NPV(Entrep Interest) = PV(Entrep Interest − Entrep Invstmt) This completes the model of the entrepreneur’s interest. Clearly, a more complex model could be developed by specifying the determinants, in equation form, of some of the terms in the preceding equations; however, the returns from adding complexity diminish rapidly. Although it is useful to think about the complex relationships that drive success or failure, a parsimonious model that is focused on key relationships is likely to yield results that are just as useful.
Step 4: Specifying the Assumptions and Describing the Uncertainties For the simulation to work, each variable in the model must be specified as an assumed value, mathematical expression, or statistical process that will generate a value. No matter how carefully you model the venture, the result will only be as good as its assumptions. Assumptions should be based on data, experience, and/or careful reasoning. If the model is to be shared with outside parties, each assumption must be defensible. As forecasting is the subject of the next two chapters, we will defer discussion of the bases for our assumptions. Figure 6.6 shows our assumptions for the model of the Large Facility.10 These assumptions parallel the more simplified assumptions used in Chapter 5 for the decision tree analysis. For example, we allow the average price and cost of a transaction to be subject to uncertainty so that the actual average price can be different from the expected value. The expected price is $10 per transaction,
Developing Venture Strategy Using Simulation 223 Fi g u r e 6 . 6 Assumptions and statistical processes of the large-facility model
Variable
Assumption
PV unit price of a meal
Normal distribution (μ = $10, σ = $1)
PV unit cost of a meal
Normal distribution ( μ = $5, σ = $0.6)
Market size estimate (after first year)
Triangular distribution (6, 2.6, 1 million units)
Market size
Normal distribution (μ = estimate, σ = 100,000)
Market share estimate (after first year)
Normal distribution ( μ = 10%, σ = 1%)
Market share
Normal distribution ( μ = estimate, σ = 0.3%)
Capacity
500,000 meals
PV fixed costs
Normal distribution (μ = $500,000, σ = $50,000)
Total investment
Normal distribution (μ = $750,000, σ = $25,000)
Entrepreneur investment
$400,000
Percent equity per dollar invested
1% per $10,000 of outside investment
μ = mean or average, σ = standard deviation.
and we assume that the uncertainty can be characterized as a normal distribution with a standard deviation of $1. Similarly, the expected cost is $5, with a standard deviation of $0.60. For reasons that will be important later, we use a two-step process to determine market size. During the first year, the entrepreneur receives a preliminary estimate of the actual market size. To characterize market size in the simulation model, we use a triangular distribution with a maximum of 6.0 million transactions (over the life of the facility), a minimum of 1.0 million, and a most likely size of 2.6 million.11 The entrepreneur then receives an update of the market size estimate. Specifically, we assume that the actual size is learned and is equal to the initial estimate plus a random error. We assume the error to be normally distributed with a mean of zero and a standard deviation of 100,000 transactions. We determine market share using a similar two-step process. Demand equals the product of market size and market share. If simulated demand exceeds capacity, then sales volume is equal to capacity. Otherwise, demand determines unit sales volume. In the model for the Large Facility, capacity is 500,000 transactions over the life of the facility. We allow uncertainty about both the level of fixed costs and the size of the total investment that is required to construct the facility. Because the entrepreneur’s investment is $400,000, the amount of outside investment is uncertain. The investor receives 1.0% of the equity for each $10,000 of capital invested. You probably can think of other ways of setting up the model and may question some of our assumptions. For the model to be useful, it is important to give careful thought to the assumptions. If they are specified arbitrarily, no one will have confidence in the results. The entrepreneur can use a variety of information
224
Chapter Six
sources to improve the quality of assumptions about uncertainty. We review information sources relevant to new venture forecasting in subsequent chapters. In addition, breaking assumptions into finer components is sometimes useful. Doing so can allow you to substitute variables that are easier to estimate for those that are difficult to estimate directly. For example, it may be easier to estimate population growth of an area and per capita transactions than to estimate market size directly. You can then derive expected market size, together with the uncertainty of market size, as the product of the two underlying variables.
Step 5: Running the Simulation With the model complete and the assumptions specified, the simulation is ready to run. To illustrate the usefulness of simulation, we selected five variables in the model, which are retained by @RISK for each trial and stored in an Excel file: market size, unit sales, PV of the venture, the entrepreneur’s ownership share, and NPV to the entrepreneur. Figure 6.7 is based on the @RISK results and shows the simulation statistics for these five variables based on running 5,000 iterations of the model.12 For each iteration, the computer makes a random draw from each of the statistical distributions that describe the uncertainty in the model. Thus, simulation improves upon using discrete scenarios Panel A
Percentages Average
Median
Standard Deviation
Skewness
3,212,694 316,821 1,079,180 0.649 299,515
3,065,070 301,949 986,546 0.649 238,803
1,054,064 104,730 647,450 0.026 419,340
0.303 0.189 0.557 0.061 0.555
Output Market size Unit sales (life time) Total present value Entrepreneur's ownership share Entrepreneur's NPV
Panel B
Panel C
NPV to Entrepreneur (Estimated based on 5,000 trials) 100%
300
90%
250
2,415,842 235,401 577,137 0.630 (23,065)
50% 3,065,070 301,949 986,546 0.649 238,803
75% 4,045,972 401,659 1,498,539 0.666 567,702
Maximum 5,977,817 500,000 3,415,528 0.740 1,908,014
Distribution of Entrepreneur’s NPV (Based on 5,000 trials)
200
70%
Frequency
Cumulative Percentage
80%
60% 50%
150 100
40%
50
30% 20%
0
10% 0%
($1,000,000) ($500,000)
908,109 78,866 (469,267) 0.562 (710,893)
25%
$0
$500,000 Value
$1,000,000
$1,500,000 $2,000,000
($710,893) ($603,999) ($497,104) ($390,210) ($283,316) ($176,422) ($69,528) $37,366 $144,261 $251,155 $358,049 $464,943 $571,837 $678,731 $785,625 $892,520 $999,414 $1,106,308 $1,213,202 $1,320,096 $1,426,990 $1,533,884 $1,640,779 $1,747,673 $1,854,567
1 2 3 4 5
Minimum
Value
Fi g u r e 6 .7 Unconditional simulation results Panel A shows the results generated from @Risk. Results are compiled using the model for the large facility. Panel B shows the cumulative distribution of the entrepreneur’s NPV. Panel C is a histogram showing the dispersion of the entrepreneur’s NPV estimates from the trials.
Developing Venture Strategy Using Simulation 225
for sensitivity analysis by allowing us to examine the impact of changing a number of variables at the same time. We begin by considering the total PV of the facility, shown in row 3 of Figure 6.7, Panel A. This is calculated for each trial as Total PV = PV(Cash Flow) = PV(Rev) − PV(Cash Exp) The average is $1,079,180, with a standard deviation of $647,450. The minimum value is negative $469,267. As all of the cash invested in the venture went into building the facility, nothing is left over to fund this shortfall. A loss of more than the initial investment is possible only if the entrepreneur makes subsequent investments or makes personal guarantees to investors or suppliers beyond the $400,000 investment. Otherwise, the loss would accrue to others who have provided resources to the venture before being paid, such as investors and creditors. The entrepreneur’s NPV is shown in row 5 of Figure 6.7, Panel A. The average value from 5,000 trials is $299,515, demonstrating that even as an accept/reject decision (i.e., before building in real options), the venture is worth pursuing. The summary table of the @RISK results shows wide variability in the entrepreneur’s NPV, including negative values for at least 25% of the trials. The cumulative distribution in Figure 6.7, Panel B shows about a 27% chance that the entrepreneur’s NPV will be negative. Figure 6.7, Panel C shows the full distribution as a histogram. The distribution is skewed toward the right, indicating the potential for some very high-valued outcomes, whereas the downside is more limited. Figure 6.7, Panel A shows that the entrepreneur’s expected ownership share is 65%, with a range of 56% to 74%. Thus, despite the uncertainty about the initial investment, the entrepreneur would always end up with a controlling (majority) interest. Recall that we use a two-step process to simulate market size. First, we use a triangular distribution to generate a preliminary estimate. Then we add a normally distributed random error (mean = 0, standard deviation = 100,000 transactions), which allows us to find the true size of the market. Figures 6.8, Panel A and Panel B are histograms that illustrate the net effect of the two-step process. The basic shape of the distribution is still triangular and not symmetrical, so that the peak is below the mean. In Panel A, we ran only 300 iterations of the model, and the resulting shape of the sample distribution is quite irregular. Panel B shows the result of running 5,000 iterations. The interplay of the triangular and normal distributions is clearer in this panel. As we discussed earlier, the realized number of transactions is the lesser of demand (market size × market share) or the capacity constraint of 500,000 transactions. Figure 6.9 shows the distribution of transactions and illustrates
Panel A 12%
Percentage of trials
10% 8% 6% 4% 2%
5,827,144
5,639,090
5,451,036
5,262,982
5,074,928
4,886,874
4,698,820
4,510,766
4,322,712
4,134,658
3,946,604
3,758,550
3,570,496
3,382,442
3,194,388
3,006,334
2,818,280
2,630,226
2,442,172
2,254,118
2,066,064
1,878,010
1,689,956
1,501,902
0%
Market size
Panel B 12%
Percentage of trials
10% 8% 6% 4% 2%
6,004,275
5,602,740
5,201,204
4,799,668
4,398,133
3,996,597
3,595,061
3,193,525
2,791,990
2,390,454
1,988,918
1,587,383
1,185,847
0%
Market size Fi g u r e 6 . 8 Distribution of market size estimates generated by simulation Panel A shows the results of 300 iterations of the simulation of market size. Panel B shows the effect of increasing the number of iterations to 5,000. In both panels, the solid vertical line represents the expected value from the sampling distribution.
Developing Venture Strategy Using Simulation 227 Fi g u r e 6 . 9
450
Histogram of unit sales simulation results
400 350 300 Frequency
The figure illustrates the effect of the capacity constraint at 500,000 transactions on total unit sales of the larger facility. Results are based on a simulation of 5,000 trials.
250 200 150 100
491,405
474,216
457,027
439,838
422,649
405,460
388,271
371,082
353,892
336,703
319,514
302,325
285,136
267,947
250,758
233,568
216,379
199,190
182,001
164,812
147,623
130,434
113,245
96,055
0
78,866
50
Transactions
Fi g u r e 6 .1 0
$450,000
Convergence of the entrepreneur’s average NPV
$400,000 $350,000 $300,000 $250,000 $200,000 $150,000 $100,000 $50,000 1 137 273 409 545 681 817 953 1089 1225 1361 1497 1633 1769 1905 2041 2177 2313 2449 2585 2721 2857 2993 3129 3265 3401 3537 3673 3809 3945 4081 4217 4353 4489 4625 4761 4897
The figure shows the rate of convergence of the entrepreneur’s NPV of the large facility, based on 5,000 trials. After about 500 iterations, the simulated estimate of NPV is quite stable, even though individual iterations are subject to considerable uncertainty.
Iterations
how capacity constrains total unit sales. In about 8% of the trials (about 400 of 5,000), the constraint is binding and only 500,000 transactions can occur. How many iterations of the model are needed to provide a reliable basis for an investment decision? One way to find out is to look at a graph of the rate of convergence of a variable of interest. Figure 6.10 shows the convergence of estimates
228
Chapter Six
of the entrepreneur’s NPV. Convergence is illustrated in the figure by plotting the average value of the variable for all the iterations up to a given number. The point on the far left reflects only the first trial from the model, while the point on the right reflects the average of all 5,000 trials. After about 500 iterations, the average entrepreneur’s NPV does not change very much. Thus, in this instance, even a fairly small number of trials yields a reliable estimate of projected NPV. If the entrepreneur’s choice is either to invest in the Large Facility now or to do nothing, then 500 iterations seem sufficient to determine that the venture is worth pursuing. But if the entrepreneur is trying to compare different alternatives, such as choosing between the Large Facility and the Small Facility, or is particularly interested in the tails of the distribution, more trials may be needed. Using the standard deviation and number of trials information from Figure 6.7, we compute that the standard error of the estimate for the NPV of the entrepreneur’s investment is about $5,900.13 This means there is about a 65% probability that the true mean NPV is between roughly $293,600 and $305,400 (i.e., $299,500 ± $5,900), and about a 95% probability that the true mean is in the range of $287,700 to $311,300 (i.e., $299,500 ± 2 × $5,900).14 Adding more trials reduces the standard error of the estimate of the mean, so that better decisions can be made even when the differences between values of alternative strategies are small. Had we limited the simulation to 500 trials, the standard error would have been about $18,600 instead of $5,900.
Step 6: Analyzing the Results The final step is to use the simulation results as a basis for making a decision. If the choice were simply between building the Large Facility and doing nothing, the positive NPV would be sufficient to conclude that the investment should be made. Most real decisions, however, are more complicated. They involve comparing several different alternatives. Many require drawing inferences about alternative scenarios that have not been formally analyzed (such as an intermediate-size facility). For such decisions, it may be necessary to develop several simulation models with alternative assumptions and to compare the results. This is the focus of the next section.
6.8 Valuing Real Options and Comparing Strategic Choices We turn now to our primary objective—use of simulation to compare strategic alternatives and to examine the values of real options. To do so, we use the same set of alternatives for establishing the IoT retailing venture as in Chapter
Developing Venture Strategy Using Simulation 229
5. The branches of the decision trees from that chapter represent alternative scenarios concerning the entrepreneur’s decisions. To recap, we consider the following possibilities: • Build a Large Facility immediately. • Build a Small Facility immediately. • Wait for more information on demand and build whichever size is best. • Build the Small Facility now and expand if demand is sufficient. • Build the Large Facility now and abandon if demand is insufficient. • Build the Small Facility now and abandon if demand is insufficient. To this list, we can add more complex alternative scenarios that combine the options to wait, expand, and/or abandon. We begin by comparing the Large Facility to the simple alternative of investing in the Small Facility. The Small Facility is modeled easily by making a few modifications to the Large-Facility model.15 Specifically, first we reduce the expected PV of fixed costs to $400,000 and the standard deviation to $40,000. Second, we reduce the capacity constraint to 260,000. Third, we reduce the expected cost of acquiring the facility to $600,000 and the standard deviation to $20,000. Simulating the Small Facility, we found that the expected NPV of the entrepreneur’s investment is $249,606. This is materially less than the $299,515 value of the Large Facility. Based on NPV, the entrepreneur should select the Large Facility. Anticipating the examination of real options to come, for the Small Facility the entrepreneur’s NPV is negative 17.3% of the time and the facility is capacity constrained 67.5% of the time. The comparable numbers for the Large Facility are 27.0% and 8.0%. Finally, the entrepreneur has a larger stake in the Small Facility, 80% on average versus 65% for the Large Facility. Are there any considerations that would shift the balance in favor of the Small Facility? One possibility is that the entrepreneur does not want to accept the downside risk of the large project. Although the initial investment is $400,000, in the event of a loss the entrepreneur may be compelled to draw on resources beyond those originally committed. For the Large Facility, the simulation showed that additional investment would be needed to cover operating losses in about 1.0% of the trials; for the Small Facility, this drops slightly, to about 0.7%. To gain additional insight about relative risk exposure, the standard deviations of the NPV of the two facilities can be compared. The Small Facility has a smaller standard deviation of the entrepreneur’s NPV. But, as its expected return is also lower, this may not make the small project any more appealing.
230
Chapter Six
A desire to minimize the downside would favor the Small Facility but is offset on the high end, where the Large Facility does substantially better. Financial theory implies that the choice should be made based on NPV (assuming you have correctly valued all of the cash flows and real options). So comparing the fractiles in the distributions is not defensible unless you think there is something that the NPV calculations do not take into account. Here is where qualitative considerations may influence the decision. For example, maybe the entrepreneur cares about the potential differences in control that are implied by the two models. The expected ownership share favors the Small Facility, as does the worst case of fractional ownership: 73% for the small versus 56% for the large. Conceivably, the difference in expected ownership share is enough to lead the entrepreneur to favor the Small Facility, despite its lower NPV. Offsetting the smaller share of ownership, the Large Facility generates more cash for the entrepreneur. Looking back at the fractiles, however, we see that the times when control is likely to be most important to the entrepreneur are the scenarios in which the venture does not do as well as expected. The Small Facility tends to have a higher NPV than the large one in these underperformance scenarios. Thus, the higher share and higher cash flow of the Small Facility in bad states of nature could be something the entrepreneur would want to consider as mitigating the NPV difference. How much is the Small Facility’s larger ownership stake worth to the entrepreneur? Simple NPV comparisons cannot directly address qualitative considerations such as these.
The Option to Abandon The analysis thus far has examined two strategic scenarios that do not incorporate any real options: a onetime investment in either a Large Facility or a Small Facility. No real-world venture is that simple. It is usually possible, for example, to abandon a venture if results are discouraging enough. How does the abandonment option change the values of the two facilities? In the case of the large one, we assume that the building has an alternative-use value of $600,000.16 In the terminology of finance, the owners of the facility have a put option with an exercise value of $600,000. As the simulation model is constructed, the exercise date is the date when true demand becomes known. The simulation model can be used to calculate the expected PV of continuing to operate the venture, conditional on the true state of demand. If this turns out to be less than $600,000, then an entrepreneur seeking to maximize the value of the investment will exercise the abandonment option.
Developing Venture Strategy Using Simulation 231
To estimate the value of the abandonment option, we run 5,000 iterations of the Large-Facility model, modified to include the option. For each trial we compare the venture’s PV to the abandonment value of $600,000. If the PV is lower, we exercise the option, convert to office space, and realize $600,000, which is shared between the entrepreneur and the investor based on their ownership fractions. The resulting estimate of NPV to the entrepreneur is $331,455. Comparing this to the earlier value of $299,515 for the Large Facility, it appears that the option is worth about $31,940 to the entrepreneur.17 When we evaluate the abandonment option of the Small Facility, with an assumed abandonment value of $300,000, we find that it is worth about $5,738 to the entrepreneur, increasing the entrepreneur’s expected NPV from $249,606 to $255,344. The Small Facility’s abandonment option has a relatively low value for two reasons: first, the probability of a state of nature being sufficiently bad to put the option in the money is lower; second, the alternative-use value of the Small Facility is low. Based on the simulation results, the abandonment option should not alter the initial decision to invest in the Large Facility. In fact, it reinforces the relative value of the Large Facility. However, investment in either size is more attractive when it includes an abandonment option. As the problem is structured, the abandonment option is costless to the entrepreneur. But what if it were not? Suppose some locations have high values as office space but others do not. Locations that afford valuable alternative uses are likely to sell for more because even real options are usually not costless to acquire. To see how much the entrepreneur should be willing to pay (in terms of a location premium) for the option to abandon, we would need to run the model again. The answer as to what the entrepreneur should be willing to pay is not obvious, since the entrepreneur’s contribution is capped at $400,000 regardless of the location choice. Thus, the investor would contribute the full location premium and yet would have to split the increase in the venture’s value with the entrepreneur based on fractional ownership. Also, because the option raises the expected value of the venture, shouldn’t the entrepreneur be able to convince the investor to take a smaller equity position per dollar of capital contributed? If your intuition does not lead quickly to the answers to these questions, then you should begin to recognize the value of simulation. The model can be used to value options that are more complex than the simple onetime option to abandon. In reality, the facility owner never knows demand with certainty. Each year is different from the one before. The abandonment option does not disappear just because it is not exercised at the end of the first year. With a more elaborate simulation model that explicitly covers several years, we could, in principle, estimate the value of a complex abandonment option that
232
Chapter Six
would give the entrepreneur the option to abandon at the end of each year.18 If the option is exercised, the process ends; if not, the option for that year expires, but options to abandon in the future continue to contribute positively to value. Let’s look in more detail at the values of the Small Facility and the Large Facility with options to abandon. Figure 6.11, Panel A plots the individual outcomes of 600 iterations of the model for the Small Facility. The horizontal axis represents the number of transactions over the life of the venture. You can see that the outcomes are dispersed around the upward-sloping line drawn in the figure and that there is a “floor” on the entrepreneur’s NPV at around negative $175,000. This floor reflects the downside protection against losses provided by the abandonment option. To show the option quality of the abandonment strategy more clearly, in Figure 6.11, Panel B we remove the uncertainty about prices and costs by using their expected values. The only random variable is the level of demand. In addition, we model demand so that all of the uncertainty is resolved after the first year. The pattern in the figure can be represented as a combination of three securities: (1) an underlying asset (the entrepreneur’s claim on the venture), (2) a put option to abandon the venture for $300,000 that is exercised if demand is low, and (3) a call option the entrepreneur has “sold” by not building a facility large enough to handle high demand. The upward-sloping portion of the value function in Figure 6.11, Panel B shows that by building the facility, the entrepreneur acquires a long position in the market demand for the venture’s products. The floor reflects the abandonment (put) option and the ceiling represents the call option that is implicit in the capacity constraint. Figure 6.11 shows that the entrepreneur is effectively hedged against low demand by acquiring an abandonment (put) option, which is valuable at demand levels below about 140,000 transactions. By building a facility that is too small to serve high levels of demand, the entrepreneur has effectively sold a call option on demand in excess of what the facility can serve. The implicit proceeds from the sale of the call option on high demand are reflected in the figure as a lower facility cost compared with the alternative of building a much larger facility that can meet the highest conceivable level of demand. To assess the differences between the Small Facility and the Large Facility, in Figure 6.12 we overlay the NPV functions based only on variations in demand. It may surprise you to see that when demand is low, the Large Facility is more valuable to the entrepreneur than the small one (when demand is less than about 175,000 total transactions). This occurs because the Large Facility requires $150,000 more in outside investment but increases abandonment value by $300,000.19 Although the investor pays the entire incremental cost, most of the increase in abandonment value accrues to the entrepreneur, who has a majority ownership
Developing Venture Strategy Using Simulation 233
Panel A shows the sampling distribution of the entrepreneur’s NPV from 600 iterations of the small-facility model with abandonment option. The effect of the option is reflected by the lower bound, or “floor,” of negative NPVs. Panel B shows the combined effects of capacity constraints and the abandonment option for the small facility, leaving out the other sources of uncertainty.
Panel A
Small-facility NPV to entrepreneur $1,400,000
$1,200,000 $1,000,000 NPV to entrepreneur
Small facility: NPV to entrepreneur
$800,000 $600,000 $400,000 $200,000 $0 –$200,000 –$400,000 Total demand (transactions)
Panel B
Small-facility NPV to entrepreneur
$1,000,000 $800,000
NPV to entrepreneur
Fi g u r e 6 .11
$600,000
Capacity $400,000 $200,000 $0 –$200,000 –$400,000 Total demand (transactions)
stake. Over the demand range from 175,000 to about 300,000 transactions, the Small Facility is more valuable for the entrepreneur. Generally, over that range the Small Facility can meet nearly all of the demand but at lower cost than the large one; this is due to its lower fixed costs. Beyond demand of 300,000 total transactions, the extra capacity of the Large Facility makes it more valuable than the small one.
Chapter Six Fi g u r e 6 .12
$1,000,000
Large facility overlaid with small: NPV to entrepreneur The figure shows the combined effects of capacity constraints and abandonment options for the large and small facilities over a range of market demand. The figure removes uncertainty by using the expected value for all variables except those related to demand quantity.
$800,000 Large facility NPV to entrepreneur
234
$600,000 Small facility
$400,000 $200,000 $0
0
100,000
200,000
300,000
400,000
500,000
600,000
–$200,000 Number of transactions
Figure 6.12 makes the choice appear simple. As long as we know true demand and can strip away the uncertainties about other factors such as prices and costs, it is obvious which of the two facilities should be built. Unfortunately, we cannot simply remove those uncertainties, nor can we ever be certain about the level of demand. We can, however, use simulation to help determine which of the two facilities has the higher expected NPV.
The Learning Option Another choice that is available to the entrepreneur is to wait to build either size facility until the preliminary estimate of market size is learned. We assume that delaying investment invites competitive entry, so that the expected market share for the entrepreneur’s venture would be reduced. Further, we assume that putting off the investment for even longer—until demand is known with certainty—would result in loss of the opportunity to invest. Because the entrepreneur invests only $400,000 but can do so in a facility of either size, the entrepreneur has a complex call option involving mutually exclusive facility size choices on an uncertain share of the value of a venture. The option is a call on the expected value to the entrepreneur of either size venture, whichever value is greater. The exercise price is $400,000, and the option expires shortly after the preliminary estimate of demand is revealed. If the entrepreneur decides to invest, she also acquires an option to abandon the venture once the true level of demand becomes clear. She will exercise the abandonment (put) option if the expected value of the venture is less than the value of the facility in alternative use.
Developing Venture Strategy Using Simulation 235
The learning option is not costless. Delaying entry reduces the PV of future cash inflows. Moreover, the delay increases the chance that competitors will enter and reduce expected market share. To reflect these costs, we assume that potential demand is 10% less than if entry were not delayed. Figure 6.13 is the branch of the decision tree (i.e., the subtree) that the entrepreneur faces if she decides to wait and learn before investing. With simulation, we no longer describe the states of nature (demand) as a few discrete possibilities (low, moderate, high). Instead, as in the figure, we describe the decision rule the entrepreneur will use to respond to nature’s choice, whatever it may be. The three-pronged choice after an estimate of demand is received reflects a complex call option on either a Large Facility or a Small Facility. The binary choices, once true demand is learned, reflect the abandonment (put) options. To simulate this complex structure, we modified the new venture model. After the estimate of demand is received, the simulation uses the result to estimate the entrepreneur’s NPV for both the Large Facility and the Small Facility. If both NPVs are negative, the entrepreneur will not invest in a facility of either size; if both are positive, the entrepreneur selects whichever size has the higher expected NPV. Then the entrepreneur learns actual demand and recalculates PV. Based on the actual demand, she compares the PV of the selected facility size against the PV of the abandonment option and decides whether to continue or abandon.
Wait Large facility Continue Actual demand Abandon Expected demand Small facility Continue Actual demand Abandon Do not invest
Receive Estimate of Demand: Estimate NPV of entrepreneur’s interest. (1) If resulting expected NPV > $0 and > NPV of small facility, invest in large facility, (2) If resulting expected NPV > $0 and > NPV of large facility, invest in small facility, (3) If resulting expected NPV < $0, do not invest.
Learn Actual Demand and Cost: Calculate PV of entrepreneur’s interest. (1) Large facility, Continue if PV > $600,000, otherwise abandon. (2) Small facility, Continue if PV > $300,000, otherwise abandon.
Fi g u r e 6 .13 Subtree for venture learning option with simulated uncertainty Prepared using PrecisionTree™, Palisade Corporation. The entrepreneur’s decisions are represented by squares; information received by the entrepreneur is represented by circles.
236
Chapter Six
At the starting point, the entrepreneur cannot know which, if any, of the options should be exercised. The only decision is whether waiting is more valuable than the highest-NPV immediate alternative (i.e., more valuable than building the Large Facility with option to abandon). When we simulated the model represented by the decision tree in Figure 6.13, the resulting expected NPV for the entrepreneur was $306,409. Thus, the learning option’s expected value is $25,046 lower than the expected value of investing today in the Large Facility with an abandonment option (NPV = $331,455). Hence, in this case the learning option has no value—it is better to proceed now. It may seem counterintuitive that one strategy containing the same and in fact more options than another strategy could have a lower NPV. The main reason the option to wait does not add value relative to the Large Facility/abandon alternative is that waiting encourages competitive entry, which reduces the entrepreneur’s expected market share conditional on entering with the Large Facility. As mentioned, the expected loss of market share can be thought of as the cost of acquiring the option to delay investing. For the Large Facility, the cost of acquiring the option, measured in terms of the expected loss of future business, is more than the increase in value that results from learning more about actual demand. As it turns out, the abandonment option is not very valuable in conjunction with the option to delay. This is because most of the uncertainty about future sales is resolved when the initial estimate is received. Accordingly, the option to abandon is almost never exercised when the option to wait is employed. Generally, when combinations of real options are used, they involve mutually exclusive or interdependent choices so that their values are not additive. This is an important difference compared to financial options, where exercise decisions can generally be made independently of each other. Techniques such as simulation or other numerical evaluation methods are extremely valuable for assessing structures involving mutually exclusive choices. Because the values of individual nonindependent real options diminish as more choices are added, we do not need to employ overly complicated models or to completely describe and model the strategic alternatives. A parsimonious model that captures the main strategic choices can generate a reliable estimate of the expected value of a venture.
The Expansion Option Finally, if the Small Facility is built immediately, the entrepreneur acquires an option to expand in the event that realized demand justifies the Large Facility. Initially, the Small Facility is built, and afterward the entrepreneur receives an
Developing Venture Strategy Using Simulation 237
estimate of demand; based on the estimate, the entrepreneur decides whether to expand. The option to expand is a call option on additional capacity, with a $200,000 exercise price (the cost to increase capacity). As we have styled this example, the entrepreneur does not need to invest anything further to exercise the option; all of the $200,000 comes from the investor. However, the expansion does reduce the entrepreneur’s ownership share. This structure, where additional financing is provided by an investor in exchange for an ownership percentage that reduces the entrepreneur’s share, is typical of VC financing arrangements. Finally, the model is structured to reflect the realistic expectation that the cost of building the Large Facility in stages ($600,000 + $200,000) is higher than that of building it at once ($750,000). When demand is allowed to vary continuously and the initial estimate of demand is uncertain, the best course of action (Large Facility or Small Facility) is not clear. Simulation can help evaluate the range of expected demand levels over which exercising the option to expand would add value. Figure 6.14 shows the subtree facing the entrepreneur who initially invests in the Small Facility. The option to expand is evaluated by comparing the expected
Small facility Expand Continue Actual demand Abandon Branch #2 Do Not expand
0 Continue Actual demand
Receive Estimate of Demand: Learn actual cost and ownership share. Estimate present value of entrepreneur’s interest. (1) Expand if expected PV of entrepreneur’s share > critical value (2) Do not expand if expected PV of entrepreneur’s share < critical value
Learn Actual Demand: Calculate PV of entrepreneur’s interest.
Abandon
(1) Large facility: Continue if PV > $600,000, otherwise abandon. (2) Small facility: Continue if PV > $300,000, otherwise abandon.
Fi g u r e 6 .14 Subtree for venture expansion option with simulated uncertainty Prepared using PrecisionTree™, Palisade Corporation. The entrepreneur’s decisions are represented by squares; information received by the entrepreneur is represented by circles.
238
Chapter Six
PV of the expanded facility with the expected PV of the small one over a range of critical values for expected unit sales. Waiting to expand until more information about market demand is obtained reduces the risk of the outside investment that is required for expansion. Because the risk of investing at this point is lower, we assume that the investor receives relatively less equity compared to the earlier investment round: 1.0% of the equity for each $20,000 invested in expansion. Again, the entrepreneur also has an option to abandon once true demand becomes known. Our purpose at this point is to determine the value of a strategy of investing in a Small Facility and waiting to see the market response before a larger investment is made. When we evaluated the option to delay investing in Chapter 5, we set up the analysis to always select the choice with the highest expected value. This time we search for the best decision by examining different critical values for the decision to expand and comparing the values for this strategy to the alternative of investing immediately in the Large Facility. Using the simulation model, we evaluated options to expand at critical values of expected demand ranging from 200,000 units to 500,000 units, in 20,000-unit increments. Exercising the option to expand benefits the entrepreneur over the entire range when compared to keeping the facility small. However, the benefit to the entrepreneur from expanding when expected demand is at the low end of the range comes at the expense of the investor. This is because at low levels of sales it is actually better to abandon the venture. Bringing in an investor under such conditions effectively subsidizes part of the entrepreneur’s losses. An investor who recognizes this conflict is unlikely to enter into such an arrangement. Accordingly, we limit the option to expand to the range of expected demand levels where the investor derives a positive expected NPV from the project. With this constraint, the option can only be exercised if expected demand is at least 300,000 units. At this level, the option to expand increases the value of the entrepreneur’s position to about $432,000, almost $100,000 higher than building the Large Facility initially. Using the model, we found that expected NPV is roughly constant up to a critical value of about 340,000 transactions. Accordingly, we settled on a strategy of exercising the option only if expected demand exceeds 340,000 transactions. Why does the option to expand create so much value for the entrepreneur even though it increases the overall cost of getting a Large Facility? There are two reasons: First, the expansion option allows the entrepreneur to avoid the higher costs of having a Large Facility when demand turns out to be low. Second, the option reduces the uncertainty of the return on the second-stage investment for the outside investor, prompting the investor to accept a smaller equity stake in exchange for contributed capital.
Developing Venture Strategy Using Simulation 239
This is an important lesson and one that will be explored in greater detail later. By staging the needs for outside capital, the entrepreneur can offer the investor a safer bet and can retain a larger share of the venture as a result.
6.9 Summary The objective of strategic planning for a new venture is to develop a framework for maximizing value. While we examine strategic choices from the perspective of the entrepreneur, a similar analysis could be done from the perspective of the investor. By the value-maximization criterion, success depends on making good assessments of the risks and uncertainties and on developing a strategy that anticipates the need to adapt to new information as it arrives. An effective framework is one that helps the entrepreneur decide whether to undertake the venture, promotes effective negotiation with providers of financing, and encourages value-maximizing decisions. Simulation is a powerful tool for evaluating the critical decisions that an entrepreneur faces. It is especially valuable for new venture strategies, which typically involve numerous real options. It can add substantial value to the venture and to the entrepreneur’s ownership stake. There are six steps to implementing a simulation for strategic purposes. First, identify the important strategic alternatives, possibly by representing them in a decision tree. Second, decide on the criteria for evaluating the choices, such as the NPV of the entrepreneur’s investment. Third, for each strategy, develop a model that can be used to evaluate the various options facing the entrepreneur, and specify mathematically how the decisions of the entrepreneur contribute to the evaluation criteria. Fourth, specify the assumptions of the model and describe the uncertainties. Fifth, run the simulation. Sixth, interpret the results. The best way to appreciate the value of simulation is to work through specific examples.
Review Questions
1. What value did Merck see in using simulation in its R&D process? 2. What are some ways that simulation can foster better decision making for new ventures? In each case, explain how simulation could be beneficial. 3. What are the steps involved in using simulation to evaluate a strategy? Explain each using a specific example of your own or use the DCF
240
Chapter Six
illustration in Section 6.4 to identify the steps and explain how they are useful in evaluating the key decisions in your example. 4. What potential problems do you see with using normal distributions to project revenues or stock prices in simulation modeling? What are some ways of addressing these problems? 5. If you were trying to simulate the value of a call option, why would you first simulate the value of the underlying stock rather than simulating the option risk directly? 6. Give some examples from new venture development that fit the distributions depicted in Figure 6.5 (e.g., arrival probabilities on a website could be modeled as a binomial distribution). 7. Describe how you could use simulation to evaluate an expansion option. How could you determine the value of the option? 8. Describe how you could use simulation to evaluate an abandonment option. How could you determine the value of the option? 9. If you simulate distributions of NPVs for two different strategies, why would it usually be inappropriate to compare the distributions and choose the one that is less risky? 10. When evaluating risk, why is it better to simulate the distribution of possible NPVs rather than taking the expected value of each risky factor and calculating the expected NPV based on the expected values? Notes
1. See Nichols (1994). Merck has continued to use simulation in their R&D budget allocation and continues to post jobs with simulation analysis in the titles. Many Fortune 500 companies make extensive use of simulation packages such as @Risk and Oracle Crystal Ball. 2. Available at http://hbr.org/1994/01/scientific-management-at-merck-an -interview-with- cfo-judy-lewent/ar/1. 3. Drawn from Nichols (1994). 4. The @Risk introduction tutorial provides a number of simulation examples, including discounted cash flow, portfolio construction, and real options. 5. While Excel has a menu of financial functions for computing such things as NPV, use of these functions can obscure the logic of the calculations. Accordingly, in general, we rely on the fundamental mathematics of time value rather than the Excel functions. In Figure 6.1, we compute the NPV mathematically.
Developing Venture Strategy Using Simulation 241
6. If you have access to @Risk, we recommend that as you read the rest of this section, you open Excel, launch @Risk, and open the Figure 6.2 file that is available on the companion website. 7. Other Excel-based simulation software packages such as Oracle Crystal Ball have different syntax, but the general logic is the same. 8. Figure 6.4 is part of an Excel file (available on the companion website) that also contains the model used for the simulation and the summary output table from @Risk. The Figure 6.4 and Model tabs require @Risk to open correctly. The other worksheets in the file do not. 9. Based on a sample of companies in the U.S. that received VC funding between 1987 and 2008, Hall and Woodward (2010) find 2,015 (13.2%) IPO exits, 5,625 (37.0%) M&A, and 7,572 (49.8%) that either were confirmed to have failed or had no record of successful exit after 5 or more years. In more recent years, as VCs have moved more into angel-stage deals and IPO exits have declined relative to M&A, these percentages have declined. In PitchBook data from the 2017 PitchBook-NVCA Venture Monitor, the average annual number of companies receiving VC funding in 2006 through 2011 was 4,447. During the subsequent 5 years (2012 through 2016), the average annual number of IPOs was 78 (1.8% of the 4,447 average), and the average number of M&A exits was 845 (19.0%), suggesting that around 79.2% had not achieved successful exits. 10. The companion website contains a copy of the Excel simulation model of the venture incorporating the assumptions described here. Separate simulation models are constructed to reflect the various real option structures for the Large Facility and the Small Facility. All can be studied on the website. 11. It is up to the user to select and calibrate a distribution that provides an accurate representation of future uncertainty. We use a triangular distribution to describe market size partly because it is an easy way to provide for a high degree of uncertainty but avoid the possibility of negative simulated values. 12. The NPV estimate converges quickly to a stable value. We use a large number of trials because we also want good information on the rest of the distribution. In a model that includes real options that are likely to be exercised only infrequently, a large number of trials will give more reliable information on the options. 13. The standard error equals the standard deviation of the entrepreneur’s NPV (from Figure 6.7) divided by the square root of the number of iterations. 14. Confidence intervals are derived using the properties of a normal distribution: approximately 65% of the distribution is within one standard
242
Chapter Six
eviation of the mean and about 95% is within two standard deviations. A d more accurate approach to estimating confidence intervals would rely directly on the trials data. 15. You can open the Excel file on the companion website to review the modifications. 16. We do not allow for any uncertainty about this value, but with simulation we easily could do so. 17. Because the option changes the riskiness of the project, it could also affect the discount rate that is appropriate for valuing cash flows. At this early stage, we have not taken account of this in the simulation assumptions. However, conceptually we can think of the facility with the option as a two-asset portfolio consisting of the underlying Large Facility (that is valued at the original discount rate), and a put option (valued at the risk-free rate). 18. This is comparable to a model for new drug development, where the pharmaceutical company has the option to abandon the effort at numerous points in the clinical testing and FDA approval stages. 19. If the investor is astute, the terms of the deal will be different for the Large Facility than for the small one, in part because the entrepreneur gets more benefit from the abandonment option of the Large Facility.
References and Additional Reading
Bell, D., and A. Schleifer Jr. 1995. Decision Making Under Uncertainty. Cambridge, MA: Course Technology. Copeland, T., and V. Antikarov. 2003. Real Options, Revised Edition: A Practitioner’s Guide. New York: Texere. Hall, R. E., and S. E. Woodward. 2010. “The Burden of Nondiversifiable Risk of Entrepreneurship.” American Economic Review 100: 1163–94. Hertz, D. B. 1968. “Investment Policies That Pay Off.” Harvard Business Review 46 (1): 96–108. ———. 1979. “Risk Analysis in Capital Investment.” Harvard Business Review 57 (5): 169–81. Mun, J. 2016. Real Options Analysis: Tools and Techniques for Valuing Strategic Investments and Decisions with Integrated Risk Management and Advanced Quantitative Decision Analytics, 3rd ed. CreateSpace. Nichols, N. A. 1994. “Scientific Management at Merck: An Interview with CFO Judy Lewent.” Harvard Business Review 72 (1): 89–99. Stevenson, H. H., D. F. Muzyka, and J. A. Timmons. 1987. “Venture Capital in Transition: A Monte Carlo Simulation of Changes in Investment Patterns.” Journal of Business Venturing 2: 103–21.
C h a p t e r S e ven
R e ve n u e Fo r eca sti n g
F i n a n c i a l f o r ec a s t i n g i s a critical element of planning for a new
venture. The principal benefits of a good forecast include: • Forecasting provides a basis for estimating value so entrepreneurs and investors can make objective comparisons between pursuing the venture and other opportunities. • Forecasting is a disciplined way to evaluate how much cash the venture is likely to require and how much it might need if it develops more quickly or slowly than expected. • Forecasting helps entrepreneurs and investors compare strategic alternatives and select the one with the highest expected value. • Forecasting helps the entrepreneur and investors understand the strengths and weaknesses of the venture. • A forecast provides a benchmark against which to compare actual performance, thereby providing a means to test hypotheses about the venture and to provide early warning of potential problems. We use the first part of this chapter to introduce the basics of financial forecasting, beginning with forecasting revenue and extending to integrated financial statement forecasting. Revenue forecasting is important because revenue is a key driver of value and for determining how much to invest in the inputs that are essential for responding to demand as it develops. To develop the revenue forecast, we begin by specifying the assumptions that drive revenue and revenue growth—assumptions about such things as market size, market share, and price. Clear, explicit, and well-supported assumptions 245
246
Chapter Seven
make the forecast more than just an exercise and make it credible to investors. In this chapter, we present a variety of forecasting techniques and review information sources that can serve as foundation for key assumptions. We then use the forecasting methods to design and construct forward-looking pro forma financial statements that allow us to forecast venture performance in an integrated way. In subsequent chapters, we use the integrated pro forma financial statements as a basis for valuation and to assess the cash needs of the venture through time.
7.1 Principles of Financial Forecasting It is useful to begin with an overview of the principles that guide the forecasting process. • Build and support a schedule of assumptions. The assumptions may come from fundamental analysis of the opportunity, information about comparable firms, or expert judgment. These assumptions guide construction of a financial model of the venture. • Begin with a forecast of revenue. It is usually easiest to start with a revenue forecast that is developed in light of the planned scale, scope, and nature of the selling effort. Most other aspects of the model are linked to revenue. For example, the targeted level of sales or rate of sales growth can determine the investment in production capacity and the necessary level of inventory. It can be useful to develop an aggregate revenue projection by combining projections for appropriate segments of targeted customer/user groups (e.g., age, region, incomes, propensity to adopt new products) and to separately estimate the numbers and growth rate of each segment. • Decide whether to develop the forecast in real or nominal terms. Nominal forecasts should include an explicit forecast of inflation, whereas realterm forecasts are in constant dollars. If you expect selling prices and input costs to track the inflation rate, then forecasting in real terms can simplify the model. If prices and costs are unrelated to inflation, it can be better to forecast in nominal terms and make explicit adjustments for price changes. Interest rates on debt and discount rates used in DCF valuation are usually quoted in nominal terms. • Integrate the financial statements. Using formulas to integrate the pro forma balance sheet, income statement, and cash flow statement is essential for (1) testing sensitivity to assumptions, (2) performing scenario analysis, and (3) simulating the uncertainty of future performance.
Revenue Forecasting 247
• Choose an appropriate time span for the forecast. The span covered by the forecast depends on the purpose. If the forecast is to be used for valuation, the period must be long enough to carry the venture to a point where harvesting opportunities are likely. If it is to be used to determine cash needs, it should extend to the point where the venture would be in a position to attract follow-on financing based on its track record. • Choose an appropriate forecasting interval. The appropriate interval depends on the planning period of the venture. For assessing financial needs of an early-stage venture, intervals of one year are too long. To arrange financing on a timely basis, the entrepreneur and investors must project cash needs over much shorter intervals. If the forecast is to be used for control, the important performance milestones are unlikely to be annual. On the other hand, daily or even weekly forecasts are unlikely to be of much value. Departures from projected results over very short intervals are largely random. Generally, for an early-stage venture an interval of about a month to a quarter provides a sensible balance of timeliness and reliability for use in cash needs assessment. • Assess the reasonableness of the model. Think through the relations among assumptions and line items within and across statements. Do they make sense? Are they internally consistent? Try a basic “what if” analysis to see if the results are internally consistent and conduct stress tests to make sure the model is robust to extreme outcomes.
7.2 Forecasting Revenue For valuation and to anticipate financing requirements, we need to link product-market performance to financing needs. The revenue forecast is the customary link. This is because financing supports the start-up investments that are necessary before revenue generation can begin. Once the venture is producing and selling a viable product, sales growth is the primary driver of financing needs. For any given forecast of future revenue, we can work back to estimate the free cash flows from operations that are expected to be available or the additional financing needed to support the growth.
Forecasting the Revenue of an Established Business A reliable revenue forecast for an established business can sometimes be based on prior experience. For example, we might project that revenue will grow at the average rate of the previous five years. A more sophisticated approach
248
Chapter Seven
could take into consideration the trend in the rate of revenue growth (as opposed to the average), or the forecast could be tied to changes in underlying economic and demographic factors. One simple approach to forecasting revenue is to extrapolate the average historical nominal growth rate. A possible way to improve the revenue forecast may be to make it in real (inflation-adjusted) rather than nominal terms. Another way to improve forecast accuracy can be to weight historical data so that the more recent experience receives greater weight. Weighting is based on the idea that the future will probably be more like the recent past than the more distant past. With enough historical data and appropriate software, one could determine the weights that yield the (statistically) best estimates of sales in later years. Exponential smoothing is a weighting scheme that is easy to apply and can work well when historical data are limited. The simplest form of the exponential smoothing model is ForecastT+1 = α × ActualT + (1 − α) × ForecastT (7.1) where α is a weighting factor between zero and one. For example, suppose that the revenue forecast for period T was 100 and the actual was 110. If α is set to 0.2, the forecast for T + 1 is 0.2 × 110 + 0.8 × 100 = 102 whereas, if α is set to 0.8, the T + 1 forecast is 0.8 × 110 + 0.2 × 100 = 108 Although the equation does not refer to periods before T, they are reflected in the forecast implicitly. Because the period T forecast is estimated in the same way as the T + 1 forecast, actual results of earlier periods are implicit in the T + 1 forecast, through the previous forecast. The term “exponential smoothing” signifies that when Eq. (7.1) is used, the weights applied to earlier results decrease exponentially. The relative importance of each year’s data to the forecast is determined by the weight factor, α. A high value of α is used if recent results are believed to be an important predictor of future results. A lower value increases the weight on older results. If α is high, the forecast adjusts quickly to new results. In the extreme, if α is equal to 1.0, then the next period’s forecast is equal to this period’s actual result. If α is low, the forecast adjusts gradually. The approaches to revenue forecasting that we have discussed, including exponential smoothing, are examples of “naïve forecasting” methods. They extrapolate existing trends without considering underlying economic forces that drive demand and revenue. More elaborate naïve models are available to deal
Revenue Forecasting 249
with factors like seasonality or with a growth rate that is expected to decline systematically over time. In some cases, it may be more reliable to forecast growth rate percentages instead of revenue levels. Beyond that, we could consider underlying economic forces that affect the level of sales or the growth rate. These forces might be macroeconomic variables, such as the growth rate of GDP, or demographic factors, such as the population growth rate or the average age of the population.1 Finally, they could be industry-specific factors, such as the revenue growth rate for the industry, emergence of new competitors, or product innovations. We caution, however, that it is not helpful to identify relationships between revenue and other factors that are themselves difficult to forecast.
Forecasting Revenue of a New Venture Developing a revenue forecast for a venture with no track record is more difficult, and the result is likely to be less certain. How can we forecast revenue for a product that does not yet exist, where the full scope of applications and customers is not yet known, and where actions and reactions of competitors are yet to be seen? Rather than allowing these concerns to overwhelm us, it is important to search for simplicity. We consider two approaches: (1) yardsticks and (2) fundamental analysis. Yardsticks. An approach that is sometimes useful is to identify reasonable “yardstick” companies for which public (and possibly nonpublic) data are available. A yardstick is a firm that is comparable to the venture on some dimensions that are important to the forecast. Actual comparability may not need to be very close. Depending on the kind of information we wish to forecast, it is not even necessary that a yardstick firm be producing the same product. Comparability can be based on factors such as the expected market, distribution channels, adoption rates, usage rates, repurchase rates, uniqueness of the product, or technology. Data availability is one advantage of the yardstick approach. Many small companies go public every year. In the process, they often supply a great deal of information about the period before the company was public. Companies that go public are ideal candidates for assessing revenue growth potential. However, revenue estimates based on public companies are by nature optimistic for a new venture. Public firms are success stories and have already survived for longer than the typical new venture. Moreover, issuing shares publicly suggests that the firms have grown rapidly enough that outside equity financing has become important for growth. By studying the experiences of yardstick companies,
250
Chapter Seven
the entrepreneur can obtain information on the various stages of new venture growth and their changing financing needs. Sometimes the IPO prospectus contains enough historical data to measure revenue growth over a number of years. In some cases, there is enough historical information to infer the length of the development period before the company was able to market a product successfully and to infer the costs involved in the process. Financial information from yardsticks has value beyond forecasting. The prospectus also contains information on how those companies met their financing needs before going public. Each company can serve as a case study, providing insight into the financing choices the entrepreneur faces. In addition, the companies’ financial statements can aid in formulating the assumptions the entrepreneur needs for projecting the financial statements of the new venture. IPO prospectuses are available from various sources, including the issuing company and the underwriter. The SEC maintains an electronic database of prospectuses and other corporate submissions on its EDGAR website.2 How can we use the sales information from yardsticks to generate a revenue forecast? The techniques are fundamentally the same as for an established business using its own historical revenue. For an existing business, its own historical experience is simply one yardstick that is likely to be particularly good. For a new venture, this convenient sales record is not available, but the task is the same—to use historical sales experience (in this case, of other firms) to generate a forecast.
A simple example. Suppose you are considering opening a coffee shop, Morebucks, that would be similar to, and compete with, other retail coffee shops. As a means of estimating revenue, as shown in Table 7.1, you have collected data from the SEC reports of several publicly held businesses that operate companyowned coffee shops. From the public filings, you have collected information on revenue per company-owned store. The data come from a variety of reports required by the SEC. Forms SB-2, SB-2A, and 10SB12G/A are original and amended securities registration forms that are required of publicly held small businesses; F-1 is a registration form used by foreign firms with securities that are traded on U.S. exchanges; S-1, S1A, and 10-12G are original and amended registration forms with publicly traded securities; and a 10-K is an annual report for a firm with publicly traded shares.3 From these reports, to minimize the effects of economies of scale and revenues from nonretail store operations, we use the earliest information where data on both revenue and number of shops are available.
Revenue Forecasting 251 Tab le 7.1 Yardstick company data
Company
Coffee People, Inc. David’s Tea, Inc Diedrich Coffee, Inc. Einstein Noah Restaurant Group Java Detour, Inc. Peabody’s Coffee, Inc. Peet’s Coffee & Tea, Inc. Starbucks, Inc. Tully’s Coffee Corp.
SEC Filing
SB-2A 9/27/1996 F-1 5/26/2015 S-1A 8/12/1996 S-1 5/18/2007 SB-2 5/2/2007 10SB12G/A 3/15/2000 S-1A 1/23/2001 10-K 9/29/1996 10-12G 7/27/1999
Data Fiscal Year
Number of Shops Owned
Revenue ($ Millions)
Revenue/Shop ($)
2016 Adjusted Revenue/Shop ($)
1993
6
5.466
911,000
1,596,100
2012
105
73.058
695,791
731,660
1995
7
7.591
1,084,429
1,707,812
2006
416
363.699
874,276
1,040,835
2006
14
6.318
451,282
537,256
1999
26
1.795
69,032
99,449
1995
22
33.252
1,511,455
2,380,313
1994
403
248.500
616,625
998,613
1999
59
20.207
342,495
493,404
Notes: The following companies reported substantial negative net income and/or operating losses in and around the data year: David’s Tea, Java Detour, Peabody’s Coffee, and Tully’s Coffee. Peabody’s was operating only kiosks and Tully’s was operating a mix of stores and kiosks.
Of these nine companies, Peabody’s and Tully’s are poor yardsticks for the proposed coffee shop since many or all of their venues are kiosks so that revenue per shop is low. Moreover, all of the companies that are unprofitable in and around the data measurement period all have revenue per shop that is materially below $1 million in 2016 dollars. The average and median revenues of profitable coffee shops in the table are both somewhat above $1.5 million. Based on the data in the table, it appears that revenue at this level would likely be sufficient for profitable operation. Based on the revenue-per-shop information for the yardstick companies, it seems unlikely that Morebucks, as a new coffee shop with a single store, could expect to do better than the median company. It seems reasonable to infer that $1.5 million would be an optimistic estimate of expected annual revenue for Morebucks when it becomes established as a single coffee shop. You should be able to improve on this estimate in several ways. For example, if you were to explore franchising opportunities with the companies shown in Table 7.1, you could ask the franchiser to provide information on average revenues from franchise outlets and on the range of performance for franchise outlets, as well as revenue by age of outlet. You could also contact other companies that
252
Chapter Seven
operate privately owned coffee shops and might be able to convince the owners to share some information with you. Many industries have trade associations that provide representative financial and operating data on member firms. Notice that the table does not include information about the growth rates of store revenue. This is because, except for the first year or two, there is little reason to expect that same-store revenues will grow at a rate very different from the inflation rate, or perhaps inflation plus population growth in the immediate market area. For this kind of retail venture, if you want to grow significantly, you need to add stores. The corporate reports to the SEC include data on growth in numbers of shops over time. Fundamental analysis. Fundamental analysis is an alternative to reliance on yardsticks. Approaches to fundamental analysis can vary. For a venture such as a coffee shop that is similar to others already in operation, the fundamental approach might be mainly empirical, such as observing the traffic at other coffee shops, analyzing their product offerings and pricing, and talking to customers. Consider Morebucks. Our yardstick estimate of $1.5 million of annual revenue depends on numerous factors. Total revenue is a function of days and hours of operation, customers per hour, and average transaction size—all factors driven by product mix, advertising, competition, and location. For example, suppose you have researched two different coffee shops—one in a neighborhood that is active from early morning until late in the evening, including weekends, and one in a location such that it primarily serves employees of nearby businesses that are open standard hours. Your estimates of revenue are as follow:
Comparable Type
Days per year
Business/Entertainment Center Business Only
360 300
Hours per day Customers per hour
18 12
40 45
Revenue per customer
Annual revenue
$6.50 $8.00
$1,684,800 $1,296,000
As the evidence indicates, revenue is sensitive to location. While the business and entertainment location generates significantly more revenue, it is also likely to incur higher rent and entails longer hours of operation, resulting in higher expenses. By comparing different locations and constructing a complete financial model for each, the entrepreneur can refine the location decision.
Demand and Supply Considerations Revenue estimates for a venture can be generated from either the demand side or the supply side. The demand-side approach assesses consumer willingness
Revenue Forecasting 253
and ability to buy the product, assuming that the venture has adequate capacity to supply all that is demanded. The approach begins with an estimate of the venture’s market share that depends on such demand-related factors as number of competitors, pricing, location, and intensity of marketing efforts. For unique products, initial market share is easy to estimate—100%—but the size of the market is difficult to judge, and the rate of market share erosion due to competitive entry depends on defensibility of the entrepreneur’s position. For a more traditional venture, the potential market size may be easy to estimate (published estimates may exist), but market share is more uncertain and the reactions of competitors may be important. In contrast, the supply-side approach seeks to determine how fast the venture can grow given managerial, financial, and other resource constraints. Possible supply-side constraints include limits on access to raw materials, financing, and technology; there also may be constraints on the ability to hire and train employees. In other words, even if demand increases rapidly, the venture’s growth rate may be limited on the supply side. Supply shortages are common in product markets where new models are regularly introduced. When Samsung introduced its Galaxy S6 Edge, for example, it was unable to keep up with demand partly because of the technical challenges of producing the phone’s curved screen. In 2013, Tesla was unable to keep up with demand for its recently introduced Model S because of a shortage of batteries. Commenting on the shortage, CEO Elon Musk stated, “We really are production-constrained, not demand-constrained,” and indicated that the supply shortage was likely to continue into 2014. Combining the supply- and demand-side approaches, a venture’s expected rate of growth is whichever is lower. Slow-growth scenarios normally are constrained by the limits of market demand, whereas rapid-growth scenarios normally are constrained by the organization’s ability to manage growth. One advantage of the yardstick approach is that it uses the actual experiences of other firms and therefore implicitly considers both supply- and demand-side factors. Whether a forecast is based on yardsticks, fundamental analysis, or a combination of the two, it is important that projections be realistic and credible. Fundamental analysis is subject to the greatest potential for wild speculation. Consider the sales forecast of a proposed fast-food chain, the Bunny Hutch: “Assume each person in an area with a population of 10 million eats one bunny burger one night per week, at $1.00 per bunny burger. That’s $10 million per week.”4 Projections made by individuals with established reputations and industry experience or by objective third parties are more likely to be realistic and credible than estimates made by an entrepreneur who is enamored of the idea in the first place. The best substitute for relying on independent expert
254
Chapter Seven
projections is to base the analysis on solid reasoning and well-supported and well-documented assumptions.
7.3 Estimating Uncertainty For a new or early-stage venture, efforts to forecast revenue and other results may seem to be of little value. After all, the probability that actual performance will turn out to be much like the forecast is quite low. Nonetheless, forecasting is probably more important for a business with an uncertain future than for one that has experienced steady growth. For a venture with an uncertain future, the forecast of expected performance is simply an anchor for estimating uncertainty. For many purposes, the forecast of uncertainty is far more important. For example, the financial needs of a venture depend heavily on uncertainty. Failure to allow for possibilities such as development delays, lower-than-expected profitability, or higher-than-expected demand can result in critical financing errors. In this section, we introduce some simple approaches for estimating uncertainty.
Assessing Risk Using Historical Data The approaches discussed previously for revenue forecasting can also be used to estimate uncertainty. One simple approach for an established company is to generate a baseline historical trend for a key variable, such as sales or the sales growth rate, and then estimate uncertainty as the historical standard deviation of differences between actual and expected values. For a venture without a track record, uncertainty is greater and more difficult to estimate. One approach is to base the estimate on the experiences of other companies that are similar in important respects. For a new venture, such as the proposed Morebucks coffee shop, we could use the dispersion of revenue across yardstick companies and over time for those companies to estimate the standard deviation. Another is to envision alternative realistic scenarios for the venture and develop projections consistent with each. An estimate of uncertainty can be developed by applying probabilities to the different scenarios.
Sensitivity Analysis In sensitivity analysis, we vary the assumptions of the model one at a time or a few at a time and observe the impact on the forecast. Used effectively, sensitivity analysis can clarify which parameters are most important in the forecast.
Revenue Forecasting 255
Once we identify the critical input parameters, we need to develop reasonable descriptions of their uncertainty. History—from the firm’s experience or the experiences of comparable firms—can provide guidance, but always with the caveat that the past may not be predictive of the future. Another shortcoming of sensitivity analysis is that varying individual model inputs over predetermined ranges ignores possible interdependencies among variables. Finally, the forecaster’s choice of maximum and minimum values for a given input may be subjective and result in a biased forecast.
Developing Alternative Scenarios Some of the limitations of sensitivity analysis can be overcome by considering specific scenarios. Scenarios allow several assumptions to be evaluated simultaneously and can incorporate correlations among variables. Prospectuses of public companies and other public data sometimes can help you make a reasonable forecast of a success scenario. The success scenario may be the one that is most important for estimating how much financing the venture will need. For other purposes, it is important to develop a more comprehensive forecast of uncertainty. Value, for example, depends on expected performance and uncertainty, not just on performance in a success scenario. How can scenarios be developed that reasonably represent the uncertainty of a new venture? For simple businesses like retail shops and restaurants, the realistic range of performance may be narrow. For those, it may be possible to rely on information from public sources, some of which we describe shortly. But how can scenarios be developed for a business with tremendous potential and great uncertainty? Consider, for example, the difficulties of forecasting the value of a speculative, but promising and potentially important, new medical treatment for cancer. The answer depends partly on expected future revenues and the uncertainty of those revenues. Defining a success scenario for the venture’s revenue forecast is not difficult. Public information is sufficient to accurately forecast the number of potential candidates for treatment. Using the techniques discussed earlier, we can forecast market share and sales volumes. Data on prevailing prices for proprietary treatments for other life-threatening ailments can be used to forecast revenues. However, it is hard to imagine a project that is subject to more uncertainty. How likely is it that the development efforts will succeed and the patent holder can sell the product for a price unconstrained by government intervention? Moreover, how many competing research projects are under way, and how likely are they to succeed? What effect would innovation by a competitor have on market share and selling price?
256
Chapter Seven
One way to come to terms with the uncertainty is to define a small number of realistic scenarios in addition to the success scenario, such as the following: • Development efforts are successful and the product faces weak competition from other successful efforts. • Successful development efforts are offset by strong competing products. • Development efforts are not successful and the project is abandoned. The entrepreneur’s challenge is to construct alternative scenarios with realistic assumptions about their effects on product price and quantity and realistic assessments of their relative probabilities. Defining a small number of realistic scenarios helps focus the research and reveals the kinds of information that will be most useful.
Incorporating Uncertainty with Simulation Simulation is the final technique we consider for incorporating uncertainty into a revenue forecast. We first identify the assumptions behind the forecast. Then we assign a probability distribution to each key assumption and estimate correlations among the variables. These assumptions can be developed using historical data, evidence drawn from other companies, and/or fundamental evaluation of the market and potential demand for the product. We already have seen that IPO prospectuses can contain much useful information. Reports from Wall Street equity analysts often include descriptions of target customers and estimates of market size, selling prices, and time to market. Agencies like the Food and Drug Administration (FDA) provide voluminous data on drug and device approvals, clinical trial timetables and results, and similar information. All such sources can be used to approximate probability distributions for key variables and to establish relationships and estimate correlations among variables.
7.4 Building a New Venture Revenue Forecast: An Illustration Consider a start-up medical device venture called NewCo. We have done the background research described earlier. Based on our research, we have generated the assumptions shown in Figure 7.1. We first use these assumptions to forecast NewCo’s expected revenue and then will introduce uncertainty into the forecast. In Chapter 8, we will build an integrated financial model based
Revenue Forecasting 257 Fi g u r e 7.1 NewCo revenue assumptions
1.
Development will require six quarters, during which period no sales will be made.
2.
Initial quarterly sales of 300 units at a price of $200 beginning at the start of Quarter 7.
3.
From Quarter 7, unit sales will grow 25% per quarter for three years (through Quarter 19) and then remain constant.
4.
The sales price will increase each quarter at the inflation rate.
5.
Inflation at 4% per year (modeled as 1.0% per quarter).
on this revenue forecast but extended to include a complete set of forecasted financial statements. Because NewCo is a new venture, we decided to use a forecasting interval of one quarter. Based on a (hypothetical) study of similar ventures, we assume that the first six quarters are expected to be required for product development and testing. The venture will initiate sales at the start of Quarter 7. Based on the study and the characteristics of the market for NewCo’s product, if development happens when anticipated, we expect initial quarterly sales of 300 units with a selling price of $200 per unit (total revenue in Quarter 7 of $60,000). Following initiation of sales in Quarter 7, we expect unit volume to grow at 25% per quarter for three years (through Quarter 19). After Quarter 19, unit sales will remain constant. In addition to growth in volume, revenue will also increase with price inflation. We estimate that inflationary price increases will average 1.0% per quarter. Note that in this example, by using unit volume (real growth) and inflationadjusted selling prices, we are forecasting in nominal terms. In Chapter 8, we will add expense and other assumptions, some that are fixed in nominal terms and others fixed in real terms. Thus, we cannot escape dealing with inflation in some way. Figure 7.2 shows the expected pattern of NewCo’s revenue for the 6-quarter development stage, initiation of sales in Quarter 7, 3 years of rapid growth, and 2 years of inflationary growth. The total forecast window is 26 quarters, or 6.5 years. As shown at the top of the figure, the revenue forecast in Quarter 7 is simply the 300-unit quantity times the $200 price, or $60,000 in total revenue. In each of the next 12 quarters, unit sales increases by 25% over the prior quarter and price increases by 1.0% over the prior quarter. Total revenue in each quarter is the product of price and quantity. The rapid-growth period ends in Quarter 19, after which the level of unit sales is constant but price continues to increase at
0
Quarte r Development Success Sales (units) Selling Price/unit Revenue Unit Growth per Quarter Inflation per Quarter
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
$0
$0
$0
$0
$0
$0
300 $200 $60,000
375 $202 $75,750 25% 1%
469 $204 $95,685 25% 1%
586 $206 $120,751 25% 1%
733 $208 $152,553 25% 1%
916 $210 $192,545 25% 1%
1145 $212 $243,088 25% 1%
1431 $214 $306,845 25% 1%
1789 $217 $387,446 25% 1%
2236 $219 $489,096 25% 1%
2795 $221 $617,484 25% 1%
3494 $223 $779,629 25% 1%
4368 $225 $984,394 25% 1%
4368 $228 $994,238 0% 1%
4368 $230 $1,004,181 0% 1%
4368 $232 $1,014,222 0% 1%
4368 $235 $1,024,365 0% 1%
4368 $237 $1,034,608 0% 1%
4368 $239 $1,044,954 0% 1%
4368 $242 $1,055,404 0% 1%
NewCo revenue forecast $1,200,000
Expected Revenue for the Quarter
$1,000,000
$800,000
$600,000
$400,000
$200,000
$0
1
2
3
4
5
6
7
8
9
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26
Quarter (0 = quarter of launch)
Fi g u r e 7. 2 NewCo revenue forecast The chart shows the revenue trajectory for NewCo using the “base case” assumptions in Figure 7.1. Quarter 0 is the date of initial investment; sales begin in Quarter 7 after development is completed. The rapid-growth period for sales runs for three years (through Quarter 18). After that, through Quarter 26, unit sales growth is zero and the subsequent increases in revenue are due to the impact of inflation on selling price.
Revenue Forecasting 259
the inflation rate. The bottom of the figure is a plot of quarterly revenue over all 26 quarters. The expected rapid-growth rate of unit sales means NewCo must be prepared to supply about 4,400 units per quarter by Quarter 19. To achieve this, the venture must quickly develop the capacity to manufacture, distribute, and support this level of sales.
7.5 Introducing Uncertainty to the Forecast: Continuing the Illustration The forecast of expected sales is only the beginning of the exercise. The forecast in Figure 7.2 is just one possible outcome from a distribution that includes phenomenal success, complete failure, and many possibilities in between. At the outset, it is unclear whether NewCo’s development efforts will be successful, or if they are, when success will be achieved. It is also not clear how successful the product will be, how rapidly demand will grow, or how quickly the growth rate will slow to a steady state. The financing needs of the venture can be sensitive to outcomes that are different from expected. Factors such as development timing, rate of sales growth, and duration of the rapid-growth period are difficult to predict with confidence but are likely to be among the most important determinants of financing need and value. They depend on unknowable factors such as competition, customer reaction, and future macroeconomic conditions. Understanding and planning for the range of possible outcomes is essential to good decision making about the venture. Some outcomes that will require more financing than in the expected scenario are still worth pursuing, whereas others are not. Effective financial planning requires awareness of the margins of performance where it would make sense to continue and those where it would make sense to abandon or modify the strategy. How can we incorporate uncertainty into the forecast and use the uncertainty results to assess financing needs and strategic choices? In this section, we examine ways of incorporating uncertainty into a revenue forecast. Subsequently, we extend the uncertainty forecast to the full financial model of the venture.
Sensitivity Analysis In sensitivity analysis, we vary key assumptions across a range of values. Impacts of uncertainty are examined by altering one variable at a time or a few
260
Chapter Seven
at the same time. The NewCo revenue model, though simple, raises a number of issues. The key assumptions include (1) initial unit sales, (2) initial selling price, (3) length of the development phase, (4) length of the rapid-growth phase, (5) rate of unit sales growth during rapid-growth, (6) rate of unit sales growth at maturity, and (7) rate of price inflation. To determine reasonable ranges to use in sensitivity analysis, we can use historical data, publicly available forecasts, data from comparable firms, and/ or informed judgment. For example, inflation estimates are available from numerous sources and often include not only averages but also medians and the range of forecasts.5 Suppose, based on such data, it appears that the inflation rate over the venture’s forecast period could range from 0.5% to 2.0% per quarter around our base estimate of 1.0%. The following table shows the impact of varying the rate of inflation on Quarter 26 revenue, the average revenue per quarter for Quarters 7 through 26, and the cumulative revenue for the entire 26-quarter forecast period.
Inflation scenario
Low Expected High
Annual inflation
Quarterly inflation
Revenue Quarter 26
Average revenue per quarter
Cumulative revenue
2.0% 4.0% 8.0%
0.5% 1.0% 2.0%
$960,435 $1,055,404 $1,272,670
$546,784 $583,862 $665,970
$10,935,688 $11,677,240 $13,319,394
The table holds growth in unit sales and all other assumptions constant at expected levels. By the end of the forecast period, even these modest changes in the inflation rate can result in Quarter 26 revenue that is about 9% lower or 21% higher than expected revenue. Such differences can be important if, for example, the venture is financed with debt at a fixed nominal interest rate. For growth in unit sales, another key variable, we might analyze adoption rates for similar products or study the sales histories of comparable firms. Sensitivity analysis can be used to assess the impact of a more modest but realistic growth rate during the rapid-growth phase, say 10% per quarter, or a more successful scenario, such as 40% growth per quarter. The following table shows the impact of varying the growth rate of quarterly unit sales while holding the other variables constant at their expected levels. Growth scenario
Low Expected High
Unit growth (quarterly)
10% 25% 40%
Quarter 26 units
941 4,368 16,999
Average units per quarter
697 2,567 8,887
Cumulative units Cumulative revenue
13,937 51,213 137,742
$3,126,823 $11,677,240 $40,851,396
Revenue Forecasting 261
Assuming that the selected ranges for inflation and unit sales both reasonably reflect the true uncertainty, revenue is much more sensitive to variation in unit sales growth than to variations in the inflation rate. Moreover, variations associated with inflation are important only to the extent that product price increases are different from inflation rates. In contrast, variations in revenue due to unit sales growth are real. If the growth rate is too slow, the venture can fail to achieve a level of sales that is sufficient for viability. Even if it does eventually begin to generate positive free cash flow, if growth is slow then the venture may need more financing to cover its operations until that point is reached. Unit sales growth is also important because it has implications for infrastructure investment. In the preceding table, we assume that the variations in unit sales growth rates are symmetrical around the expected growth rate (25% per quarter ± 15%). However, the resulting differences in unit sales per quarter are not symmetrical around the expected level. Under the low-growth assumption, ultimate unit sales is 78% below the expected number, whereas under the high-growth assumption, ultimate unit sales is 289% above the expected number. Given the asymmetry, even if the venture would not be profitable at the expected growth rate, the upside might be sufficiently attractive that it would make sense to pursue the opportunity. Many VC investments are like this—the expected outcome is not sufficient to justify continuing the venture, but some possible outcomes are so attractive that accepting the high probability of failure is warranted. Finally, we can allow both inflation and unit sales growth to vary at the same time and observe the impact on a particular variable, such as cumulative revenue. In the following table, differences in each column are due to the inflation rate assumption. Those in each row are due to the unit growth rate assumption. Cumulative revenue over forecast period Unit sales growth rate
10% Inflation rate
25%
40%
0.5%
$2,951,662
$10,935,688
$38,107,969
1.0%
$3,126,823
$11,677,240
$40,851,396
2.0%
$3,512,841
$13,319,394
$46,942,340
Sensitivity analysis is simple to implement and can provide valuable information, but it has some shortcomings. First, there is little guidance as to what constitutes an appropriate range for any given variable. Second, it is difficult to test sensitivity to two or more variables at once. Third, even in this simple example, we have not considered the effects of changing the development period,
262
Chapter Seven
the rapid-growth period, or growth during the harvest period. Finally, variables are often correlated, and sensitivity analysis does not readily accommodate the correlations.
Scenario Analysis Scenario analysis is one way to address the limitations of sensitivity analysis. The analyst develops scenarios that incorporate reasonable values for the important variables and recognize the relationships that may exist among them. For NewCo, consider the following two scenarios: • Scenario 1. Product development proceeds more quickly than expected. The venture’s sales start at 300 units in Quarter 5 rather than Quarter 7. The new product does very well in the market and NewCo is able to patent important aspects of the technology. This keeps competitors at bay and enables NewCo to increase the initial selling price to $220. Unit sales grow at 35% each quarter for two years and then 30% quarterly for one year. For the balance of the forecast period, quarterly unit sales are assumed to be constant so that revenue grows at the 1.0% per quarter inflation rate. • Scenario 2. Product development hits numerous roadblocks and a competitor beats NewCo to the market. When NewCo finally begins to sell (in Quarter 9), the market only supports a $180 price. Unit sales start at 300 per quarter and grow at 15% each quarter for six quarters and then 10% for one year before falling to zero. Expected inflation is 1.0% per quarter. The following table shows revenue and unit sales numbers for both scenarios: Revenue
Scenario 1 Scenario 2
Quarter 26
Quarterly average
$2,563,227 $217,438
$1,565,980 $139,769
Unit sales Cumulative
$31,319,590 $2,795,373
Quarter 26
9,454 1,020
Quarterly average
6,146 779
Cumulative
122,913 14,022
These scenarios yield widely disparate forecasts, which is not surprising given their very different assumptions. They also provide a rough picture of the uncertainty about NewCo’s future. Although scenario analysis overcomes some of the limitations of sensitivity analysis, it has its own shortcomings. Practical considerations limit the number of scenarios we can develop and ana-
Revenue Forecasting 263
lyze. Moreover, for scenario analysis to be valuable, we need accompanying assumptions about the probability of each scenario. Moreover, no matter how carefully we construct each scenario, the subsequent reality will inevitably differ considerably.
Simulation Simulation overcomes some of the difficulties of sensitivity and scenario analysis. As discussed in the previous chapter, simulation requires that uncertainty be described in terms of probabilities or statistical distributions. When the simulation model is run, a random draw is made from each distribution and the results are used to construct a “trial.” Simulation can be used to quickly generate thousands of trials based on the underlying uncertainty assumptions. In essence, each trial is a scenario, and because they are generated from statistical distributions, each trial is equally likely to occur. By analyzing the trials data, we can make inferences about the probabilities of good or bad outcomes and can evaluate choices related to the exercise of real and financial options. In the NewCo example, the most important risk factors are development timing, the length of the rapid-growth period, and the rate of growth during the
Fi g u r e 7. 3 NewCo revenue simulation assumptions
1.
The earliest that successful development can occur is Quarter 3. Starting in Quarter 4, the probability of development success is exponentially distributed with a mean of 6 quarters. However, if development is not completed within 15 quarters, then it is clear that successful development of a valuable product is no longer feasible.
2.
If development is successful, the rapid-growth stage is expected to end around Quarter 20, after which it is expected that unit sales growth will fall to zero. The uncertainty about when the rapid-growth stage will end is normally distributed with a mean of 20 and a standard deviation of 1.5 quarters.
3.
Sales begin the quarter when development is successful. The initial sales level is expected to be 300 units if development occurs by Quarter 6, and to decline by 10 units for each quarter that development is delayed after Quarter 6.
4.
The initial selling price is subject to uncertainty depending on the quality of the development result and competitive factors. This uncertainty is normally distributed with a mean of $200 and a standard deviation of $20. After the first quarter of sales, the selling price increases at the rate of inflation each quarter.
5.
During the rapid-growth period, quarterly unit sales growth is normally distributed with a mean of 25% and a standard deviation of 6%.
6.
Inflation is forecast to be 1.0% per quarter.
264
Chapter Seven
Fi g u r e 7. 4 Discrete probability approach to simulating development timing The figure shows the simulation structure and a randomly generated outcome.
rapid-growth period. Figure 7.3 shows how we have modified the static NewCo assumptions to reflect uncertainty and accommodate simulation. The first assumption in the figure captures uncertainty related to development timing and allows for the possibility that NewCo never succeeds in creating a valuable product. The second assumption captures uncertainty related to the length of the rapid-growth period. Assumptions 3 and 4 describe the starting unit sales and selling price in the quarter after development is successful. Assumption 5 describes the uncertainty of demand growth. There are many ways to build timing uncertainty into a simulation model. We could assign a specific probability to each quarter, so that the probabilities (including the probability of failure) sum to 100%; then, for each trial, we could make one draw from the distribution to determine whether and in which quarter development is successful. For example, suppose that the timing for an event of interest (completion of a prototype, launch of a website, etc.) for a venture could occur in any of three months (Month 1, 2, or 3) and that the respective probabilities are 20%, 40%, and 30%. Because these sum to 90%, there is a 10% chance of failure. Figure 7.4 is a screen shot that illustrates use of a discrete probability distribution in @RISK to simulate a success month.
Revenue Forecasting 265
In the NewCo example, we model a more complex timing structure. Because we expect that development will take at least three quarters and that the probability of success will be highest in Quarter 4 and taper off thereafter, we use the exponential distribution to model development success. To reflect the potential for increasing intensity of competition if development is delayed, we reduce the starting unit sales volume by 10 per quarter for each quarter after Quarter 6. We also assume that if success is not achieved by Quarter 15, then it is not worth continuing to pursue the venture. Failure may occur because a competitor beats us to the market or because of failure of our R&D efforts to achieve a breakthrough. Based on Assumptions 1 and 2 in Figure 7.3, the cumulative probability of successful development is about 50% by Quarter 7 (start-up in Quarter 8) and about 86.5% by Quarter 15 (the last quarter when development would be successful). This leaves a residual probability of failure of about 13.5%. The @RISK cell formula to generate the development quarter outcome for each trial is as follows: =INT(4+RiskExpon(6)) where 4 is the first quarter when sales could begin and 6 is the mean of the exponential distribution after that point. As discussed in Chapter 6, the function RiskExpon(μ) is the @RISK macro for drawing from an exponential distribution with a mean of μ, and INT is an Excel function that truncates the simulation result to an integer. For example, trial outcomes of 10.1 and 10.9 both return a value of Quarter 10. In a separate step, we test whether the result is greater than 15 and if it is, we assume that development fails. Figure 7.5 shows the @RISK histogram from a simulation of 100,000 trials of the first quarter of revenue generation (the quarter after development success). If revenue generation starts after Quarter 15 in the simulation, the effort is classified as a failure, and for convenience in the modeling, we reassign Quarter 27 as the start of revenue generation. Quarter 27 is beyond the time range of the simulation model so that effectively the 13.5% of observations that are shown in Quarter 27 represent unsuccessful development efforts. Keep in mind that successful development is only the first stage along the way to commercialization. Some of the success trials could still prove to be unprofitable. At this point, we assume that the venture has enough cash to fund the development effort for as long as needed. Later, we will change that assumption to allow for the possibility that the entrepreneur would need to raise additional capital to continue the venture.
266
Chapter Seven
86.5%
16%
13.5%
14%
Percentage of trials
12% 10% 8% 6% 4% 2% 0%
0
5
10
15
20
25
30
First quarter of revenue generation Fi g u r e 7. 5 NewCo simulated distribution of the start of revenue generation The figure shows the simulated distribution of development timing based on 100,000 trials of the revenue model. The simulation is based on an exponential distribution with a mean of six quarters after the initial three (during which development success is assumed not to be feasible). The model also reflects the assumption that revenue generation must begin within 15 quarters or the venture fails. Failure percentage is shown in Quarter 27.
Figure 7.6 incorporates the other revenue projection assumptions from Figure 7.3 and shows plots of randomly generated revenue forecasts from five simulated trials. The figure illustrates a wide range of possible paths for NewCo revenue. Trial 1, a moderately successful outcome, has development completion in Quarter 7, with revenue generation beginning in Quarter 8, but a relatively low rate of growth through Quarter 18. Trial 2, the most successful result in the figure, has development success at the same time but a higher growth rate, and a longer time before growth ends. Trial 3 has the earliest development success and a high rate of growth, but the rapid growth period ends early. Trial 4 has very late development success, low initial revenue, and a low rate of revenue growth. Even though the rapid growth period ends the latest of the trials shown, ultimate revenue is quite low. Trial 5 is a failure scenario where development never occurs. The results shown in Figure 7.6 represent only five of an infinite number of paths NewCo could take based on our starting variables and assumptions about probabilities and statistical distributions. The simulation results incorporate many more possibilities than either sensitivity or scenario analysis and align more closely with what a new venture’s path might look like. They also dramatically reinforce the reality that the future of any new venture is highly uncertain.
Revenue Forecasting 267 Fi g u r e 7. 6
$1,400,000
NewCo revenue forecast: Sample trial results
Revenue (Trial 2) Revenue (Trial 3)
$1,000,000 Quarterly revenue
The figure shows five iterations of the NewCo revenue forecasting model. The model incorporates uncertainty with simulation, using the assumptions outlined in Figure 7.3. Each trial shows when/ if development is successful, how rapidly revenue grows, and how long the rapid-growth period lasts. Quarter 0 is date of initial investment.
Revenue (Trial 1) $1,200,000
Revenue (Trial 4) Revenue (Trial 5)
$800,000
$600,000
$400,000
$200,000
$0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 Quarter
7.6 Calibrating the Development Timing Assumption: An Example When product development efforts are expensive and timing of development success is uncertain, the financial needs of a venture depend critically on development timing. A venture that runs out of cash before reaching a significant milestone in the development process (completion of a prototype, completion of preliminary testing, etc.) is likely to face great difficulty lining up additional financing. The difficulty arises because failure to achieve the milestone may suggest to investors that the entrepreneur’s projections are overly optimistic in other ways as well. While there is no general rule as to how long development of a new product takes, it is possible to estimate expected time and timing uncertainty. In some cases, engineering studies or regulatory reports are helpful, as are experiences of comparable ventures. Innovation in new drugs is an example of products with protracted, variable, and unpredictable development horizons, due, in part, to technological and regulatory hurdles. New drugs make a good example because the FDA keeps records of all new drug applications. Since passage of the 1992 Prescription Drug User Fee Act, development times for new drugs (the combined time for clinical trials and regulatory approval) have fallen significantly. This trend is valuable information for anyone who is attempting to forecast financial information for a new
268
Chapter Seven
pharmaceutical venture. However, the general trend reveals nothing about the uncertainty of the development process, which is critical for forecasting cash flow and assessing cash needs. The FDA, however, does report, on a case-by-case basis, information regarding the times required for completion of each stage of testing prior to approval. The breakdown of development timing sometimes includes long response times from the drug developer. The detailed data allow us to better understand the approval process and can help us estimate the expected development time and the variation, or standard deviation, of that estimate.6
7.7 Summary Selling a product or service is at the core of any company’s business model. Although estimating sales is especially challenging for a new venture, financial forecasting disciplines the way an entrepreneur thinks about the venture. Not only can forecasting be used to develop and test hypotheses about the opportunity and help determine cash needs and the timing of those needs, it can also affect the value of the venture. A forecast can be an important fund-raising tool if it convinces prospective investors of the merits of the project and provides some specific performance benchmarks. The starting point for any financial forecast is an estimate of revenue. The venture’s strategy for getting customers to purchase its goods or services is of critical importance for any business plan. If the forecast is for an existing business, we can use historical data as the basis for projections. Numerous statistical techniques are available to improve forecast accuracy using historical data. Tradeoffs of selection criteria include data availability, model complexity, and uncertainty. For a new venture, yardsticks and fundamental analysis can help overcome the lack of an historical track record. Yardsticks are comparable firms—either public or private—that have a product and strategy reasonably close to that of the venture. Data on yardsticks can come from IPO prospectuses, 10-K filings and annual reports, and other public sources. Even if good yardsticks are available, it is useful to develop a sales forecast based on fundamental analysis. With fundamental analysis, the entrepreneur collects data on customer demographics, macroeconomic factors, substitute products, and so on and then uses the data to develop a set of assumptions for forecasting revenue. The forecast model is most useful if it reflects a solid understanding of the product and industry and if the forecast links expectations for performance in the product market to implications for financial performance. The assumptions are most useful if they are defensible and backed by research and objective data.
Revenue Forecasting 269
Any financial forecast must reflect the uncertainty implicit in the key assumptions. Uncertainty can be analyzed using sensitivity analysis, scenarios, and simulation. With sensitivity analysis, we vary key assumptions over some reasonable range to see how revenue or some other financial metric is impacted. Sensitivity analysis is limited, as we can vary only one or two variables at a time and correlations between variables are not recognized. Scenario analysis involves developing several stories about how the venture might progress and then matching assumptions to each story. Using scenarios overcomes the limitations of sensitivity analysis, but we are still limited by the number of scenarios that can reasonably be evaluated and the need to assign probabilities to each outcome. With simulation, an infinite number of possible paths can be generated and analyzed. This method allows for the most realistic assessment of the venture’s future, but reliable results are dependent on the validity of the assumptions. The new venture revenue forecast forms the basis for forecasting the rest of the financial statements, with an ultimate goal of estimating the venture’s cash flows and valuing the opportunity.
Review Questions
1. In what ways can forecasting help entrepreneurs? 2. Why is forecasting revenue the logical starting point for preparing a new venture’s financial statements? 3. What factors are important in establishing the appropriate forecast period and forecasting interval? 4. How can naïve forecasting methods be used to develop a revenue forecast for an existing business where management believes that more recent historical years are more predictive of future performance? 5. What are some of the considerations when deciding whether to forecast in real or nominal terms? 6. What are the strengths and shortcomings of sensitivity analysis and of scenario analysis? 7. Why is simulation particularly suited to forecasting revenue of new ventures? 8. What techniques are available to introduce uncertainty into a revenue forecast? Describe the pros and cons of each approach. 9. How might you use publicly available information to calibrate an assumption about the development time needed to build a drone delivery service?
270
Chapter Seven
10. What can you learn from a prospectus? Are the data in an IPO prospectus relevant to a new venture just getting off the ground?
Notes
1. The Federal Reserve Economic Database (FRED) is a freely available source of U.S. and international macroeconomic data that can be useful for developing new venture projections. Time series data on macroeconomic variables are available at https://fred.stlouisfed.org/. 2. The EDGAR website can be accessed at http://sec.gov. Prospectuses of new issues normally are posted to the site within a few days of filing. 3. Company filings are available from the SEC’s Edgar database at https://w ww.sec.gov/edgar.shtml. 4. Paraphrased from Silver (1994), p. 20. 5. The Philadelphia Federal Reserve Bank conducts a quarterly survey of professional forecasters and provides mean and median economic estimates for key macroeconomic indicators but no detail on individual forecasts, at http://w ww.phil.frb.org/research-and- data/real-time- center/survey-of -professional-forecasters. 6. See https://www.fda.gov/downloads/AboutFDA/ReportsManualsForms/ Reports/UserFeeReports/PerformanceReports/UCM548128.pdf.
References and Additional Reading
Copeland, T., T. Koller, and J. Murrin. 2015. Valuation: Measuring and Managing the Value of Companies, 6th ed. New York: Wiley. Doll, D., and M. Heesen. 2009. “NVCA 4-Pillar Plan to Restore Liquidity in the U.S. Venture Capital Industry.” Presentation at the annual meeting of the National Venture Capital Association. http://w ww.slideshare.net/ NVCA/nvca- 4pillar-plan-to -restore -liquidity-in-the -us -venture - capital -industry-1360905. Drucker, P. F. 1985. Innovation and Entrepreneurship. New York: HarperCollins. Hitchner, James R. 2017. Financial Valuation: Applications and Models, 4th ed. New York: Wiley. Shane, S. A. 2008. The Illusions of Entrepreneurship. New Haven, CT: Yale University Press. Silver, D. A. 1994. The Venture Capital Sourcebook. Chicago: Probus.
C h a p t e r Ei g ht
Fi nan cial M o d e li n g
O n c e a r e v e n u e forecast is completed, it can be used as a baseline for
developing pro forma financial statements. Ultimately, we are concerned with forecasting cash flows from operations. Operating cash flow is a key factor in determining a venture’s financing needs. Cash flow available to investors is central to valuation. To project cash flows, we first forecast the income statement and balance sheet and then use those statements (and period-to-period changes in the statements) to develop the pro forma statement of cash flows. For a new venture, we confront the same issue here as we did in forecasting revenue: that of estimating income statement and balance sheet relationships for a company that has no operating history. Here again, it is useful to rely on data from comparable public and/or private companies. However, the yardsticks and dimensions of comparability that are most useful for projecting income statement and balance sheet relationships are not necessarily those that are most useful for projecting revenue. Our main objective in this chapter is to illustrate how to build and integrate financial statements into a model that can be used in cash needs assessment and valuation. There are two aspects to this; the first is developing the assumptions that drive the financial model, and the second is the process of building and integrating the financial statements, including the interactions among them. Mastering the chapter should enable you to build your own integrated model. If the financial statements in a model are properly integrated, the full impact of changing any assumption will be reflected automatically and consistently in all of the statements and all of the forecast periods. This makes it easier to perform sensitivity analysis and is necessary for incorporating uncertainty with simulation. 271
272
Chapter Eight
We begin with an overview of the three main financial statements and how they are linked. With this background, we develop the notion of statement integration, beginning with an overview of the cash conversion process. Next, we provide an example of how working capital policy choices act in an integrated fashion to affect cash needs. We then describe several standard sources of information and demonstrate their use for developing forecast assumptions. Finally, we work through the construction of two integrated financial models, the first for Morebucks and the second for NewCo.
8.1 An Overview of Financial Statements Pro forma financial statements used for business forecasting of a new venture can be simple, but they require enough detail to capture the important activities of the enterprise. They are focused on the assumptions and financial statement items most important to the venture. The model for a new gaming software venture might incorporate development time and cost of bringing a viable product to market. For a coffee retailer, cost of goods sold and inventory may be the most important factors to model carefully. The first step in building an integrated model is to identify the critical assumptions. This list determines the accounts and line items that need to be incorporated into the model. In an integrated model, changes to any assumption will immediately update all three financial statements—the income statement, balance sheet, and cash flow statement—through the formulas and links in the spreadsheet. The income statement (also called a profit and loss [P&L] statement or statement of operations) describes revenues and expenses over a period of time, such as a quarter or a fiscal year. It answers the question of whether the business is profitable. The balance sheet (also called the statement of financial position) presents a picture at a point in time of what the firm owns (its assets), how much it owes (its liabilities), and what is left for shareholders (equity, owner’s equity, or net worth). Once we have created the pro forma income statement and balance sheet, the cash flow statement is a simple derivation.
The Income Statement For the income statement, a general form sufficient for many purposes is shown in Table 8.1. The line items in the pro forma income statement depend on the type of business. In this simplified statement, cost of goods sold (COGS) includes the costs of raw materials and expenses associated with manufactur-
Financial Modeling 273 Tab le 8 .1 Income statement (year ended 12/31/2018)
Revenue − Cost of goods sold (COGS) Gross profit − Operating expenses (cash) Earnings before interest, taxes, and noncash expenses (EBITDA) − Operating expenses (noncash) Operating profit (EBIT) − Interest expense (net of any interest income) Earnings before tax (EBT) − Income tax Net income (NI) note: Operating expenses includes both cash and noncash expenses like depreciation and amortization. Normally, these are not reported separately. They are separated here to show the distinction between EBIT and EBITDA.
ing the product. A retailer or wholesaler would show COGS as the acquisition cost of the products they resell. For a service company, COGS is the labor and machinery cost associated with providing the service. Operating expenses are other expenses related to the productive activities of the venture. These can include selling, general, and administrative (SG&A) expenses, R&D, overhead expenses, lease expenses, depreciation expenses, and similar items. Sometimes it is useful to include more detail on specific operating expenses that are central to success, especially if their inclusion makes forecasting easier, more reliable, and more convincing. In particular, noncash expenses, such as depreciation, are useful to break out because explicit listing can facilitate construction of the cash flow statement. Earnings before interest, taxes, and noncash expenses like depreciation and amortization (EBITDA) is not generally separately reported in the income statement. EBITDA is a measure of the short-run cash flow that would be available for such things as debt service, taxes, and new investment. Depreciation and amortization expenses depend on asset investment decisions that were made in earlier periods and are reflected in the balance sheet. Operating profit (EBIT) is a core measure of performance, as it is calculated before interest and tax expenses, two items that are usually outside the control of operating managers. For a business that is in steady state, it is also a measure of cash flow available to all investors and for taxes. The next deduction, interest expense, is a result of the financing decision that is reflected in the balance sheet. Sometimes, in developing the pro forma forecast, it is convenient to build the model so that surplus cash is assumed to
274
Chapter Eight
be retained in the venture. If so, it is appropriate to assume that the surplus cash is invested to earn a risk-free rate of interest and is reported as interest income. Income tax expense is determined by the profitability of the venture and prevailing tax rules. When developing the pro forma income statement, it is important to consider the impact of fixed and variable costs. Fixed expenses will not change with increases or decreases in revenue (over some range), which introduces the concept of operating leverage. Operating leverage describes the relationship between percentage revenue growth and percentage profitability growth. Firms with a larger proportion of fixed expenses have higher operating leverage and will see profits increase (decrease) faster as revenue goes up (down). However, overly simplistic classification of expenses as either fixed or variable can yield incorrect results. It is a good idea to study actual expense levels of businesses of different sizes. You may find that your intuition about fixed and variable expenses is not supported by the evidence. Few expenses are truly fixed, and others may vary more than proportionately with revenue. Consequently, assuming that variable expenses will change in proportion to revenue and that fixed expenses will not change at all can overstate the profit potential associated with revenue growth.
The Balance Sheet The balance sheet depicts the venture’s financial position at a point in time, such as at the end of a fiscal year. The left side reports what the company owns—its assets. The right side reflects how the ownership is financed— liabilities and equity. In the modeling, it will be more convenient for us to stack the balance sheet information vertically—assets first, then liabilities, then equity. This will enable us to keep all of the accounting numbers for a given period in a single spreadsheet column. A general form of a balance sheet sufficient for many purposes is shown in Table 8.2. Asset accounts are presented in order of declining liquidity, starting with cash. Accounts receivable (AR) and inventory are both current assets and components of working capital. The fixed asset account represents the property, plant, and equipment (PP&E) of the venture and is the basis for computing depreciation expense on the income statement. For a capital-intensive manufacturing business, PP&E might be the largest balance sheet line item. Liabilities, or what the firm owes, are ordered by the closest date of maturity. Current liabilities, such as accounts payable (AP) and wages payable, are considered working capital items and are tied to the venture’s day-to-day operations. Other liabilities are incurred by raising debt capital (notes payable, long-term debt). Many technology ventures receive annual payments for services they are
Financial Modeling 275 Tab le 8 . 2 Balance sheet (year ended 12/31/2018)
Assets
Liabilities and Equity
Current assets
Current liabilities
Cash and cash equivalents
Accounts payable (AP)
Accounts receivable (AR)
Wages payable
Inventory
Deferred revenue
Other current assets
Notes payable
Total current assets
Other current liabilities Total current liabilities
Fixed assets (PP&E)
Equity
Gross fixed assets
Preferred stock
Less: accumulated depreciation
Common stock
Net fixed assets
Retained earnings
Intangible assets
Total equity
Total assets
Total liabilities and equity
note: Other current liabilities include current portion of long-term debt and capital leases. Long-term debt includes debt convertible to equity.
expected to deliver in the future over time, giving rise to a deferred revenue account, which is another current liability. Explicit borrowing or leasing can also create a current liability such as a note payable or the current portion of longterm debt. An important distinction is that liabilities that arise from financing activities usually carry explicit interest payments while other working capital liabilities usually do not. Long-term debt and other funded debt such as notes payable drive the interest expense calculation in the income statement. For many entrepreneurial firms, long-term debt may consist of debt that is convertible to equity, as discussed in Chapter 4. Equity represents the shareholders’ position and is made up of two or three main components. As discussed in Chapter 4, entrepreneurial ventures often issue preferred stock to early investors, which stock is convertible to common under certain conditions. Preferred stock plus common stock is the cumulative amount investors paid for the venture’s stock when it was sold by the company. The other component of equity is retained earnings; the retained earnings balance is the accumulated profit (or loss) of the venture from its inception to the date of the balance sheet, less any dividends the firm has paid. The retained earnings account does not indicate money available to shareholders. Over any given period, retained earnings will increase by the net income for the period and be reduced by dividends paid to shareholders. The result of the income statement is net income, or the “bottom line.”
276
Chapter Eight
The Cash Flow Statement Net income is not cash flow. To determine financing needs and as a basis for valuation, we need to calculate cash flow. The cash flow statement connects the income statement and balance sheet changes but also takes account of noncash expenses, new investments, and new financing transactions. We use the “indirect” method to prepare the pro forma cash flow statement. This approach is mechanically straightforward and requires only the income statement and the beginning and ending balance sheets for the period. A general format for the cash flow statement is shown in Table 8.3. The cash flow statement is organized into three categories: operating cash flow, investing cash flow, and financing cash flow. The entry “(Increase) decrease” denotes whether the change in the balance sheet account represents a negative or positive cash flow. Increases in noncash assets like A/R and investments in fixed assets are uses of cash (outflows), while decreases are sources of cash (inflows). For liabilities and equity, increases are sources of cash, while decreases are uses. For example, because inventory must be purchased, an increase in inventory is Tab le 8 . 3 Cash flow statement (year ended 12/31/2018)
Net income
+
Depreciation and amortization
+
(Increase) decrease in AR
+
(Increase) decrease in inventory
+
Increase (decrease) in AP, wages payable, etc.
+
Increase (decrease) in deferred revenue
Operating cash flow +
Proceeds from sale of fixed assets
–
Gross investments in fixed assets
Cash flow from operations and investing +
Increase (decrease) in notes payable
+
Increase (decrease) in long-term debt
+
Increase (decrease) in common stock
–
Dividends paid
∙ ∙
∙
Operating cash flow
Investing cash flow
Financing cash flow
Net cash flow +
Beginning cash and cash equivalents
Ending cash and cash equivalents note: Free cash flow (FCF) is a non-GAAP financial performance measure, calculated as operating cash flow minus capital expenditures. FCF represents the cash that a company is able to generate from operations after making investments required to maintain or expand its asset base. If FCF is negative, the company must obtain additional financing to pursue desired investments. If FCF is positive, funds can be used to redeem debt early or can be distributed to equity investors.
Financial Modeling 277
a use of cash. Analogously, when subscriptions are paid in advance, an increase in deferred revenue is a source of cash. The total of operating cash flow, investing cash flow, and financing cash flow is net cash flow for the period. In forecasting, adding net cash flow to the beginning cash balance tells us how much cash (or cash and equivalents) the firm is expected to have on hand at period end.
8.2 Working Capital, Growth, and Financial Needs To introduce the concept of financial statement integration, we begin with how the working capital policies of a growing venture affect its financial needs. The most important components of working capital usually are inventory (raw materials and finished goods), A/R, A/P, and cash. Normally, because of the cash flow cycle, a relationship exists between sales and the levels of the various working capital accounts. A business that sells from inventory must have enough on hand to fill orders as they arrive. Because timing of demand is not perfectly predictable and there is often an inventory “pipeline,” the business normally will carry enough inventory to supply expected demand for several days, weeks, or even months. Purchases of inventory, either finished goods for resale or raw materials, are often made on credit in the form of A/P. Purchasing on credit reduces the need for cash, since payment is deferred. Conversely, if the business sells on credit, the balance of A/R will be driven by the terms it offers to credit customers, the extent of credit sales, and how quickly receivables are collected. Offering customers the option to delay payment requires additional cash; the business is deferring collection of payment but has already incurred the cost of goods. Offering credit also means facing the reality of customer defaults. Some working capital transactions generate financing that is referred to as “spontaneous” in the sense that it arises naturally through the normal business practices of the venture. Since its inception, Amazon.com, an example we have previously mentioned, has taken advantage of the unique working capital characteristics inherent in its business model. At year end 2000, when Amazon was still focused on book distribution, its A/R balance was equivalent to 3.9 days of sales and its inventory was equivalent to 30.3 days of cost of goods sold. The company’s combined investment in those two accounts was $204.2 million, which had to be financed in some way. Amazon was also taking advantage of the long payment terms that normally are offered by book publishers. These terms were originally offered to
278
Chapter Eight
encourage brick-and-mortar book sellers to stock large and diverse inventories for customers to peruse and were an industry convention long before Amazon .com existed. At the end of 2000, Amazon’s A/P balance stood at 84.1 days of cost of sales, or $485.4 million. This was far more than was needed to finance the A/R and inventory balances. The contribution of these accounts to net working capital was negative $281.2 million. This negative contribution to net working capital functions as spontaneous financing that Amazon could use to fund its operations. Spontaneous financing is financing that comes about through the ongoing operations and growth of the firm and supplants the need for formally raising cash flows through explicit financing activities. We can define the level of spontaneous financing as the excess of current nonfinancial liabilities over current nonfinancial assets other than the level of cash required for operating the enterprise. Many highly successful ventures have, like Amazon, relied heavily on spontaneous financing. For example, at the time of its IPO, salesforce.com had nonfinancial current liabilities totaling $73.3 million and noncash current assets totaling $14.0 million, resulting in spontaneous financing of $59.4 million. The main contributor to this total was $52.3 million in deferred revenue. Salesforce is a customer relationship management (CRM) cloud-based platform that offers subscription-based applications for sales, service, and marketing. These prepaid subscriptions are a current liability—deferred revenue—and hence are a source of funds similarly to A/P. Other examples of companies with important deferred revenue include Zendesk, Dropbox, and Box. The extent of spontaneous financing can be affected in deliberate ways by changing working capital practices. A business can deliberately change its reliance on accounts and wages payable as financing sources by changing the rate at which it pays for its inventory and wages. In a competitive industry, however, normal practices are guided by competitive pressure, and it can be difficult to gain a sustainable advantage by manipulating working capital practices.
Working Capital Financing Net working capital is the difference between the sum of the current asset categories of working capital and the spontaneous current liabilities.1 If the balance is positive (current assets exceed current liabilities), the difference constitutes a net investment in working capital and the resulting cash deficit must be financed in some way. If the balance is negative, as in the salesforce.com example, then not only are the firm’s operations self-financing but they also generate cash that is available to finance other investments or cover other expenses. For most businesses, however, the net working capital balance is posi-
Financial Modeling 279
tive and additional financing is required. The larger and faster the firm grows, the more financing it requires for net working capital. Figure 8.1 is a portion of an integrated set of financial statements that is designed to illustrate how working capital policy choices contribute to the funding required for net working capital. The policy choices are represented in the dark shaded text boxes in columns 1 and 4 of the figure. The unshaded text boxes in columns 2 and 3 are directly related to the income statement. The light shaded text boxes represent the current asset (column 5) and current liability (column 6) accounts from the balance sheet. Shaded numerical boxes are those that can be changed to evaluate alternative working capital policy choices. Dollar amounts shown in columns 5 and 6 and as Income Contribution are the financial results of the policy choices. The amount of net working capital (for which financing is required) is shown at the bottom right side of the figure. The numbers included in the figure as an illustration are representative of a manufacturing venture that carries significant inventories, offers trade credit, purchases raw material inventory on terms, and pays employees one week in
Fi g u r e 8 .1 Working capital policy template This figure is a template for assessing the impact of working capital policies on financial needs. It can be used to examine how working capital policies interact to determine the balances of current asset and current liability accounts. Net working capital is the excess of current assets over current liabilities and represents a funding need (or source) for the venture.
Column:
1
2
3
4
5
Pricing Policy Price/unit $10.00
Quantity per day 100
Revenue/ day $1,000
Credit Policy Days 45
Accounts receivable $45,000
Inventory Policy Days 12
Materials inventory $4,800
Purchasing Policy Materials cost/unit $4.00
Quantity per day
Materials cost/day
100
$400
Payables Policy Days 10
Inventory Policy Wage Policy Labor cost/unit $2.50
Days 5 Quantity per day
Labor cost/day
100
$250
Accounts payable $4,000
Finished goods inventory $3,250
Payroll Policy Days 7
Wages payable $1,750 Current assets
Income Contribution
$350
6
$53,050
Current liabilities $5,750
Net working capital $47,300
280
Chapter Eight
arrears. The result of the specific assumptions is a positive financing need of $47,300, or approximately 47 days of sales. For other types of businesses, the shaded numerical fields in the template can be modified to incorporate such things as deferred revenue.
Working Capital Policy As Figure 8.1 shows, a company’s working capital position is the result of several policy choices. Some are easily recognizable as aspects of working capital policy, whereas others are not. Furthermore, as we explain, the effects of individual policy choices are interdependent in ways that are not emphasized in the figure. Pricing policy. Pricing policy is a generic descriptor of decisions related to product positioning, pricing, and marketing. Collectively, these decisions determine the expected sales volume. We emphasize pricing in the figure because the effective price (along with quantity of sales) is influenced by credit policy. If, for example, a venture sells on terms that include a discount for prompt payment, then the price used in the calculation should be the average expected price net of anticipated trade discounts and collection losses. With cash sales, the price is not discounted and there is no risk that a buyer will default. On the other hand, the quantity of sales is likely to be higher if the venture offers credit. Credit policy. Given that credit policy influences the effective average price and quantity of sales, how does a business establish an appropriate policy? Because selling on credit gives rise to A/R, Figure 8.1 suggests that a policy of selling only on cash terms would reduce the amount required to finance net working capital. But since selling only for cash could result in lower sales, simply minimizing the need for financing is usually not consistent with maximizing value. It could make sense to evaluate the effect of some alternative credit policies on profitability and value: If you were to sell only for cash, what would be the likely effect on sales? If you were to accept credit cards, how much would unit sales be expected to increase and what would be the expected effect on average net price? Credit card sales are quickly converted to cash, so the A/R period would be very short. The credit provider also discounts the payments to the vendor as a fee for providing credit. As a practical matter, new and small businesses often have limited ability to determine their own credit policies. Smaller and less well-established businesses are more likely than others to offer terms that include delayed payment. Custom-
Financial Modeling 281
ers may demand opportunities to verify the quality of the seller’s product before they pay. Also, although trade credit terms vary greatly across industries, they tend to be similar within an industry. Thus, in an industry where offering credit is common, a new or small business generally must offer terms consistent with the general practice. In an industry where cash payment is common, a new or small business still may find it important to offer credit as a means of initially attracting customers and signaling product quality.2 Purchasing and inventory policies. Purchasing policy relates to choices that affect cost of materials. Choices of materials and negotiations with suppliers affect the cost of materials per unit produced. Inventory policy relates to the average number of days of inventory the business seeks to maintain in raw materials and in finished goods. Inventory policy affects average cost of goods indirectly, because maintaining a large average inventory may enable the business to take advantage of purchase quantity discounts. This benefit is offset by the costs of holding inventory, including storage, spoilage, and obsolescence. Purchasing and inventory policies interact with sales quantity to determine the cost of goods and represent another link between the balance sheet and income statement. Payables policy. Payables policy concerns the ability to defer payment for materials as a means of financing. Vendors generally require cash payment, especially for a new business with no track record of paying suppliers, or offer credit terms either with or without a discount for prompt payment. Normally, a purchaser would choose to delay payment, provided that doing so did not affect the effective price. The policy decision is whether to take advantage of discounts for prompt payment. Doing so reduces the effective cost of materials, as suggested by the feedback loop in Figure 8.1, but also reduces the A/P balance, giving rise to a need for additional financing. Often, discounts for prompt payment are large enough that a business with access to other sources of financing would routinely choose to pay in time to get the discount.3 Wage and payroll policies. Wage policy relates to decisions such as whether to offer a higher wage rate to motivate employees and limit turnover or pay a lower wage. Depending on the importance of experience on the job, motivation, availability of new employees, and similar considerations, average labor cost per unit of sales could be lower by either approach. The average labor and materials costs per unit of finished goods make up the direct cost of a unit. These costs interact with finished goods inventory policy to determine the value of finished goods inventory, which appears on the balance sheet.
282
Chapter Eight
Payroll policy relates to the frequency and timing of payroll, relative to the timing of the actual work done by the employees. The balance of wages payable as a source of spontaneous financing is increased by paying in arrears and paying less frequently (e.g., biweekly instead of weekly). Payroll policy affects wages indirectly. Wage payments that are substantially in arrears may necessitate higher wage rates. Wage and payroll policies interact with sales quantity to determine the balance of wages payable.
Evaluating Working Capital Policy Choices For many ventures, the need for working capital places a significant demand on cash. It is important to understand the working capital needs and to factor them into the pro forma financials and estimates of financial needs. As shown in Figure 8.1, the elements of working capital policy are interrelated and impact many aspects of venture operations, making even small changes important. Working capital policy choices can be evaluated in the same way as any other investment or financing decision. A policy change that increases net working capital requires additional investment. Assuming that the change is expected to be permanent, Figure 8.1 can be used to determine the daily increase in profitability from the change. Annualizing the daily increase and dividing by the increase in net working capital is effectively an internal rate of return that can be compared to a reasonable estimate of cost of capital to assess the merits of the possible policy change. For example, a change that increases profit per day by $5 will increase it by $1,825 per year (in perpetuity or as long as the policy remains in place). If the change would increase net working capital by $20,000, the annual return on that additional investment would be 9.125%. The change is profitable in an accounting sense, but the IRR is lower than many reasonable estimates of cost of capital, so the change might represent a negative NPV investment in additional working capital. Conversely, a reduction in net working capital is effectively a financing choice since it reduces the necessary investment to support the level of working capital. If the action would reduce accounting profitability per day, that reduction can be annualized and compared to cost of capital to evaluate whether the change would be beneficial based on economic profit. If the cost of the reduction is below cost of capital, the change is effectively a low-cost financing action. Working capital is only one example of why the integration of income statement, balance sheet, and cash flows is important. Other important interdependencies are related to fixed assets and depreciation and to debt financing and interest and debt service.
Financial Modeling 283
8.3 Developing Assumptions for the Financial Model A new venture financial model is only as good as its underlying assumptions. Assumptions that are evidence based and well-reasoned are important for reliable forecasting and for convincing investors of the validity of the projections. There are many good places to search for data useful for projecting revenue, expenses, working capital, fixed asset intensity, and cash flow. In this section, we introduce some of them and illustrate their use for making the assumptions we will use for building the Morebucks financial model.
Information Sources One valuable source of information about small, revenue-generating, nonpublic companies is the Risk Management Association (RMA) publication Annual Statement Studies. RMA compiles financial statistics from information supplied by credit customers of banks.4 Reporting companies are classified by North American Industry Classification System (NAICS) code so that, for a given venture, it may be possible to locate one or more industry groupings as benchmarks. An attractive aspect of the RMA data is that the sample includes many small, nonpublic companies, which data are otherwise difficult to obtain. Although tax minimization strategies of closely held businesses can distort some of the income statement relationships, much of the information is quite useful. Other sources with similar information include the Almanac of Business and Industrial Financial Ratios (CCH, Inc.); Industry Norms and Key Business Ratios (Dun & Bradstreet, Inc.); the annual corporate income and unincorporated income publications of the Internal Revenue Service (available at http://w ww .irs.gov/taxstats/); and the Quarterly Financial Report for Manufacturing, Mining, and Trade Corporations (U.S. Department of Commerce). These sources contain a myriad of financial information summarized at the industry level and sometimes broken out by firm size. PrivCo provides business and financial data on major nonpublicly traded companies, family firms, and private equity. Standard & Poor’s Compustat and Capital IQ are comprehensive databases of financial and market information for public firms worldwide. In addition, trade associations and investment services such as the Value Line Investment Survey and the Standard & Poor’s Analysts’ Handbook publish industry-level data. Moody’s and Standard & Poor’s report data for individual public companies. As discussed in Chapter 7, financial reports of individual companies are available on the SEC’s EDGAR database and through online services such as Yahoo! Finance, Hoover’s, and LexisNexis. Press releases, conference
284
Chapter Eight
call transcripts related to periodic reports of public companies, and other information is available on the Factiva database. Many industries have one or more specialized publications that contain valuable information and may be available online. To locate other sources, consult the Gale Cengage Learning Encyclopedia of Business Information Sources. SEC filings of comparable firms may include financial data from before the firm went public. If comparable firms can be identified, their income statement relationships can be used to calibrate the assumptions underlying a financial model. If the income statement benchmark ratios (such as gross margin) are consistent, even with only a few comparable firms, then we can be more confident when using them in a forecasting model. If there is substantial variation across the comparable firms, it is important to try to understand why the financial ratios for apparently similar companies differ.
Using Industry Data and SEC Filings to Develop Assumptions How might we use industry data and information on comparable companies to develop the assumptions for Morebucks, the coffee shop venture we introduced in Chapter 7? To illustrate, we use information compiled by Dun & Bradstreet (D&B) and reported in its key business ratios statistics. Table 8.4 shows the ratios reported by D&B for eating and drinking establishments. The last three columns are data for the subset of smaller eating establishments, those with $500,000 to $1 million in total assets ($560,000 to $1.112 million in 2015 inflation-adjusted dollars). Consistent with what we would expect of a coffee shop that sells for cash or by credit card, D&B reports A/R collection periods of only a few days—the medians range from 1.8 to 4.8 days’ sales.5 The financial ratio reported by D&B for measuring inventory is the ratio of sales to inventory, also known as inventory turnover. Median values range from 60.1 to 87.4 times per year.6 These numbers imply that eating and drinking establishments typically maintain inventories sufficient for only 4 to 6 days. This is not surprising given that their inventories are highly perishable. Table 8.4 also reports information on other ratios, including A/P to sales, some measures of financial solvency, and some measures of profitability. The A/P to sales percentage is in the 2.0–3.0% range, which implies an A/P balance between 7 and 11 days. The short payables period suggests that the ventures are probably paying cash for some supplies and getting short credit terms for others. We do not use the profitability measures because the D&B data include private companies, where the owner may be seeking to reduce taxes by doing such things as paying high salaries to family members. We do not use the solvency measures because we intend to decide separately on financing.
Financial Modeling 285 Tab le 8 . 4 Key business ratios for eating and drinking establishments
SIC Code Line of Business Asset Size
5813
5812
5812
Drinking Places
Eating Places
Eating Places
All Asset Ranges
All Asset Ranges
$500,000 to $1,000,000
Sample Size
Statement Sampling: 12
Statement Sampling: 202
Statement Sampling: 42
Solvency
Upper
Median
Lower
Upper
Median
Lower
Upper
Median
Lower
Quick Ratio (times) Current Ratio (times) Current Liabilities/Net Worth (%) Current Liabilities/Inventory (%) Total Liabilities/Net Worth (%) Fixed Assets/Net Worth (%)
3.1 6.0 6.7 86.4 40.3 80.7
0.9 1.3 31.1 415.0 127.6 107.4
0.4 0.7 90.3 777.3 260.9 193.2
1.0 2.0 23.0 326.9 38.5 63.4
0.5 0.9 49.2 768.8 101.7 112.9
0.2 0.6 103.5 999.9 275.2 194.9
1.7 2.2 19.7 263.2 24.7 29.4
0.7 1.4 36.9 506.5 49.9 75.9
0.3 0.7 99.7 929.6 157.8 117.2
Efficiency Collection Period (days) Sales/Inventory (times) Assets/Sales (%) Sales/Net Working Capital (times) Accounts Payable/Sales (%)
2.6 96.8 23.5 144.6 1.3
3.5 60.1 70.4 24.1 1.8
6.2 46.1 112.2 10.4 2.5
1.5 128.4 24.5 29.4 1.8
4.8 87.4 45.7 14.4 3.0
11.7 47.0 71.4 7.8 4.1
0.7 125.9 21.2 28.6 2.2
1.8 70.6 28.5 20.7 3.0
7.0 36.2 46.7 6.8 3.7
8.2 19.3 19.5
3.8 5.1 11.6
–3.0 –4.1 –35.9
5.6 13.0 28.3
2.1 4.9 12.0
–0.6 –1.0 –0.3
6.8 19.2 38.7
3.3 10.5 20.6
0.8 2.3 5.3
Profitability Return on Sales (%) Return on Assets (%) Return on Net Worth (%) http://kbr.dnb.com/K BR _ Main .asp
What can you learn from a prospectus or annual report? Prospectuses of young companies often contain historical information on early-stage financing, revenue growth, development time, margins, and relationships with suppliers and customers. In Chapter 7, we collected data from comparable public companies to use as yardsticks for estimating the revenue of Morebucks. Given its size and orientation at the time of IPO, we consider Coffee People to be the most useful yardstick. For confirmation, we also use Peet’s Coffee but do not report the statistics. Consideration of multiple yardsticks can be more informative and reliable. The prospectus of a yardstick company such as Coffee People contains both quantitative and qualitative information that can be useful. Coffee People filed its prospectus (Form SB-2/A) with the SEC in September 1996 and went public shortly thereafter. It was acquired by Diedrich Coffee in 1999, which was itself acquired by Green Mountain Coffee Roasters in 2009. The prospectus describes Coffee People as a seller of coffee beverages and other food products through company-owned retail stores. At the time of the IPO, the company sold through 17 company-owned stores, up from 7 stores 5 years earlier.
286
Chapter Eight
The prospectus includes the income statement and balance sheet information shown in Table 8.5, along with data on the number of stores. The statements for the most recent two years before the offering are provided in more detail than those for earlier years. For example, instead of a detailed breakout of current asset and liability accounts, the early-year information shows only a total for net working capital. We use the financial data to compute the common size income statements and balance sheets shown in Table 8.6. In a common size income statement, each revenue or expense line item is stated as a percentage of revenue. In a common size balance sheet, each asset or liability account is stated as a percentage of total assets. Common size analysis allows us to compare firms of different size and also makes it easier to spot trends. For example, the gross margin of Coffee People varies somewhat between 1991 and 1995, but with no discernible trend, which could indicate that there are no important economies of scale in purchasing or production as the company adds stores. Over the same period, store operating expenses increase somewhat more slowly than revenue, as do other expenses that reduce operating profit. Overall, the operating profit margin does not vary systematically with firm size. This suggests that profitability is not very sensitive to scale over the range of firm sizes represented by Coffee People. Table 8.7 presents some standard financial ratios calculated using the data from Coffee People. In contrast to the common size statements, financial ratios often combine information from the income statement and balance sheet. For most measures, we have data for only the last two years before the IPO. The decline in the asset turnover ratio in the table in later years is likely to reflect the company’s expansion in the number of stores. When using yardstick data, it is important to be cognizant of differences in the business models. For example, Coffee People, which is most similar to Morebucks, has very high A/R turnover (less than one day of sales) and less than 20 days of sales in inventory. Peet’s, in contrast, is more vertically integrated and has about four days of sales in receivables and around 100 days in inventory. Peet’s is not simply a multistore coffee shop business; rather, its business model evolved over a number of years. Peet’s started with stores that sold only roasted coffee beans. Just before its IPO, only about half of its store revenue was from beverages and pastries, about 20% was from online sales and through specialty grocers, and the rest was sales of other items. In addition to quantitative information, the Coffee People prospectus contains qualitative information that could be relevant to the Morebucks financial model. For example, the risk factors described in the prospectus include such things as risk associated with the company’s growth strategy, uncertainty about
Tab le 8 . 5 Coffee People, Inc. financials prior to IPO
Fiscal Year ($000) 1991
1992
1993
1994
1995
Income Statement Total revenue
3,512
4,498
5,466
7,708
11,257
Cost of sales
1,797
2,154
2,499
3,788
5,388
Gross profit
1,715
2,344
2,967
3,920
5,869
Store operating expenses
1,173
1,418
1,733
2,314
3,451
Other operating expenses General and administrative expenses Depreciation and amortization Total operating costs and expenses Income (loss) from operations Interest expense Other income Income (loss) before income taxes Income tax provision (benefit) Net income (loss)
0
10
25
40
63
291
775
799
1,210
1,550
60
75
119
175
391
3,321
4,432
5,175
7,527
10,843
191
66
291
181
414
10
8
43
88
134
2
1
0
39
43
183
59
248
132
323
0
17
0
16
112
183
42
248
116
211
86
157
253
472
260
Balance Sheet Assets Current assets: Cash and cash equivalents Accounts receivable
12
9
Inventories
204
264
Other current assets
115
125
Total current assets
803
658
(517)
(517)
1,613
2,155
Net working capital
42
(24)
(217)
Fixed and intangible assets: Property and equipment, net Intangible and other assets, net
96
23
2,512
2,836
Accounts payable
941
775
Accrued liabilities
180
196
Other current liabilities
101
55
Short-term borrowings
152
322
Total current liabilities
1,374
1,348
Total assets
565
664
888
Liabilities and Shareholders’ Equity Current liabilities:
Long-term liabilities: Long-term borrowings, less current portion
469
633
298
355
1,086
1,843
1,981
669
855
Total shareholders’ equity
267
309
(198)
669
855
Total liabilities and shareholders’ equity
565
664
888
2,512
2,836
Total liabilities Shareholders’ equity: Common stock issued an outstanding
Number of Stores in Operation Beginning of year
7
8
9
12
17
Store openings
1
1
3
5
0
End of year
8
9
12
17
17
Tab le 8 . 6 Coffee People, Inc. common size statements
Fiscal Year ($000) 1995
1996
1997
1998
1999
100.0%
100.0%
100.0%
100.0%
100.0%
51.2% 48.8% 33.4% 0.0% 8.3% 1.7% 94.6% 5.4% 0.3% 0.1% 5.2% 0.0% 5.2%
47.9% 52.1% 31.5% 0.2% 17.2% 1.7% 98.5% 1.5% 0.2% 0.0% 1.3% 0.4% 0.9%
45.7% 54.3% 31.7% 0.5% 14.6% 2.2% 94.7% 5.3% 0.8% 0.0% 4.5% 0.0% 4.5%
49.1% 50.9% 30.0% 0.5% 15.7% 2.3% 97.7% 2.3% 1.1% 0.5% 1.7% 0.2% 1.5%
47.9% 52.1% 30.7% 0.6% 13.8% 3.5% 96.3% 3.7% 1.2% 0.4% 2.9% 1.0% 1.9%
15.2%
23.6%
28.5%
7.4%
–3.6%
–24.4%
18.8% 0.5% 8.1% 4.6% 32.0% –20.6%
9.2% 0.3% 9.3% 4.4% 23.2% –18.2%
100.0%
64.2% 3.8% 100.0%
76.0% 0.8% 100.0%
37.5% 7.2% 4.0% 6.1% 54.7%
27.3% 6.9% 1.9% 11.4% 47.5%
Income Statement Total revenue Operating expenses: Cost of sales Gross profit Store operating expenses Other operating expenses General and administrative expenses Depreciation and amortization Total operating costs and expenses Income (loss) from operations Interest expense Other income Income (loss) before income taxes Income tax provision (benefit) Net income (loss) Balance Sheet Data Assets Current assets: Cash and cash equivalents Accounts receivable Inventories Other current assets Total current assets Net working capital Fixed and intangible assets: Property and equipment, net Intangible and other assets, net Total assets
100.0%
100.0%
Liabilities and Shareholders’ Equity Current liabilities: Accounts payable Accrued liabilities Other current liabilities Short-term borrowings Total current liabilities Long term liabilities: Long term borrowings, less current portion Total liabilities Shareholders’ equity: Common stock issued an outstanding Total shareholders’ equity. Total liabilities and shareholders’ equity
52.7%
53.5%
122.3%
18.7% 73.4%
22.3% 69.9%
47.3%
46.5%
–22.3%
26.6% 26.6%
30.1% 30.1%
100.0%
100.0%
100.0%
100.0%
100.0%
Financial Modeling 289 Tab le 8 .7 Coffee People, Inc. financial ratios
Fiscal Year Financial Ratio
Asset Turnover Fixed Asset Turnover Accounts Receivable Turnover Days’ Sales in Accounts Receivable Inventory Turnover Days’ Cost of Sales in Inventory Sales/Inventory Accounts Payable/Cost of Sales Days’ Cost of Sales in Accounts Payable Compensation and Other Payable/Cost of Sales Cash/Revenue Days’ Revenue in Cash
1995
1996
1997
1998
1999
6.22
6.77
6.16
3.49% 12.74
4.63% 16.89
3.07 4.78 642.3 0.57 18.57 19.7 37.78 24.8% 90.7 4.8% 6.12% 22.35
3.97 5.22 1250.8 0.29 20.41 17.9 42.64 14.4% 52.5 3.6% 2.31% 8.43
2.45% 8.94
Definitions
Asset Turnover = Sales/Assets Fixed Asset Turnover = Sales/Net Fixed Assets Accounts Receivable Turnover = Sales/Accounts Receivable Days’ Sales in Accounts Receivable = Accounts Receivable/(Sales/365) Inventory Turnover = Cost of Sales/Inventory Days’ Cost of Sales in Inventory = Inventory/(Cost of Sales/365) Days’ Cost of Sales in Accounts Payable = Accounts Payable/(Cost of Sales/365) Days’ Revenue in Cash = Cash/(Revenue/365)
market acceptance, and dependence on a single supplier of coffee beans. Some of the risk factors provide important insights into specific risks facing the company. How can we evaluate whether Coffee People is a reasonable yardstick company? One test is whether the ratios of A/R and inventory to sales are similar to those from the D&B data for eating and drinking establishments. From Tables 8.4 and 8.6, we can see that the A/R and inventory days are within the ranges of the D&B data for small eating places. No matter what firms we select as yardsticks, we may need to make appropriate adjustments for differences in scale and scope of the venture.
Developing Assumptions Based on Fundamental Analysis Industry data and yardsticks can be valuable, particularly when underlying business models are similar to those of the subject venture. But for in-depth planning, fundamental analysis is useful as a way to test the merits of assumptions that are based on yardsticks. Sometimes fundamental analysis is the only
290
Chapter Eight
reasonable choice for developing a specific assumption. For example, information for estimating operating leverage or the proportions of fixed and variable expenses are rarely available from yardstick data. As an example of the importance of fundamental analysis, consider the assumptions used to forecast fixed asset investment. If we were to use data from Coffee People, we might divide the 1995 property and equipment balance of $2.16 million by the number of stores (17 during 1995), producing an estimate of about $127,000 in fixed assets per coffee shop ($142,000 in 2016 dollars). But can we really trust this to be accurate for Morebucks? The investment in fixed assets will be a function of the store location and depends on what leasehold improvements the venture will require. Moreover, the yardstick information is for net fixed assets (which are partially depreciated) and could be impacted by whether stores are purchased or leased. For reasons such as these, there is more variability in balance sheet estimates developed using yardsticks than for items on the income statement. For a venture like Morebucks, fundamental analysis can yield a more reliable forecast of the fixed asset investment and related depreciation expenses than can a yardstick approach. When Morebucks opens, all of its furniture and equipment will be newly acquired. Suppose the space will be leased but that leasehold improvements, furniture, and equipment are estimated to cost $150,000, all of which can be depreciated. If we assume five-year, straight-line depreciation, the annual depreciation expense will be about $30,000. Because the furniture and equipment are fully paid for at the start, there is no cash flow directly tied to depreciation. However, depreciation expense reduces pretax income and the tax the firm pays each year. At a 35% tax rate, the $30,000 depreciation expense would reduce income tax by $10,500 each year, which does impact cash flow. We could, of course, do more with fundamental analysis. The data from Coffee People show only a single line for store operating expenses, which includes store salary and benefits. However, the operating expenses line also includes other items. We could make a better forecast by developing salary forecasts for Morebucks based on a specific staffing plan. We could collect information on local wages to help estimate the prevailing level of wages and benefits. We could develop a specific marketing plan—which we would need in any case—that would enable us to estimate how much will be spent on advertising and promotion. And we could estimate raw materials costs based on local market prices, the anticipated product mix, and volumes implicit in the revenue forecast. Fundamental analysis is always useful when preparing pro forma financials. The questions that drive the process are a standard part of any business planning exercise. Answering these questions forces the entrepreneur to undertake
Financial Modeling 291
significant due diligence and better understand the industry, the market, and the new venture strategy. Yardsticks can then be used to validate forecast assumptions that are developed by fundamental analysis.
8.4 Building a Financial Model of the Venture We can now develop the integrated financial statements for Morebucks. We base the statement assumptions on the information and methods discussed in the previous section. By building the model in stages, we demonstrate how the financial statements interact and how the cash flow statement can be derived from the pro forma balance sheet and income statement. Of course, Morebucks will not achieve full capacity instantly; for this reason the financial model needs to incorporate the anticipated growth trajectory. To reflect this reality in the dynamic relationships among the financial statements, we introduce a time dimension into the forecast. As a basis for this, we might do further research using yardstick and industry data, perform additional fundamental analysis, and meet with local coffee shop owners about their start-up experiences. Recall from Chapter 7 that the estimate of annual revenue for viable coffee shops was approximately $1.5 million. Suppose our new research and analysis supports an assumption that, as a new venture with no established brand name or reputation, Morebucks revenue will reach one third of the $1.5 million during the first year, two thirds during the second year, and the full estimate in the third year. After the third year, annual revenue is expected to stay constant in real terms at $1.5 million. Table 8.8 shows a partially completed template of the three financial statements we will construct. It combines elements of the basic statements discussed in Section 8.1 as well as line items from the Coffee People’s (CP) financials. Because revenue is expected to take three years to reach steady state, we have structured the pro forma statements to cover Time 0 and four years of operation. This allows us to illustrate the impact of start-up events on the financial statements and to cover two years of steady-state operation. In Table 8.8, we introduced only income statement data and have not yet incorporated depreciation expense in the income statement. The table is constructed using Excel and includes the cell formulas needed to link the assumptions to the financial statements and the financial statements to each other. The columns to the right summarize the numerical assumptions and describe from where they originate. Where possible, we link the assumptions directly to the spreadsheet so the effects of changing an assumption will
Tab le 8 . 8 Step 1 of pro forma financial model for Morebucks: income statement assumptions
Pro Forma Income Statement
Time 0
Year 1
Year 2
Year 3
Year 4
Net revenue Cost of sales
0
500,000 242,000
1,000,000 484,000
1,500,000 726,000
1,500,000 726,000
Gross profit Operating expenses General, administrative, and other expenses Depreciation and amortization expenses
0
258,000 472,500 69,500 0
516,000 472,500 139,000 0
774,000 472,500 208,500 0
774,000 472,500 208,500 0
Income from operations Interest income (expense), net
0
(284,000)
(95,500)
93,000
93,000
Income before income taxes Income tax provision
0
(284,000) (99,400)
(95,500) (33,425)
93,000 32,550
93,000 32,550
Net income
0
(184,600)
(62,075)
60,450
60,450
Time 0
Year 1
Pro Forma Balance Sheet
Year 2
Year 3
Assumption
Year 4
From revenue forecast 48.4% From CP common size statement (average) 31.5% From CP common size statement (at steady state) 13.9% From CP common size statement (at steady state) 5 Years, straight line—on fixed assets, gross
35% Effective rate—applies to all income
Assumption
Assets Current Assets Required cash Surplus cash Accounts receivable Inventory Total current assets Fixed assets, gross Less: Accumulated depreciation
0 0
0 0 0
0 0 0
0 0 0
0 0 0
Net fixed assets
0
0
0
0
0
Total assets
0
0
0
0
0
Liabilities Current Liabilities Accounts payable Wages and other payables
0.0% 0.0%
Total current liabilities Long-term debt
0
0
0
0
0
Total liabilities Equity Common stock Retained earnings
0
0
0
0
0
0
0 (184,600)
0 (246,675)
0 (186,225)
0 (125,775)
Total equity Total liabilities and equity
0 0
(184,600) (184,600)
(246,675) (246,675)
(186,225) (186,225)
(125,775) (125,775)
Basis for Assumption
Basis for Assumption
Pro Forma Cash Flow Statement Operating Cash Flow Net income Plus: Depreciation (Increase) decrease in accounts receivable (Increase) decrease in inventory Increase (decrease) in accounts payable Increase (decrease) in wages payable Operating cash flow Investing Cash Flow (Increase) decrease in gross fixed assets
Time 0
Year 1
Year 2
Year 3
Year 4
0 0
(184,600) 0
(62,075) 0
60,450 0
60,450 0
0
(184,600)
(62,075)
60,450
60,450
Investing cash flow Financing Cash Flow Increase (decrease) in debt Increase (decrease) in common stock Dividend paid Financing cash flow
0
0
0
Net cash flow Beginning cash
0 0
(184,600) 0
0
(62,075) 0
0
60,450 0
60,450 0
Ending cash
0
(184,600)
(62,075)
60,450
60,450
294
Chapter Eight
be automatically reflected in all of the financial statements. This is standard practice for any spreadsheet model we might want to build.
Modeling the Income Statement The revenue forecast is taken from Chapter 7 and reflects the growth trajectory discussed above. The expense ratio assumptions—expressed as percentages of revenue—are based on data from the Coffee People prospectus. We decided to use an average expense ratio from the 1991–1995 data, which represents a five-year period when Coffee People was relatively small and before it approached the point of going public. Coffee People does not provide information on the breakdown between fixed and variable expenses in its income statement. In our base case model, we assume that expenses other than depreciation are fixed percentages of revenue for a single shop. When we add uncertainty to the model, we will relax this assumption. The final assumption required in the income statement is the effective combined state and federal tax rate, which we have assumed to be 35%. There are various ways to model the effect of taxes in years when the venture has a net loss. The simplest, the one we apply in this model, is to assume that the entrepreneur has other earnings against which losses can be offset so that the tax benefit of a loss is realized in the year of the loss. Alternatively, we could accumulate the losses and apply them against future positive earnings. In light of the added complexity, we do not use that approach here. Because depreciation expense is driven by the amount invested in facilities and improvements, we defer making this assumption until we construct the balance sheet. The tax and net income totals will not be correct until the depreciation expense line is complete. Note, however, that the net income line also now appears as the first item on the cash flow statement, where it is one of the determinants of operating cash flow. This is one of many links between the financial statements.
Modeling the Balance Sheet In Table 8.9, we turn to the balance sheet, beginning with the current asset and current liability accounts. Again, the numerical assumptions and their sources are in the right-hand columns. In some cases, the assumptions come directly from the Coffee People prospectus. When the data from Coffee People differ substantially from industry norms or the conclusions of our fundamental analysis, we have established the assumption based on judgment and the evidence.
Financial Modeling 295
At this point, the only initial investment shown in the Time 0 column is the $150,000 investment in fixed assets, which is assumed to be made with equity capital provided by the entrepreneur. Note that the balance sheet in Table 8.9 is in balance each year, that is, total assets equal total liabilities and equity each year. To accomplish this, the balance sheet includes a line item we call “Surplus Cash.” This is a modeling choice we made for this example. To force the statements to balance, we allow surplus cash to show negative balances. We use the surplus cash account as a device for assessing financial need. If the balance is negative in any period, the company has less resources than it needs to fund the operation and additional financing is needed. There are other ways to address this in an integrated financial model. One is to constrain surplus cash to be nonnegative but allow the balance sheet to be out of balance and then make subsequent adjustments. For example, suppose the model’s assumptions produce assets greater than liabilities plus equity. Balance can be achieved either by reducing assets or by increasing liabilities or equity (or some combination of the two). For example, if in the Morebucks model we were to assume an initial purchase of $150,000 of fixed assets but only $50,000 of equity investment, the balance sheet would be out of balance by $100,000. Balance could be achieved by reducing the fixed asset assumption to $50,000, increasing the equity investment to $150,000, adjusting both accounts, or increasing some other source of financing. Generally, it is easier to force the statement to be in balance and keep track of cash shortages or surpluses, as we have done here. In this approach, the surplus cash “plug” becomes a line item on the asset side. In Table 8.9, the Year 1 balance sheet is balanced only because the surplus cash balance is extremely negative. This is attributable primarily to the negative financial performance in the first year, with no commensurate increase in liabilities or equity (i.e., financing). This funding shortfall causes surplus cash to be negative. This approach shows the financing shortage or surplus in the surplus cash line, which represents an allocation of the total cash balance or an indication of a need for additional funding. Another good alternative is to shift the shortage or surplus to the liabilities and equity side of the statement. Were we to do this here, we would first establish a minimum cash level on the balance sheet and force the cash balance to equal the cash required. The model already has assumptions about all other asset line items, so this would effectively determine the asset (left) side of the balance sheet and produce a total assets number. We can then add a “plug” item to the right side of the balance sheet, calling it “additional funding needed” or “additional financing.” This makes the right side of the balance sheet a column of numbers, including the unknown plug, that sums to a known number (total
Tab le 8 . 9 Step 2 of pro forma financial model for Morebucks: Balance sheet assumptions
Pro Forma Income Statement Net revenue Cost of sales Gross profit Operating expenses General, administrative, and other expenses Depreciation and amortization expenses
Time 0 0 0
Year 1
Year 2
Year 3
Year 4
500,000 1,000,000 1,500,000 1,500,000 242,000 484,000 726,000 726,000 258,000 516,000 774,000 774,000 472,500 472,500 472,500 472,500 69,500 139,000 208,500 208,500 30,000 30,000 30,000 30,000
Income from operations Interest income (expense), net Income before income taxes Income tax provision
0
(314,000) (125,500)
63,000
63,000
0
(314,000) (125,500) (109,900) (43,925)
63,000 22,050
63,000 22,050
Net income
0
(204,100)
40,950
40,950
Pro Forma Balance Sheet Assets Current Assets Required cash Surplus cash Accounts receivable Inventory Total current assets Fixed assets, gross Less: Accumulated depreciation Net fixed assets Total assets
Time 0
0 0 0 0 0 150,000 150,000 150,000
Year 1
(81,575) Year 2
Year 3
Year 4
19,000 38,000 57,000 57,000 (163,611) (204,697) (123,257) (52,307) 2,466 4,932 7,397 7,397 25,641 51,282 76,923 76,923 (116,504) (110,483) 18,063 89,013 150,000 150,000 150,000 150,000 (30,000) (60,000) (90,000) (120,000) 120,000 90,000 60,000 30,000 3,496 (20,483) 78,063 119,013
Liabilities Current Liabilities Accounts payable Wages and other payables
0 0
47,432 10,164
94,864 20,328
142,296 30,492
142,296 30,492
Total current liabilities Long-term debt
0 0
57,596 0
115,192 0
172,788 0
172,788 0
Total liabilities Equity Common stock Retained earnings Total equity Total liabilities and equity
0
57,596
115,192
172,788
172,788
150,000 150,000 150,000
150,000 (204,100) (54,100) 3,496
150,000 150,000 150,000 (285,675) (244,725) (203,775) (135,675) (94,725) (53,775) (20,483) 78,063 119,013
Assumption
Basis for Assumption
From revenue forecast 48.4% From CP common size statement (average) 31.5% From CP common size statement (at steady state) 13.9% From CP common size statement (at steady state) 5 Years, straight line—on fixed assets, gross
35% Effective rate—applies to all income
Assumption
Basis for Assumption
3.80% Based on CP cash/revenue ratio (average) 2 Days in A/R based on industry A/R turnover ratio 19.5 Based on CP sales/inventory ratio (average) 150,000 Based on fundamental analysis
19.6% Based on CP AP/cost of sales ratio (average) 4.20% Based on CP wages and other payables/cost of sale
150,000 Selected to cover start-up investments
Pro Forma Cash Flow Statement Operating Cash Flow Net income Plus: Depreciation (Increase) decrease in accounts receivable (Increase) decrease in inventory Increase (decrease) in accounts payable Increase (decrease) in wages payable Operating cash flow Investing Cash Flow (Increase) decrease in gross fixed assets Investing cash flow Financing Cash Flow Increase (decrease) in debt Increase (decrease) in common stock Dividend paid Financing cash flow Net cash flow Beginning cash Ending cash
Time 0 0 0 0 0 0 0 0
Year 1
Year 2
(204,100) 30,000 (2,466) (25,641) 47,432 10,164 (144,611)
(81,575) 30,000 (2,466) (25,641) 47,432 10,164 (22,086)
Year 3
Year 4
40,950 30,000 (2,466) (25,641) 47,432 10,164 100,439
40,950 30,000 0 0 0 0 70,950
(150,000) (150,000)
0 0
0 0
0 0
0 0
0 150,000
0
0
0
0
150,000 0 0 0
0 0 (144,611) 0 (144,611)
0 0 0 0 (22,086) 100,439 (144,611) (166,697) (166,697) (66,257)
0 0 70,950 (66,257) 4,693
298
Chapter Eight
assets). The spreadsheet would be designed so that the plug figure causes assets to equal liabilities plus equity. If the model indicates that the enterprise requires additional funding, the plug will be positive; if the venture has more resources than needed, the plug will be negative. Note that the balance sheet in Table 8.9 is in balance each year; that is, total assets equal total liabilities plus equity. The venture might be able to distribute cash to investors or pay down a loan. A negative balance of additional funding needed serves the same purpose as a positive surplus cash balance in the first approach. A more elegant approach is to combine these last two alternatives. This third approach recognizes the reality that a negative balance in either the external funding or the surplus cash line items does not make sense. If, before forcing the balance, assets exceed liabilities, then the firm needs more investment capital and the Excel model can be structured so that additional financing is the plug; if assets are less than liabilities, then the firm has more capital than it needs and the Excel model can make surplus cash the plug—so there are only positive plugs. This approach provides greater flexibility than the others. For example, surplus cash might be credited with earning the risk-free rate of interest and additional financing required might be assumed to be in the form of debt, with a risky interest rate, or in the form of equity. With the balance sheet completed, the changes in the various current account balances from one year to the next appear in the cash flow statement in the operating cash flow section. When the venture reaches steady state after Year 3, the net working capital balance does not change. As a result, the Year 4 cash flows from changes in current assets and liabilities are zero. Table 8.9 also shows a $150,000 investment in fixed assets, which is based on fundamental analysis. This investment is depreciated over five years, which means we can calculate the depreciation expense as $30,000 for each year. This completes the income statement, with the possible exception of any interest expense associated with financing. Accumulated depreciation on the balance sheet is linked to the income statement and adds the depreciation expense each year to the prior year’s accumulated depreciation balance. This has the effect of reducing the net fixed assets balance over time. Because depreciation is a noncash expense, it appears in the cash flow statement, where it is added back to net income as part of operating cash flow. The negative cash surplus shown in Table 8.9 implies that additional financing is needed. This can be resolved by adding an assumption that the venture will be entirely equity financed and structuring the model to automatically add the needed new equity. Table 8.10 shows the resulting financial statements. To cover the cost of the initial investment in fixed assets and working capital, we start with an estimate of $375,000 of equity, which appears on the Time 0 bal-
Financial Modeling 299
ance sheet as $375,000 in common stock that is divided between the investment in fixed assets and surplus cash. The equity investment also shows up on the statement of cash flows as a positive financing cash flow. Based on our initial assumptions, the coffee shop is expected to incur significant losses in Years 1 and 2, due to start-up levels of revenue, and to be profitable each year thereafter. With no distributions or additional equity infusions, Table 8.10 shows the accumulation of net income (or loss) from each year moving to the retained earnings balance in the equity section of the balance sheet. The retained earnings balance at the end of Year 4 reflects the projected accumulated loss from Morebucks’ first four years of operation. The balance sheet in Table 8.10 shows a positive surplus cash balance each year, with the Year 2 ending balance being the lowest, and increasing thereafter. Rather than assuming that the entrepreneur withdraws the cash as it accumulates, we have allowed it to remain in the venture. The surplus cash could earn a return (by being held in an interestbearing account such as a money market fund), but to keep the model simple we ignore this for now. The positive surplus cash balances raise the question of how the assumption of a $375,000 initial investment was determined. Does the positive projected cash surplus in Year 2 mean that the entrepreneur has put too much cash into the venture up front? The projections seem to suggest that it does. However, we should not draw this conclusion without further analysis. The current model does not capture important timing issues during the first few years and does not allow for uncertainty of performance. To accurately estimate cash needs, we might want to prepare pro forma statements on a monthly or quarterly basis and also make an assessment of uncertainty of performance.
Deriving the Cash Flow Statement The cash flow statement allows us to see the venture’s sources and uses of cash. This information may appear to be readily available from the income statement, but that is not the case. First, income statement items such as depreciation do not translate directly to cash flow; second, many cash transactions don’t appear on the income statement. In addition, the income statement and balance sheet provide little information about the venture’s ability to generate cash to fund growth. The cash flow statement translates the income statement and balance sheet information into cash flows and separates them into three categories: operating, investing, and financing. The coffee shop cash flow statement in Table 8.10 is straightforward. When the venture begins operation and is growing, operating cash flow reflects increases in working capital accounts. When the venture
Tab le 8 .1 0 Step 3 of pro forma financial model for Morebucks: investment assumption
Pro Forma Income Statement Net revenue Cost of sales Gross profit Operating expenses General, administrative, and other expenses Depreciation and amortization expenses Income from operations Interest income (expense), net Income before income taxes Income tax provision Net income Pro Forma Balance Sheet
Time 0 0 0
0 0 0
Year 1
Year 2
Year 3
500,000 1,000,000 1,500,000 242,000 484,000 726,000 258,000 516,000 774,000 472,500 472,500 472,500 69,500 139,000 208,500 30,000 30,000 30,000 (314,000) (125,500) 63,000 (314,000) (109,900) (204,100)
(125,500) (43,925) (81,575)
63,000 22,050 40,950
Year 4 1,500,000 726,000 774,000 472,500 208,500 30,000 63,000 63,000 22,050 40,950
Time 0
Year 1
Year 2
Year 3
Year 4
0 225,000 0 0 225,000 150,000
19,000 61,389 2,466 25,641 108,496 150,000 (30,000) 120,000 228,496
38,000 20,303 4,932 51,282 114,517 150,000 (60,000) 90,000 204,517
57,000 101,743 7,397 76,923 243,063 150,000 (90,000) 60,000 303,063
57,000 172,693 7,397 76,923 314,013 150,000 (120,000) 30,000 344,013
0 0
47,432 10,164
94,864 20,328
142,296 30,492
142,296 30,492
0 0 0
57,596 0 57,596
115,192 0 115,192
172,788 0 172,788
172,788 0 172,788
375,000 (204,100) 170,900 228,496
375,000 (285,675) 89,325 204,517
375,000 (244,725) 130,275 303,063
375,000 (203,775) 171,225 344,013
Assumption
Basis for Assumption
From revenue forecast 48.4% From CP common size statement (average) 31.5% From CP common size statement (at steady state) 13.9% From CP common size statement (at steady state) 5 Years, straight line—on fixed assets, gross
35% Effective rate—applies to all income
Assumption
Basis for Assumption
Assets Current Assets Required cash Surplus cash Accounts receivable Inventory Total current assets Fixed assets, gross Less: Accumulated depreciation Net fixed assets Total assets Liabilities Current Liabilities Accounts payable Wages and other payables Total current liabilities Long-term debt Total liabilities Equity Common stock Retained earnings Total equity Total liabilities and equity
150,000 375,000
375,000 375,000 375,000
3.80% Based on CP cash/revenue ratio (average) 2 Days in A/R based on industry A/R turnover ratio 19.5 Based on CP sales/inventory ratio (average) 150,000 Based on fundamental analysis
19.6% Based on CP AP/cost of sales ratio (average) 4.20% Based on CP wages and other payables/cost of sales
150,000 Selected to cover start-up investments
Pro Forma Cash Flow Statement Operating Cash Flow Net income Plus: Depreciation (Increase) decrease in accounts receivable (Increase) decrease in inventory Increase (decrease) in accounts payable Increase (decrease) in wages payable Operating cash flow Investing Cash Flow (Increase) decrease in gross fixed assets Investing cash flow Financing Cash Flow Increase (decrease) in debt Increase (decrease) in common stock Dividend paid Financing cash flow Net cash flow Beginning cash Ending cash
Time 0 0 0 0 0 0 0 0
Year 1
Year 2
(204,100) 30,000 (2,466) (25,641) 47,432 10,164 (144,611)
Year 3
(81,575) 30,000 (2,466) (25,641) 47,432 10,164 (22,086)
Year 4
40,950 30,000 (2,466) (25,641) 47,432 10,164 100,439
40,950 30,000 0 0 0 0 70,950
(150,000) (150,000)
0 0
0 0
0 0
0 0
0 375,000
0
0
0
0
0 0 100,439 58,303 158,743
0 0 70,950 158,743 229,693
375,000 225,000 0 225,000
0 0 (144,611) 225,000 80,389
0 0 (22,086) 80,389 58,303
302
Chapter Eight
reaches steady state, between Years 2 and 3, operating cash flow includes only net income and the add-back for (noncash) depreciation expense. Because the working capital account balances are constant, they have no impact on cash flow. The statement next provides for changes in cash due to investing. The gross fixed assets balance represents the cumulative original cost of the firm’s PP&E. The change in gross fixed assets from the balance sheet is shown as an investment cash flow on the cash flow statement. When fixed assets are purchased, the increase appears as a negative cash flow. If fixed assets are sold, the sale produces a cash inflow. The last section of the cash flow statement covers financing activities. The only event in this example is the equity investment of $375,000 at Time 0. With all of the financing provided at Time 0, there are no changes in debt or equity for Years 1 through 4. Return of the surplus cash to the equity investor can be treated in two ways. One way is to pay a dividend, which would appear as a negative financing cash flow and would reduce retained earnings on the balance sheet. Another is to repurchase shares, which would be a negative financing cash flow and would reduce common stock outstanding on the balance sheet. Finally, at the foot of the cash flow statement, we use the net cash flow plus the cash balance from the prior period to compute the ending cash balance. Both beginning and ending cash balances include required and surplus cash. This information is used in the balance sheet to fill in the surplus cash amount.
An Illustration of Financial Statement Integration Table 8.10 is a fully integrated Excel file of the Morebucks financial model. It contains all important assumptions, along with source documentation. The financial statements are linked to the assumptions, to one another, and also over time. This allows us to immediately see the impact of changing any key assumptions. For example, suppose it may be possible to acquire the needed fixed assets for $320,000 rather than the original estimate of $350,000 and that Year 1 revenue might only be $300,000. These changes can be made quickly. The lower revenue causes net income to turn more negative. This is partially offset by the reduction in depreciation expense arising from the smaller starting balance in gross fixed assets. Operating cash flow goes down, reflecting both the lower net income and reduced investment in working capital. Surplus cash is also lower and becomes negative in Year 2. Because the model is fully integrated, the two changes to the assumptions are immediately reflected in affected line items on all three statements. This allows the entrepreneur or an investor to focus on the operational implications of the changes.
Financial Modeling 303
8.5 Adding Uncertainty to the Financial Model The Morebucks financial model incorporates the static assumptions shown in Table 8.10. The resultant pro forma financial statements are useful but do not reflect the uncertainty inherent in any forecast. To introduce uncertainty, we modify the Morebucks model in two ways. Further research into the performance of new coffee shops would reveal a high variability in first-year revenue and many different three-year trajectories of revenue growth. For example, 10-K data from yardstick companies contain cross-sectional information on per-store revenue. Fundamental analysis and discussions with other coffee shop owners would provide additional data that can be used to estimate the variability of revenue. To incorporate uncertainty, we modify the model by replacing the static revenue assumption each year with statistical distributions. We also introduce uncertainty about operating expense and the size of the initial fixed investment. The specific changes are shown in the following table. The revenue assumptions are designed to model an uncertain initial revenue and rate of revenue growth over the first two years, and to reach a specific expected ultimate revenue level by Year 3. Variable
Distribution assumption
Year 1 revenue Year 2 revenue
Triangular with a mode of $500,000, a minimum of $400,000, and a maximum of $700,000 Triangular with a mode of 2 times the Year 1 result, a minimum that is $50,000 lower, and a maximum that is $50,000 higher Normally distributed with a mean of $1.5 million and a standard deviation of $100,000 The same as Year 3 revenue Normally distributed with mean of 31.5% of revenue and standard deviation of 2% Normally distributed with a mean of $150,000 and standard deviation of $5,000
Year 3 revenue Year 4 revenue Operating expense Fixed asset investment
We used @Risk to track the initial assumptions and each year’s net income, and cash flow. Table 8.11 shows one trial from a simulation model that incorporates these elements of uncertainty into the pro forma statements. The cells modified to reflect these uncertainty assumptions are shaded and can be reviewed in the Excel file. Table 8.11 is identical to Table 8.10 except for the effects of random draws from the assumed distributions. Table 8.12 shows a summary of some of the results from simulating the model over 1,000 trials. The Inputs section of the table shows that in some trials, the rate of revenue growth is slower than expected and in some it is faster. The ultimate level of revenue is reached in Year 3 and ranges from about $1.2 to $1.8 million. The table also shows distribution results for the operating expense percentage
Tab le 8 .11 Pro forma financial model for Morebucks with simulation
Pro Forma Income Statement Net revenue Cost of sales Gross profit Operating expenses General, administrative, and other expenses Depreciation and amortization expenses Income from operations Interest income (expense), net Income before income taxes Income tax provision Net income Pro Forma Balance Sheet
Time 0 0 0
0 0 0
Year 1 526,010 254,589 271,421 400,571 73,115 30,000 (232,265) (232,265) (81,293) 150,972)
Year 2
Year 3
1,083,672 1,486,566 524,497 719,498 559,175 767,068 400,571 400,571 150,630 206,633 30,000 30,000 (22,027) 129,864 (22,027) (7,709) (14,317)
Year 4 1,486,566 719,498 767,068 400,571 206,633 30,000 129,864
129,864 45,452 84,412
129,864 45,452 84,412
Year 2
Year 3
Year 4
Time 0
Year 1
0 0 0 0 0 150,000
19,988 (109,938) 2,594 26,975 (60,380) 150,000 (30,000) 120,000 59,620
41,180 (82,556) 5,344 55,573 19,540 150,000 (60,000) 90,000 109,540
56,489 40,308 7,331 76,234 180,362 150,000 (90,000) 60,000 240,362
56,489 154,719 7,331 76,234 294,774 150,000 (120,000) 30,000 324,774
49,899 10,693 60,592 0 60,592
102,801 22,029 124,830 0 124,830
141,022 30,219 171,240 0 171,240
141,022 30,219 171,240 0 171,240
150,000 (150,972) (972) 59,620
150,000 (165,290) (15,290) 109,540
150,000 (80,878) 69,122 240,362
150,000 3,533 153,533 324,774
Assumption
Basis for Assumption
From revenue forecast 48.4% From CP common size statement (average) 26.9% From CP common size statement (at steady state) 13.9% From CP common size statement (at steady state) 5 Years, straight line—on fixed assets, gross
35% Effective rate—applies to all income
Assumption
Basis for Assumption
Assets Current Assets Required cash Surplus cash Accounts receivable Inventory Total current assets Fixed assets, gross Less: Accumulated depreciation Net fixed assets Total assets Liabilities Current Liabilities Accounts payable Wages and other payables Total current liabilities Long-term debt Total liabilities Equity Common stock Retained earnings Total equity Total liabilities and equity
150,000 150,000
0 0 0 0 0 150,000 150,000 150,000
3.80% Based on CP cash/revenue ratio (average) Plugged to make statement balance 2 Days in A/R based on industry A/R turnover ratio 19.5 Based on CP sales/inventory ratio (average) 150,000 Based on fundamental analysis
19.6% Based on CP AP/cost of sales ratio (average) 4.20% Based on CP wages and other payables/cost of sales
150,000 Selected to cover start-up investments
Pro Forma Cash Flow Statement Operating Cash Flow Net income Plus: Depreciation (Increase) decrease in accounts receivable (Increase) decrease in inventory Increase (decrease) in accounts payable Increase (decrease) in wages payable Operating cash flow Investing Cash Flow (Increase) decrease in gross fixed assets Investing cash flow Financing Cash Flow Increase (decrease) in debt Increase (decrease) in common stock Dividend paid Financing cash flow Net cash flow Beginning cash Ending cash
Time 0 0 0 0 0 0 0 0
Year 1 (150,972) 30,000 (2,594) (26,975) 49,899 10,693 (89,949)
Year 2 (14,317) 30,000 (2,750) (28,598) 52,902 11,336 48,573
Year 3
Year 4
84,412 30,000 (1,987) (20,661) 38,220 8,190 138,174
84,412 30,000 0 0 0 0 114,412
(150,000)
0
0
0
0
(150,000)
0
0
0
0
0 150,000
0
0
0
0
0
0
0
0
150,000 0 0 0
0 (89,949) 0 (89,949)
0 48,573 (89,949) (41,377)
0 138,174 (41,377) 96,797
0 114,412 96,797 211,209
306
Chapter Eight
and the fixed asset investment. Among the outputs, it is noteworthy that in some trials, net income is still negative in Year 4. These are clear cases of failure since no further growth is expected. Even some positive net income numbers would also effectively be failures since the returns are too low to justify the investments. Cash flow is negative over all trials in Year 1 and over more than half of the trials in Year 2. It is positive over all trials in the last two years. Although we began the simulation with an investment of $375,000, it is apparent from the surplus cash balances that this is sometimes not enough. This is most apparent in Year 2, where over 25% of the trials have negative surplus cash balances. The worst case also occurs in Year 2, where the minimum surplus cash balance is negative $173,000. Apparently, increasing the initial equity investment to $550,000 would be sufficient to cover all of the outcomes. However, it may not be desirable to invest so much since the negative balances in Year 2 may be associated with outcomes that should be abandoned—possibly slow-growth, low-profit trials. We cannot evaluate the additional investment or abandonment questions on the basis of the summary statistics in Table 8.12. Rather, we need to investigate the trials data. Figure 8.2 shows the results of simulating the Morebucks model. The figure overlays results for Year 4 profitability (the year when Morebucks’s revenue is in steady state) with results for surplus cash in Year 2 (the year when surplus cash is usually lowest). The trials data are sorted by net income. Based on 1,000 trials, the figure shows that Year 4 net income is negative in a few cases.
Tab le 8 .12 Morebucks: Summary of simulation results (based on 1,000 trials)
Name
Mean
Minimum
25%
50%
75%
Maximum
577,399 1,156,060 1,567,213 32.8% 153,363
696,151 1,412,934 1,822,649 37.8% 168,351
Inputs Net revenue/Year 1 Net revenue/Year 2 Net revenue/Year 3 Operating expenses Fixed assets, gross
533,339 1,066,675 1,500,012 31.5% 150,001
403,496 794,352 1,190,968 25.0% 134,094
486,538 970,318 1,432,526 30.1% 146,616
526,654 1,057,105 1,499,982 31.5% 149,998
Outputs Net income/Year 4 Net cash flow /Year 1 Net cash flow/Year 2 Net cash flow/Year 3 Net cash flow/Year 4 Surplus cash/Year 1 Surplus cash/Year 2 Surplus cash/Year 3 Surplus cash/Year 4
40,964 (134,463) (3,769) 96,522 70,965 70,270 46,234 126,289 197,254
(21,873) (239,497) (129,608) 27,439 9,319 (35,134) (173,254) (147,509) (138,189)
27,538 (156,158) (34,137) 80,596 57,384 48,936 (1,470) 73,609 134,482
40,562 (133,296) (3,706) 96,282 70,745 71,331 47,863 128,409 197,640
54,278 (112,584) 26,045 112,745 84,055 91,199 93,224 179,076 260,319
102,296 (36,906) 128,127 172,728 132,020 162,835 258,024 377,067 498,951
Financial Modeling 307
While we defer a formal analysis of valuation until a later chapter, we can use Figure 8.2 to look informally at the choice of how much to invest, presuming that negative surplus cash would lead to abandonment. Based on the figure, if we want to cover all possible outcomes, we would need to invest an additional $175,000 ($550,000 total) so that the Year 2 surplus cash balance would always be positive. But, as noted above, this would fund a number of bad outcomes. If we limit the total investment to $500,000, it appears that all trials with Year 4 income above about $35,000 would be fully funded. But even this investment may not be warranted. If we limit the investment to a total of $450,000, it appears that all trials with at least $60,000 in income would be fully funded. This rudimentary examination shows that much more can be done with simulation to fine-tune the analysis and sharpen the investment and abandonment decisions. Introducing uncertainty with simulation in the Morebucks example can add significant richness. We could continue to refine the assumptions or consider other decision variables. We can also do more with simulation to understand the risks of the venture. We could, for example, use the trial results to study the
$300,000 $250,000 $200,000 $150,000 $100,000 $50,000
1 20 39 58 77 96 115 134 153 172 191 210 229 248 267 286 305 324 343 362 381 400 419 438 457 476 495 514 533 552 571 590 609 628 647 666 685 704 723 742 761 780 799 818 837 856 875 894 913 932 951 970 989
$0 –$50,000 –$100,000 –$150,000 –$200,000
Net income/year 4
Surplus cash/year 2
Fi g u r e 8 . 2 M o reb u ck s sim u lat io n of p rofit a bilit y a n d s u rp l u s ca s h Results are shown from a 1,000-trial simulation using @Risk. Simulated variables include net income and surplus cash.
308
Chapter Eight
relation between profitability and the operating expense result or profitability and the realized sales level. Most new ventures would require more frequent pro formas (e.g., quarterly or monthly) and a longer forecast horizon. The current simulation model also allows ending cash to be negative (although, given our assumptions, this does not occur). Negative cash would require more initial equity or additional cash investment after start-up.
8.6 NewCo: Building an Integrated Financial Model We now return to our earlier example, the medical technology venture called NewCo. In Chapter 7, we developed a static revenue forecast and then extended the revenue forecasting model to incorporate uncertainty. We follow a similar trajectory here in developing an integrated financial forecast. We begin with a static version of the NewCo financial model. The static model is based on the assumptions that are detailed in Figure 8.3. These assumptions reflect the expected value of each forecast. After we complete the static model, we will modify the assumptions to introduce uncertainty. You should assume that we developed the assumptions in Figure 8.3 on the basis of yardsticks, industry norms, and fundamental analysis, as we did for Morebucks. The first five assumptions were used to develop the static revenue forecast in Chapter 7. The others are the bases for individual line items in the pro forma income statement and balance sheet. To avoid the complexities of dealing with depreciation, we assume that all fixed assets are leased, with lease payments made each quarter as part of SG&A Expense. The final 3 assumptions specify the initial investment and provide a mechanism for bringing the balance sheet into balance, either by increasing financing or by building up surplus cash. We establish a minimum required cash balance of the greater of $15,000 or 10% of the quarter’s sales. We introduce a line of credit (debt) as a device that will allow NewCo to maintain the minimum cash balance. Consistent with the earlier discussion, if the venture is short of cash in any period, it is assumed to draw automatically on the credit line; if it generates free cash flow, the line is paid down. If the line is fully paid off and the venture generates free cash flow, the excess is retained as surplus cash. In contrast to the Morebucks model, we include an explicit inflation forecast in the assumptions so that we are forecasting in nominal terms. Inflation will impact the selling price and most of the expenses, because they are calculated as percentages of revenue or cost of sales. Some expenses are fixed in nominal terms.
Financial Modeling 309 Fi g u r e 8 . 3 N ewCo integ rated fin a n cia l m o d el a ss u m pt io n s
1.
All assumptions shown are expected values.
Development will require 6 quarters, during which period no sales will be made.
2.
Initial quarterly sales of 300 units at a price of $200 beginning in at the start of quarter 7.
3.
After quarter 7, unit sales will grow 25% per quarter for three years (through quarter 19) and then remain constant.
4.
The sales price will increase each quarter at the inflation rate.
5.
Inflation at 4% per year (modeled as 1.0% per quarter).
6.
Operating expenses during the 6-quarter development period are projected to be $60,000 per quarter plus inflation.
7.
Cost of sales is projected to be 45% of revenue.
8.
Beginning in quarter 7, the venture is expected to incur fixed Selling General and Administrative (SG&A) expenses of $90,000 per quarter, growing at the inflation rate. This includes the entrepreneur’s salary. Variable SG&A expenses are projected to be 20% of sales.
9.
A production facility will come on line at the end of quarter 6 and is expected to be adequate for the ensuing 5 years of operation (through quarter 26). Quarterly lease payments for the facility and production equipment will begin in quarter 7 and are included in fixed SG&A expenses.
10.
The effective corporate tax rate is projected to be 35% on positive income with no loss carry-forward; that is, any loss in a given period gets no tax credit and cannot accumulate to offset future profits.
11.
Net working capital other than cash, including accounts receivable (A/R) and inventory, less accounts payable (A/P), is 50% of sales in the quarter plus 50% of cost of sales in the subsequent quarter.
12.
The company needs to maintain a minimum cash balance equal to either 10% of the prior quarter’s sales or $15,000, whichever is greater.
13.
Initial equity investment by the entrepreneur is $500,000. Additional funding, if needed, will come from a hypothetical line of credit with no limit. Interest on the credit line is 2% quarterly (8% annually).
14.
Free cash flow in any period will first be used to reduce the balance of the line of credit, and then will be accumulated as surplus cash.
Table 8.13 shows the pro forma financial statements on a quarterly basis from the start of product development through the first five years of sales. With an expected development period of six quarters, the total forecast horizon is 26 quarters. The financial statements are simplified but appropriate for the venture. While only selected quarters are shown in Table 8.13, the Excel file on which the table is based contains all quarters. We selected the quarters to include in the printed table because they correspond to major milestones: development, start of revenue, initiation of external financing, profitability, positive operating cash flow, and the end of five years of sales operation. We now work through the statements chronologically, highlighting the quarters shown.
310
Chapter Eight
Modeling the Development Stage Under the heading for Quarter 0, Table 8.13 shows the beginning balance sheet. Based on the assumptions in Figure 8.3, NewCo has only one asset at start-up—cash—all of which the entrepreneur invests in the venture as equity. For now, we assume an initial investment of $500,000. In later chapters we will use simulation to estimate the appropriate level of initial financing. Development activity begins in Quarter 1 and is shown as the $60,000 of development expenses on the income statement. To keep the model simple, we do not consider tax-loss carryforwards or tax credits on negative net income. The cash flow statement reflects this loss as the first line in operating cash flow. Because the company has no depreciable assets and (at this point) no activity that would create working capital, the loss is operating cash flow. The negative operating cash flow reduces the company’s cash balance to $440,000, far in excess of the required minimum. Except for inflationary increases in development expenses, the statements for Quarters 2 through 6 are similar to those for Quarter 1. Over this period, losses continue to erode NewCo’s cash. At the end of the development period, Quarter 6, there is expected to be $117,379 remaining in cash.
Modeling the Start of Sales In Quarter 6, development is completed, and by the end of the quarter, in anticipation of the initiation of sales, NewCo has an initial investment in net working capital (inventory less A/P) on its balance sheet. Sales commence in Quarter 7, with $60,000 of sales revenue appearing on the income statement. With all sales made on account, much of the revenue goes to A/R, which increases net working capital. At this point, the venture is not profitable, with a loss in Quarter 7 of $69,000. Operating cash flow is even more negative, attributable to the increase in net working capital that is needed to support sales growth.
Modeling External Funding At the end of Quarter 7, NewCo’s projected cash balance is approaching zero. The venture is not yet profitable and has a negative operating cash flow. To maintain the minimum cash balance, NewCo must begin to draw on its line of credit. The credit line carries 2% quarterly interest, with interest expense computed on the basis of the prior quarter’s balance. The balance sheet shows the outstanding balance of the credit line beginning in Quarter 7; this is the
Tab le 8 .13 NewCo pro forma financial statements
Quarter Income Statement
0
1
Unit sales Selling price
6
7
18
19
26
$
300 200.00
3494 $ 223.13
$
4368 4368 $ 234.52 $ 236.86
$
$ $ -
$ $ -
$ 60,000 $ 27,000
$ 779,629 $ 350,833
$ 984,394 $ 442,977
$ 994,238 $ 447,407
$ 1,024,365 $ 460,964
$ 1,034,608 $ 465,574
$ 1,055,404 $ 474,932
Gross profit Development expense SG&A expense
$ $ 60,000 $ -
$ $ 63,061 $ -
$ 33,000 $ $ 102,000
$ 428,796 $ $ 256,336
$ 541,417 $ $ 298,293
$ 546,831 $ $ 301,276
$ 563,401 $ $ 310,405
$ 569,035 $ $ 313,509
$ 580,472 $ $ 319,811
Operating profit Interest income (expense), net
$ (60,000) $ (63,061) $ $ -
$ (69,000) $ 172,460 $ 243,124 $ $ (12,308) $ (13,025)
$ 245,555 $ 252,996 $ 255,526 $ 260,662 $ (12,450) $ (4,044) $ (975) $ -
Profit before income tax Tax expense
$ (60,000) $ (63,061) $ $ -
$ (69,000) $ 160,152 $ $ 56,053
$ 230,098 $ 80,534
$ 233,105 $ 81,587
$ 248,952 $ 87,133
$ 254,551 $ 89,093
$ 260,662 $ 91,232
Net income
$ (60,000) $ (63,061)
$ (69,000) $ 104,099
$ 149,564
$ 151,518
$ 161,819
$ 165,458
$ 169,430
$
4368 227.62
24
0 $ -
-
4368 225.37
23
0 $ -
Revenue Cost of sales
20
4368 241.62
Balance Sheet Cash Net working capital (excluding cash)
$ 500,000 $ -
$ 440,000 $ -
$ 117,379 $ 13,500
$ 15,000 $ 47,044
$ 61,748 $ 611,303
$ 77,963 $ 715,901
$ 98,439 $ 723,060
$ 101,422 $ 744,969
$ 210,692 $ 752,419
$ 532,751 $ 767,543
Total current assets Fixed assets, net
$ 500,000 $ -
$ 440,000 $ -
$ 130,879 $ -
$ 62,044 $ -
$ 673,052 $ -
$ 793,864 $ -
$ 821,499 $ -
$ 846,391 $ -
$ 963,111 $ -
$ 1,300,293 $ -
$ 440,000
$ 130,879
$ 62,044
$ 673,052
$ 793,864
$ 821,499
$ 846,391
$ 963,111
$ 1,300,293
Total assets Long-term debt (credit line)
$ -
$ -
$ -
$ 165
$ 651,266
$ 622,514
$ 498,631
$ 48,739
$ -
$ -
Total liabilities Equity
$ 500,000
$ $ 440,000
$ $ 130,879
$ 165 $ 61,879
$ 651,266 $ 21,786
$ 622,514 $ 171,350
$ 498,631 $ 322,868
$ 48,739 $ 797,653
$ $ 963,111
$ $ 1,300,293
Total liabilities and equity
$ 500,000
$ 440,000
$ 130,879
$ 62,044
$ 673,052
$ 793,864
$ 821,499
$ 846,391
$ 963,111
$ 1,300,293
(continued )
Tab le 8 .13 (continued )
Quarter
Statement of Cash Flows
0
Operating Cash Flow Net income Less: Increase in net working capital (excluding cash)
1
6
7
18
19
20
23
24
26
$ (60,000) $ (63,061) $ $ 13,500
$ (69,000) $ 104,099 $ 33,544 $ 127,145
$ 149,564 $ 104,598
$ 151,518 $ 7,159
$ 161,819 $ 7,376
$ 165,458 $ 7,450
$ 169,430 $ 7,599
Operating Cash Flow Investing Cash Flow Change in gross fixed assets Financing Cash Flow Change in long-term debt (credit line) dividend
$ (60,000) $ (76,561)
$ (102,544) $ (23,046) $ 44,966
$ 144,359
$ 154,443
$ 158,008
$ 161,831
$ -
$ -
$ -
$ -
$ -
$ -
$ -
$ -
$ -
$ -
$ -
$ 165
$ 35,885
$ (28,752)
$ (123,883) $ (153,439) $ (48,739) $ -
Financing cash flow
$ -
$ -
$ 165
$ 35,885
$ (28,752)
$ (123,883) $ (153,439) $ (48,739) $ -
Net Cash Flow Beginning cash Ending cash
$ (60,000) $ (76,561) $ 500,000 $ 193,940 $ 440,000 $ 117,379
$ (102,379) $ 12,839 $ 117,379 $ 48,910 $ 15,000 $ 61,748
$ 16,215 $ 61,748 $ 77,963
$ 20,477 $ 77,963 $ 98,439
$ 1,004 $ 100,418 $ 101,422
$ 109,270 $ 101,422 $ 210,692
$ 161,831 $ 370,920 $ 532,751
$ $ -
$ $ -
$ 165 $ -
$ 35,885 $ -
$ $ 28,752
$ $ 123,883
$ $ 153,439
$ $ 48,739
$ $ -
$ $ 425,000 $ $ (60,000) $ (60,000)
$ $ 102,379 $ $ (63,061) $ (76,561)
$ 60,000 $ $ 165 $ (69,000) $ (102,544)
$ 779,629 $ $ 651,266 $ 104,099 $ (23,046)
$ 984,394 $ $ 622,514 $ 149,564 $ 44,966
$ 994,238 $ $ 498,631 $ 151,518 $ 144,359
$ 1,024,365 $ $ 48,739 $ 161,819 $ 154,443
$ 1,034,608 $ 108,255 $ $ 165,458 $ 158,008
$ 1,055,404 $ 428,255 $ $ 169,430 $ 161,831
0
0
$ 500,000
Financing Activity New financing needed Debt repayment Revenue Surplus cash Cumulative financing needed Net income Operating cash flow Test: Cash out month
$ $ 485,000
0
0
0
0
0
0
0
Financial Modeling 313
start of a period of three years during which NewCo is expected to increase its borrowing every quarter. Not until Quarter 19, after the period of rapid sales growth ends, does the credit line balance begin to decrease as cash flow is sufficient to pay down the line of credit. By Quarter 7, the company has exhausted the entrepreneur’s initial investment, and the balance sheet shows that, from an accounting “book value” perspective, owner’s equity is almost gone. If we were to construct financial statements in terms of economic value, the statements would look very different. Assuming that the development activities are progressing, the expenditures during the first six quarters are actually capital investments in an intangible asset. The value of this asset, though not reflected in book value, is a critical element of the venture’s economic value. In fact, without it, NewCo would have little ability to attract funding. We have assumed that perceived economic value is sufficient to convince a lender to make a long-term loan in the form of a line of credit. In Quarter 8 (not shown in the table but viewable in the Excel file on the website), the book value of equity turns negative, which persists until it becomes positive in Quarter 18. The negative balance simply means that all of the owner’s initial capital has been expended, plus part of what was borrowed from the line of credit. The equity is negative because of the long period of unprofitability coupled with the increasing net working capital needed to support growth. If economic value is high enough, the negative balance in the equity account should not concern us. In fact, many new ventures with negative book equity still have positive economic value.
Achieving Profitability Thought not shown in Table 8.13, NewCo reaches profitability in Quarter 14. As significant as this is as a milestone, operating cash flow continues to be significantly negative because of the continuing working capital investment. This difference between profit and cash flow is important because even though NewCo is profitable, it continues to need to access the line of credit to offset the negative operating cash flow.
Operating Cash Flow and Stable Growth Unit sales grow at 25% per quarter through Quarter 19, when the expected growth rate of unit sales drops to zero. In Quarter 19, NewCo’s accounting profits exceed its working capital needs for the first time, resulting in positive operating cash flow. This is a critical milestone, because it reverses the trend of q uarterly
314
Chapter Eight
borrowing. With positive operating cash flow and reduced working capital needs due to slower growth, the company begins to repay the credit line. Repayment continues through Quarter 23, when a final payment reduces the credit line to zero. From this point through Quarter 26, operating cash flow accrues to each quarter’s ending cash balance and is assumed to be retained as surplus cash.
Forecasting Financing Needs Figure 8.4 graphically depicts NewCo’s pro forma expected financial performance over the 26-quarter forecast period. Surplus cash starts at $485,000, which is the entrepreneur’s Time 0 equity contribution minus the $15,000 minimum cash balance. By Quarter 7, surplus cash reaches zero and remains there until Quarter 24. In Quarter 7, NewCo begins to borrow on the credit line. The need for external financing continues for 11 quarters, with the credit line peaking at $651,266 in Quarter 18. Under our expected assumptions, the venture is in sound financial shape at Quarter 26. It is profitable, is generating operating cash flow, has no debt, and has over $480,000 in surplus cash. However, to reach this point, the venture is expected to need to borrow over $651,000 along the way.
Uncertainty in the NewCo Model The NewCo financial model provokes a number of questions. For example, how might things change if some of the assumptions prove to be wrong (as Fi g u r e 8 . 4 NewCo expected financial performance
$1,200,000
$1,000,000
$800,000
$600,000
$400,000
$200,000
$0 3
6
9
12
15
18
21
Quarter
–$200,000 Revenue
Net income
OCF
Credit line
Surplus cash
24
Financial Modeling 315
they undoubtedly will)? What would change if the entrepreneur were to put more equity into the deal, or if financing were raised in the form of outside equity, or if it were raised in stages? Finally, how should the entrepreneur select from among the various financing options that may be available? More generally, how does uncertainty affect the financing needs of the venture and its economic value? We begin to explore these issues here and continue the discussion in the next chapter. In Chapter 7, we introduced uncertainty to the NewCo sales forecast with regard to development time, initial selling price, and duration and magnitude of quarterly sales growth. The simulation results in Chapter 7 illustrated how this uncertainty can produce very different trajectories for revenue over the 26-quarter projection period. We now incorporate this development timing and revenue uncertainty into the integrated financial model shown in Table 8.13 and add the following random variables to the simulation: cost of sales, quarterly development expense, and variable SG&A expense: percentage of sales. Specifically, we make the following assumptions: Variable
Distribution assumption
Cost of sales Quarterly development expense Variable SG&A expense
Uniform distribution with a minimum of 40% and maximum of 50% Normally distributed with a mean of $60,000 and standard deviation of $600 Triangular distribution with a minimum of 18%, most likely of 20%, and maximum of 30%
Again, you should assume that these distributions were developed using some combination of yardstick information, industry data, and fundamental analysis.
Linking Assumptions to the Financial Model To facilitate use of the NewCo model for simulation, we have specified all of the assumptions in a single worksheet that is linked directly to the financial statements. Table 8.14 shows the assumption sheet, including one random trial for the simulated variables. In this trial, it turns out that development success is achieved in Quarter 9 and the rapid-growth period extends for nine quarters through Quarter 18. The initial unit sales level is 270, initial selling price is $226.74, and cost of sales is 43.66% of price. Development expense begins at $58,934 per quarter and SG&A is 25.65%. The shaded cells can be changed to test for such things as the effects of a larger initial investment and different expectations about cost percentages. The unshaded cells are values derived by simulation.
Tab le 8 .14 NewCo simulation assumptions
Revenue Assumptions Development Completion Quarter (lognormal distribution)
Preliminary quarter Development completion quarter Development failure (1 = yes) Rapid Growth Period (normal distribution) Standard deviation of rapid growth period (quarters) Realized length of rapid growth period (quarters) Initial Unit Sales per Quarter Initial units/quarter Unit Sales Growth during Rapid Growth (normal distribution) Expected growth/quarter Standard deviation of growth/quarter Realized growth rate per quarter Initial Selling Price (normal distribution) Expected initial selling price Standard deviation of selling price Realized initial selling price Inflation Rate per Quarter Inflation/quarter
9 9 0 1.5 9 270 25.00% 6.00% 24.28% $200.00 $20.00 $226.74 1.00%
Income Statement Assumptions
Cost of Sales (uniform distribution) Minumum cost of sales Maximum cost of sales Realized cost of sales Quarterly Development Expense (normal distribution) Quarterly development expenses (expected) Quarterly development expenses (standard deviation)
40.00% 50.00% 43.66% $60,000 $600
Realized development expense SG&A Expenses (fixed + triangular distribution) Quarterly fixed SG&A expense Minimum variable SG&A (expected % of sales) Most likely SG&A expense Maximum SG&A expense
$58,934
Realized variable SG&A percent of sales Interest Income and Interest Expense Interest expense per quarter
25.65%
Interest income on surplus cash per quarter Income Tax Expense Income tax rate (on positive income)
$90,000 18% 20% 30%
2.00% 0.00% 35%
Financial Modeling 317 Tab le 8 .14 (continued )
Balance Sheet Assumptions
Cash Balance Minimum cash balance Continuing cash percentage of prior quarter sales
$15,000 10.0%
Accounts Receivable Policy Percentage of current quarter sales
50%
Inventory and Payables Policy Percentage of next quarter cost of sales
50%
Initial Investment Initial equity investment
$500,000
The table represents the assumptions page of the NewCo integrated financial model. Shaded cells are inputs that can be changed; other cells are generated using the assumptions. Information in the assumptions page links directly to the NewCo financial model.
Results of the Simulation The NewCo simulation model is designed so the venture has an open-ended line of credit. If the venture runs short of cash in any quarter, borrowing is automatic and unlimited. Anytime the venture generates free cash flow, repayment of the loan is automatic. The quarterly credit line balance, which reflects NewCo’s cumulative cash needs, is affected by how quickly development occurs, the growth and profitability of subsequent sales, and the amount of investment required for working capital.7 It may seem that cumulative financing needs would be lowest when performance is good and highest when it is not. However, the relationship is not so simple. For example, in one simulated trial, where development fails and the venture is assumed to keep trying through Quarter 26 and $500,000 of initial equity funding is assumed, the credit line balance reaches $1.46 million. If this project is abandoned after Quarter 15, the balance is $497,000. In a high-profit trial where development success occurs early, the maximum credit line balance is $438,000, but in another high-profit trial with a similar development success date, the credit line balance reaches $1.89 million by Quarter 23 before turning down. The high need in this case occurs because the revenue growth rate is very high. The point is that large credit line balances can occur for good or bad reasons and are not necessarily evidence of NewCo’s success or failure. In the next chapter, we consider how operational performance impacts the venture’s need for financing. We also consider how staging, milestones, and financing decisions can help distinguish ventures that are likely to be successful from those that are probable failures.
318
Chapter Eight
We use a simulation of 1,000 trials to examine some of the important outcome variables from NewCo. Revenue in Quarter 26 ranges from $0 to more than $17.4 million, though half of the trials (the interquartile range) have revenue between about $258,800 and $1,713,500. Quarter 26 net income is negative in more than 30% of the trials but also has the potential to be very high. While the median is $78,800, the mean is $205,400, because there are a small number of very good outcomes. Ending cash flow is similar to net income, but somewhat lower, with a median of $72,700 and mean of $194,700. The maximum balance on the line of credit ranges from a low of $339,400 to almost $3.5 million but three-quarters of the time is less than about $1.4 million. Ideally, we would like to invest in the good outcomes but to cut off investment in the bad ones as quickly as we can identify them. In Chapter 9, we explore the use of simulation to assess financing needs and opportunities to stage financing for the purpose of creating more value from investment in the venture.
8.7 Summary Pro forma financial statements are important to any new venture and a key component in any entrepreneur’s or investor’s toolkit. Financial forecasting adds discipline to the way an entrepreneur or investor thinks about the venture. It provides estimates of future cash flows, which drive financing decisions and are important determinants of value. A forecast can also be important for convincing prospective investors of the merits of the project and can provide specific performance benchmarks around which financing and incentive contracts can be designed. A well-constructed integrated financial model of a new venture accomplishes two things. First, it reflects the important aspects of the venture’s business model in the three main financial statements: income statement, balance sheet, and cash flow statement. Second, it provides reliable estimates of the venture’s future cash flows. We use the venture’s cash conversion cycle to better understand the important links between day-to-day operations and cash flow. The cash conversion cycle visually portrays how cash moves through the firm, as well as in and out of the firm from and to capital providers. Working capital policies are important determinants of a venture’s operating cash flows. These include the granting of credit to customers, levels of inventory on hand, and the ability to generate spontaneous financing in the form of A/P. Preparing a credible and useful financial model requires well-researched and defensible assumptions. Yardsticks based on public or private companies and
Financial Modeling 319
industry statistics can be used to develop benchmark performance metrics and to build the new venture’s financial model. Public SEC documents, primarily 10-K and prospectus filings, are one source of information on comparable firms. The yardstick approach is straightforward to implement and can provide robust estimates of revenue growth, resource requirements, and industry practices related to working capital management. No new venture will perform exactly like the yardstick companies, however, and financial projections should also reflect the use of fundamental analysis to develop the schedule of assumptions. An integrated financial model might use assumptions developed using yardstick data and then validated with fundamental analysis or the converse. With a schedule of the key assumptions, spreadsheet modeling can be used to develop an integrated financial model for preparing pro forma financial forecasts. An integrated model features an income statement, a balance sheet, and a cash flow statement, all linked to one another and across time. An integrated financial model allows the entrepreneur or investor to conduct “what if” analysis and facilitates the use of simulation to assess the effects of uncertainty. Both scenario analysis and simulation analysis are valuable tools for assessing the prospects and potential future value of any new venture.
Review Questions
1. What information does each of the three main financial statements convey about the venture’s operations? Describe three important links across the statements. 2. Describe how to generate spontaneous financing by managing A/R and A/P. 3. Explain the cash flow cycle in Figure 8.1. 4. What are the main components of a venture’s working capital policy? How does each impact the firm’s operations and cash flows? How much control does a new venture have over its working capital policies? 5. Describe several common sources of industry and firm data that can be used as bases for developing forecast assumptions. 6. Under what conditions is the yardstick approach to preparing pro forma financials likely to produce accurate statements? When is fundamental analysis a preferable method? 7. Describe three different approaches to making the balance sheet balance when building an integrated financial model using spreadsheets.
320
Chapter Eight
8. Explain the concept of depreciation as a “noncash” expense. Does this mean it has no impact at all on a venture’s cash flows? 9. What can cause the retained earnings account to change from one period to the next? 10. What is the accounting impact of negative shareholders’ equity? How can a venture with negative book equity continue to operate and attract capital?
Notes
1. Net working capital is usually defined as all current assets less all current liabilities. However, some current assets, such as marketable securities, may not be central to operation of the venture, and some current liabilities, such as notes payable, which are related to financing, are not considered part of the operating activities of the business. 2. See Ng, Smith, and Smith (1999) for supporting evidence. 3. Smith (1987) describes credit terms with significant discounts for prompt payment as a device the seller can use to gain timely information about the financial health of its customers. 4. See http://w ww.rmahq.org/annual-statement-studies/ for more information. 5. When developing assumptions based on comparables or industry statistics, medians can be more meaningful than means because medians are not affected by extreme outlier values. 6. To compute days in inventory, divide 365 by the inventory turnover ratio. 7. The simulation results discussed here are not shown, but the model is available online for @Risk and Crystal Ball users.
References and Additional Reading
Mian, S. L., and C. W. Smith Jr. 1992. “Accounts Receivable Management Policy: Theory and Evidence.” Journal of Finance 47: 169–200. Ng, C., J. K. Smith, and R. L. Smith. 1999. “Evidence on the Determinants of Credit Terms Used in Interfirm Trade.” Journal of Finance 54: 1109–29. Smith, J. K. 1987. “Trade Credit and Informational Asymmetry.” Journal of Finance 42: 863–72.
C h a p t e r Nine
A ss e ssi n g Ca s h N e e ds
T h e o b j ec t i v e o f t h i s c h a p t e r is to enable the entrepreneur to
answer the question, “How much money do I need and when do I need it?” For this, we build on the forecasting tools from Chapters 7 and 8 and focus on methods of assessing financial needs. An entrepreneur must have a good sense of how much cash will be required to carry the venture to the point where it becomes self-sustaining or capable of attracting additional funding, as well as a good sense of when cash infusions are likely to be needed. An entrepreneur who does not anticipate the cash needs over the life of the venture assumes unnecessary risk. The venture may fail, not because the idea is bad but because the entrepreneur did not anticipate the cash need far enough in advance to do anything about it. An entrepreneur’s failure to anticipate the need for financing can also result in an adverse negotiating position with investors—either because the need is urgent or because the original financing agreement impedes the ability to raise cash in the future. “Do not run out of cash” is a common admonition to entrepreneurs. But having too much cash can also be problematic. An entrepreneur who is overly cautious may find that raising “enough” cash up front is not feasible. Even if substantial early-stage financing can be arranged, it may come at a high price, and the entrepreneur may be compelled to give up more of the venture than is necessary or desirable. Although a venture cannot survive without cash, the objective is not merely survival; rather, it is to finance the venture in a way that yields the highest expected value for the parties. As a general principle, an entrepreneur who is more confident of success than are potential investors can benefit by raising only enough cash to carry the venture to the next milestone. At that point, the lower risk of failure will be more 321
322
Chapter Nine
apparent to investors. Reduced risk can foster competition among investors and should enable the entrepreneur to raise capital on more favorable terms than in the earlier round. There is, of course, a trade-off. Raising less cash in an early round increases the probability that the venture will run out of cash before the next milestone. Running short of cash before reaching an important milestone can suggest to investors that the venture is not on track for success. A cash flow breakeven analysis is one tool for assessing financial needs. It can be used to identify the cash flow breakeven point (BEP) and to estimate cash needs for a firm that is operating below its BEP. The BEP is the level of sales at which a venture would be able to maintain operations without additional funding, although the venture still could need funding for growth. Hence, combining cash flow breakeven analysis with projections of sales growth can help the entrepreneur assess the amount of investment a venture would need in order to achieve a level of sales sufficient to maintain its operations. The sustainable growth model is another useful tool. The model seeks to identify the conditions under which the growth of a venture can be sustained solely by the initial investment. We refer to this as the “sustainable growth rate.”1 While the model is intended more for established businesses, it can also be helpful for estimating the financing needs of early-stage ventures that are expected to grow. If the entrepreneur anticipates or aspires to a growth rate that is higher than the sustainable growth rate, additional financing is required and alternatives for adding investment capital must be evaluated. Later in the chapter, we use scenario analysis and simulation to focus on how uncertainty affects the need for financing. These methods can be used to assess how cash needs may be affected by uncertainty about development timing, sales levels, fixed and variable costs, and other factors. Assessing financial need only on the basis of expected performance can expose the entrepreneur to avoidable risk of venture failure or loss of control. When a venture’s prospects are uncertain, staging the investment around milestones can be of great value, but uncertainty of performance also makes the funding need in a given financing round uncertain. In the last part of the chapter, we use simulation to fine-tune the funding decision in the context of staged financing.
9.1 Cash Flow Breakeven Analysis Much is written about breakeven analysis in accounting and finance textbooks. Not all of it is flattering, primarily because the accounting net income approach to breakeven analysis ignores the time value of money and focuses
Assessing Cash Needs 323
on accounting net income rather than cash flow. In the accounting net income approach, the BEP is the quantity of sales or amount of revenue where the total contribution margin over all units sold equals total fixed cost. The contribution margin is the difference between price and variable cost. So, for example, if the price is $20 and variable cost is constant at $16 per unit, the contribution margin is $4. With total fixed cost of $175,000, the breakeven point would be 43,750 units, or $875,000 in revenue. The accounting net income approach can make sense for long-run analysis if it is expected that depreciation expense will be offset by the new investment necessary to maintain the capital stock. But the approach does not work well if one desires to assess the short-run BEP or if the investment needed to maintain the fixed assets is systematically different from depreciation expense. Moreover, even as a long-run concept, because accounting approaches do not factor in the opportunity cost of capital, they are not well suited for investment decision making. For assessing financial needs, cash flow breakeven analysis can provide insight. It addresses the question, “What level of sales generates operating cash inflows that are sufficient to cover operating cash outflows?” The cash flow BEP is where the venture achieves a level of sales high enough to maintain its operations at the current level. At the cash flow BEP, cash inflows from operations are sufficient to maintain and replace current assets but not to fund growth. This is the minimum level of revenue the venture needs to survive without additional funding. In conjunction with a forecast of revenue, finding the cash flow BEP can help the entrepreneur assess initial financing needs. Once a breakeven model is constructed, the entrepreneur can use it to estimate how initial cash needs depend on sales levels, sales growth, product prices, fixed costs, variable costs, and noncash revenues and expenses. Breakeven analysis can also be used to conduct a variety of “what ifs” or sensitivity analyses.
An Illustration of Net Income and Cash Flow BEPs Consider an entrepreneur, Trinity Matrix, who has developed a low-cost virtual reality (VR) headset that she plans to market on Amazon.com for $10 per unit. Before she can do so, she needs to invest in a production facility. The fixed investment in the facility is expected to be $3 million. The investment can be depreciated over five years, straight line (i.e., $600,000 per year in depreciation expense). After that point, a fixed annual expenditure of $100,000 will be sufficient to maintain the facility. She will also need to make an investment in net working capital, expected to be 10% of annual sales. Variable cost,
Panel A Income Statement Unit sales volume Unit price
0
1 50,000 $10
2 250,000 $10
3 250,00 0 $1 0
4 62,50 0 $1 0
5 62,50 0 $1 0
6 62,500 $10
7 62,500 $10
$500,000 $350,000 $600,000 $950,000 $125,000 $1,075,000 ($575,000)
$2,500,000 $350,000 $600,000 $950,000 $625,000 $1,575,000 $925,000
$2,500,00 0 $350,00 0 $600,00 0 $950,00 0 $625,00 0 $1,575,00 0 $925,000
$625,00 0 $350,00 0 $600,00 0 $950,00 0 $156,25 0 $1,106,25 0 ($481,250)
$625,00 0 $350,00 0 $600,00 0 $950,00 0 $156,25 0 $1,106,25 0 ($481,250)
$625,000 $450,000 $0 $450,000 $156,250 $606,250 $18,750
$625,000 $450,000 $0 $450,000 $156,250 $606, 250 $18,750
($575,000)
$350,000
$1,275,000
$793,750
$312,500
$331,250
$350,000
0
1 $7.50 $950,000 126,667 $1,266,667
2 $7.50 $950,000 126,667 $1,266,667
3 $7.50 $950,000 126,667 $1,266,667
4 $7.50 $950,000 126,667 $1,266,667
5 $7.50 $950,000 126,667 $1,266,667
6 $7.50 $450,000 60,000 $600,000
7 $7.50 $450,000 60,000 $600,000
0
1 $375,000 $25,000 $50,000 ($25,000) ($3,025,000)
2 $1,875,000 $1,525,000 $200,000 $1,325,000 ($1,700,000)
3 $1,875,000 $1,525,000 $0 $1,525,000 ($175,000)
4 $468,750 $118,750 ($187,500) $306,250 $131,250
5 $468,750 $118,750 $0 $118,750 $250,000
6 $468,750 $18,750 $0 $18,750 $268,750
7 $468,750 $18,750 $0 $18,750 $287,500
1 $7.50 $350,000 46,667 $466,667
2 $7.50 $350,000 46,667 $466,667
3 $7.50 $350,000 46,667 $466,667
4 $7.50 $350,000 46,667 $466,667
5 $7.50 $350,000 46,667 $466,66 7
6 $7.50 $450,000 60,000 $600,000
7 $7.50 $450,000 60,000 $600,000
$1,325,000 $1,095,041
$1,525,000 $1,145,755
$306,250 $209,173
$118,750 $73,734
$18,750 $10,584
$18,750 $96,217
Sales Fixed cash expenses Depreciation expense Total fixed expenses Variable costs Total expenses Net income Cumulative net income Panel B Net Income BEP Contribution margin per unit Total fixed costs Unit BEP Dollar BEP Panel C Cash Needs Assessment Total contribution to fixed costs Contribution after fixed cash expenses Change in net working capital Net cash flow Cumulative cash need Panel D Cash Flow BEP Contribution margin per unit Cash fixed costs Unit BEP Dollar BEP Panel E NPV Investment Assessment Net cash flow for period PV of cash flow (10%) Net present value
($3,000,000) 0
($3,000,000) ($3,000,000) ($392,223)
($25,000) ($22,727)
Fi g u r e 9.1 Net income and cash flow break-even points, cash needs assessment, and investment value The figure shows annual net income, net income and cash flow break-even points, annual net cash flows, cash flow BEPs, cumulative cash needs, and a valuation of the virtual reality venture.
Assessing Cash Needs 325
including production, marketing, and distribution, is estimated to be 25% of sales. During the first five years, the cash portion of annual fixed costs (i.e., excluding depreciation) is expected to total $350,000. The entrepreneur projects that the venture can achieve sales volume of 50,000 units in Year 1 and 250,000 units in each of Years 2 and 3. After this demand peak, annual sales volume will stabilize at around 62,500 units. Before deciding to move forward with the project, Ms. Matrix would like to develop a better understanding of the cash needs of the venture. We begin the analysis by constructing annual income statements for the venture and using the cost information to compute the net income BEP for each year. Panel A of Figure 9.1 shows the annual income statements. Net income is negative in the low-sales-volume years during the time when the initial investment is being depreciated but is positive when sales volume is high and after the investment is fully depreciated. Cumulative net income over the period of the explicit forecast is $350,000, suggesting that the project may be worth pursuing. Panel B shows the net income BEP for each year. During the first five years, because of depreciation expense, the net income BEP is high, at $1.267 million over this period. Starting in Year 6, when depreciation is zero but fixed costs to maintain the facility increase by $100,000, the net income BEP falls to $600,000. However, the analysis in Panels A and B is misleading because it is not focused on cash flow. Note that changes in net working capital that are driven by sales growth affect the cash flow BEP but do not affect the net income BEP.
Using Breakeven Analysis to Project Financial Needs Using breakeven analysis to assess financial needs requires another step. Financial need depends on two things; one is the time until breakeven is reached and the other is the amount of financing required to cover the shortfalls until that time. Thus, to estimate the amount of financing required for the VR venture, we can combine breakeven analysis with a sales forecast. We can use the cash flow BEPs in conjunction with the sales forecast to estimate cash needs. In Panel C, we use the financial information to project annual net cash flow, beginning with the initial $3 million investment. The first row of Panel C is computed as unit sales volume times the $7.50 contribution margin per unit (75% of the $10 price). In the second line, cash fixed expenses from Panel A are deducted. The third line is computed from balance sheet information on working capital as 10% of the annual change in sales revenue. In years when revenue increases, an additional investment in working capital is required; in years when revenue declines, the decline in net working capital is a source of cash. In the fourth line, the change in working capital is subtracted from the
326
Chapter Nine
contribution after fixed expenses to get annual net cash flow. The final line in the panel shows cumulative net working capital. The cumulative total is positive but below cumulative net income. In contrast to net income, net cash flow is positive in all years after the first. Also, net cash flow is higher than net income in Years 1 through 5 because depreciation is noncash and cash flow is equal to net income in Years 6 and 7 because all expenses are cash expenses. We can see from Panel C that beyond the initial investment at Year 0, the venture is expected to need another $25,000 in Year 1, after which net cash flow is expected to be positive. Panel D shows the cash flow BEP annually. This BEP of $467,000 in Years 1 through 5 is well below the net income BEP, and, at $600,000, is equal to the net income BEP in Years 6 and 7. One problem with the cash flow BEP analysis is that it is static in that necessary changes in investment are ignored. That is, the static BEPs in Panel D do not take account of the necessary changes in net working capital. The net cash flow projections in Panel C do reflect working capital changes. The analysis in Panels A through D does not tell us whether pursuing the venture would be a good idea (even though cumulative net income and cumulative net cash flow through Year 7 are both positive). For that, we would need an analysis of net present value. Panel E shows a simple investment analysis, discounting each year’s free cash flow (FCF) at an assumed rate of 10%, with theYear 7 cash flow being capitalized as a perpetuity.2
Present Value Breakeven Analysis Cash flow breakeven analysis enables the entrepreneur to get a better feel for the venture by providing a way to assess how sales levels, prices, and fixed and variable costs affect cash needs. However, the analysis does not contribute much insight to the investment decision since it does not use the PV concept or risk assessments that are central to investment decision making. A modified form of breakeven analysis can determine the level of sales where the PV of revenues is sufficient to cover the PV of cash outflows (both cash expenses and investment outlays). Present value breakeven analysis helps answer the question, “What level of sales is needed to justify investing?” It is best suited for projects where revenue and expense streams can be described as level annuities. The PV approach is most helpful for capital budgeting and investment decisions. It is unlikely to be of much help for assessing cash needs. For example, we could ask what level of perpetual free cash flow (FCF) would justify an investment of $3 million if the cost of capital is 10%. The answer
Assessing Cash Needs 327
is $300,000 (because $300,000/0.10 = $3 million). FCF is cash flow above the amount needed to fund expected growth and to deal with uncertainty. In this case, if cash fixed costs are $500,000 and the contribution margin is 75%, the annual revenue level to achieve the cash flow breakeven point is $800,000/.75 = $1.067 million. That is, the contribution from $1.067 million of sales is $800,000, of which $500,000 covers fixed cost, and the balance of $300,000 is the FCF available to investors.
9.2 Sustainable Growth The breakeven approach to estimating cash needs is based on the assumption that the investment in fixed assets does not vary over a relevant range of sales levels. An alternative approach is to assume that the investment in fixed assets must increase in proportion to sales. With that as a starting point, we can use the sustainable growth model to estimate the level of resources that can be generated through the continuing operations of the venture and, if a higher rate of growth is projected, the additional investment that would be required. The model, while it is designed for established businesses that are expected to grow in fixed proportion with constant profitability over a relevant range of asset levels, can be modified to estimate cash needs of an asset-driven venture, where sales are expected to grow in proportion to assets, but profitability is subject to economies of scale. Sustainable growth, in the standard application, starts with the assumption that as the venture grows, assets, debt, equity, sales, and net income all grow in fixed proportion to sales.3 This means the sustainable growth rate for an established enterprise depends on four factors: 1. Asset turnover (“turnover”)—the amount of sales revenue that can be supported per dollar of assets, including fixed assets and net working capital 2. Financial leverage (“leverage”)—the ratio of the venture assets to its equity, where the difference between assets and equity represents debt financing 3. Return on sales (“ROS”)—the profitability of sales in terms of after-tax net income per dollar of sales 4. Dividend policy (“retention”)—the fraction of each dollar of net income that is retained in the venture as opposed to being paid out as dividends
328
Chapter Nine
An Established Business Example Figure 9.2 illustrates the concept of sustainable growth, denoted as g*, and incorporates the following assumptions: Factor
Definition and value
Asset turnover Financial leverage Return on sales (ROS) Dividend retention (R)
Sales / total assets = 3.0 Total assets / equity = 1.5 Net income / sales = 10% Fraction of net income retained = 2/3
Suppose an entrepreneur makes an initial investment of $100 in the form of equity and wishes to calculate the rate of growth this can sustain. The leverage ratio of 1.5 (assumed to be appropriate for the venture) implies that the venture has $1.50 of assets for each $1 of equity. The difference of $50 is debt financing. These amounts are reflected in the starting balance sheet shown in Figure 9.2. Reinforcing the notion of statement integration from Chapter 8, the turnover ratio in the figure shows that each $1 of assets (in the balance sheet) is expected to support $3 of sales (in the income statement). This results in sustainable sales in the first year of $450. The assumed 10% return on sales (shown in the income statement) implies Year 1 net income of $45. In the figure, we assume that the venture has adopted a policy of retaining two thirds of the net income
Starting Balance Sheet
Year 1 Income Statement
Ending Balance Sheet g* = 30.0%
Initial equity investment
Sales
$100
$450
Initial equity + retained earnings $130
×
×
×
Leverage 1.5
Turnover 3
= Initial total assets $150 E = $100 D = $50
ROS 10%
= Net income $45
Retention 66.7% $30
Leverage 1.5
Dividend payout 33.3% $15
Ending total assets $195 E = $130 D = $65
Year 2 sales $585
=
Fi g u r e 9. 2 Sustainable growth model template This figure is an Excel template that illustrates the key variables, relationships, and results in the sustainable growth model. Assumptions of the model can be changed in the template to assess their impact on g* and the levels of assets, debt, and sales.
Assessing Cash Needs 329
and distributing one third as a dividend; hence, $30 goes to retained earnings in the balance sheet, increasing equity to $130. Because the leverage ratio remains constant at 1.5, $130 of equity supports $195 of assets at the beginning of the second year. This level of assets, based on the turnover ratio of 3.0, will support an increase in sales to $585. This 30% increase in sales is the sustainable growth rate of the venture, given its leverage and dividend policies, its profitability, and the efficiency of use of its assets.4
Financial Policy Choices Figure 9.2 shows that the sustainable growth rate, g*, is equal to the percentage change in equity. That is: g* = ΔE/EBeginning (9.1) where E is equity and ΔE is the change in equity due to retaining earnings. Because the sustainable growth model excludes the possibility of issuing new equity, the change in equity comes only from net additions to retained earnings; that is, ΔE = NI − Div = NI × R (9.2) where NI is net income, Div is the dividend, and R is the retention ratio. For a venture seeking the maximum sustainable growth rate, R would be 100%. Of course, if income is negative, as it is for many new ventures, then so is the sustainable growth rate, and the venture will depend on new investment of debt or equity, both to sustain its current size and to grow. Substituting Eq. (9.2) into Eq. (9.1) gives: g* = ΔE/EBeginning = (NI × R)/EBeginning = NI/EBeginning × R = ROE × R (9.3) where ROE is return on beginning equity, a simple measure of the venture’s profitability. We can separate ROE into three components as follows: ROE = NI/E = NI/S × S/A × A/E = ROS × Turnover × Leverage (9.4) where S is sales and A is total assets. Substituting Eq. (9.4) into Eq. (9.3) yields g* = ROE × R = ROS × Turnover × Leverage × R (9.5) Thus, the venture’s sustainable growth rate is the product of (1) its profit margin or return on sales (NI/S or ROS), (2) the asset turnover ratio (S/A or turnover), (3) financial leverage (A/EBeginning or leverage), and (4) the retention ratio or dividend payout policy (R). The equation reflects the assumption that sales, net income, assets, and debt all increase proportionately as the venture grows.
330
Chapter Nine
Equation (9.5) suggests that the sustainable growth rate of a venture can be increased in several ways. The entrepreneur can try to improve the profit margin on sales, generate more sales from its asset base, or rely more heavily on financial leverage. If the venture is already operating efficiently, then increasing profitability and asset turnover are not feasible. This leaves leverage and the retention ratio as the policy choices that can influence g*.
Interest Expense, Taxes, and Economies of Scale Because of the tax deductibility of interest payments, net income is not independent of the leverage policy. To see the interdependence, as well as the tax shelter, we restate net income as NI = EBIT − I − T = [EBIT − r(A − E)](1 − t) (9.6) where EBIT is earnings before interest and taxes, I is interest expense, T is taxes, A − E is assets minus equity (the amount of debt financing), r is the interest rate on debt, and t is the corporate tax rate. The term in square brackets is simply income before taxes. Substituting Eq. (9.6) into Eq. (9.5) yields g* = [EBIT − r(A − E)](1 − t)/S × Turnover × Leverage × R (9.7) Equation (9.7) can be used to see how the sustainable growth rate varies in response to leverage (defined as assets/equity) and payout policy choices. Further, to adapt the model to early-stage ventures, we can express EBIT as a function of sales. Suppose that the venture is subject to economies of scale. If the relationship is linear, EBIT can be expressed as EBIT = a + b × S, where a and b are parameters reflecting the scale economies of the venture, so that Eq. (9.7) can be expanded as follows: g* = [(a + b × S) − r(A − E)](1 − t)/S × Turnover × Leverage × R (9.8) The explicit incorporation of scale economies makes the sustainable growth rate dynamic. As the venture grows, it becomes more profitable and internally generated funds can support a higher rate of growth. Realistically, at lower levels of sales, the venture would be more dependent on external financing to support a targeted growth rate.
An Early-Stage Venture Example Gill Bates is considering a venture to develop and support an online virtual world, iFree. Bates is prepared to make an initial equity investment of $500,000. Consistent with what is normal for early-stage ventures, he plans to use no debt financing. For strategic reasons, he believes the iFree venture must
Assessing Cash Needs 331
reach sales of $2.0 million by the start of the sixth year and seeks to grow the venture at a constant rate of 14.87% per year to achieve this. Bates is willing to retain all earnings in the venture to reach this goal. In the iFree business model, we assume, reflecting the scale economies of the venture, an operating margin (EBIT/S) of –$200,000 + 20% of sales. That is, the venture has $200,000 in fixed costs and variable cost of 80% of sales, resulting in a 20% contribution margin. In addition, the asset turnover ratio is expected to be 2.0, and the corporate tax rate is 35%. Bates’s decision to avoid debt means the leverage ratio (A/E) is initially 1.0. We can quickly establish that the initial investment of $500,000 will not be sufficient to achieve the Year 6 sales target. The equity investment of $500,000 will support $1 million in sales based on the turnover ratio (S/A) of 2.0. The projected EBIT at that level of sales is $0 (i.e., 20% of sales, less the base level loss of $200,000). At this level, the venture will just break even on profitability, so the targeted 16.5% growth of assets would need to come from an additional investment of equity of $82,500 (i.e., 16.5% of the initial $500,000 investment). With no debt, assets and equity are equal. We can use Eq. (9.8) to determine the additional equity investment needed each year for iFree to achieve the growth rate that will enable it to achieve the sales target of $2.0 million by the sixth year. Figure 9.3 shows iFree’s projected assessment of cash needs over the first six years. In the first year, the targeted growth rate is achieved entirely from new outside equity. After the first year, because of scale economies, profitability increases and the need for additional outside equity declines. The lower panels in the figure show key assumptions in the model, the sustainable growth rate
Year 1 2 3 4 5 6
Beginning Equity $500,000 $574,347 $659,748 $757,848 $870,534 $999,977
Assets $500,000 $574,347 $659,748 $757,848 $870,534 $999,977
Assumptions Assets/equity Sales/assets EBIT intercept EBIT slope Interest rate Tax rate Dividend payout pct. Target growth rate
Sales
EBIT
$1,000,000 $0 $1,148,693 $29,739 $1,319,496 $63,899 $1,515,695 $103,139 $1,741,069 $148,214 $1,999,954 $199,991
1.0 2.0 –200,000 20.00% 10.00% 35.00% 0 .0 0 % 14.87%
Interest $0 $0 $0 $0 $0 $0
Taxable Income $0 $29,739 $63,899 $103,139 $148,214 $199,991
Year 1 2 3 4 5 6
Tax $0 $10,409 $22,365 $36,099 $51,875 $69,997
Retained Net Dividend Earnings Income $0 $0 $0 $0 $19,330 $19, 330 $0 $41,534 $41,534 $0 $67,040 $67,040 $0 $96,339 $96,339 $0 $129,994 $129,994
Sustainable Growth Rate 0.00% 3.37% 6.30% 8.85% 11.07% 13.00%
Ending Equity $574,347 $659,748 $757,848 $870,534 $999,977 $1,148,666
New Equity $74,347 $66,071 $56,565 $45,646 $33,103 $18,696
Entrepreneur Ownership Value Percent $500,000 100.00% $516,828 89.99% $549,365 83.27% $597,962 78.90% $664,137 76.29% $750,473 7 5 .0 5 %
Fi g u r e 9. 3 Dynamic sustainable growth for an early-stage venture The figure shows the relation of sales level to EBIT, with implications for the sustainable growth rate, the amount of outside financing required to achieve a targeted growth rate, and the fraction of entrepreneur ownership.
332
Chapter Nine
each year, and the cumulative percentage of book valued equity held by the entrepreneur. Given the assumptions, the sustainable growth rate asymptotically approaches 20% as the venture grows and the entrepreneur’s ownership fraction gradually declines but still remains above 75% in the sixth year. Without knowing the market value of the venture at each year, we cannot determine the fraction of equity Bates would have to give up in exchange for the outside equity. That will depend on the value the investor places on the venture at each point. Nor can we determine the value of the remaining equity to the entrepreneur. The sustainable growth model, even with modifications for interest expense and taxes, is still only an approximation approach. In the example, the venture never has negative income, so we do not confront the question of how taxes would be affected by negative operating income. Also, the model is focused on earnings and not on cash flow. If the assets are depreciable, net income will understate the cash flow available to support growth. In that case, the amount of external financing required would be less than what is estimated in the model. To address those concerns, a fully integrated financial model would do a better job.
9.3 Planning for Financial Needs When the Desired Growth Rate Exceeds the Sustainable Rate Product-market growth that is either too rapid or too slow is problematic. Growth that is too slow threatens venture survival by encouraging competition or hastening technological obsolescence. As suggested by the prior example, growth that is too rapid threatens the entrepreneur’s control and the value of the entrepreneur’s ownership share. Long-run survival depends on achieving a level of sales that is sufficient for financial viability. In some product markets, survival and profitability depend on rapidly attaining a substantial market position. Computer software and some Internet ventures are good examples of markets in which long-run survival can depend on the ability to achieve a critical mass of users quickly. In the case of software, this is because many software packages benefit from important “network externalities.” Demand for the product increases as more users affiliate with the network. Because of these network externalities, a software manufacturer that does not quickly achieve substantial market share may be driven from the market, even if the product is good. There are also significant switching costs to software users, making them more likely to adopt and continue using the market leading product.
Assessing Cash Needs 333
In some cases, network externalities are sufficiently important that a dominant product can hold competitors at bay, even when products that are technically superior are available at lower prices. In others, rapid growth is not essential for survival but still may be an element of the entrepreneur’s aspiration. Without careful analysis, it is easy to equate rapid growth with financial success. But as the sustainable growth model implies, this is not necessarily correct. There are countless examples of rapid growth ultimately destroying a venture. Webvan was one of the companies that launched in the mid-1990s to offer home delivery of groceries ordered over the Internet. Webvan’s management team pursued a “get big fast” strategy, launching distribution in many communities at the same time or in rapid succession. Webvan raised almost $800 million from private and public equity sales in 1999 and had a market capitalization of $8 billion immediately after its IPO. The company entered into a $1 billion contract with Bechtel for construction of 26 state-of-the-art distribution centers across the country.5 The firm was betting that a significant fraction of the population in each market would switch to ordering groceries on the Internet. Unfortunately, consumer adoption was slower than Webvan had hoped, so that profitable operation was never achieved. The company filed for bankruptcy in July 2001. Contrast Webvan with FreshDirect, another Internet-based home-delivery grocery business. FreshDirect sells in only one market, the more densely populated and wealthier areas of New York City and surroundings, and has declined to move into other, less dense, markets. FreshDirect introduced its service to NYC consumers in 2002 and has chosen to remain a privately held company. It reportedly is a highly profitable business that has succeeded on the basis of its focused strategy.6 There are other examples where the venture survives but high growth results in loss of control for the entrepreneur. The housing and financial system crisis that began in 2007 provides many good examples. Developers, who hoped to capitalize on the surging demand for housing, dramatically overbuilt in a number of markets. When demand slowed, many of these developers failed even though they held valuable inventories of fully or partially constructed housing. The developers lost their equity investments, and property ownership reverted to those who had provided credit for the development. Commercial and investment banks encountered a similar problem. At the time, these companies were operating with very high financial leverage. When the market values of their assets declined, institutions such as Bear Stearns, Lehman Brothers, Merrill Lynch, Washington Mutual, and Wachovia were compelled to write down the values of their housing-related assets. In some cases the drop was sufficient to wipe out the existing stockholders.
334
Chapter Nine
Ironically, during this period regulation that was intended to protect those who traded with the institutions dramatically impeded the institutions’ ability to raise additional capital, even though the low values of some of their housingrelated assets were likely to be transitory. Institutions such as Goldman Sachs that barely managed to skirt bankruptcy or government takeover have now recovered most of their losses. From a high of $236 in October 2007, Goldman Sachs’s stock price declined by 77.5% to $53 in November 2008. By October 2009, the price was back up to $189; Goldman’s investors had recovered most of their loss. By year-end 2016, it was trading at slightly above its 2007 high and had paid dividends throughout the entire period. In contrast, the shareholders of Lehman Brothers, which was slightly more aggressive than Goldman in taking risks related to the housing market, lost everything. The point is that not providing a financing solution to cover unexpected bad outcomes can result in permanent loss for the stockholder (or the entrepreneur) even if the lack of financing is temporary.
9.4 Planning for Product-Market Uncertainty Financial planning prompts the entrepreneur to assess the benefits and threats of high-risk ventures. Financing considerations and implications for value must be assessed before committing to a strategy. Financing considerations can lead the entrepreneur to reject what may appear to be the best productmarket strategy in favor of one that is less ambitious but more valuable. For iFree, the conjectural product-market strategy implied a sales goal of $2.0 million by the sixth year. Given the amount Gill Bates was able to invest, the strategy could not be achieved without outside financing. Whether it would be advantageous for the entrepreneur to pursue outside financing or scale back the Year 6 sales target depends on which approach would maximize value for the entrepreneur, a question we take up in the next few chapters.7 For now our focus is on how uncertainty of future product-market performance affects the amount of “financial slack” the entrepreneur might want to provide. Financial slack is liquidity that would enable the venture to deal with surprises without the need to raise additional risk capital. Financial slack is available in various forms, including, for example, cash flow from operations, excess cash, or an unused line of credit. Lehman Brothers failed because declining asset values wiped out its financial slack. Any entrepreneur of a risky venture faces a similar prospect of short-run financial adversity for an otherwise healthy venture.
Assessing Cash Needs 335
Planning for Success Ironically, unexpected success in the product market can be a threat to new venture survival and control. Consider, for example, VC funds. Kaplan and Schoar (2005) find that managers of successful funds are able to attract capital for the next fund more easily than are their competitors. They also, however, find that persistence of performance from one fund to the next is inversely related to the growth of funds under management. They interpret their findings as evidence that fund managers who try to grow too rapidly after an early success sacrifice performance, possibly because they are unable to grow their management teams fast enough or because they have limited ability to identify attractive investment opportunities. The same can be expected for any entrepreneur who is confronted with rapidly growing demand. The entrepreneur may face such challenges as being unable to (1) grow the management team rapidly enough, (2) acquire the assets necessary for production, (3) maintain consistent quality, or (4) arrange financing to support the rate of demand growth. The last point is the focus of this chapter. For assessing long-run financial needs, it is important to account for the risk of unexpected success in the product market. If a venture is profitable and is generating cash in excess of capital replacement requirements, it can finance some growth internally. This is the lesson of the sustainable growth model. Excess operating cash flow is an important source of investment capital. For a venture that is debt-free and pays no dividends, the ability to finance growth internally is approximately equal to the venture’s after-tax rate of return on beginning assets. This was demonstrated in Eq. (9.7). If the return on assets (ROA) is, for example, 8%, then the venture is capable of growing assets and revenue at a rate of approximately 8% by relying exclusively on internally generated funds. A higher growth rate would require external funding. Conversely, if the ROA exceeds the growth rate of sales, the organization generates FCF. FCF that is retained by the venture does not contribute to its value but can be retained in an interest-bearing account so that retaining surplus cash does not diminish PV. To enable investors to make the best use of their resources, a venture that generates cash flow and has no strategic reason for holding cash reserves should distribute the surplus funds. Dividends, share repurchase, and reducing leverage are all means of distributing FCF.8 Unexpected product-market success compels the entrepreneur to explore alternatives for outside financing. Arranging new financing, however, takes time, and the need to devote attention to seeking financing comes at a critical point for the entrepreneur, a point when rapid growth is likely to generate a
336
Chapter Nine
host of organizational challenges. With effective planning, the entrepreneur can prepare in advance for scenarios involving growth that is faster than expected. In structuring initial financing the entrepreneur must anticipate that growth will require additional financing and have a sense of how the necessary capital would be raised. In negotiating the investment terms for the initial round and other early rounds, the entrepreneur will want to preserve the option of raising additional funds if the growth rate justifies doing so. We have already seen in Chapter 4 that current financing decisions sometimes cause problems if the venture needs more capital at a later stage. Loan contracts and preferred stock, for example, sometimes contain covenants or other provisions that preclude raising additional funds without the lender’s or preferred stockholder’s approval. If that time comes, the lender’s interests may conflict with the entrepreneur’s. Some equity financing structures can give rise to similar difficulties. Antidilution provisions sought by equity investors can impair a venture’s ability to raise capital, particularly where the venture has run into problems and needs cash to get through them. Agreeing to these constraints on future fund-raising can lower the apparent cost of financing to the entrepreneur. But if the venture grows rapidly or runs into difficulties, those earlier decisions can be problematic. Financial modeling of the venture, including modeling of the financing terms, is a means of identifying, and designing around, these threats before they become problems.
Planning for Failure Once a venture is established and is generating positive cash flows, unexpectedly slow growth may not pose a very serious problem. If the actual growth rate is lower than expected but still high enough to assure viability, the venture may actually be able to reduce its reliance on external capital. If g* is greater than actual growth, then the resulting FCF can be used to fund new investments or can be distributed to investors. A more serious problem arises if unexpectedly poor performance is encountered before the organization has reached financial viability. This may occur, for example, if product development takes longer than expected or if sales growth is slower than expected and the attained level of sales is below that needed to generate funds for capital replacement. Either circumstance means the organization must depend on external financing to a greater extent than expected. This is a particularly acute problem because a venture that has not achieved viability and has not met growth expectations will have difficulty raising capital. A forwardlooking entrepreneur can manage the risk by maintaining financial slack and preserving the ability to raise additional capital if growth is slower than expected.
Assessing Cash Needs 337
High-Tech, High-Growth Innovation Some of the most challenging financing problems are associated with a product innovation that requires a long and expensive development period, which, if successful, is followed by rapid sales growth. Negative cash flows, in such cases, can extend over many years and can last through both the development and rapid-growth stages. The venture will not begin to generate FCF for investors until the growth rate slows to a point where operating cash flows are more than sufficient to fund growth. NewCo has these characteristics and we extend the model later in this chapter. The pattern of a long development lead time followed by rapid growth is characteristic of many high-tech innovations (pharmaceuticals, biotechnology, and some electronics, for example). It is not surprising, therefore, that large, well-established companies undertake much of the development activity. Such companies can draw on their current financing capabilities without the need to convince investors of the merits of a specific project. They often also have existing infrastructure that reduces the investment required during development and can handle the operational demands of rapid growth.
9.5 Assessing Financial Needs with Sensitivity/Scenario Analysis Because the future is unknown, financing decisions should depend on both the expected future of the venture and uncertainty. Methods of assessing financial needs under uncertainty are particularly valuable. The financial information for iFree in Figure 9.3 is based on the “expected” scenario. But what happens if the venture is more or less successful than expected? Sensitivity analysis is a simple way to construct scenarios that incorporate uncertainty into projections of financial need. A scenario describes one possible version of the future and specifies the key assumptions associated with the scenario. An entrepreneur can use alternative scenarios, together with their associated assumptions, to generate alternative projections of financial needs. One might, for example, consider “expected,” “best case,” and “worst case” scenarios, each with its own assumptions and resultant estimate of financing need. For ventures with long product development periods, it can be helpful to consider one scenario in which product introduction is delayed and sales grow slowly and are not initially profitable, and a second in which sales growth is rapid and more profitable but product development is still slow.
338
Chapter Nine
The key to forecasting long-run financial needs is developing a model that links product-market performance to financing requirements. We saw how this can be done in the NewCo model in Chapter 8, as well as through the cash flow breakeven analysis and sustainable growth model in this chapter. The first step to forecasting financial needs is to specify a set of assumptions about productmarket performance: for example, expected sales, costs, and resulting profitability. Incorporating the assumptions into a financial model will determine the cash flows we can anticipate the venture to generate. We can then forecast financing needs by using estimates of asset efficiency, such as the levels of fixed assets and working capital needed to support projected sales. Scenario analysis can be developed in much the same way as a decision tree, but with nature making the decisions. For example, Figure 9.4 describes the use of sensitivity analysis to construct 27 scenarios of sales growth, cost, and asset turnover for iFree. The key assumptions in the model are the rate of sales growth, the profitability of sales (the variable cost percentage), and the efficiency of asset utilization to generate sales (the sales to asset ratio). We assume that asset turnover could be 15% higher (i.e., 2.3 times) or 10% lower (i.e., 1.8 times) than the expected turnover of 2.0 times, the variable contribution percentage could be 10 percentage points higher (30%) or lower (10%) than the expected level of 20%, and the rate of sales growth could be 5 percentage points higher (20%) or lower (10%) than the expected rate of 15%. The most optimistic scenario would be one where asset turnover is high, profitability is high, and the rate of sales growth is high. This would be consistent with a case where the product is well received so that it can command a high margin and where demand grows rapidly. The most pessimistic is one where the product is not well received so that demand is low, resulting in lower pricing and slower growth. The full set of scenarios generated from the sensitivity analysis is shown in Figure 9.4. Suppose Bates is faced with the issue of how much initial financing to secure and wants to apply scenario analysis to a cash flow breakeven study. Although there are many possible scenarios in the figure, some are more realistic than others. The scenarios differ substantially in the total amount of additional capital required after the initial $500,000 investment. The high-turnover, high-profit, low-growth scenario needs the least total investment, whereas the low-turnover, low-profit, high-growth scenario needs the most—over $1 million in additional investment. The shaded cells in the figure are the points where, for each scenario, the cumulative cash need reaches a peak. The low-growth scenarios generally reach their maximum cash needs sooner and begin to generate FCF. Should Bates plan to provide funding up to the maximum amount shown in the scenarios, or would it be better to commit to a lower level and plan to
Assessing Cash Needs 339
abandon the venture if it appears not to be doing well? To help address this, the figure also shows Year 6 net income. Some high-cash-needs scenarios are profitable by Year 6, but others are not, and some are only slightly profitable. To decide on the appropriate amount to invest, it is useful to estimate the net present value of each scenario. For this, we have assumed that the cost of capital for the venture is 10% and that after Year 6, net income will grow at an annual rate of 4%.9 The NPV column in Figure 9.4 is computed by valuing the growing perpetuity plus the present value of any FCFs generated during the first 6 years, less the present value of any investments. As can be seen in the table, all of the low-profit scenarios have negative NPVs, whereas almost all of the others have positive NPVs. The highest cash-need, positive NPV scenario requires a total investment of about $1.1 million. Most of the negative NPV scenarios require more. In this case, because the main driver of acceptability is the variable contribution to profit percentage, and that percentage is assumed to be constant, a simple decision rule is to invest the initial $500,000 to learn what the profit percentage will be and to abandon any scenario where the low percentage (10%) is realized. Because all of these cases have NPVs that are more negative than the positive $500,000 initial investment, Bates will be better off by abandoning the venture in any of those scenarios. Figure 9.4 provides a simple example of how staging of financing can be used to increase the value of an opportunity. By only investing $500,000 at the beginning, the entrepreneur can learn whether to invest more in subsequent years or abandon the venture. For example, if we assume the scenarios are equally likely and all are pursued fully, the average NPV would be about $714,000. However, if we assume that the low-profitability scenarios are dropped after the first year and none of the initial investment is recoverable, the average NPV would be about $909,000. Staging the investment increases the NPV by $195,000. There are important limitations of using scenario analysis to assess cash needs. Most importantly, scenarios probably are not equally likely so that simply averaging the results over scenarios would not give an accurate estimate. If, for example, the probabilities of low contribution margins are low (high), the gain from using scenarios to identify outcomes that should not be pursued will be less (more) valuable. Also, the assumptions in scenario analysis are discrete. We cannot, for example, use Figure 9.4 to determine whether scenarios with 11% contribution percentages would be worth pursuing. Also, scenario analysis cannot be used reliably to evaluate staging options involving continuous data. Simulation is designed to address both of these problems. The different trials that are produced in a simulation are designed to be equally likely, and the full dispersion of possible results is evaluated instead of just a few discrete scenarios. If the probability distributions of key assumptions in the model are
Equity Book Value (Cumulative Cash Need) S/A
Profit Growth
Time 0
Time 1
Time 2
Time 3
Time 4
Time 5
Time 6
Year 6 NI
NPV
High High High High High High High High High Exp. Exp. Exp. Exp. Exp. Exp. Exp. Exp. Exp. Low Low Low Low Low Low Low Low Low
High High High Exp. Exp. Exp. Low Low Low High High High Exp. Exp. Exp. Low Low Low High High High Exp. Exp. Exp. Low Low Low
$500,000 $500,000 $500,000 $500,000 $500,000 $500,000 $500,000 $500,000 $500,000 $500,000 $500,000 $500,000 $500,000 $500,000 $500,000 $500,000 $500,000 $500,000 $500,000 $500,000 $500,000 $500,000 $500,000 $500,000 $500,000 $500,000 $500,000
$505,750 $480,750 $455,750 $580,500 $555,500 $530,500 $655,250 $630,250 $605,250 $535,000 $510,000 $485,000 $600,000 $575,000 $550,000 $665,000 $640,000 $615,000 $554,500 $529,500 $504,500 $613,000 $588,000 $563,000 $671,500 $646,500 $621,500
$486,650 $439,113 $394,075 $651,100 $599,825 $551,050 $815,550 $760,538 $708,025 $551,000 $502,000 $455,500 $694,000 $641,750 $592,000 $837,000 $781,500 $728,500 $593,900 $543,925 $496,450 $722,600 $669,700 $619,300 $851,300 $795,475 $742,150
$437,730 $371,729 $313,233 $709,820 $631,299 $560,655 $981,910 $890,868 $808,078 $544,200 $473,300 $410,050 $780,800 $699,013 $625,200 $1,017,400 $924,725 $840,350 $615,180 $541,014 $474,595 $828,120 $744,155 $668,230 $1,041,060 $947,296 $861,865
$353,026 $274,739 $211,306 $754,284 $647,994 $558,221 $1,155,542 $1,021,248 $905,135 $510,040 $420,795 $347,055 $858,960 $745,364 $648,720 $1,207,880 $1,069,934 $950,385 $614,716 $518,166 $437,555 $928,744 $810,278 $709,053 $1,242,772 $1,102,391 $980,552
$225,381 $143,700 $86,186 $781,641 $647,693 $542,543 $1,337,900 $1,151,686 $998,899 $443,048 $340,914 $264,761 $926,752 $779,169 $661,592 $1,410,456 $1,217,424 $1,058,424 $588,159 $472,391 $383,810 $1,023,493 $866,820 $740,958 $1,458,826 $1,261,249 $1,098,107
$46,207 ($26,495) ($64,445) $788,469 $627,846 $512,297 $1,530,730 $1,282,188 $1,089,039 $336,658 $229,551 $161,237 $982,102 $798,544 $662,751 $1,627,547 $1,367,537 $1,164,266 $530,291 $400,249 $311,691 $1,111,191
$428,006 $321,047 $231,157 $242,004 $170,698 $110,771 $56,002 $20,349 ($9,614) $355,222 $262,215 $184,049 $193,482 $131,476 $79,366 $31,741 $738 ($25,317) $306,700 $222,993 $152,645 $161,133 $105,329 $58,430 $15,567 ($12,336) ($35,785)
$4,214,861 $3,168,633 $2,276,441 $1,777,566 $1,155,753 $622,881 ($659,729) ($857,127) ($1,030,680) $3,261,137 $2,380,984 $1,629,396 $1,141,750 $630,654 $191,517 ($977,637) ($1,119,677) ($1,246,362) $2,625,321 $1,855,885 $1,198,032 $717,873 $280,587 ($96,059) ($1,189,575) ($1,294,710) ($1,390,150)
High Exp. Low High Exp. Low High Exp. Low High Exp. Low High Exp. Low High Exp. Low High Exp. Low High Exp. Low High Exp. Low
$912,343
$763,054 $1,692,092 $1,424,437 $1,214,417
Fi g u r e 9. 4 iFree scenario analysis The figure shows 27 scenarios for iFree, varying the asset turnover (S/A = 0.8, 1.0, 1.3), the variable contribution margin to profit (Profit = 10% 20%, 30%) and the sales growth rate (Growth = 10%, 15%, 20%). Annual cumulative cash needs (Equity Book Values) are shown together with net income in Year 6. Maximum cumulative cash need in each scenario is highlighted. NPV is estimated assuming a 10% cost of capital and 4% growth rate after Year 6.
Assessing Cash Needs 341
well supported, the decision criteria for making follow-on investments can be designed with considerable precision.
9.6 Assessing Financial Needs with Simulation To see how simulation can be used to assess financial needs, we reconsider NewCo. Refer again to Table 8.13, which contains the pro forma financial statements for the venture in selected quarters. The statements are generated on the basis of what is expected to occur. As we discussed in Chapter 8, if we assume that the venture begins with $500,000 of equity and raises additional funding as needed by drawing on a line of credit, the company’s cumulative need for debt financing peaks at about $651,000 in Quarter 18, making the cumulative (undiscounted) need from all sources about $1,151,000. But what if NewCo does not develop as expected? We demonstrated at the end of Chapter 8 that, given our assumptions about the uncertainty of development timing, sales growth, length of the rapid-growth period, and profitability, the actual financial needs of the venture could be much higher or lower than implied by the expected performance assumptions. It would not make sense for the entrepreneur to try to fully fund all of the possible scenarios with capital raised at the beginning. Doing so would involve raising large amounts of capital at the point where uncertainty about ultimate success is greatest and would include coverage of scenarios in which early abandonment would have been a better course of action. Ultimately, we would like to approach the decision of how much initial financing to raise on the basis of NPV; that is, we want to select the level of financing that results in the greatest NPV for the entrepreneur (or the investor, depending on who is making the decision). As we illustrate in the balance of this section, a structured analysis of the results of simulating the financial model can dramatically improve the financing decision. Figure 9.5 provides a comprehensive view of NewCo’s future cash needs based on 1,000 iterations of the simulation model. The results from the simulation present information on the distribution of the initial $500,000 cash investment plus the credit line balance at the end of each of NewCo’s first 6 years (24 quarters) and the last quarter of the simulation. The graph shows the cumulative distribution of financing needs as of each year. The vertical axis in the graph represents the amount of total (undiscounted) financing NewCo needs, expressed in dollars, and the horizontal axis represents the cumulative percentage of trials from the simulation. To generate the table and graph, we set the initial investment in NewCo at $500,000.
342
Chapter Nine
$5,000,000 $4,500,000 $4,000,000 $3,500,000 $3,000,000
Quarter 4 Quarter 8 Quarter 12 Quarter 16 Quarter 20 Quarter 24 Quarter 26
$2,500,000 $2,000,000 $1,500,000 $1,000,000 $500,000 $0 0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Percentage of trials Fi g u r e 9. 5 NewCo total cash need as of quarter end The figure is based on a simulation of the NewCo financial model with an initial investment of $500,000. Plotted lines are total cash needs measured as the sum of the initial investment and the maximum value of the line of credit over the 26 quarters of the simulation.
For example, the trials data underlying Figure 9.5 show that an initial investment of $500,000 is sufficient to cover all of the simulated trials through the first year (Quarter 4). By the end of the second year, almost all trials need additional funding but the amounts are generally small, resulting in an average total funding need of $550,500. Although the distribution is skewed, the maximum is still only $654,100. From Figure 7.5, we can determine that the probability of development success by Quarter 4 is about 15.4% and the cumulative probability of development by Quarter 8 is 56.5%. This means that for most of the trials, the initial investment represents the cost associated with 4 quarters of development effort plus maintaining the $15,000 minimum cash balance, and that an additional investment of $154,100 would cover the additional development cost to raise the development success probability to 56.5%. From Figure 7.5, we can determine that for an additional 30.0% of the trials this level of funding would not be enough to reach successful development. Moreover, for the cases where development is completed by Quarter 8, we do not know whether the outcomes are high-valued, and we do not know whether
Assessing Cash Needs 343
investing more would be valuable. From Figure 9.5, we can see that investing $1 million would cover almost all of the trials for up to 12 quarters. We can also see that beyond this point the trials begin to separate into some that are producing positive FCF so that the line of credit is being paid down, and others that continue to require greater and greater investment. By Quarter 26, the end of the simulation, Figure 9.5 shows that 36.5% of the trials have produced enough FCF to have fully paid down the credit line and may have substantial surplus cash balances.
Developing a Decision Rule for Staged Investment Our goal here is to use simulation to develop a decision rule about how much to invest initially in the venture so that we can make an NPV-maximizing decision of whether to continue to invest or abandon the venture. We begin by rerunning the NewCo simulation model, setting the initial investment to zero so that the line-of-credit balance shows the cumulative financing need as of each quarter end. To structure the analysis, we have extended the model to include some basic present value information. Specifically, we assume an annual cost of capital of 10% (2.41% per quarter). Further, we assume that any positive surplus cash balance is distributed to investors at the end of Quarter 26, and that the Quarter 26 value of cash flows after that quarter can be estimated as 10 times the annualized net income in Quarter 26 (i.e., a 10X multiple), which value is discounted back to time zero at cost of capital. Running the simulation model with 1,000 trials, we estimate that the average NPV of investing thorough Quarter 26 in all trials would be $3.447 million and that the average investment over all trials would be $1.828 million. Given the positive NPV, apparently the venture would be worth pursuing, even without any staging. However, Figure 9.6 is a plot of the NPVs from the trials sorted from lowest to highest. While the average NPV is positive, the distribution is highly skewed, with a much lower median of $1.553 million. Further, NPV is negative in over 45% of the trials, and the positive average is driven by a small number of extremely positive trial outcomes. Given that many trials have substantially negative NPVs, we should be able to use staging to improve the expected NPV. We begin in Figure 9.7 by sorting the trial results by the first quarter of revenue generation and then counting the numbers of development success trials with positive and negative NPV. For example, from a total of 1,000 trials, 15.4% began to generate revenue in Quarter 5. Of those, 150 trials (15% of all trials) resulted in positive NPV and 4 trials (0.4%) resulted in negative NPV.
344
Chapter Nine
Fi g u r e 9. 6
$100,000,000
NewCo NPV results from simulation The figure shows the distribution of estimated NPV from 1,000 trials of the NewCo simulation model, with initial investment set to zero. NPV is estimated using a 10% annual discount rate and capitalizing Quarter 26 income by annualizing and applying a 10X multiple.
$80,000,000
$60,000,000
$40,000,000
$20,000,000
95.0%
90.0%
85.0%
80.0%
75.0%
70.0%
65.0%
60.0%
55.0%
50.0%
45.0%
40.0%
35.0%
Percentage of trials, sorted by NPV
18%
Fi g u r e 9.7 NewCo simulation results for development completion timing (based on 1,000 trials)
16%
Trials with development success that result in positive NPV
14%
Trials with development success that result in negative NPV
12%
Percentage of trials
The figure plots the first quarter of revenue generation from a simulation of the NewCo financial model with 1,000 trials. Each bar in the figure shows the percentage with negative NPV and the percentage with positive NPV.
100.0%
($20,000,000)
30.0%
25.0%
20.0%
15.0%
5.0%
10.0%
$0
10% 8% 6% 4% 2% 0%
5
6
7
8
9
10
11
12
13
14
15
16
17
Fail
First quarter of revenue generation
As reflected in the model assumptions, the value of development success is expected to be greatest if success occurs quickly. Earlier success results in more rapid demand growth and other benefits of being a first mover. Figure 9.7 illustrates the impact of development timing on the probability that the trial
Assessing Cash Needs 345
will result in positive NPV. The probability of positive NPV is very high if revenue generation begins in Quarter 5. This probability diminishes the longer development takes. Although the model is designed to allow development efforts to continue through Quarter 17, in the simulated trials, there are no positive-NPV trials that begin revenue generation after Quarter 14. Based on the trials data, there is no reason to fund development efforts beyond Quarter 14. Also, the probability of a positive NPV outcome for an effort that begins to generate revenue in Quarter 14 is very low—only one of the 34 trials that began revenue generation in Quarter 14 is NPV-positive. Thus, it seems unlikely that funding development efforts in Quarter 14 would be valuable. More systematically, we can use the trials data to study how discontinuing funding after any given quarter would affect the estimated NPV of the venture. Table 9.1 provides a summary of results of alternative decision rules of investing in all trials through Quarter X and abandoning any trials that have not achieved development success by that point. Because we know there are no successful trials with positive NPV after Quarter 14, we abandon all remaining trials after that quarter. So we can construct a correct decision rule, when a trial is abandoned, we replace the computed NPV with the (negative) amount invested in that trial up to that point. As shown in the table, investing through Quarter 14 and not investing thereafter in trials that have not achieved development success results in a higher average NPV ($4.041 million) than does continuing to invest, entirely because the investor does not waste resources by investing in the trials that have not Tab le 9.1 NewCo cumulative NPV by first quarter of revenue generation
Quarter
Mean
Median
Cumulative Trials with Development Success
Cumulative Trials with Positive NPV
5 6 7 8 9 10 11 12 13 14 All trials
$1,751,300 $3,028,335 $3,781,240 $4,144,837 $4,255,877 $4,300,509 $4,244,371 $4,187,066 $4,116,698 $4,040,740 $3,477,094
($334,606) ($403,054) ($472,912) ($543,001) ($243,299) $357,612 $498,449 $538,526 $555,722 $562,733 $562,733
154 284 393 487 565 632 688 736 777 811 885
150 272 369 441 487 521 534 542 545 546 546
The table shows the results from the NewCo simulation trials data of implementing a decision rule to invest fully in any trial that has Quarter X or earlier as its first quarter of revenue generation and to abandon any trial where development has not been completed by Quarter X.
346
Chapter Nine
yet achieved development success (with negative NPV) or may never achieve development success. At this stage in the analysis, investment continues without limit in all trials that have been successful by the indicated quarter. The least desirable result in Table 9.1 is to discontinue investing in any trial that has not begun revenue generation by Quarter 5. As of that point, in only 154 trials has development been successful, and 4 of those are trials with negative NPV. Because so many potentially good outcomes are abandoned with Quarter 5 as the decision criterion, the average NPV is only $1.751 million and abandonment of so many trials results in the median NPV over all 1,000 trials being negative $335,000. Our goal with this decision rule is to select the critical quarter based on average NPV over all trials. As shown, the maximum is reached if the critical quarter is Quarter 10, at an average NPV of $4.301 million. The median at this critical value is also positive because most of the trials are being pursued, including over 95% of the trials where NPV is positive (521 of 546).
Modifying the Decision Rule to Also Consider Revenue Growth With the decision rule of investing fully in all trials that begin revenue generation in Quarter 10 or sooner, we would end up investing fully in 111 trials where development is successful but NPV is negative. These are the 632 successes minus the 521 with positive NPV from the Quarter 10 row in Table 9.1. We should be able to further improve the value if we can identify factors that can be recognized early and are key drivers of NPV. Conceptually, the value of the development effort depends on the quality of the result and the emergence of competing products. By inspections, we determined that, in the model, these considerations are manifested mainly through the rate of sales growth during the rapid growth period. That is, among the trials with development success in any given quarter, NPV is strongly related to the realized rate of revenue growth. Since the drivers of growth should be observable quickly after development is achieved, we decided to refine the earlier decision rule by abandoning very-low-growth trials in each development success quarter. Abandoning low-growth trials, while it reduces investment in some negative NPV trials, also results in abandoning some where NPV is positive. We cannot simply focus on NPV since the realized NPV would not be observable by investors until much later. Accordingly, using the trials data, we selected critical growth rate values for abandonment in each development success quarter. For example, in Quarter 5, we decided to abandon any trial where the realized growth rate was less than 14%.
Assessing Cash Needs 347 Tab le 9. 2 NewCo cumulative NPV by first quarter of revenue and revenue growth rate
Quarter
Mean NPV Based on Critical Quarter
Adjusted Mean NPV Adding Critical Growth Rate
Increase in NPV
Critical Growth Rate
5
$1,751,300
$1,752,005
$705
14
6
$3,028,335
$3,032,139
$3,804
15
7
$3,781,240
$3,788,735
$7,495
15
8
$4,144,837
$4,163,002
$18,165
16
9
$4,255,877
$4,296,719
$40,842
16
10
$4,300,509
$4,360,445
$59,937
20
The table shows the gain in average NPV that results from modifying the “critical quarter” decision rule for abandonment to also incorporate a critical growth rate for abandonment.
As shown in Table 9.2, applying this additional criterion to the trials data in Quarter 5 modestly increases the average NPV (by $705). The increase is modest because there are only 4 negative NPV trials with development success as of Quarter 5, 3 of which are abandoned, and there is one positive NPV trial with a growth rate below 14%. Moving to Quarter 6, we retain the abandonment criterion for the Quarter 5 trials and, by inspection, use 15% as the critical value of the Quarter 6 growth rate. In general, the critical value for positive NPV outcomes is higher the later development occurs. The result is a modest $3,804 increase in average NPV. Continuing this stepwise process through to Quarter 10 development, the increase in average NPV that results from incorporating the growth rate criterion is estimated to be about $60,000. We do not continue the process beyond Quarter 10 since the drop-off in baseline average NPV in Quarter 11 is too large to be offset by the gain from modifying the decision rule. In conclusion, the highest NPV is achieved by exercising the abandonment option if revenue generation does not begin by Quarter 10 or if the revenue growth rate in any quarter (5 through 10) is below the critical value for that quarter, as indicated in Table 9.2.
Additional Considerations and Caveats The preceding discussion demonstrates the process of using simulation to develop a decision rule related to abandonment. It should be possible to further improve the decision by examining a larger number of simulated trials to improve the precision of abandonment criteria and by adding additional considerations related to the decision. Further, there may be additional choices that should be taken into consideration. For example, abandonment is only one option. It may also be possible to pivot or refocus the effort in response to the learning that occurred during the development stage.
348
Chapter Nine
9.7 Summary A number of analytical techniques are available for assessing financial needs. They range from the simple approaches like cash flow breakeven analysis and sustainable growth model to more complex methods that involve scenario analysis and simulation to study how uncertainty affects financial needs. Cash flow breakeven analysis considers financial needs in a different way. On one level, the technique helps determine the level of sales a venture must achieve to finance its operations from cash flow. At that point, the venture is viable on a cash flow basis, but growth beyond the breakeven point would require additional capital. On another level, by combining cash flow breakeven analysis with a sales forecast, the entrepreneur can estimate the investment needed to sustain the venture until the breakeven point is reached. The value of the sustainable growth model is that it links a venture’s ability to finance growth from operations to a few policy decisions. For any given growth objective, the model enables the entrepreneur to understand when growth can be financed from operations and how much will need to be funded externally. Scenario analysis is an approach to studying the effects of uncertainty. When the technique is applied to assessing financial needs, the objective is to understand how financial needs vary over a range of realistic scenarios. Scenario analysis can be combined easily with other methods of assessing financial needs; that is, the cash flow breakeven point or the sustainable growth rate can be evaluated over a range of scenarios to improve understanding of financial needs. Simulation is by far the most powerful analytical tool for evaluating financial needs. Beginning with sound assumptions and an integrated financial model, the effects of key aspects of uncertainty can be evaluated simultaneously. Skillful application of simulation can help the entrepreneur or an investor to design a financial structure that preserves the potential for success but does so with a limited amount of investment capital. At the same time, the financial structure enables the entrepreneur and investors to avoid overinvesting once it becomes apparent that success is unlikely. Simulation can be used effectively to study the effects of subtle but significant changes in financing provisions like staging and to devise decision criteria for exercising real options.
Review Questions
1. Describe the main factors that determine a venture’s sustainable growth rate. What are the key assumptions in the sustainable growth model?
Assessing Cash Needs 349
2. List four ways in which managers can increase a venture’s sustainable growth rate. Which do you think is easiest to implement? Which is most difficult? 3. Explain how a venture’s sustainable growth rate is related to its financing needs. 4. What is the difference between accounting breakeven and cash flow breakeven analysis? Which is more important for a new venture and why? 5. Explain the difference between marginal and average contribution margins. Why is it important to distinguish between them when making operational decisions? 6. Why is it important for an entrepreneur to plan for unexpected productmarket success? What are the potential consequences of failing to do so? 7. How might you determine the number of scenarios needed to analyze a new venture’s uncertainty? Explain how you might estimate reasonable probabilities for each scenario. 8. What benefits does simulation bring to the analysis of venture uncertainty? 9. Explain how simulation and staging can be combined to estimate an optimal level of initial funding for a new venture. 10. Describe how you might use simulation with different assumptions about the amount of initial financing to help determine the investment that would help assure that good outcomes are not missed but bad outcomes do not receive funding? Notes
1. See Higgins (2009), chap. 4, and Donaldson (1991) for additional discussion of sustainable growth. 2. In this chapter, we assume a basic knowledge of present value analysis and an assumed cost of capital. Later in the book, we formally develop new venture valuation. 3. This is equivalent to assuming that asset turnover, financial leverage, and return on sales are constant. 4. Perhaps you are wondering why, in this model, we have not mentioned cash flow. The reason is that the sustainable growth model assumes that depreciation expense each period is equal to the investment required to replenish the assets. Thus, in order to prepare a cash flow statement, we would add back
350
Chapter Nine
the depreciation expense. But to sustain operations at the same level as in the prior year, the entrepreneur would need to reinvest the same amount back in the venture. Doing so would maintain the balance sheet at its initial level. 5. Information from Webvan Form 10K-405, year ended December 31, 1999; and “Technology Journal: Webvan IPO to Stir Grocery Industry,” Wall Street Journal (Europe), November 9, 1999, p. 10. 6. See https://www.freshdirect.com/index.jsp. 7. The same question could be examined from the perspective of an investor, and similar issues would arise but the conclusions might be different. 8. Note that cash from operations that can be used to support an acquisition strategy is not FCF. Moreover, new ventures frequently raise money to buy out a competitor to gain market share, access to a particular market, or intellectual property. SeatGeek.com has made several acquisitions. Zipcar acquired several small competitors before, itself, being acquired by Avis. 9. In later chapters, we will focus more systematically on cost of capital and valuation. For now, we use these simple assumptions to illustrate the approach.
References and Additional Reading
Donaldson, G. 1991. “Financial Goals and Strategic Consequences.” In Strategy: Seeking and Securing Competitive Advantage, ed. C. Montgomery and M. Porter, 113–34. Cambridge, MA: Harvard Business School Press. Higgins, R. C. 2009. Analysis for Financial Management, 9th ed. Boston: McGraw-Hill Irwin. Kaplan, J. 1994. Startup. New York: Penguin. Kaplan, S. N., and A. Schoar. 2005. “Private Equity Performance: Returns, Persistence and Capital Flows.” Journal of Finance 60: 1791–823. Ross, S. A., R. W. Westerfield, and J. Jaffe. 2008. Corporate Finance, 8th ed. New York: McGraw-Hill Irwin. Sahlman, W. A. 1999. “The Financial Perspective: What Should Entrepreneurs Know?” In The Entrepreneurial Venture, ed. W. A. Sahlman and H. H. Stevenson, 2nd ed., 238–61. Cambridge, MA: Harvard Business School Press.
C h a p t e r Ten
Fo u n datio n s o f N e w Ve ntu r e Valuatio n
H ow d o VC s and other investors select the projects in which they invest?
And how do they settle on the ownership stakes, terms, and conditions they require in exchange for investing? There are no simple answers to these questions. Certainly, the perceived value of the entrepreneur’s concept and capabilities are critical factors. No venture will be funded unless an investor sees the merits of the concept and regards the entrepreneur as capable of implementing it or of working with the investor to build a capable team. In addition, there are issues of fit and timing: Can the investor contribute value to the project? Does the investor have the financial and organizational capacity to take on another project? The answers to these questions depend on the particular expertise of the investor and on whether, at the time, the investor is looking for new projects. In this chapter, we introduce the foundations of valuation and present the most widely used valuation methodologies, identifying the pros and cons of each. Chapter 11 concerns implementation and provides guidance on how to ensure that the valuation approach is internally consistent and how to find the information necessary to perform the valuation. In Chapter 11, we use a single example to illustrate the use of the various valuation methods that we present in this chapter. An important lesson is that if the assumptions are all consistent, then, in a perfect world, every valuation method will produce the same estimate of value. Of course, in reality they do not. The data and assumptions in each valuation effort are estimates of the “true” underlying values. Differences in estimated value across methods therefore reflect the effects of estimation errors. The art of valuation is in deciding how to weight the information from each approach and use it to reach a conclusion about value.
353
354
Chapter Ten
10.1 Perspectives on the Valuation of New Ventures The value of any investment depends on its ability to generate future cash flows, as well as on investor assessments of and tolerance for the riskiness of those cash flows. Two aspects of valuation make new venture investment decisions particularly difficult. First, the future cash flows are volatile and difficult to forecast. Second, discount rates appropriate for new venture investments can be challenging to estimate. In spite of the near impossibility of precision, earnings or cash flow forecasts appear in most business plans, and forecasts are made and studied by VCs and other investors who are shopping for deals. In addition, VCs and other investors address the problem of determining the appropriate discount rate routinely. We show in this chapter that financial economic theory provides considerable guidance for estimating appropriate discount rates. In an area as competitive and complex as investing in new ventures, the importance of good decision making about the terms for investing cannot be overemphasized. One indicator of the potential for good decision making to add value is the variation in investment performance of VC funds. Several sources compile information on the returns to investors in VC funds. In Figure 10.1, we use information reported in the Preqin Private Equity database to produce a histogram of annualized internal rates of return (IRRs) for 1,715 funds that were launched between 1969 and 2013.1 The returns range from almost a total loss (–79.2% per year) to one fund with annual returns of more than 500%. To put this in perspective, a fund that returns 500% per year for a five-year period would return more than $1,200 for each $1 invested. Based on its net multiple of 19.82 (i.e., each $1 invested returned $19.82), this particular fund, however, had a weighted average life of only about 1.5 years so the actual return is much lower that implied by the 19.82 net multiple. Such high returns are a rare occurrence; only 0.4% of the funds in the sample had IRRs over 200%. The average IRR in the sample is 13.2% per year, and the median is 8.7%. Weighting by fund size reduces the average IRR to 9.95%. The compound average annual return on the S&P 500 over approximately the same period (though not strictly comparable) was 8.6%, which is higher than almost half of the VC funds. Based on this evidence, historically VC investing has generally done better than investing in the market but the return differential has been small. Apparently, the small return premium is sufficient to compensate for the risk and illiquidity as VC has continued to attract funding. Low rates of return for some VC funds can result from numerous factors: unfortunate timing, bad luck, lack of skill or access to deal flow, and unforeseeable negative events. However, two important reasons for low rates of return
Foundations of New Venture Valuation 355 Fi g u r e 1 0.1
25%
Distribution of VC fund average annual IRRs (vintage years 1969–2013) Preqin Private Equity Database.
20%
Average fund IRRs are computed for 1715 funds as of returns reported though 2016.
Percentage of observations
source:
15%
10%
245%
235%
225%
215%
205%
195%
185%
175%
165%
155%
145%
135%
125%
115%
95%
105%
85%
75%
65%
55%
45%
35%
25%
5%
15%
–5%
–15%
–25%
–35%
–45%
–55%
–65%
0%
–75%
5%
Fund internal rate of return
are valuation mistakes and deal-structuring mistakes. Both of these can be avoided, or at least minimized, by using decision-making methods that give the investor a competitive advantage over its rivals in both project selection and deal structuring. The arrival of professional investment managers to VC investing has changed the market in fundamental ways. Most important for our purpose is the changing way in which investment opportunities are valued by investors. In time, the changes will affect the ways in which entrepreneurs evaluate projects. Because the market is changing, some of the approaches and rule-of-thumb heuristics that have been used historically for investing in new ventures no longer can be relied on to identify projects that are likely to yield acceptable rates of return.
10.2 Myths About New Venture Valuation Past practices have generated four myths about new venture investing.
Myth 1: Beauty Is in the Eye of the Beholder Three decades ago, Gordon Baty (1990) wrote, “Pricing a new company’s stock is much like pricing any other glamour item (such as perfume, paintings, rare coins) where appeal is based on emotional, as well as analytical considerations.”2
356
Chapter Ten
While it is reasonable to expect that the entrepreneur may care about qualitative factors, it would not be a good idea for a VC fund manager to propose to prospective investors that financial return be traded off against other factors. Professional investors understand that the fundamental trade-off between cash flow and risk must drive valuation. Here, we present theoretically sound valuation tools while at the same time recognizing the implementation challenges.
Myth 2: The Future Is Anybody’s Guess This is a more reasonable-sounding version of Myth 1. The claim is that even though cash flow is what matters, future cash flows of new ventures are so uncertain that forecasting them is of little value. Although new venture financial forecasts are subject to great uncertainty, such uncertainty—rather than making the forecast worthless—makes forecasting critical. In particular, it is important to try to understand the extent, nature, and implications of the uncertainty. It is true that a single-scenario forecast for a new venture is unlikely to be of much value. Scenario analysis and simulation are, however, of considerable practical value for understanding and dealing with the risks, establishing strategy, assessing cash needs, and valuing the venture.
Myth 3: Investors Demand Very High Rates of Return to Compensate for Risk New ventures are high-risk investments that tie up the investor’s capital for several years with no easy means of exit. This has led to a broadly held perception that required rates of return for VC investing must be very high. On this subject Michael Roberts and Howard Stevenson write, “In order to compensate for the high risk of their investments, give their own investors a handsome return, and make a profit for themselves, venture firms seek a high rate of return. Target returns of 50% or 60% are not uncommon.”3 As summarized in the following table, Jeffrey Timmons and Stephen Spinelli (2007, p. 449) provide a more comprehensive summary that echoes the same point: Stage
Annual ROR (%)
Total expected holding period (years)
Seed and start-up First stage Second stage Expansion Bridge and mezzanine LBO Turnaround
50–100% or more 40–60% 30–40% 20–30% 20–30% 30–50% 50%+
More than 10 5–10 4–7 3–5 1–3 3–5 3–5
Foundations of New Venture Valuation 357
Scholars and others who remark on the high rates of return base their statements on historical practice or statements of “sought for” returns cited in PPMs, which generally are indicating the rates of return VC investors apply when discounting the entrepreneur’s projected cash flows. Approaching the question in this way, they find that the rates are typically quite high. For example, Gompers, Gornall, Kaplan, and Strebulaev (2016) report, based on a survey of a large sample of VCs, that the median “required” IRR for investment selection was 30% and the “required” net multiple (cash-on-cash return) was 5.5 times. They also report that in marketing to potential investors VCs indicate seeking investments that can offer IRRs of 24% and net multiples of 3.5 times. These high returns, however, are not supported by historical evidence; an examination of the actual average realized returns for investing in new ventures tells a very different story. In addition to the information in Figure 10.1, numerous studies by academics and industry participants over more than four decades report average annual rates of return to investors in VC funds in the mid- to high teens.4 For the recent 20-year period, ending with the 2013 vintage year, the average annual return from a sample of U.S. VC funds was 17.3%. Figure 10.2 shows estimates of VC IRRs since 1981 by vintage year: the series exhibits very high volatility, including vintage years with very high returns and others with negative or near-zero returns.5
100
Fi g u r e 1 0. 2 VC returns to limited partners by vintage year Thomson Reuters, Venture Economics, VentureXpert database, extracted August 2009; PitchBook data are from the PitchBook 2017 PE VC Fund Performance Report and Preqin
90 80
sources:
60 50 40 30 20 10 0 –10 –20 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Annualized IRR (percentage)
Data through 2006 are pooled average VC returns by vintage year compiled by Thomson Reuters and measured through 2009. Data from 2007 on are VC fund medians reported by PitchBook/NVCA and measured through 2017.
70
Vintage year
358
Chapter Ten
How can the common practice of using very high discount rates to value projects be reconciled with the evidence that over many years the average realized return has been much lower? Do investors suffer from chronic unfounded optimism, so that actual returns have been disappointingly low? Or is there another interpretation of the evidence, one that does not rely on biased decision making? Surely, if rates of return in the neighborhood of 30% could reasonably be expected, capital would flood the new venture market, driving the returns to levels that are more consistent with those for other forms of investment. True required rates of return are much closer to the range documented in the empirical studies than to the very high rates that are sometimes sought by investors when they evaluate individual projects. Later in this chapter, we offer a reconciliation of the preceding statements and evidence that places the statements in a more useful context and demonstrates that actual required rates are much lower. The point for now is that the contradiction is more apparent than real.
Myth 4: The Investor Determines the Value of the Venture A common contention is that it is pointless for the entrepreneur to undertake a valuation. The argument is that investors do not accept the entrepreneur’s valuation anyway, so the entrepreneur’s efforts are better spent in other ways. The problem with this view is that it fails to recognize the pivotal role that valuation plays in reaching agreement between the entrepreneur and the investor, as well as the role it can play in helping the entrepreneur decide whether to undertake the venture. VC investors report in the Gompers et al. (2016) survey that less than 2% of the companies they consider for investment result in completed funding rounds and that many of their negotiations fail because of concerns about the management team, which can arise partly from unrealistic perceptions of value. In an interview, Sonja Hoel, a managing director at Menlo Ventures, cited valuation issues as an important factor. We almost always get it right if we turned down a deal because there wasn’t a market. Where we don’t always get it right is valuation. If we turned it down because of valuation, we had a 10% error rate. Of all the decisions we made because of valuation, 90% were good but 10% were bad.6 Good working knowledge of valuation can help the entrepreneur avoid a breakdown of negotiations. It is true that investors commonly prepare valuations based on their own research and assumptions. However, there is more to new venture financing negotiations than a simple exchange of cash for a percentage of the equity. In the context of a financing negotiation, valuation is important
Foundations of New Venture Valuation 359
to the entrepreneur for at least three reasons. First, the entrepreneur can better understand how the venture is likely to be valued by prospective investors. Second, the entrepreneur can better understand what the venture should be worth to him and how that differs from its value to the investor. Third, the entrepreneur needs to understand how alternative deal structures affect overall value and the values of the financial claims of investors and the entrepreneur. Even for very early-stage ventures, where “unpriced” rounds are common, it would be difficult to reach agreement on deal terms without a sense of the value of the opportunity. The entrepreneur might expect that competition among prospective investors can eliminate the need to value complex financial claims. This expectation is incorrect. Even if several investors are vying to participate in a venture, they may have varying views of the venture’s strategy and will probably seek different structures of ownership claims and financing commitments. Without studying the valuation consequences of different proposals, choosing the best alternative can be problematic for the entrepreneur. A solid understanding of basic valuation techniques can ensure that entrepreneurs better understand how investors perceive the opportunity and help to reach a mutually beneficial agreement.
10.3 An Overview of Valuation Methods For an investor who cares only about financial return, the value of any investment is the PV of its future cash flows. Although a variety of methods exist for estimating value, ranging from explicit discounting of future cash flows to valuation based on simple multiples to valuation based on comparable firms, they all are attempting to measure, directly or indirectly, the PV of the right to receive future cash flows. As discussed in Chapter 1, valuation is guided by two fundamental principles: that a dollar today is worth more than a dollar received in the future, and that a safe dollar is worth more than a risky one, that is, a safe dollar is more valuable than a gamble with an expected payoff of one dollar. Thus, the PV of any investment depends on the timing and riskiness of expected future cash flows. In theory, if a person could correctly identify expected cash flows, risk, and cost of capital, the result of a discounted cash flow analysis would be the “true” PV. In practice, however, there is considerable judgment involved in valuation. Because we must rely on estimates, we can be sure our calculations yield only approximations of true PV. Rather than despair, we should recognize that imperfect PV estimates provide opportunity. There are potentially large rewards
360
Chapter Ten
for doing a better job of estimating PV than do your rivals. Other things being equal, the entrepreneur or investor who consistently does a good job of estimating value will outperform one who is right on average but makes large over- and underestimation errors. There are many theoretically sound and empirically validated valuation tools for analyzing new ventures. In this chapter, we introduce the following methods and highlight their strengths and weaknesses: 1. Discounted cash flow (DCF) methods
a. The risk-adjusted discount rate (RADR) approach b. The certainty equivalent (CEQ) approach 2. Relative value (RV) method 3. The Venture Capital (VC) Method 4. The First Chicago Method These methods vary in their complexity and the directness of their connection to underlying economic theory, but all are used in practice and therefore are important to entrepreneurs and investors.
The Discounted Cash Flow (DCF) Methods We consider two approaches for estimating PV by DCF. The first is the RADR approach and the second the CEQ approach. The difference between them lies in how the adjustment for risk is incorporated into the calculation. If the approaches are applied in a consistent manner and the underlying assumptions are consistent, they will yield identical estimates of value. However, the information needed to implement the two approaches differs. Consequently, availability and quality of information are important determinants of which to use. The RADR approach is the more widely known and used DCF approach. In RADR, an expected future cash flow is converted to PV by applying a discount rate that reflects both the time value of money and the riskiness of the future cash flow. This rate is known as the “risk-adjusted discount rate” because the effect of risk on value is built into the discount rate. The RADR approach is used most commonly in corporate finance because it is convenient and because the information requirements are easily satisfied by using data on comparable public firms. With the CEQ approach, instead of adjusting the discount rate, the risk adjustment is made directly to cash flow. This adjustment yields a risk-adjusted (or certainty equivalent) cash flow that is converted to PV by discounting at the risk-free rate. It turns out that for new ventures, it can be easier to estimate the
Foundations of New Venture Valuation 361
CEQ cash flow than the RADR. In subsequent discussion, we examine both approaches.
The Relative Value (RV) Method Relative valuation uses market data on other companies or other transactions as bases for inferring the value of the subject venture. This method is sometimes referred to as “comparables” or “multiples valuation.” It is widely used in practice, especially for established private companies, and can provide a quick and easy ballpark estimate of value. The underlying logic is arbitrage. If two different companies are expected to produce identical future cash flows and are subject to identical risks, they should have the same value. If they did not, then investors should want to sell the higher-valued one and buy the lower-valued one. By inference, if firms have similar characteristics, the arbitrage reasoning should provide a good approximation. Relative valuation is thus an effort to finesse questions of discount rates and explicit cash flow projections. If RV and DCF could be correctly and perfectly applied to a given venture, then they should yield identical values. Suppose, for example, that we find a perfect comparable for a venture we want to value and that the comparable has been the subject of a recent transaction that implied a specific value. Under RV, we could simply say that the values are the same; under DCF, in contrast, we could use the expected future cash flows and implied value of the comparable to infer the discount rate that we should use to value the cash flows of the subject venture. We would get the same answer either way. In practice, when RV and DCF yield different estimates of value, there is something inconsistent about the assumptions. Either the assumptions need to be revised or judgment must be used to assess the significance to place on each estimate. In RV, market prices or prices from public or private transactions are collected along with information on observable characteristics of the comparable firms. This information may include accounting ratios and operating data that can be used to estimate the value of a private firm or new venture. While the logic behind RV is straightforward, there are many challenges to valid implementation. Appropriate comparables for a new venture (with transactions at the same stage of development) can be difficult or impossible to find or to verify, and common metrics such as P/E ratios are not useful for a start-up that has not reached profitability. Because some multiples are based on accounting data, the usual caveats about quality of earnings and adjustments to ensure comparability apply. More generally, while the idea of valuation based on comparisons to other firms sounds simple, it is not. There are many dimensions of comparability, including industry, business model, stage, size, and accounting ratios. Ideally,
362
Chapter Ten
a comparable firm has expected cash flows and risk that are similar to those of the new venture. Certainly, no public company can be a direct comparable for an early-stage venture. Moreover, there are no readily accessible databases of transactions related to early-stage private companies that have sufficient data to be useful in a valuation. Use of an RV approach is more common and more defensible for a company that is being valued at the time of an assumed harvesting event, such as an IPO or acquisition. At that stage, information on comparable firms and transactions is likely to be more plentiful and the necessary analogies are easier to make and defend. In fact, RV is often the approach that is used in DCF methods to estimate the harvest-date cash flow. That said, even for a mature private company there are likely to be no truly comparable public companies. Companies that are similar in some key respects but not identical to the subject company are likely to yield a wide range of value estimates. Deciding how to select the companies and how to weight the evidence are part of the art of valuation.
The Venture Capital (VC) Method The VC Method combines elements of DCF and RV and has been popular in the private equity arena. It is computationally straightforward and parallels the way some VCs have traditionally approached their investments.7 The method starts with an estimate of future value conditional on success— possibly the projections in the entrepreneur’s business plan, which typically seek to demonstrate the potential success of the venture. Success scenario projections incorporate an assumed timetable for exit, typically three to five years, and a conjecture as to the form of exit and resulting future value. The estimate of exit value conditional on success is sometimes developed using RV multiples to a projected performance measure such as net income or EBIT. The success-scenario exit value is then discounted to PV, typically using a very high annual discount rate such as the “sought for” rates we presented earlier. This hurdle rate is intended to take account of (1) time value, (2) risk, (3) the bias associated with discounting only the success-scenario cash flows, and (4) the dilutive effect of subsequent financing rounds.8 The result is an estimate of the venture’s PV. In spite of its widespread use and intuitive appeal, the VC Method has theoretical and practical shortcomings. Most important, the hurdle rate selection implicitly combines so many conditions that are difficult to specify so that there is no reliable way to determine the rate that should be used. Because only a success scenario is used, the assumed timing of the harvest cash flow can impart
Foundations of New Venture Valuation 363
systematic bias to value estimates. We discuss these shortcomings in greater depth below. Because of these, we regard the VC Method as more of a device for negotiating with the entrepreneur than a reliable valuation method.
The First Chicago Method The First Chicago Method is also used widely by VC and PE practitioners and represents an improvement on the VC Method.9 The goals of the First Chicago Method are to provide a simple way of performing DCF valuation and to mitigate the valuation biases of the VC Method. Rather than limiting the analysis to a success scenario, the First Chicago Method uses probability-weighted scenarios to come up with a more reliable estimate of expected (in the statistical sense) cash flows, rather than just the optimistic cash flows used in the VC Method. These expected cash flows are then discounted using a more realistic cost of capital, rather than the high hurdle rates used in the VC Method. If the scenario probabilities are correctly weighted, the appropriate discount rate is identical to the one that would be used in DCF valuation by the RADR approach. A benefit of the First Chicago Method is that it requires the analyst to think about the range of possible outcomes for the venture and their probabilities. The question of which valuation tool is best for a particular situation does not have a clear answer. Like any quantitative methodology, the results of each approach depend on the availability and quality of the information it requires. There is a benefit to using as many of the techniques as the data will allow. Doing so will inevitably produce a range of value estimates. The final step is to compare the results, seeking to understand the reasons for material differences. On reflection, you may conclude that in a given instance, one approach is more reliable than are the others. More likely, each provides some useful information, and your task is to decide how best to weight the different estimates.
10.4 Discounted Cash Flow Valuation As noted earlier, we consider two approaches to DCF valuation, the RADR approach and the CEQ approach. The main difference between them is in the way we adjust for risk. Most introductory finance texts cover only RADR or only briefly describe CEQ. Because of its different information requirements, however, the CEQ method can be better suited than RADR for new venture valuation.
364
Chapter Ten
The Risk-Adjusted Discount Rate Approach Under RADR, expected future cash flows are discounted to PV using a discount factor that reflects the time value of money and the riskiness of the future cash flows. The present value, PVj, of an investment that offers a series of expected future cash flows, Cjt, is given as: C jt PV j = Σ t (10.1) (1 + rt )t In Eq. (10.1), rt is the risk-adjusted discount rate that is appropriate for computing the PV of an expected time t cash flow. The expression is general in that it allows for cash flows to be received at any time and for the cost of capital to be specific to the period in which the flow is expected. Note that PVj is the value today of all of the cash flows that are generated by the venture’s operations, whether they are positive or negative. It is not the same as net present value (NPV), which is net of the investment required to receive the venture’s future cash flows. To illustrate, consider a delivery business such as Postmates. The PV would include positive cash flows (e.g., revenue) and negative cash flows (e.g., compensation to independent contractors, lease payments on vehicles, salaries, and fuel). Let’s assume that discounting all of the positive and negative cash flows using Eq. (10.1) produces a PV of $150,000. If the owner agrees to sell you the venture for $130,000, the resultant NPV is $20,000. That is, NPV is equal to PV minus the investment required to acquire the venture. Applying Eq. (10.1) requires that we (1) identify the cash flows that are appropriate to include in the valuation model and (2) determine the appropriate discount rate for valuing each cash flow. Identifying relevant cash flows. Conceptually, determining the cash flows to include in an RADR valuation is straightforward. They are the cash flows the investor can expect (statistically) to receive in exchange for investing. To identify the relevant cash flows, we need to understand exactly what asset is being valued. The asset may be the entire venture or a particular financial claim on the venture, such as a portion of the common stock, preferred stock, or debt. A share of stock, for example, yields a stream of cash flows in the form of dividends. An investor who owns the share for a finite period, of course, does not receive the entire dividend stream but receives dividends paid by the firm during the holding period and a lump-sum payment when the share is sold. The payment when the share is sold depends on the stream of dividends that are expected to be received from that point forward. Implicitly, then, the stock price at any point reflects the value of all subsequent dividend cash
Foundations of New Venture Valuation 365
flows.10 The expected proceeds from selling the stock at a future date can be thought of as the “continuing value” of the stock at that point, based on cash flows (dividend payments) expected after the sale. Because we normally project dividends in simple ways, such as by assuming that they are expected to grow at a constant rate, the continuing value approach (of an actual or hypothetical sale in the future) is a shortcut approach to discounting a long stream of future cash flows. Debt is expected to yield a stream of interest payments and the eventual repayment of principal. If the debt is risky, the interest payments and principal repayment to be valued are not those specified in the debt contract. Rather, they are the cash flows that are expected to be received, recognizing that the borrower may prepay or default on the obligation. If the asset to be valued is a venture, the relevant cash flows are the (positive and negative) periodic free cash flows generated by the venture and available to all capital providers. Depending on your purpose, it may be advantageous to measure these cash flows net of any retention for the purpose of capital replacement or growth and (for consistency) to measure investment as only the cash infusions that investors make in the venture. Like a share of stock, a venture can, in principle, last forever. Conceptually, we would like to know the value of all future cash flows. However, at some point after the first few years it is more convenient (and possibly no less accurate) to assume a hypothetical sale where the selling price is based on the value (at the time of sale) of all free cash flows from that point forward. We refer to this value as the continuing value of the venture. The “explicit value period” is the span over which periodic cash flows are estimated explicitly and each is discounted to its PV. Total PV thus consists of two parts: (1) the PV of the expected cash flows from the explicit value period and (2) the PV of the expected continuing value. Determining the discount rate (cost of capital). The next step in valuation by RADR is determining the discount rate for valuing each cash flow. Because in the RADR approach the discount rate takes account of both time value and cash flow risk, we can think of the discount rate used in Eq. (10.1) as having two components. The first is the return for investing in a risk-free asset that would pay off at the same time as the project cash flow. The second is a risk premium that depends on the riskiness of the expected future cash flow. For a particular project j that yields a single uncertain cash flow at time t, the appropriate discount rate can be stated as follows: rjt=rFt+RPjt (10.2)
366
Chapter Ten
where rFt is the required rate of return for investing in a risk-free asset that would pay off at time t, and RPjt is a risk premium or risk adjustment. The risk adjustment depends, in some fashion, on the riskiness of the time t cash flow. Because the cash flows being valued are expected cash flows, the appropriate discount rate is the opportunity cost (or rate of return) the investor could expect to realize by investing in an alternative financial claim with the same holding period and risk. In contrast to the ad hoc determination of the hurdle rate in the VC Method, the financial economic theory underlying determination of the discount rate for RADR valuation is well established. In summary, the steps for using the RADR approach are as follows: (1) forecast expected future cash flows, (2) estimate the risky discount rate, rjt, for each cash flow, (3) discount each future cash flow to its PV, and sum the PVs. Equation (10.2) is useful because it is sometimes easier to estimate rFt and RPjt than to estimate rjt directly. A simple way of estimating rFt is to examine currently available returns on zero-coupon government bonds of similar duration to the cash flow that is being valued. Estimating the risk premium is more difficult: first, we need a measure of risk; second, we need a metric to determine the associated risk premium. The measure of risk. The measure of risk that has become the norm for investment valuation is the standard deviation of holding-period returns.11 A holding-period return is a rate of return that is measured from the point of investment to the point when the return is realized. That is, the holding-period return is expressed as a percentage based on the amount invested. Take, for example, a common stock. Its holding-period return consists of two parts, a dividend yield and a capital gain (or loss). Uncertainty about the size of the dividend and the price at which the shares can be sold in the future are the risks faced by investors. Assume that a share of stock that has a market value of $20 today has the following equally likely payoffs and returns in one year and pays an annual dividend (in one year from the time of investment): State of the economy
Dividend (D1)
Price (P1)
Good Moderate Bad
$2.00 $1.00 $0.00
$24.00 $22.00 $18.00
Expected
$1.00
$21.33
Dividend yield
Percent appreciation
Total return
10% 5% 0%
20% 10% (10%)
30% 15% (10%)
5%
6.67%
11.67%
With equally likely outcomes, the expected return over the holding period is the average of the three possible returns, which is 11.67%, and the standard deviation of the holding-period return is 16.5%.12 Note that we calculated the
Foundations of New Venture Valuation 367
holding-period returns and standard deviation based on today’s market price of $20. We can use this calculation to infer cost of capital because the stock is traded in a market such that buying the stock today and holding it for one year is expected to be a zero-NPV transaction. Here, we find an important distinction between new ventures and publicly traded stock. Because new ventures are not like shares of stock that trade in a market, we cannot assume that investing is a zero-NPV choice. We will see later that this feature makes RADR difficult to apply unless we can find some way to infer the appropriate discount rate. The price for bearing risk. How do we go from measuring risk to inferring a price for bearing risk? To answer, we must first distinguish between the entrepreneur and prospective investors in a new venture. For investors, we follow the standard corporate finance approach to estimating a risk premium, which relies on the following assumptions: • There is active competition to invest in new ventures. • Investors view new venture investing as an alternative to other investment opportunities. • Investors assess project risk based on its contribution to the risk of a diversified portfolio. • Illiquidity does not affect the investors’ valuation of new venture investment. These assumptions allow us to distinguish between market and nonmarket risk. The total risk, measured as the standard deviation of holding-period returns, is the sum of two categories of risk. The first—called market, systematic, or nondiversifiable risk—consists of marketwide risk factors that affect all ventures. Economic recession and boom, changes in inflation, and changes in the flow of funds into the capital markets are among the factors that tend to affect asset values marketwide. Because market risk affects all assets, it cannot be eliminated by diversifying the investor’s portfolio. The second category—called firm-specific, idiosyncratic, nonmarket, or diversifiable risk—includes all risk other than market risk. Examples of diversifiable risk include success or failure in achieving a technology milestone and idiosyncratic product demand that is higher or lower than expected. Because of diversification, the unexpected positive and negative nonmarket components of returns to different investments are random so that, in diversified portfolios, they tend to cancel each other out. In a portfolio with even a
Chapter Ten Fi g u r e 1 0. 3 How portfolio risk depends on the number of assets in the portfolio The figure shows how diversification reduces total risk as the number of assets in the portfolio increases. Risk is measured by the standard deviation of holding-period returns. The risk that remains is shown as σM and is the market, or nondiversifiable, risk.
Risk
σPortfolio
Standard deviation
368
Diversifiable risk σM
Nondiversifiable risk Number of assets in portfolio
modest number of assets, the idiosyncratic component of risk can be substantially reduced. Diversification cannot be used, however, to reduce exposure to the component of risk that is due to the overall market. Figure 10.3 demonstrates the diversification principle that the total risk of a portfolio, σPortfolio, varies as a function of the number of randomly selected securities. As shown, the total risk of the portfolio is composed of diversifiable (nonsystematic) and nondiversifiable (systematic) risk. By holding more securities, diversifiable risk approaches zero and total risk approaches the risk of the market, σM. Any investor can diversify by investing in an equity mutual fund that is designed to match the performance of a standard market index, such as the S&P 500. Because diversification is easy, investors cannot expect to be compensated if they choose not to take advantage of the simple opportunity to diversify. Thus, the expected return on a risky asset depends only on risk that is not diversifiable. The nondiversifiable component of risk is known as beta (β) risk, or market risk. By convention, the beta risk of the market portfolio is defined as one unit of risk. The risk-free asset, by definition, has no risk and therefore has a beta of zero. An asset with nondiversifiable risk that is half as great as the market has a beta of 0.5, and so on. The notion that investors care only about time value and nondiversifiable risk is reflected in the Capital Asset Pricing Model (CAPM).13 Figure 10.4 shows the relationship between beta risk and expected return that is implied by the CAPM. The CAPM tells us that different investments with different amounts of total risk but equal amounts of beta risk will have the same expected return. Accordingly, all assets and portfolios with the same beta risk plot at the same
Foundations of New Venture Valuation 369
Expected return Security market line (SML)
rm
Market portfolio
rf
1.0
Beta (b )
Fi g u r e 1 0. 4 The capital asset pricing model (CAPM) The CAPM describes the required return on an asset as a function of its nondiversifiable (market) risk as measured by beta. The market portfolio by definition has a beta of 1.0 and is expected to earn rM. The riskless asset has a beta of zero and earns rF, the risk-free rate. The difference between rM and rF is called the market risk premium. If the CAPM correctly describes investor behavior, then all market assets offer risk and expected return combinations that plot on the security market line (SML).
point in the figure. The sloping line in Figure 10.4 is known as the security market line (SML) because all risky assets must plot on the line. The result is a single price, or expected return, for bearing a unit of nondiversifiable, or beta, risk. The difference between the expected return on the market portfolio and the return on a risk-free asset is called the market risk premium. In algebraic form, the CAPM is: rj = rF + βj(rM − rF) (10.3) Comparing Eq. (10.2) and Eq. (10.3), we can see that the CAPM defines the risk premium for an asset as βj (rM − rF), which has two components. The term in parentheses, (rM − rF), is the market risk premium and is computed as the difference between the expected return on the market portfolio and the return on a risk-free asset. This market risk premium is then scaled by βj, the beta or systematic risk of asset j relative to the risk of the market portfolio. The product of the two is the risk premium for asset j. The beta of asset j is the measure of its nondiversifiable risk. Specifically, βj is computed as follows: Covr ,r ρ r ,r σ j j M j M βj = = (10.4) 2 σ σM M In Eq. (10.4), Covr ,r is the covariance of holding-period returns of asset j with the holding-period returns of the market, ρrj , rM is the correlation coefficient of holding-period returns between the asset and the market, σ M2 is the variance of market returns, and σM and σj are the standard deviations of returns of the market and of asset j. j
M
370
Chapter Ten
Equation (10.3) identifies the information necessary to use the CAPM to estimate the required rate of return on an investment. Estimating the risk premium for use in the RADR model requires estimating both the beta of the asset and the market risk premium. Equation (10.4) shows two different ways to estimate beta. As an alternative to estimating beta directly, we can calculate it from estimates of the standard deviations of holding-period returns for both the asset and the market and an estimate of the correlation between asset returns and market returns. In Chapter 11, we examine methods of estimating the information needed to use the CAPM in RADR valuation. Given its focus on nondiversifiable risk, the CAPM is appropriate when investors are able to diversify at low cost. The typical investors in new ventures and in VC funds (pension plans, endowments, and insurance companies) are able to do so. The CAPM suggests that these investors should not require an increase in expected return for bearing venture-specific risk that is diversifiable. Other kinds of investors in new ventures may find diversification more difficult to achieve. In particular, private corporations and high-net-worth individuals, if they want to invest in venture capital, may be compelled to hold portfolios that are not well diversified. If such investors hope to be compensated for sacrificing diversification, they must either find opportunities that well-diversified investors do not recognize or be able to contribute unique value to the venture in other ways that well-diversified investors cannot.
The Certainty Equivalent Approach It sometimes is difficult to use the RADR form of the CAPM to value real investment opportunities with risky cash flows. The model requires that risk be measured as the standard deviation of holding-period returns for an asset that is correctly valued—an equilibrium holding period return. However, even if we can estimate the expected cash flows of a project and their riskiness (i.e., standard deviation), we cannot correctly determine the expected holding-period return or the standard deviation of holding-period returns without knowing the PV of the project. When it is possible to rely on market assets to infer cost of capital, this is not a problem because market assets are zero NPV by definition (cost and present value are the same). However, this does not work so simply for new ventures, where we cannot easily refer to market assets. We can demonstrate the difficulty of using RADR by making a couple of substitutions to Eq. (10.3). Equation (10.5) makes the appropriate substitutions for rj and βj in Eq. (10.3), where Cj is the expected cash flow of project j, σC j is the standard deviation (risk) of this cash flow, and ρC j , rM is the correlation between the project j cash flow and the market return. For convenience, we will define
Foundations of New Venture Valuation 371
rates of return in the preceding expressions over the holding period for Cj (one year, two years, etc.). σC ρC j , rM j Cj PV j − 1 = rF + ( rM − rF ) (10.5) PV j σM In this modified CAPM, the left-hand side of the equation is the expected return and the right-hand side is the expression for the required rate of return. In other words, we must find the required returns at the point where the expected return equals the required return. Note, however, that PVj appears on both sides of the equation. The two sides will be equal only at the point where PVj is the PV of Cj. That is, Eq. (10.5) can be used to search for the PV of a future cash flow by using trial and error to find the value of PVj that equates the two sides of Eq. (10.5). Although this obviates the need to assume a standard deviation of holding-period returns (using, instead, the standard deviation of cash flows), it does not resolve the fundamental problem that the correct risky discount rate depends on the (unknown) value of the cash flows. An illustration of the difficulties of using the RADR approach. To see the nature of the problem, consider a wager that will pay either $1.00 or $2.00, with equal probability. We can easily determine that the risk of the bet (in terms of the standard deviation of payoff cash flows) is $0.50.14 But what about the standard deviation of holding-period returns? Suppose the wager can be acquired for $1.25. If so, the $2.00 payoff is a return of 60% and the $1.00 payoff is a return of −20%. This makes the expected return 20% and the standard deviation of holding-period returns 40%.15 But what if the wager costs $1.50? In this case, the $2.00 payoff yields a return of 33.3% and the $1.00 payoff yields −33.3%. At this price, the expected return is zero, and the standard deviation of holding-period returns is 33.3%. Thus, the standard deviation of holding-period returns depends on the cost of the wager. To find the PV of the wager, we need to know the standard deviation of holding-period returns that corresponds to investing exactly the PV of the expected payout. If the required investment is different from the PV of the expected payout, we can then correctly measure NPV as the difference between the actual investment and the investment that would yield a zero NPV. You can see that applying the conventional RADR form of the CAPM is challenging when it is being used to value real assets. The problem is the inherent simultaneity and the lack of a zero-NPV market asset to use as a basis for inferring the correct discount rate. Valuing the expected cash flows requires knowing the discount rate, but the discount rate calculation requires a beta, which depends on the
372
Chapter Ten
standard deviation of holding-period returns, which, in turn, depends on the PV of the project. In corporate settings, these problems are finessed by analogizing the investment decision to an existing market asset that is publicly traded (and, by inference, fairly priced). If that is possible, a two-step approach can be used: first, estimate the beta of the market asset; second, discount project cash flows using that beta. Unfortunately, convincing analogies are hard to find if the project is a new venture or a financial claim on a new venture. In Chapter 11, we seek to overcome this problem by providing some empirical estimates of betas for publicly traded corporations that are similar to entrepreneurial ventures. If those estimates are reliable, then using the RADR approach is straightforward, just as it is for public corporations. If not, the CEQ approach can be easier to use. Advantage of the CEQ method. The certainty equivalent of a risky cash flow is the certain cash flow that, if received at the same time as the risky cash flow, would be equally valuable to the investor. For risk-averse investors, the certainty equivalent is less than the expected risky cash flow. Thus an investor might regard a risky cash flow with an expected value of $200 and a standard deviation of $50 as being equal in value to a certain $125 cash flow. In the CEQ approach of DCF valuation, instead of adjusting the discount rate, the risk adjustment is made to the cash flow: CE(Cjt) = Cjt − RDjt (10.6) where CE(Cjt) is the certainty equivalent of the period t expected cash flow, Cjt, and RDjt represents the dollar-valued discount to Cjt that is required to convert the risky expected cash flow to its certainty equivalent. Once we have this stream of certainty equivalent cash flows, we can convert them to PV using a risk-free discount rate as follows:
PV j = Σ t
CE (C jt )
(1 + rFt )t
(10.7)
where rFt is the risk-free rate for period t cash flows. To apply Eq. (10.7), we need a way to compute the CEQ cash flow. Because the CEQ form of the CAPM uses the standard deviation of cash flows instead of the standard deviation of holding-period returns, it avoids the simultaneity problem of the RADR form. The CEQ approach is general in that it does not impose any particular tradeoff between risk and return. If we assume that the CAPM is the correct asset pricing model, then Eq. (10.5), the expanded version of the RADR form of the CAPM, can be solved for PVj to yield the CEQ form of the CAPM.
Foundations of New Venture Valuation 373
PV j =
C jt
ρC
j , rM
σ
Cj
(rM − rF ) σM (10.8) 1 + rF
The numerator in Eq. (10.8) is the CAPM-based certainty equivalent of the risky cash flow, Cj, which corresponds to Eq. (10.5). The denominator is a discount factor that is used to determine the PV of a riskless cash flow.16 When the CEQ form of the CAPM is used, the risky cash flow is adjusted by a factor that makes the PV of the CEQ cash flow, when discounted with a risk-free rate, exactly equal to the value derived by discounting the uncertain expected cash flow at the appropriate risky rate. To see how the CEQ approach simplifies finding PV, let’s revisit the wager that will pay either $1.00 or $2.00 with equal probability. We assume that the risk-free rate is 4%; the market risk premium is 6%; the standard deviation of holding-period returns of the market portfolio is 20%; and the correlation between the payoff of the bet and the return you could earn by investing in the market portfolio is 0.6. Given this information, it is easy to use Eq. (10.8) to determine the PV of the wager.
PV j =
cj
pc j , rM σ C j
σM 1 + rF
( rM − rF )
=
$1.50 −
0.60$0.50 (0.06) $1.41 0.200 = = $1.356 1 + 0.04 1.04
The CEQ cash flow of the expected (risky) $1.50 is $1.41, which when discounted at the riskless rate yields a PV of $1.356 for the wager. You can verify that if this value is used in Eq. (10.5), then the two sides are equal (i.e., required return equals expected return at a PV of $1.356). Given the PV of the wager, the NPV of the opportunity to acquire it for $1.25 is $0.106 (i.e., $1.356 − $1.25 = $0.106). We also can determine that the correct risk-adjusted discount rate (the cost of capital), given the PV of $1.356, is (Cj /PVj ) − 1, or 10.62%. Because Eq. (10.8) adjusts for risk by using the correlation between project cash flows and market returns, it circumvents the need to determine the riskiness of the holding-period returns. However, it does raise another question: How can we estimate the correlation between project cash flows and market returns? We address this problem later in the chapter. A similar information requirement exists if the RADR form is used, but in that case, the need is to estimate the correlation between project returns and market returns. The problem for the RADR form is sometimes finessed by analogizing the project to a publicly held corporation and using estimates of beta from published information.
374
Chapter Ten
10.5 The Relative Value Method Relative valuation (RV) uses data from public companies and public and private market transactions to estimate value. It is widely used in the real estate industry, especially for residential real estate, for which it is particularly well suited. Owner-occupied housing does not produce observable cash flow streams that can be valued by DCF. Moreover, there are many trades of highly similar properties, where differences that affect value tend to be associated with observable factors such as square footage and number of bedrooms. For ventures, the RV method is commonly used to value companies at the time of IPO or acquisition.
Relative Valuation and New Ventures We can apply RV to new ventures by gathering information on the value drivers of comparable firms. A firm’s value is driven by its profitability, expected growth, and risk. In RV analysis, we can classify multiples into two categories: (1) multiples based on the capitalized value of equity and (2) multiples based on enterprise value, where enterprise value is defined as the market value of equity, plus the market value of interest-bearing debt, minus excess cash. For consistency of comparison, before computing the multiples, excess cash (cash that is not needed for operation of the enterprise) should be subtracted from both the subject asset and the comparables. After the multiples are applied to estimate the enterprise value of the subject, any excess cash that is expected to be accumulated by the subject can be separately discounted to PV at the risk-free rate, analogous to distributing the cash as a dividend. For a firm such as Apple, which in 2017 had over $74 billion of cash and short-term investments (24% of total assets), this is not a trivial adjustment. We can also classify multiples as either (1) accounting based or (2) non-accounting-based. Accounting-based approaches. Accounting-based multiples relate the value of venture equity or total capital to reported accounting information. The most familiar of these, the P/E ratio, is a way to estimate the value of equity on the basis of reported net income. Other accounting-based approaches to valuing equity include the following: • Price to cash income, which is measured as net income plus noncash expenses • Price to levered free cash flow, which differs from the previous measure by subtracting increases in net working capital and fixed assets
Foundations of New Venture Valuation 375
• Price to book value of equity (more commonly referred to as market value to book value of equity) Accounting-based approaches can also be used to estimate the combined value of equity and debt, or enterprise value. The commonly used measures are similar to those for valuing equity, except that they are usually based on cash flow measures including interest expense. Such approaches include the following: • Enterprise value to EBITDA (earnings before interest, taxes, and depreciation and amortization) • Enterprise value to unlevered free cash flow, which is EBIT minus theoretical taxes as if the company were entirely equity financed • Enterprise value to book value of debt and equity • Enterprise value to sales While accounting data are readily available for public firms, accountingbased valuation multiples have two major problems. First, they are subject to accounting choices that can make comparison across firms inaccurate. Second, the data in public financial statements are based on historical results, whereas value today depends on expected future performance. While accounting-based comparisons can work well for valuing established private businesses with simple business formats, such as dry cleaners or donut shops, they are not easily applied to early-stage ventures. Because new ventures are often unprofitable at the valuation date, may not have initiated sales, and may also have negative net worth, many of the commonly used accounting measures may not be meaningful. The VC Method, discussed next, is designed to get around this problem while retaining the simplicity of RV. Problems related to differences in accounting practices can be overcome with careful analysis. We can also overcome the backward-looking nature of accounting data by performing the RV analysis using forward projections for the comparable firms. For example, instead of calculating a P/E ratio based on the prior year’s earnings, we can base the value estimate on multiples tied to earnings forecasts. Of course, simple ratio-based comparisons are only helpful if the subject venture and the comparables have similar expected growth rates. There are other, more formal, approaches, as well, for incorporating growth expectations into multiples analysis. The following are two of the most common:
376
Chapter Ten
• The PEG ratio (or P/E/growth), where growth is the expected growth rate of earnings • Enterprise value to EBITDA/growth, where growth is the expected future growth rate of EBITDA The growth-adjusted approaches are ways of dealing with differences in growth expectations between the subject venture and the comparables. Non-accounting-based approaches. The other category of multiple we consider is industry-specific nonfinancial metrics. An online subscription-based venture could be valued on the basis of the projected number of subscribers or estimated web pages of advertising. A proxy for the value of a biotechnology venture might be the number of patents, while a pharmaceutical venture’s value might be estimated using the number of products in various stages of FDA approval. Internet ventures have been valued based on the number of website visits and time spent on the site.17 Depending on the venture’s stage of development, the focus of non-accounting-based valuation might be on the drivers of revenue. Revenue-related metrics include, for example, the number of visitors per month, the fraction of visitors who become customers, the percentage of visitors who linger on the website instead of leaving quickly, average order size, monthly active users, revenue per active user, recurring (subscription) revenues, and annualized monthly revenue. For a more seasoned venture, useful metrics might include contribution margin per order or customer, the average cost of acquiring a customer, the customer churn rate (lost customers/total customers), the cash burn rate per month, and the average lifetime value of a customer. A number of websites provide benchmarking data that can be used to estimate value relative to other companies with established values. See, for example, SimilarWeb.com, JumpShot.com, and SemRush.com. There is some evidence that industry-specific non-accounting-based measures can be better value predictors than accounting-based multiples.18 For example, for early-stage Internet companies, commonly used valuation metrics include monthly unique visitors, conversion rates (visitors into customers), bounce rates, and average order value. Industry-specific multiples may be used to estimate either equity or enterprise value. The important consideration is consistency. If the comparables used in the valuation are materially debt financed but the venture is entirely financed with equity, then the comparables ratios need to be based on enterprise value.
Foundations of New Venture Valuation 377
An illustration: initial public offering. Here is a simple example to illustrate the concepts. Suppose your social media company is considering an IPO. The company’s most recent annual EBIT is $2.2 million on revenue of $35 million, coming primarily from the sale of advertising. The company currently has 15 million users per day and the growth rate of users per day in the most recent year was 10%. You have also collected the following accounting-based and non-accounting-based data on recent IPO transactions: Comparable transaction
Value at IPO (million)
Revenue (million)
EBIT (million)
Company A Company B Company C Company D Average
80 120 135 55 100
$36.360 $38.710 $79.410 $13.750 $42.058
$2.96 $1.88 $3.75 $0.71 $2.325
Users per Value/ day Growth rate Revenue at (million) of daily users IPO
33.10 11.00 16.80 9.20 17.53
6.0% 12.0% 9.0% 21.0% 12.0%
2.20 3.10 1.70 4.00 2.75
Value/ EBIT at IPO
Value/ Daily user at IPO
27.03 63.83 36.00 77.46 51.08
2.42 10.91 8.04 5.98 6.83
There are different ways to use the information from the comparables to estimate the enterprise value of the subject company. One simple approach, the one we illustrate here, is to consider each factor one at a time and compare the averages to the subject. If the comparables include large outliers, it can also be better to focus on medians instead of averages. A better approach, if sufficient data are available, is to fit a statistical model to the data, for example, by using multiple regression. The model can be used to assess the marginal effect on value of each factor in conjunction with the effects of the others. The last row of the preceding table shows the averages of the key variables and enterprise value (EV) multiples. Averaging the results across the four comparable companies yields the following estimates of enterprise value: Mkt Value ) × RevenueSubject = 2.75 × $35.0 = $96.25 million Revenue Mkt Value = Avg ( ) × EarningsSubject = 51.08 × $2.2 = $112.38 million Earnings Mkt Value = Avg ( ) × UsersSubject = 6.83 × 15.0 = $102.52 million Users
EV Est .subject = Avg ( EV Est .subject EV Est .subject
All three approaches yield similar estimates. Averaging the three produces an estimated enterprise value of $103.72 million. How reliable is this value? One positive indication is that the 10% growth rate in average users per day is similar to the 10.5% average over the comparable companies. Another indication is that the valuation based on the averages of comparable companies is almost the same as the median (not shown) for the
378
Chapter Ten
comparables. The implied value based on the medians of each factor would be $105.10. We can also look separately at each comparable. Company A yields the lowest implied value, $57.57 million based on the average of the three measures. But Company A has the lowest user growth rate of any of the comparables. Company B yields the highest implied value, $137.52 million, but has a higher growth rate of users and a high value per existing user. Overall, it appears that our valuation is reasonable. Private value versus market value. It is important to be cognizant of the assumptions implicit in the choice of comparables. If, for example, a private venture is valued on the basis of public company multiples, then the result is an estimate of the value as if the venture were public. If the comparables are from private transactions, such as acquisitions, then the estimate is as if the venture were to be acquired. In principle, if everything is properly controlled, there may be no difference between the values implied by the two different kinds of comparables data. However, the choice to go public or exit via acquisition is not arbitrary. Some companies are better suited for IPO and others are better suited for acquisition. The valuation is likely to be more credible if the comparables represent the exit choice that is most appropriate for the subject venture. The main drawback of relative value multiples is that current financial measures are linked primarily to assets in place, which may bear little relation to likely future performance.19 Furthermore, comparable firm and comparable transaction data can be difficult to collect.20 That said, practitioners frequently use multiples, so we include them here in our discussion and offer suggestions to help overcome their shortcomings.
10.6 Valuation by the Venture Capital Method The VC Method is the traditional approach for VC investment valuation. It is also the simplest approach for valuing early-stage ventures.21 In the VC Method, value is estimated on the basis of projected harvest-date cash flows under the assumption that the venture meets its performance objectives. The scenario in which the venture meets its objectives is referred to as a “success scenario.” The procedure is as follows:
Foundations of New Venture Valuation 379
1. Select a terminal year for the valuation by determining a point where, if the venture is successful, harvesting by acquisition or IPO would be feasible. Estimate net income or other cash flows in that year based on the success scenario. 2. Use the appropriate P/E or other multiple and the harvest-date earnings or cash flow projection to compute continuing value. The multiple should reflect the expected capitalization of earnings or cash flow for a company that has achieved the level of success reflected in the scenario. 3. Convert the continuing value estimate to PV by discounting at a hurdle rate that is high enough to compensate for time value, risk, the probability that the success scenario will not be achieved, and dilution from anticipated future funding rounds. 4. Based on estimated PV, compute the minimum fraction of ownership an investor would require in exchange for contributing a given amount of capital. The VC Method is problematic for several reasons. Most fundamentally, it is based on an optimistic forecast of the future.22 While the hurdle rate is intended to compensate for the optimistic cash flow estimate, there is no indication of the likelihood that the success scenario will be achieved. To compensate for the optimistic forecast of continuing value, the hurdle rate must be well above cost of capital. The approach can easily be based on hurdle rates in excess of 50%. In addition, the VC Method does not consider cash flows in the explicit value period. This could make sense if the expected cash flows during the explicit value period are very small compared to continuing value. If they are not, the problem can be addressed by adding the PV of the cash flows from the explicit value period to the PV of continuing value. Because these cash flows are from a success scenario, the model implies that they again would be discounted at a high rate. To illustrate, recall the stock price example from Section 10.4. If we were to value the stock using the VC Method, we would consider only the good state (i.e., the success scenario) as shown here. State of the economy
Good
Dividend (D1)
Price (P1)
Dividend yield
Percent appreciation
Total return
$2.00
$24.00
10%
20%
30%
We know that the actual PV is equal to today’s market price of $20, and in Section 10.4, we established that the cost of capital (i.e., the expected return) taking account of all three possible outcomes is 11.67%. To get the same
380
Chapter Ten
$20 market value by the VC Method, we would need to discount the Time 1 success-scenario price of $24, plus the $2 dividend by a 30% hurdle rate, well above the actual 11.67% cost of capital. The Achilles heel of the VC Method is that intuition and experience form the basis of the hurdle rate to use. So while the method might work reasonably well for an investor who has years of experience of experimenting with different hurdle rates in different circumstances, there is no real guidance for an entrepreneur, a first-time investor, or a student who aspires to become a VC or entrepreneur. In spite of these problems and challenges, the VC Method continues to have some appeal because of its simplicity, intuitiveness, and potential usefulness in negotiating with an entrepreneur.
10.7 Valuation by the First Chicago Method Under the First Chicago Method, a user identifies a small number of discrete scenarios and values them using a discount rate that is reflective of cost of capital. The First Chicago Method is designed to be simple to use but also to address the biases of the VC Method. Usually the First Chicago Method is based on three scenarios: “success,” “sideways,” and “failure.” The success scenario might be the same as in the VC Method, possibly the entrepreneur’s projection. The failure scenario is one in which investors realize essentially no return on their investment and lose the principal. The sideways scenario is one of moderate performance in which the venture languishes with no prospect of a high-valued outcome. In this scenario, investors might earn a preferred dividend return and recoup the initial investment but nothing more. If the First Chicago Method is intended to correct the biases of the VC Method, the scenarios and their probability weights will be chosen so that opportunity cost of capital is the correct discount rate to use to value the scenario cash flows. Because the stock price example in Section 10.4 simplifies the stock price outcomes to three equally likely scenarios, we can think of it as an example of the First Chicago Method. Implementation of the method involves the following steps: 1. Select a terminal year for the valuation based on the likely harvest date in the event of success. 2. Estimate the cash flows during the explicit value period based on a small number of discrete scenarios.
Foundations of New Venture Valuation 381
3. Compute the continuing value by applying a multiplier to the financial projection. The multiplier for the success scenario should reflect the expected capitalization for a company that has achieved the level of success reflected in the scenario. The multiplier for the sideways scenario may be different, depending on differences in expected growth rates used in the capitalization. In the failure scenario, the venture probably is not sold, but the liquidation value, if any, of the assets should be incorporated in the cash flows. 4. Compute the expected cash flow in each period by appropriately weighting the scenarios. 5. Compute PV by discounting the expected cash flows, including expected continuing value at opportunity cost of capital. 6. Based on PV, determine the minimum fraction of ownership the investor should require in exchange for contributing a given amount of capital. The First Chicago Method makes fundamental sense because the goal is to value expected cash flows at opportunity cost of capital. Traditional application of the First Chicago Method, however, does not offer much guidance on how to determine the cost of capital. It might seem that the dispersion information reflected in the cash flow scenarios should be useful. Later, we will see that it is, but in the context of the CEQ model. In fact, if the CEQ model is applied to similar, discrete scenarios, it is the same as the First Chicago Method with the CAPM being used to value the cash flows. Ultimately, it is not possible to value future cash flows by either the VC Method or the First Chicago Method without relying on some asset pricing model to determine the appropriate discount rates. The asset pricing model might be based on experience and rules of thumb. It might be based on simple comparisons to other assets in the market. Or it might be derived from a formal asset pricing model, such as the CAPM.
10.8 Reconciliation with the Pricing of Options Option pricing theory implies that the values of some kinds of claims increase with risk.23 Because new ventures are, in effect, portfolios of options, you might ask whether the Option Pricing Model (OPM) would be a better valuation model for our purposes. To answer, we need to examine the principles underlying the OPM.
382
Chapter Ten
Risk contributes positively to the value of options because options partition risk into the risk of an increase in the price of an underlying asset and the risk of a price decrease. An investor who is optimistic about the future performance of a share of stock can buy a call option on the stock instead of the underlying share. If the stock value increases, the call option will increase as well. However, unlike an investor in the stock, the call option investor is protected against price declines. If, on the date the option expires, the stock value has fallen to below the exercise price of the option, then the option is worthless. Conversely, if, on the day the call option expires, the stock price is above the exercise price, the value of the option will equal the difference between the stock price and the option exercise price. Because the exposure to risk is not symmetric (only the upside matters), the value of a call increases with anything that increases the riskiness of the underlying shares. There is no inconsistency between using the OPM to value puts and calls and using the CAPM to value the underlying asset. The key to reconciliation is that an option is a derivative asset: its value is “derived” from the value of the underlying asset and from the risk characteristics of the underlying asset. Because the underlying asset does not partition risks into good and bad, an investor in the asset must bear the risk of loss to acquire the potential for gain. In that setting, risk aversion leads to the conclusion that investors will require compensation for bearing risk that is not diversifiable. Although the CAPM and OPM are consistent, certain aspects of the OPM make its application to valuation of new ventures difficult. One problem is that the OPM is derived under an assumption of market completeness and continuous trading of assets. In general terms, a complete market is one with negligible transaction costs and where there is a price for every asset in every possible state of the world. A complete market, in our narrower context, means that the underlying asset, matched pairs of puts and calls (with the same exercise price and expiration date), and riskless debt must all be continuously available, and it must be possible, at low cost, to take long or short positions in each. In standard pricing of options on freely tradable assets, by either the Black-Scholes Option Pricing Model or the Binomial Option Pricing Model, the current price of a freely traded underlying asset is established in a market so that, implicitly, there is agreement on value. Riskless debt with the same maturity as an option on the underlying asset is also freely traded. The logic of option pricing is arbitrage. Suppose, for example, that there are only two future states of the world and that a share of stock that sells for $10 will, in one year, pay $15 if the good state occurs or $9 if the bad state occurs. Riskless debt with the same maturity will return 10%. You would like to price a one-year call option with an exercise price of $10. The option will pay either
Foundations of New Venture Valuation 383
$4 or $0. In this case, we can replicate the payoffs of the call option by investing in two thirds of a share of the underlying stock on margin. The stock investment will cost $6.67. To construct the call option payoff profile, you will pay $1.22 of this price yourself and will borrow the remaining $5.45. At the time of option expiration, the stock investment will pay either $10 or $6, but you will need to pay off the margin loan and accrued interest, at a total cost of $6, leaving you with either $4 or $0, the same as if you had invested in the call option directly. Hence, by this replication argument, the current price of the call option must be $1.22. In this example, we know the current stock price, the conditional payoffs of the stock in one year, and the interest rate on riskless debt, and we can use this information to find the conditional payoffs of the call option. We do not know the probabilities of the good and bad states in one year. Nor do we know how the market determined that those conditional payoffs had a current value of $10 per share. We can, however, narrow the feasible range of state probabilities. If investors are risk-averse to any extent, since $11 is the one-year payoff on a $10 investment in riskless debt, the true probabilities from investing in the one share of stock must result in an expected payoff greater than $11. Mathematically, we can determine that a 16.7% probability of the good outcome (a payoff of $15) would make the expected payoff $11, exactly the same as the risk-free payoff. So the implied probability of the bad state must be 83.3%. These state probabilities that would make the expected payoff the same as the risk-free payoff are what are referred to in standard option valuation as “risk-neutral probabilities.” If investors are risk-averse, the true probability of the good state must then be more than 16.7%. Since standard option pricing relies on riskless arbitrage and a known set of current and contingent future prices, we can price the option even though we do not know the true probabilities. These conditions or conditions like them that hold reasonably well for publicly traded options on publicly traded stock are the core of the Black-Scholes and binomial option pricing models. They also may provide reasonable approximations for options on assets such as gold mines, the values of which are closely linked to the value of gold, a freely traded asset. Clearly, however, they do not hold for many high-risk, high-potential new ventures. Investors and entrepreneurs are likely to disagree about the current value and about future state-contingent payoffs. The underlying asset, one that can service the full potential product-market demand and does not provide for abandonment, is virtually never observed or priced. As a consequence, riskless arbitrage between options and underlying assets is not possible. The upshot is that a replicating portfolio is unlikely to exist and the standard OPM logic is likely to misvalue the real options that comprise most new ventures. Moreover, as a new venture is an amalgam of many interrelated options, some with complex exercise provisions
384
Chapter Ten
that are controlled by different parties, use of the OPM can quickly become impractical. The alternative we use is pricing by simulating the uncertainties in a decision tree that incorporates the important real options. In summary, the choice of real option valuation methods depends on what you are believe you know, and what you don’t. If you believe you know the statecontingent payoffs and that you can replicate the payoffs with a freely traded asset that has the same risk profile, but you do not know the state probabilities, you can use market data and the standard OPM replication approach to value the venture and price the real options. Alternatively, if the option structure is complex and there is no market asset that can replicate the venture, but you think you know the probabilities of option exercise and associated future values, you can use the decision tree simulation approach similarly to what was illustrated in Chapter 6 to value the venture. In a later chapter, we present a simulation approach to valuing new ventures and financial claims based on the CEQ form of the CAPM. The approach provides a convenient way to estimate the effects of embedded real options on the value of a new venture.24
10.9 Required Rates of Return for Investing in New Ventures Early in the chapter we promised a resolution of the paradoxical difference between the rates of return sought by VC investors and those realized by the investors. The CAPM helps with the explanation and offers some supporting evidence. Assuming that the CAPM is a reasonable way to estimate the rates of return investors in new ventures require, how high would those rates be? Clearly, a new venture is a risky proposition, which might suggest that the expected returns should be very high. On the other hand, the point of the CAPM is that only nondiversifiable risk matters. It seems likely that much of the risk associated with new ventures is diversifiable or firm specific. Consider, for example, a biotechnology venture. How much of the risk is likely to be nondiversifiable? In many respects, a biotechnology venture is like a lottery: by investing, you place a bet today, and at some future date, you learn whether you have a winning ticket. The future payoff from such an investment is highly unpredictable and depends on factors such as how significant the innovation proves to be. The payoff is not likely to depend very much on economic fluctuations or similar factors that would have much greater effects on such sectors as discretionary consumer durables.
Foundations of New Venture Valuation 385
Because beta risk depends on how the venture’s payoff varies with marketwide fluctuations, a biotech venture is likely to have a low beta despite its high total risk. At the same time, the cash flows from investing in a portfolio of 100 different biotech ventures might not be very risky at all. Most of the risk is specific to the individual venture and would be substantially reduced by diversifying, even within the same industry. In fact, for a large sample of public biotechnology ventures, we estimated an average beta of 0.75 but an average total risk almost five times as high.25 We will see in Chapter 11 that estimating the beta of an individual new venture requires judgment. However, we can gain information about the betas of new ventures as a group by studying the price performance of publicly traded firms that were funded by VCs. Such firms typically have betas in the range of 1.0 to 2.0. What do betas in this range imply for the required rates of return on new venture investments? We can answer by looking back at Eq. (10.3) and filling in the other pieces of the equation with reasonable values. Suppose we use 4.0% as an estimate of the riskless rate of return, rF, and 8.0% as the historical average market risk premium, rM − rF.26 Beta values in the 1.0 to 2.0 range would imply required rates of return in the 12–20% range. Such low required rates of return seem inconsistent with the earlier-mentioned claims that, for investing in new projects, VCs sometimes seek rates of return in excess of 50%. On the other hand, they are fully consistent with the historical evidence on VC returns cited earlier in this chapter. To look more specifically at the rates of return required by VC firms, the discount rate for VC investing can be disaggregated into four components as follows:
VC rproj = rF + β proj (rM − rF ) + Effort + Liquidity (10.9)
VC where rproj is the discount rate used to value the VC’s claim on the value of a project, is the beta risk of the project, Effort is a measure of the cost of effort committed to the project by the VC and expressed in returns form, and the last term is a measure of the required return differential due to illiquidity of the investment. Note that rF + β proj (rM − rF ) is just the CAPM applied to the project. With the model from Eq. (10.9) in mind, an average gross realized return as high as 25% or 30% on VC investing could be understandable. The gross return is different from the net return because, as discussed in Chapter 3, it includes the fees charged by the GP and the portion of capital appreciation (the carried interest) that is retained by the GP.
386
Chapter Ten
We previously calculated, based on historical data, that a beta of 2.0 yields a cost of capital for an entrepreneurial venture of about 20%. This corresponds to the returns realized by LPs. VC limited partnerships segregate returns to the GP (the active investor) from returns to LPs (passive investors). The GP’s return is typically 20% of fund returns after initial investment capital has been returned to the LPs, and after the annual management fee, which is equal to about 2.5% of committed and invested capital.27 A reasonable estimate, then, is that one fifth of a 25% gross return is actually compensation to the GP. This return leaves nothing to compensate for the illiquidity of the investment. However, an adjustment for illiquidity of near zero comports with recent empirical evidence on the cost of illiquidity.28 Low compensation for bearing illiquidity risk also makes sense given the fact that most investors in VC limited partnerships are institutions with long-term investment horizons, at least for some fraction of their portfolios. It is reasonable to expect that competition would drive them to demand a negligible return premium to compensate for illiquidity. Why does the CAPM accurately predict average returns of VC investors? In large part, the answer appears to be that assumptions underlying the CAPM are reasonably well satisfied. Most importantly, any venture is likely to represent only a small fraction of the asset portfolio of a large pension plan or life insurance company. For such investors, a substantial component of the total risk is diversifiable. How, then, do we account for the high sought-for rates of return? Our answer is that those returns represent hurdle rates that investors sometimes use to value cash flow projections that are developed on the presumption that the venture will be successful. VCs know that a large fraction of their deals will fail completely and others will limp along offering breakeven returns at best. These will be offset by a few home runs that will offer spectacular returns, sometimes returning 5 or 10 times the initial investment. All of the prospective ventures have rosy forecasts at the time of investment, and there is no reliable way for the VC to know ex ante which will be home runs and which will fail. To compensate for the optimism built into the cash flow projections, the investor applies a hurdle rate that is substantially above the required return for investing in the project.29
10.10 The Entrepreneur’s Opportunity Cost of Capital A unique challenge to new venture contracting is that the entrepreneur necessarily invests a large fraction of his financial wealth and human capital in the venture. The resulting underdiversification implies that the entrepreneur faces
Foundations of New Venture Valuation 387
a different risk-return tradeoff than does a well-diversified investor such as an LP in a VC fund. The entrepreneur’s decision to pursue a new venture is among the most difficult to make correctly. Yet research and books on entrepreneurial finance take the entrepreneur’s investment decision as a forgone conclusion. A common perception is that the entrepreneur’s investment problem is less a quantitative decision than a qualitative one, having to do with wanting to be one’s own boss, one’s lifestyle preferences, and one’s tolerance for risk. This focus is too narrow. The entrepreneur’s investment decision problem is fundamentally different from the investor’s. Even though the underlying financial economic theory is the same, because of necessary underdiversification, the opportunity cost of capital that is appropriate for the entrepreneur is different from that of the investor. This is true even if the parties hold identical financial claims. The entrepreneur should take a separate look at value for three compelling reasons: 1. Underdiversification causes differences in required rates of return. Because an entrepreneur is necessarily sacrificing some ability to diversify, the entrepreneur’s opportunity cost of capital is higher than that of investors. 2. Ownership claims of investors and entrepreneurs usually are not identical. As we have seen, investors often receive “sweeteners” such as options and preferences that make the deal more attractive. Sweeteners raise the value of an investor’s ownership claims and therefore reduce the value of the entrepreneur’s claim. 3. The parties may have different beliefs about expected performance and risk. The entrepreneur and the investor are unlikely to agree about the risk and expected return of the venture. Differences in expectations about such things as development timing, achievable sales, and profitability make contracting more difficult. Building on financial economic theory from this chapter, we provide a valuation framework that addresses the entrepreneur’s underdiversified risk. The framework can be applied in the context of differences in expectations and financial claims that differ in priority.
Opportunity Cost and Choosing Entrepreneurship What drives the decision to become an entrepreneur? Opportunity cost is key to answering this question, along with related questions such as: Why, compared to in the U.S., do greater percentages of the populations in Brazil, Lebanon, and South Africa, for example, engage in entrepreneurial activities,
388
Chapter Ten
and why do lower percentages in countries like France and Sweden? Why do students drop out of college to start businesses? And why is the probability of starting a new business in the U.S. positively related to wealth whereas in some other countries there is a negative relationship between per capita wealth and engagement in entrepreneurship? Research indicates that, in general, the decision to become an entrepreneur rests on rational assessments that individuals make based on two components of opportunity cost—the opportunity cost of their committed effort (their human capital) and the opportunity cost of their invested wealth (financial capital). At least intuitively, individuals try to compare the expected value of starting their own businesses with the value of the next best alternative, which for many is staying on the employment track. The patterns of entrepreneurial activity described here are explained by two factors, both reflective of opportunity cost. First, in countries with fewer good employment opportunities, individuals are more likely to turn to self-employment out of necessity. In countries such as the U.S., where attractive career opportunities are usually available, there are fewer entrepreneurs as a percentage of the population than in countries with chronically limited employment opportunities. However, the opportunity cost of pursuing an entrepreneurial opportunity is not very high for a student, who can easily return to student status if the venture fails, or for an employee, who can quickly return to similar employment. Second, the ability to diversify financial assets can significantly reduce the entrepreneur’s opportunity cost of capital and make entrepreneurship more attractive. Wealthy individuals are able to commit smaller fractions of total wealth to an entrepreneurial venture and thus can be better diversified. Other things being equal, greater wealth lowers the opportunity cost of becoming an entrepreneur. Of course, factors other than opportunity cost also affect the decision to become an entrepreneur, such as a desire to be one’s own boss or to pursue a particular lifestyle, as well as differences in risk tolerance. In this section, we set these other considerations aside to focus on opportunity cost. We analyze, at a conceptual level, how risk and the opportunity cost affect the decision. To ensure clear understanding, we provide formal analysis of the underlying foundations and derivations. To make application more practical, Chapter 11 includes discussion of shortcut approaches that make the entrepreneur’s valuation problem more tractable.
The Entrepreneur as an Underdiversified Investor Often, an entrepreneur must commit most of his time, at least for a few years, as well as a substantial fraction of his financial capital to the venture. As a
Foundations of New Venture Valuation 389
result, the entrepreneur necessarily bears not only the nondiversifiable risk of the venture but also the total risk. The focus on total risk as a determinant of the entrepreneur’s required rate of return is appropriate in part because entrepreneurial ventures are not market assets. To illustrate the differences between a well-diversified investor and the entrepreneur, consider an individual who has total wealth of $300,000, which includes both financial and human capital. The entrepreneur is considering whether to invest one third of his total wealth ($100,000) in a venture that will pay off in one year. Excluding the entrepreneur’s salary, there are three equally likely payoffs on his investment, as shown: Scenario
Probability
Year 1 payoff
Return
Success Likely Failure
1/3 1/3 1/3
$200,000 $125,000 $53,000
100% 25% (47%)
Using the required investment of $100,000, the expected return and standard deviation of returns on the investment are 26% and 60%, respectively. The 60% standard deviation represents the total risk of the new venture. Assuming a 4% risk-free rate, an expected return on the market of 12%, a standard deviation of market returns of 15%, and a correlation coefficient of the venture’s returns with the market of 0.2, our estimate the venture’s beta is
ρrj , rM σ j
0.2 × 0.60 = 0.80 0.15 σM With this beta, the CAPM return that a well-diversified investor would require in exchange for investing in the new venture is 10.4%:
βj =
=
rj = rF + βj(rM − rF) = 4% + 0.8(12% − 4%) = 10.4% The anticipated 26% return indicates a positive NPV for a well-diversified investor. However, based on the opportunity cost of bearing risk, considering his inability to fully diversify, the entrepreneur possibly should not settle for a 26% return. Suppose the entrepreneur were to use leverage to achieve the same level of risk by investing only in the market portfolio. His expected return would be higher than 10.4%. But how much higher? We can use this opportunity cost reasoning to compute the entrepreneur’s required rate of return for investing in the new venture. We start by considering the standard deviation of the entrepreneur’s portfolio, which (we assume) is divided between the new venture and the market. The standard deviation of a portfolio of two risky assets is calculated as follows:
390
Chapter Ten
σ port = xM2 σ 2M + xP2 σ 2P + 2 xM xP ρM , P σ M σ P (10.10)
The variables xM and xP are the value weights of total risky investments in the market and the project, respectively, where the weights sum to 100%. The variable ρM,P is the correlation coefficient between the market and the project. In our example, the weights in the market and the venture are 2/3 and 1/3, respectively, which yields a portfolio standard deviation of 24.1%:
σ port = (2 3) 2 (0.15) 2 + (1 3) 2 (0.60) 2 + 2(2 3)(1 3)(0.2)(0.15)(0.60) σ port = 0.241 or 24.1% The following formula allows us to calculate a portfolio’s required return, using the total risk of the portfolio as compared to the market: rport = rF + (σ port σ M ) RPM (10.11)
Using the CAPM-based approach, the entrepreneur’s required return on his risky portfolio is rport = rF + (σ port σ M )(rM − rF ) = 4% + (24.1% 15%)(12% − 4%) = 16.85% Now that we have the required return on the portfolio, we can use the fact that the required return on the portfolio is equal to the weighted average of the required return on the project and the required return on the market. That is, rport = x p rp + xM rM (10.12)
Because we know everything except the required return on the project, we can rearrange to solve for rp, the project required return:
rport =
rport − xM rM xP
=
16.8% − (2 3)12% = 26.55% (10.13) 13
Therefore, the opportunity cost of investing one third of total wealth in the new venture is a 26.55% return—slightly higher than the 26% expected return on the project. From a purely financial perspective, this is the return the entrepreneur would need in order to forgo an investment in the market portfolio that was leveraged to achieve the same total risk as the entrepreneur’s portfolio. We can conclude that if the entrepreneur would have to invest one third of total wealth in the venture, the NPV of the investment would be negative. Note, however, that if the entrepreneur could find a way to invest a smaller fraction of wealth in the venture, his required return would be lower, and possibly the project NPV would be positive.
Foundations of New Venture Valuation 391
This example makes clear that the entrepreneur’s required return, even with less than a full commitment of wealth to the new venture, is substantially higher than the return required by a diversified investor.
Factors That Offset the Entrepreneur’s Cost-of-Capital Disadvantage Based on the preceding discussion, the entrepreneur’s cost-of-capital disadvantage may seem daunting. In a competitive market for launching new ventures, we should find that, because of their lower cost of capital, public corporations can spend more resources searching for viable ventures than can entrepreneurs and that corporations can legitimately undertake ventures before they reach the threshold of economic profitability for underdiversified entrepreneurs. Offsetting their cost-of-capital advantage, however, corporations face challenges when they seek to engage in entrepreneurship. For example, because of internal equity concerns, the maximum reward a corporation can offer to an employee for perceiving and pursuing entrepreneurial opportunities can be less than what an individual can realize by acting alone. Consequently, corporate employees with good new venture opportunities frequently jump ship to pursue them on their own. As a result, the venture cash flows that are available to entrepreneurs may not be equally available to large corporations. Hence, a trade-off exists between some of the organizational efficiencies and expediencies of stand-alone entrepreneurial ventures and the cost-of-capital advantage of a public corporation. This trade-off gives insight into the kinds of ventures that are likely to be pursued by corporations rather than individuals. • The larger the scale of the venture and the more complex the organization that is required to undertake it, the more likely the venture is to be pursued by a corporation. • The higher the level of total risk, as compared to beta risk, the greater the cost-of-capital advantage of the diversified investor, and hence the more likely the venture is to be pursued by a public corporation. • With a longer expected holding period between investment and harvesting, the diversified investor’s cost-of-capital advantage compounds, making corporate entrepreneurship more likely. There are additional reasons why, in the market for new venture investing, large public corporations and diversified investors do not entirely displace entrepreneurs. As one example, for any given venture, investors may not perceive
392
Chapter Ten
the opportunity or may be less optimistic than is the entrepreneur. In such cases, the entrepreneur may need to invest more resources that are personal because money from diversified investors is not available, or the entrepreneur may need to make a larger initial investment to help convince prospective investors of the merits of the venture.
10.11 Matching Cash Flows and Discount Rates Cash Flow Definitions Many valuation errors, even by employees of well-known investment banks, arise from incorrect matching of cash flows and discount rates. The cash flow definitions used in a valuation need to be tied to the specific financial claim being valued and must be matched correctly with the appropriate discount rates. As an aid to identification, Table 10.1 includes specific definitions of the most important measures of expected cash flow that are appropriate for valuing debt and equity claims and for valuing the enterprise: • Cash flow to all investors represents the amount of cash available to all capital providers, after funding net working capital and capital expenditures. Because this measure is after actual taxes in light of the venture’s leveraging decision, the tax benefit of debt financing (i.e., tax deductibility of interest payments) is reflected in the cash flow measure. • Cash flow to creditors is the net of what the firm’s creditors expect to receive in the form of interest and principal payments less any new borrowing. If there is a risk of default or prepayment, then the expected cash flows to creditors are not the same as the contractual cash flows. • Cash flow to stockholders represents the residual cash flow available to the equity investors in the venture. This is a measure of residual free cash flow after expected debt service (principal and interest payments to creditors less new borrowing), taxes, and investments in working capital and fixed assets. • Unlevered free cash flow is the cash flow the venture would generate if it were unlevered, that is, financed entirely with equity (as is typical for high-risk, early-stage ventures). By calculating tax on EBIT, we ignore the beneficial tax effect of the actual financing choice. To offset, we need to incorporate the tax benefits of debt financing into the discount rate.
Tab le 1 0.1 Measures of expected cash flow
Cash flow to all investors (both stockholders and creditors) Cash Flow to All Investors = EBIT − Actual Taxes + D&A − ΔNWC − ΔFixed Assets Cash flow to creditors Cash Flow to Creditors = Expected INT - Expected ΔDebt Cash flow to stockholders (residual, in light of expected cash flows to creditors) Cash Flow to Stockholders = EBIT − Actual Taxes + D&A − ΔNWC – Δ Fixed Assets − Expected INT + Expected ΔDebt Unlevered free cash flow (as if financed with no debt) Unlevered Free Cash Flow = EBIT − Theoretical Taxes without Debt + D&A − ΔNWC − ΔFixed Assets EBIT = earnings before interest and taxes, or operating profit D&A = depreciation and amortization ΔFixed Assets = change in fixed assets = net capital expenditures ΔNWC = change in net working capital = NWC investment INT = net interest payments ΔDebt = net change in debt financing = proceeds from new debt – principal payments on outstanding debt
394
Chapter Ten
Consistent Discount Rates In Table 10.2, we match these cash flow measures with the appropriate discount rates. Under the RADR method, expected cash flows are discounted to PV using a discount rate that is based on the market risk of the cash flow. The tax deductibility of interest payments is a complicating factor. In a correct valuation, the tax benefit can be recognized in either the cash flow measure or the discount rate, but not in both. Therefore, it is important to make sure the cash flow definition and discount rate assumption are consistent. • Unlevered cost of equity. One way to achieve consistency is to estimate expected after-tax cash flows given the target capital structure and to discount those flows at a rate that is not adjusted for the tax deductibility of interest expense. The appropriate rate is called the “unlevered cost of equity.” Under this approach, the tax benefit (if any) is incorporated as an adjustment to cash flows and not to the discount factor. • Weighted average cost of capital (WACC). The other way to achieve consistency is to estimate theoretical cash flows as if the venture were financed entirely with equity and then discount those cash flows using a discount rate that is adjusted for the benefit of the debt tax shelter at the target capital structure. Under this approach, the tax benefit (if any) of debt financing is incorporated in the discount factor and not in the cash flow. In principle, either approach can yield a correct estimate, but the net tax advantage is difficult to determine directly. The tax advantage of debt financing depends on the aggregate supply and demand for debt and equity, the statutory corporate tax rate, the structure of personal tax rates, and the company’s profitability and nondebt tax shelters. Empirical evidence suggests that the tax advantage of debt financing is usually small as a result of low effective corporate tax rates and partially offsetting personal tax rates. Thus, in the calculation of WACC, the appropriate tax rate is probably well below the statutory rate. The preferred approach depends on ease of estimation and availability of information. There are other ways to calculate cash flows beyond these definitions. Some informal valuation approaches are based on a narrower concept of operating cash flow: EBIT or EBITDA. While EBITDA is convenient to compute, it does not provide for capital replacement or expected growth. Consequently, EBITDA may not be a good measure of the cash flows investors can expect to receive. The important point is to match the cash flow of the claim being valued with a consistent discount rate.
Foundations of New Venture Valuation 395 Tab le 1 0. 2 Matching cash flows to discount rates for various financial claims
Financial claim
Discount rate
Discount rate formula (CAPM)
Comment
Cash flows to all investors
Unlevered cost of equity
rA = rF + βA(rM − rF)
Cash flow to creditors
Cost of debt
rD = rF + βD(rM − rF)
Cash flow to stockholders
Cost of equity
rE = rF + βE(rM − rF)
Unlevered free cash flow
Weighted average cost of capital
WACC = (D/V)(1 − tc)rD + (E/V)rE
The required rate of return on assets, or the unlevered cost of equity, is used to value cash flows that are expected to be received by all claimants given the target capital structure. The effect of tax deductibility of interest payments is reflected in the cash flows. The cost of capital for debt depends on the extent to which debt service payments are subject to market risk. The cost of capital for equity depends not on the total risk of equity but on the market component of the risk. The weighted average cost of capital (WACC) is used to value hypothetical cash flows as if the venture were financed entirely with equity. D and E are market values of debt and equity, V = D + E. The tax benefit of debt financing is an adjustment to the cost of debt capital.*
rA = return on assets rD = return on debt rE = return on equity WACC = weighted average cost of capital βD = debt beta βE = equity beta βA = asset beta rF = risk-free rate rM = expected return on the market (rM − rF) = market risk premium tc = corporate tax rate D/V = market value debt/total firm value (debt + equity) E/V = market value equity/total firm value (debt + equity) *The correct tax adjustment should be the net advantage of debt financing, giving consideration to the offsetting effects of personal taxes. However, this number is unobservable and difficult to estimate, so in practice, tc , the marginal corporate tax rate, is typically used. See Miller (1977) and DeAngelo and Masulis (1980).
Consistency in the Use of Continuing Value The same issues of consistency can arise when we use continuing value in a PV calculation. Continuing value is a shortcut way of projecting future cash flows. Instead of explicitly projecting each period’s cash flow, discounting it back to PV, and then summing, with continuing value a single cash flow number is capitalized using a capitalization rate that takes account of both the riskiness of the cash flow and its expected growth rate. Consider a simple example in which, after several years of rapid growth, the cash flow available to all investors is expected to grow at a constant rate of 4%. Based on Table 10.2, if we are valuing cash flow to all investors, then the appropriate discount rate is the unlevered cost of equity. Suppose we have
396
Chapter Ten
determined that the unlevered cost of equity is 10%. We can determine the appropriate continuing value multiple in this example as Multiple =
(1 + g ) 1.04 = = 17.33 (rA − g) (0.10 − 0.04)
Thus, we would estimate continuing value by multiplying cash flow to all investors in the last year of the explicit value period by 17.33. To determine the PV of the continuing value, we would discount the continuing value back to Time 0 at rA. If we assume that the D/E ratio stays constant, we can use the same approach to find the PV of debt but would use rD to determine the multiple and to discount the continuing value. This example demonstrates the importance of consistency. If we were to use a different cash flow measure and not adjust the cost of capital to be consistent, the result would be a different (and biased) estimate of continuing value. Many of the cash flow measures in the earlier discussion of relative value do not have obvious discount rates and multiples that are theoretically consistent. Instead, the normal approach is to estimate the multiple based on data for comparable companies. The consistency point, however, is the same. If we are using comparables to value cash flow to all investors, then the same cash flow should be used for the comparable companies, and either the comparables should be selected to have similar leverage ratios or the cash flows should be adjusted to be based on similar leverage ratios. Our overarching point is that consistency is important to the valuation. This is true whether you are using explicit DCF or RV and whether you are using an approach based on cost of capital or a simplified approach such as the VC Method.
10.12 Summary Valuation is central for entrepreneurs and investors when they are making decisions about business planning, capital raising, and deal structuring. In the last three decades, private equity investing has become more competitive, increasing the role and importance of theoretically sound and empirically supported valuation tools. The valuation is a key contributor to a successful negotiation between the entrepreneur and investors. We describe several valuation methodologies that fall into two basic categories: discounted cash flow and relative value. Discounted cash flow approaches start with the premise that the value is simply the PV of all future cash flows. This is a theoretically sound approach, but DCF can also be challenging to
Foundations of New Venture Valuation 397
implement. To use DCF, we need estimates of future cash flows and the corresponding discount rates. We stress the importance of matching the financial claim with the appropriate cash flow measure and discount rate and also introduce the concept of continuing value. To estimate discount rates, the notion of nondiversifiable or market risk is highlighted. Because the investors who participate in VC investing will be those best able to diversify and hold illiquid investments, we are able to rely on CAPM-based measures for calculating the appropriate discount rates. Continuing value represents a shortcut to estimating the value of the venture’s long-term cash flows while avoiding the need for extensive, explicit cash flow forecasts. We review the RADR approach of DCF valuation and introduce the CEQ approach as a way to overcome the inherent challenges of using RADR to value new ventures. The relative valuation method uses the values of comparable firms, relative to operating or financial multiples, to provide a basis for estimating value. The challenge in using RV is in finding comparables that are truly representative of the new venture at the point when the venture is being valued. The Venture Capital Method is simple and intuitively appealing but has theoretical shortcomings that raise challenges to validity and call for caution in its implementation critical. Because the VC Method assumes a successful outcome, the corresponding annual discount rate is typically very high (30–80%) to compensate for the cash flow bias. The First Chicago Method was developed to address some of the shortcomings of the VC Method. It is a scenario-based valuation methodology, which means it incorporates a range of possible outcomes and therefore uses a more reasonable, lower discount rate. The First Chicago and RADR approaches yield similar value estimates when the scenarios in the First Chicago approach are weighted to yield expected cash flows. A key point in this chapter is to reconcile the high hurdle rates associated with use of the Venture Capital Method with reasonable estimates of cost of capital. We demonstrate that high hurdle rates are simply a heuristic device that investors can use to compensate for optimistic cash flow projections. Another key point is that because entrepreneurs cannot be fully diversified, they bear the unsystematic risk of their investment of financial and human capital in the venture. As a result, the entrepreneur’s cost of capital is higher than that of well-diversified investors. The extent of the disadvantage depends on the fraction of wealth the entrepreneur must commit to the venture. The chapter concludes with a structured discussion of the importance of matching cash flow measures to discount rates and emphasizes that the tax
398
Chapter Ten
a dvantage of debt financing can only be incorporated into the cash flow estimate or in the discount rate, but not in both. Review Questions
1. How might you respond to an entrepreneur who says, “There is so much risk and uncertainty associated with new ventures that forecasting future revenue and cash flows is a waste of time”? 2. Describe the recent performance of VC funds. Do the empirical results fit with your prior expectations? If not, how can you explain the discrepancy? 3. Explain the theoretical basis for discounted cash flow valuation. What are the key inputs required for DCF analysis? 4. Explain the intuition behind the Capital Asset Pricing Model. Is the CAPM a reasonable tool for estimating the required return for an investor? For an entrepreneur? Why or why not? 5. Why is it important to match the cash flows and discount rate when using DCF valuation? 6. What are the main differences between the RADR and CEQ approaches to DCF valuation? Why is the CEQ approach often preferable when valuing new ventures? 7. Explain the intuition behind each of the following valuation techniques: (a) relative valuation (b) the VC Method (c) the First Chicago Method. Describe the strengths and weaknesses of each approach and the caveats regarding its implementation. 8. How would you reconcile the following? (a) VCs often use very high hurdle rates when valuing potential new investments. (b) The historical average return for a large sample of VC funds is around 14%. 9. Why should an entrepreneur’s cost of capital for valuing an opportunity depend on the fraction of the entrepreneur’s wealth that must be invested in the venture? 10. How is the benefit of diversifying related to the correlation of returns or cash flows between different assets in the portfolio? Notes
1. Funds with vintage years later than 2013 do not have enough history to produce reliable IRRs.
Foundations of New Venture Valuation 399
2. Baty (1990, p. 63). 3. See Roberts and Stevenson (1992). 4. For more detail on historical VC returns in the early years of the industry, see Poindexter (1976), Ibbotson and Brinson (1987, pp. 99–100), Martin and Petty (1983), and Cochrane (2005). 5. All of these estimates are likely to be positively biased. Information on fund IRRs is not reported formally to any official source; rather, both Thomson Reuters and PitchBook obtain the information from voluntary disclosures, from third parties such as limited partners of the fund, and through Freedom of Information Act (FOIA) disclosures. Funds that have not performed well are less likely to disclose performance voluntarily. 6. See Roberts and Barley (2004, p. 8). 7. The term “Venture Capital Method” was first coined by HBS Professor William Sahlman in a 1987 publication. 8. We use the term “hurdle rate” to distinguish it from cost of capital, which is the rate used to discount expected cash flows as opposed to an optimistic projection. 9. The First Chicago Method was developed by the VC group of First Chicago Corporation. Sahlman and Scherlis (2009) describe it as a method developed to address valuation biases inherent in the VC Method. 10. See Brealey, Myers, and Allen (2017, chap. 4) for elaboration of the relation between share value and expected dividends. 11. A pioneer of modern portfolio theory, Harry Markowitz, first proposed use of the standard deviation of holding-period returns as the measure of risk; see Markowitz (1952). Markowitz was awarded the Nobel Prize in economics in 1990 for his contributions to the theory of decision making. 12. The expected return is computed as [(30% × 1/3) + (15% × 1/3) + (–10% × 1/3)] = 11.67%. The standard deviation is computed as [(30% – 11.67%)2 × 1/3 + (15% – 11.67%)2 × 1/3 + (–10% – 11.67%)2 × 1/3]0.5 = 16.5%. 13. Development of the CAPM from its roots in modern portfolio theory was the contribution of several individuals, working independently. See Sharpe (1964), Lintner (1965), and Mossin (1966). The model is based on a number of assumptions, including homogeneity of beliefs about risk and return, a one-period time horizon, availability of a risk-free asset, and quadratic utility functions for investors or normally distributed risk. Empirical evidence from testing of the CAPM is roughly consistent with the CAPM’s theoretical predictions. This suggests that, in most uses, violations of the assumptions are not very important and do not impede its application as a venture valuation tool.
400
Chapter Ten
14. The expected return is computed as [($1 × 0.5) + ($2 × 0.5)] = $1.50. The standard deviation is computed as [($1 − $1.50)2 × 0.5 + ($2 − $1.50)2 × 0.5]0.5 = $0.50. 15. The expected holding-period return is computed as [(60% × 0.5) + (−20% × 0.5)] = 20%. The standard deviation is computed as [(60% − 20%)2 × 0.5 + (−20% − 20%)2 × 0.5]0.5 = 40%. 16. For additional discussion of the CEQ valuation approach, see Brealey, Myers, and Allen (2017). 17. See Demers and Lev (2001), Kozberg (2001), and Rajgopal, Venkatachalam, and Kotha (2003) for evidence of nonfinancial value drivers for Internet companies. 18. See Amir and Lev (1996), who explore the valuation impact of nonfinancial metrics such as population growth and market penetration in the wireless industry. See also the survey results summarized by Prinz (2013). 19. Black (2003) finds no value relevance of earnings in a sample of startup and growth-phase firms. The author reports that cash flow measures are more value relevant than earnings in early stages of a firm’s existence. 20. Wright and Robbie (1996) survey VC firms in the U.K. regarding their valuation approaches. They find that capitalization of historical or prospective earnings is the most widely relied-upon approach. 21. For discussion and examples using the VC Method, see Lerner and Willinge (2002) and Sahlman and Scherlis (2009). Sahlman and Scherlis describe the VC Method and illustrate its application. See also Timmons and Spinelli (2007). 22. The relation between net income and cash flow is not explicit. In our discussion, we assume that net income is used as an approximation of steadystate cash flow. 23. Fischer Black, Myron Scholes, and Robert Merton are the original developers of the OPM. See Black and Scholes (1973) and Merton (1973). Scholes and Merton were awarded the Nobel Prize in economics in 1998 for their contributions. Black died in 1996, and the prize is not awarded posthumously. 24. Most proposals for valuing options that are not traded or where the market is incomplete involve some means of backing into completeness. For example, Mason and Merton (1985) and Kasanen and Trigeorgis (1994) propose that real options that are not traded can be valued by appealing to the existence of a “twin security” that is traded and has risk characteristics that are perfectly correlated with the real option. Short of that possibility, several researchers suggest that untraded options can be valued in the presence of nondiversifiable risk by replacing expected cash flows with their certainty equivalents and valuing the certainty equivalent cash flows at the risk-free
Foundations of New Venture Valuation 401
rate of interest. They generally do not address how to determine the certainty equivalent. See Constantinides (1978); Cox, Ingersoll, and Ross (1985); and Harrison and Kreps (1979). For an overview and summary of the literature, see Trigeorgis (1996) and Borison (2005). 25. See Kerins, Smith, and Smith (2004). 26. These are roughly the long-run historical averages over the 1928– 2016 period, as estimated by Damodaran. See http:// pages .stern .nyu .edu/ ~adamodar/New_Home_ Page/datafile/h istretSP.html. 27. See Sahlman (1990) and Gompers and Lerner (1999). 28. For evidence that illiquidity premia tend to be small, see Blackwell and Kidwell (1988), Hertzel and Smith (1993), and Smith and Armstrong (1993). 29. See Bhagat (2014), who makes a similar argument.
References and Additional Reading
Achleitner, A., and E. Nathusius. 2005. “First Chicago Method: Alternative Approach to Valuing Innovative Start-Ups in the Context of Venture Capital Financing Rounds.” Betriebswirtschaftliche Forschung und Praxis (BFuP) 57: 333–47. Amir, E., and B. Lev. 1996. “Value-Relevance of Nonfinancial Information: The Wireless Communications Industry.” Journal of Accounting and Economics 22: 3–30. Baty, G. 1990. Entrepreneurship in the Nineties. Englewood Cliffs, NJ: Prentice-Hall. Bhagat, S. 2014. “Why Do Venture Capitalists Use Such High Discount Rates?” Journal of Risk Finance 15: 94–8. Black, E. 2003. “Usefulness of Financial Statement Components in Valuation: An Examination of Start-Up and Growth Firms.” Venture Capital: An International Journal of Entrepreneurial Finance 5 (1): 47–69. Black, F., and M. Scholes. 1973. “The Pricing of Options and Corporate Liabilities.” Journal of Political Economy 81 (3): 637–54. Blackwell, D. W., and D. S. Kidwell. 1988. “An Investigation of Cost Differences between Public Sales and Private Placements of Debt.” Journal of Financial Economics 22: 253–78. Borison, A. 2005. “Real Options Analysis: Where Are the Emperor’s Clothes?” Journal of Applied Corporate Finance 17: 17–31. Brealey, R. A., S. C. Myers, and F. Allen. 2017. Principles of Corporate Finance, 12th ed. New York: McGraw-Hill. Cochrane, J. H. 2005. “The Risk and Return of Venture Capital.” Journal of Financial Economics 75: 3–52.
402
Chapter Ten
Constantinides, G. 1978. “Market Risk Adjustment in Project Valuation.” Journal of Finance 33: 603–16. Cox, J., J. Ingersoll, and S. Ross. 1985. “An Intertemporal General Equilibrium Model of Asset Prices.” Econometrica 53: 363–84. Demers, E., and B. Lev. 2001. “A Rude Awakening: Internet Shakeout in 2000.” Review of Accounting Studies 6: 331–59. Fama, E., and K. French. 1995. “Size and Book-to-Market Factors in Earnings and Returns.” Journal of Finance 50: 131–55. Gompers, P., W. Gornall, S. N. Kaplan, and I. A. Strebulaev. 2016. “How Do Venture Capitalists Make Decisions?” NBER Working Paper 22587. Gompers, P., and J. Lerner. 1999. “An Analysis of Compensation in the U.S. Venture Capital Partnership.” Journal of Financial Economics 51: 3–44. Harrison, J., and D. Kreps. 1979. “Martingales and Arbitrage in Multiperiod Securities Markets.” Journal of Economic Theory 20: 381–408. Hertzel, M. G., and R. L. Smith. 1993. “Market Discounts and Shareholder Gains for Placing Equity Privately.” Journal of Finance 48: 459–85. Ibbotson, R. G., and G. P. Brinson. 1987. Investment Markets. New York: McGraw-Hill. Jones, C. M., and M. Rhodes-Kropf. 2013. “The Price of Diversifiable Risk in Venture Capital and Private Equity.” Review of Financial Studies 26: 1854–89. Kasanen, E., and L. Trigeorgis. 1994. “A Market Utility Approach to Investment Valuation.” European Journal of Operational Research 74: 294–309. Kerins, F., J. K. Smith, and R. L. Smith. 2004. “Opportunity Cost of Capital for Venture Capital Investors and Entrepreneurs.” Journal of Financial and Quantitative Analysis 39: 385–405. Kozberg, A. 2001. “The Value Drivers of Internet Stocks: A Business Models Approach.” SSRN Working Paper Series. http://papers.ssrn.com/sol3/ papers.cfm?abstract_ id=256468. Lerner, J., and J. Willinge. 2002. “A Note on Valuation in Private Equity Settings.” Harvard Business School Note 9-297-050. Cambridge, MA: Harvard Business Publishing. Lintner, J. 1965. “The Valuation of Risk Assets and the Selection of Risky Investments in Stock Portfolios and Capital Budgeting.” Review of Economics and Statistics 47: 13–37. Markowitz, H. 1952. “Portfolio Selection.” Journal of Finance 7: 77–91. Martin, J. D., and W. Petty. 1983. “An Analysis of the Performance of Publicly Traded Venture Capital Companies.” Journal of Financial and Quantitative Analysis 18: 401–10.
Foundations of New Venture Valuation 403
Mason, S. P., and R. C. Merton. 1985. “The Role of Contingent Claims Analysis in Corporate Finance.” In Recent Advances in Corporate Finance, eds. E. I. Altman and M. G. Subrahmanyam, 9–54. Homewood, IL: Irwin. Merton, R. C. 1973. “Theory of Rational Option Pricing.” Bell Journal of Economics and Management Science 4 (Spring): 141–83. Miller, M. H. 1977. “Debt and Taxes.” Journal of Finance 32: 261–75. Mossin, J. 1966. “Equilibrium in a Capital Asset Market.” Econometrica 34: 768–83. Poindexter, J. B. 1976. “The Efficiency of Financial Markets: The Venture Capital Case.” PhD dissertation, New York University. Post, T., M. J. van den Assem, G. Baltussen, and R. H. Thaler. 2008. “Deal or No Deal? Decision Making Under Risk in a Large-Payoff Game Show.” American Economic Review 98: 38–71. Prinz, K. F. 2013. “Determinants of Valuation of Early-Stage High-Growth Start-ups.” https://ssrn.com/abstract=2398567 . Rajgopal, S., M. Venkatachalam, and S. Kotha. 2003. “The Value Relevance of Network Advantages: The Case of E-Commerce Firms.” Journal of Accounting Research 41: 135–62. Roberts, M. J., and L. Barley. 2004. “How Venture Capitalists Evaluate Potential Venture Opportunities.” Harvard Business School Case 9-805-019. Cambridge, MA: Harvard Business Publishing. Roberts, M. J., and H. H. Stevenson. 1992. “Alternative Sources of Financing.” In The Entrepreneurial Venture, ed. W. A. Sahlman and H. H. Stevenson, 171–8. Boston: Harvard Business School Press. Robichek, A. A., and S. C. Myers. 1966. “Conceptual Problems in the Use of Risk-Adjusted Discount Rates.” Journal of Finance 21: 727–30. Sahlman, W. A. 1990. “The Structure and Governance of Venture Capital Organizations.” Journal of Financial Economics 27: 473–521. Sahlman, W. A., and D. Scherlis. 2009. “A Method for Valuing High-Risk Long-Term Investments: The ‘Venture Capital Method.’” Harvard Business School Note 9-288-006. Cambridge, MA: Harvard Business Publishing. Sharpe, W. F. 1964. “Capital Asset Prices: A Theory of Market Equilibrium Under Conditions of Risk.” Journal of Finance 19: 425–42. Smith, R. L., and V. Armstrong. 1993. “Misperceptions About Private Placement Discounts: Why Market Reaction to Rule 144A Has Been Lukewarm.” In Modernizing US Securities Regulation: Economic and Legal Perspectives, ed. K. Lehn and R. Kamphuis, 175–91. Homewood, IL: Irwin.
404
Chapter Ten
Timmons, J. A., and S. Spinelli Jr. 2007. New Venture Creation: Entrepreneurship for the 21st Century, 7th ed. Chicago: McGraw-Hill Irwin. Trigeorgis, L. 1996. Real Options. Cambridge, MA: MIT Press. Venture Economics. 1997. Investment Benchmarks Report: Venture Capital. New York: Venture Economics. Wright, M., and K. Robbie. 1996. “Venture Capitalists, Unquoted Equity Investment Appraisal and the Role of Accounting Information.” Accounting and Business Research 26: 153–68.
C h a p t e r Ele ven
N e w Ve ntu r e Valuatio n i n P r ac tice
W e saw i n the previous chapter that many tools are available for valuing new
ventures. Most are conceptually straightforward and intuitive; the challenges come with implementation. In this chapter, after presenting a framework for selecting a valuation method, we address use of the continuing value approach as a way to make the valuing of going concerns tractable. We then address the challenges of making the assumptions that are needed for applying any CAPMbased approach to DCF valuation. In the last part of the chapter, we use a single example to illustrate all of the valuation approaches and to highlight the advantages and disadvantages of each. The chapter can serve as a handbook for generating the information that is required to implement each valuation approach and illustrates how the approaches are related. It can aid in selecting the approaches that will be most useful for addressing a particular valuation problem.
11.1 Criteria for Selecting a New Venture Valuation Method As we have seen, there is a discrepancy between what we frequently observe in practice among VCs in their valuation approaches and what theory would suggest. No doubt, this is attributed in part to the obvious difficulty in forecasting cash flows for risky ventures that will evolve/pivot over time and determining opportunity cost for discounting. However, we have seen that the heuristics that many VCs use (net multiples and hurdle rates of IRRs) are not incompatible with theory in that when we observe successful decisions (ex post 405
406
Chapter Eleven
correct decisions of which ventures to invest in and at what valuations) those decisions will mimic theoretically correct valuation approaches. Improving on heuristics requires the decision maker to evaluate the criteria for deciding on a valuation approach and to get as close as possible to the true intrinsic value of the venture. The following questions are relevant for assessing the relative merits of different approaches. Is cost of capital used as the discount rate? Attempting to compensate for positively biased estimates of cash flow by discounting with positively biased hurdle rates tends to cause projects with more distant payoffs to be rejected incorrectly. Similarly, discount rates based on total risk rather than nondiversifiable risk can lead to rejecting projects that should be accepted by an investor who is well diversified. How does the approach deal with cash flows that vary in risk? Different cash flow streams that occur in the same period can differ in risk. The appropriate discount rates will vary accordingly. Models that do not distinguish among cash flows that differ in risk can produce distorted estimates of value. Cash flows that occur at different times can also differ in risk. If the discount rates do not account for such differences, valuation errors can result. Can the model be used to value embedded options and complex financial claims? Complexity affects both expected returns and risk. A financial structure that includes real or financial options can alter the overall value of a venture relative to a simple accept/reject approach. The values of the options depend on both expected cash flows and the risk of the option cash flows. How difficult is it to estimate the information required for the valuation? There is virtue in simplicity. Valuation approaches that are complex or difficult to use are sometimes too costly to justify. This is true particularly if the project is clearly worth pursuing and agreements for sharing gains and losses can be reached informally. Are there sufficient data available to have confidence in relative valuation estimates? Relative value works best if the expected future cash flows (including expected growth), total risk, and market risk of the comparables are believed to be proportional to those of the subject venture.
New Venture Valuation in Practice 407
11.2 Implementing the Continuing Value Concept It is not practical to value a going concern by forecasting cash flows explicitly into the indefinite future and then discounting each periodic cash flow back to PV. Instead, the normal approach is to project explicit cash flows over a period until it can be assumed that subsequent growth is likely to stabilize and then to summarize the value of cash flows beyond that point as a single “continuing value.” The continuing value concept is used to convert cash flows after the explicit value period to a single estimate of value that is equivalent to valuing each subsequent cash flow. In other words, the cash flows after the first few years are valued implicitly by applying a multiplier to the last explicit cash flow. The multiplier is intended to take account of expected growth of cash flows from that point forward and the riskiness of those cash flows. Normally, continuing value is estimated by using a theoretically determined discount rate or is based on observed multipliers of market assets that are similar to the one being valued. For many new ventures, continuing value is an important component of total PV.1 The overall valuation is thus divided into two periods. For the first period, an explicit cash flow projection is made for each year, quarter, or other appropriate $1,600 $1,400 $1,200 Cash flow
$1,000 $800 $600 $400 $200 $0 –$200
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Year
Explicit Value Period: Compute the present value of each periodic cash flow.
Continuing Value Period: Estimate the continuing value of the stream as of year 5, and then convert the continuing value to present value.
Fi g u r e 11 .1 Using continuing value to estimate the value of a new venture A common approach used in DCF valuation is to divide the forecast into two periods. During the explicit value period, cash flows are forecasted individually and valued directly. During the continuing value period, cash flows are converted to a capitalized value at the end of the explicit value period. This capitalized value is then discounted back to Time 0. Normally, the continuing value period begins when the venture reaches a point of stable growth.
408
Chapter Eleven
interval. We refer to this as the “explicit value period.” We refer to the period after the explicit value period as the “continuing value period.”2 Figure 11.1 illustrates a forecast of the future cash flows of a venture and their segregation into explicit value and continuing value periods. In the example, Year 5 is the last explicit forecast. Projections beginning in Year 6 are implicit, based on assumptions about the growth rate from that point forward. The venture’s continuing value is an estimate as of the end of the explicit value period (Year 5 in the figure). To estimate PV, continuing value must be discounted to PV, as if it were a single cash flow that was expected to be received for selling the venture or the financial claim. Equation (11.1) describes the value of a venture in terms of explicit and continuing value components, based on annual data:
PV = Σ Tt =1
Ct CVT + (11.1) (1 + rt )t (1 + rT )T
where PV is the present value of the venture; Ct is the annual (or other periodic) cash flow for each explicit year, t; CV T is the continuing value at the end of the explicit value period, year T; and rt is the discount rate for year t cash flows. As Figure 11.1 and Eq. (11.1) suggest, continuing value is a DCF valuation method. However, it is not one that relies entirely on explicit forecasts of cash flows. Instead, the cash flow forecast during the continuing value period is implicit in the assumptions used to estimate continuing value. Continuing value is estimated based on information from the last explicit forecast along with assumptions about the expected growth rate and cost of capital or on the values of comparables. Determining the explicit value period. The first step in using the continuing value concept is to decide where to draw the line between the explicit and continuing value periods. Continuing value estimates are most reliable if they are made for a period when a firm has established a track record and has reached a point of stable growth. Continuing value does not work well for valuing the early stages of a new venture because of the expectation of temporary high growth and the volatility of revenue and cash flows. Thus, explicit cash flows are normally estimated during development, for periods when the venture has not yet achieved a profitable level of sales, and during periods of rapid growth. In Figure 11.1, the explicit value period, Years 1 through 5, includes years when expected cash flows are negative and years when they are growing rapidly. The continuing value period begins at the point where the venture is expected to be in steady state. Sometimes continuing value estimates are applied at earlier stages of development. Rapidly growing ventures with large capital needs sometimes seek
New Venture Valuation in Practice 409
public equity financing before establishing a stable track record. If comparable companies have gone public at similar stages, then underwriters are likely to use a combination of continuing value methods and explicit DCF methods to estimate the price at which shares will be offered to the public. In such a case, it could be appropriate to use data from IPOs of other early-stage ventures to estimate continuing value. Determining which multiplier to use. Multipliers can be tied to any accounting item or nonaccounting item that can be measured at the end of the explicit value period, and they can be derived from theory or based on comparables. A free cash flow measure seems like the obvious choice for estimating the multiple, since that is the source of the investor’s return. Sometimes, however, other measures, such as EBIT or sales, or even website visits, for example, can yield more reliable multipliers. We showed in Chapter 10 that multipliers implicitly include an assumed growth rate and discount rate for cash flows during the continuing value period. In that example, the multiple was based on explicit assumptions about the discount rate and growth rate of cash flows. In other scenarios, we might use a multiple based on comparable transactions. For example, a firm planning to go public at the end of the explicit value period might collect multiples from the IPO valuations of comparable ventures. Choosing the right multiplier or set of multipliers requires judgment and benefits from practical experience. Multiples of operating cash flow, net income, sales, or assets are commonly used. For example, continuing value might be capitalized at 10 times expected cash flow in the last year of the explicit value period, or 1.5 times expected sales. The relative merits of different multipliers (e.g., sales versus cash flow) are discussed in a variety of sources.3 Continuing value is sometimes estimated on the basis of sales or asset multiples, not because investors care specifically about sales or assets but because sales or asset levels at the end of the explicit value period bear a stronger relationship to expected future cash flows over the continuing value period than does cash flow at the end of the explicit value period. The continuing value approach works best when the growth rate of the accounting stream (sales, earnings, etc.) or nonaccounting stream (web clicks, site visits, etc.) on which the valuation is based has stabilized and when the relation between the stream and value is strong. The strength of the relationship between the stream and value can be assessed by comparing measures of dispersion of alternative multipliers across a sample of comparable firms. In general, a multiplier with low dispersion, standardized by the mean or median value, yields a better estimate.
410
Chapter Eleven
If you decide to use a multiple based on comparables, it is still helpful to use the theory-based approach to test whether the multiple derived from the comparables really makes sense for your venture. Whatever the multiple, it implies something about expected growth and cost of capital, and those implications should be assessed. Equation (11.2) summarizes the implicit assumptions. You can easily assess whether a multiple derived from comparables implies a sensible value for the expected growth rate and cost of capital for your venture. Determining the multiplier. Equation (11.2) describes the relation between value in one period (the end of the explicit value period) and all future cash flows:
Vt =
Ct +1 Ct (1 + g ) = (11.2) r−g r−g
where Vt is value at time t, Ct+1 is cash flow at time t + 1, r is the discount rate, and g is the expected growth rate of cash flows. Generally, Ct is the cash flow generated over a year, and Vt is value at the end of that same year (i.e., an earnings/price ratio). This is sometimes referred to as a trailing value. It does not imply that value is determined by prior earnings. Rather, it implies that prior earnings can be used in a consistent way, as shown in the equation, to predict future earnings. Vt, the value at the end of period t, is a function of the cash flows from period t + 1 onward. In the equation, Ct is increased by g, which effectively means that the numerator represents the next period’s cash flow. Equation (11.2) is the standard expression for the PV of a growing perpetuity of cash flows. The equation can be rearranged to calculate the cash flow multiplier, as shown in Eq. (11.3):
Vt (1 + g ) (11.3) = Ct r−g
where Vt /Ct is the cash flow multiplier. Equation (11.3) shows how the expected growth rate and discount rate affect the multiplier. The assumptions of a 10% discount rate and 4% annual growth rate in cash flows yield the following calculation and cash flow multiple: Cash Flow Multiple =
Vt (1 + g ) 1.04 = = 17.33 = Ct r−g (.10 − .04)
It is easy to see that a higher rate of expected growth would increase the multiplier, as it simultaneously increases the numerator and makes the denominator smaller. A higher discount rate would reduce the multiplier by increasing the denominator. This makes intuitive sense, as cash flows that are growing more quickly are worth more, and riskier cash flows are worth less. Although the connections between value and other accounting and nonaccounting streams are indirect, the same principle applies. Higher expected
New Venture Valuation in Practice 411
growth rates imply higher multiples, and larger discount rates imply lower multiples. This suggests a way to use market data to estimate a multiplier. If, relative to comparable firms, the venture being valued has a high expected growth rate, a larger multiplier is implied. If the comparable firms are selected correctly and if the cash flow used in Eq. (11.3) is the expected cash flow, there is no reason for the discount rate to be different from that for the comparables. The other determinant of Vt in Eq. (11.2) is Ct. There are two important issues to keep in mind about Ct. First, the cash flows of comparable public firms are based on publicly reported information. They have been prepared under U.S. Generally Accepted Accounting Principles (GAAP) or International Financial Reporting Standards (IFRS) and have been subject to independent audit, whereas the venture’s forecast has not been. Second, the comparable firms have survived long enough to have gone public, whereas the venture is at an earlier stage. How best to deal with these issues depends on the purpose of the valuation and on your comfort level with the financial projections. Suppose you believe that the projections of the venture were prepared consistently with GAAP and, though not audited, are unbiased. In this case, it is reasonable to assume that the projections are comparable to the reported numbers of the public companies. Suppose, instead, that the entrepreneur prepared the projections. Such projections are likely to reflect the entrepreneur’s inherent optimism that the venture will be successful. In that case, direct application of multipliers from comparable public companies could result in an overestimate of continuing value. One way to avoid this survival bias is to base the continuing value estimate on multipliers from private transactions. Such information can be useful, especially if the selected transactions are for companies at similar stages of development to the subject venture. But information about private transactions is difficult to acquire and to verify. A second solution is to adjust the public company multiplier for an estimate of the bias in the accounting projections for the venture. If, for example, you believe that the venture’s probability of failure is 30% and is not reflected in its projected cash flows, it would be appropriate to adjust the public company multiplier down by 30%. The latter solution is implicit in the actual multipliers that are frequently used in private transaction valuations. Such adjustments are often characterized (incorrectly, we believe) as “illiquidity discounts.” The true nature of the discount is not illiquidity but rather biased cash flow estimates. By recognizing this, you can do a better job of applying the data from comparable public firms to your own valuation questions. This leads to the third solution. If the lack of comparability is due to positively biased projections for the venture, then one way to solve the problem is to develop a set of projections that reflect the true expectations, including the risk
412
Chapter Eleven
of failure. You would then value the asset using the comparable market data without additional adjustment. Forecasting the multiple. A correct valuation must be based on a multiple that is expected to be accurate at the time when the continuing stream of cash flows is being capitalized. This is most obvious if it is assumed that there will be a liquidity event, such as an IPO, at the end of the explicit value period. It follows that the multiples you can observe today are not the ones you would want to use in the valuation. To illustrate, consider the data shown in Figure 11.2. In 2002, the price/earnings (P/E) ratio of the S&P 500 Index was around 37.3, a historical high relative to prior years, and the aggregate dividend yield was near a historical low. If the basis for equity valuation is the PV of expected future dividends, then either the expected rate of dividend growth must have been very high in 2002 or the cost of equity capital must have been very low. The true explanation probably involves a combination of both factors: an expectation of rising dividends combined with a low cost of equity capital. Imagine that you are back in 2002 and are planning to value a company based on its expected future earnings. You believe that the P/E ratio of the S&P 500 Index is appropriate to use as a benchmark for estimating continuing value five years later. Knowing that the 2002 P/E ratio is at a historical high, you should not base the continuing value estimate on the 2002 S&P 500 P/E ratio. Because Fi g u r e 11 . 2
90
Historical price/ earnings ratios of the S&P 500 Index
80
60 50 40 30 20 10 0
1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017
The P/E ratio each year is calculated as the monthly average of the value of the S&P 500 divided by aggregate earnings over the preceding (trailing) 12 months.
70 Prices as a multiple of earnings
source: http://w ww .quandl.com.
Year
New Venture Valuation in Practice 413
there is no plan to harvest the investment in 2002, the multiple that existed at that time is not the best one to use. A better approach is to forecast the multiple that is likely to exist at the time of a hypothetical sale of the investment. P/E multiples are influenced by factors that can affect either the numerator or the denominator. In 2002, they soared in part because earnings levels of technology-related firms declined sharply after the Internet bubble. In contrast, the 2009 multiple of 83.6—a clear outlier—was caused by a large, widespread drop in corporate earnings brought on by the subprime and financial crises. In fact, the S&P 500 Index fell almost 40% in 2008, its worst annual return since 1931, but in 2008 and 2009 earnings fell by an even greater percentage, so the 2009 S&P 500 P/E ratio increased. Suppose that in 2009 you want to select a multiple to use in estimating continuing value where harvesting is expected to occur in five years. Figure 11.2 shows that between 1980 and 2008 there was a great deal of variation in the annual S&P 500 P/E ratio, from a low of 7.9 in 1980 to a high of 37.8 in 2002. The simple average over the 53-year period is 19.6, but it appears that the most recent historical period has seen higher multiples, on average. Finally, from Figure 11.2 it appears that the change in P/E from one year to the next is not random; instead, a high value in one year tends to be followed by a high value in the next. So how can we best estimate an appropriate multiple for 2013? One valid approach is to use statistical techniques—regression analysis, exponential smoothing, or the like—to estimate future P/E multiples. This level of rigor may or may not be warranted in practice. For some valuation problems it may be sufficient to recognize that historical P/E multiples appear to be mean reverting and that the multiple in 2008 is historically high. In recognition of the uncertainty as to what the future exit multiple will be, we suggest using several different values of exit multiples and assessing their impact on the resultant valuation. If the accept/reject decision is not sensitive to multiples over a reasonable range, that alone may be sufficient to proceed with the venture. Even if the appropriate multiple for your venture is not the S&P 500 Index P/E ratio, whatever multiple you do use, if it is based on comparables, is likely to be highly correlated with the S&P Index multiple. You can update Figure 11.2 and use the data to make inferences about whether other multiples are currently likely to be too high or too low.
11.3 Implementing DCF Valuation Methods CAPM-based approaches to DCF valuation require specific assumptions about the risk-free rate, the market risk premium, and beta. In lieu of beta,
414
Chapter Eleven
they may require specific assumptions about market risk, project risk, and correlation. In the RADR form, project risk is the standard deviation of holding-period returns, and the correlation is between project returns and market returns. In the CEQ form, project risk is the standard deviation of cash flows, and the correlation is between project cash flows and market returns. There are many subtleties and challenges involved in making good and internally consistent assumptions. In practice, it is common to use some simplifying assumptions that make implementation easier. In this section, we first outline approaches to making valid and internally consistent assumptions; we then discuss shortcuts that are commonly used.
Estimating the Risk-Free Rate of Interest The appropriate risk-free rate for valuing a future cash flow is one that is available in the market as of the valuation date and is for a holding period of the same duration as the cash flow being valued. Thus, for a cash flow that is expected in five years, ideally we would use the currently available risk-free rate for an instrument that would mature in five years. We normally assume that U.S. government debt has no risk of default, so we can infer the appropriate risk-free rate directly from current interest rates of U.S. Treasury securities of appropriate maturities.4 We also must be cognizant of whether the cash flow projection is in real or nominal terms. If it is in nominal terms, then the appropriate risk-free rate is also the nominal rate that can be inferred directly from market data. If the forecast is in real terms, then the risk-free rate must be converted to an estimate of the real rate by subtracting the rate of inflation that is expected in the market for the holding period. For this, we can search for publicly available forecasts of inflation or use historical data to infer the normal relationship between risk-free interest rates and inflation. In most cases, the expected cash flows of the venture will have been projected in nominal terms. U.S. Treasury yields at the time of this writing were as follows:5 U. S . Trea s u r y yield s , J a n u a r y 2 6 , 201 8
Maturity
Yield %
Maturity
Yield %
3-month 6-month 1-year 2-year 3-year
1.50 1.87 1.83 2.04 2.21
5-year 10-year 20-year 30-year
2.47 2.61 2.74 2.94
source : Based on http://w ww.wsj.com/m dc/p ublic/p age/2 _ 3020- tstrips.html. Maturities of one year or less are from zero-coupon securities. Maturities greater than one year are based on STRIP (Separate Trading of Registered Interest and Principal) yields.
New Venture Valuation in Practice 415
These data suggest the use of a different risk-free rate in calculating each period’s discount rate. The cash flow from Year 1 would be discounted using a rate based on the 1-year U.S. Treasury bill. The cash flow in Year 5 would use a rate computed with the 5-year Treasury bond.
Estimating the Market Risk Premium The market risk premium that is used in the CAPM is the expected difference between the return on the market and the risk-free rate over the period from investment until a cash flow is received. In contrast to the current risk-free rate, which is observable, the current market risk premium is not. Because it is not, three main approaches are used to estimate the market risk premium: (1) a long-term historical average, (2) a risk premium that is implied by discounting a forecast of future dividends (i.e., the IRR that makes the PV of expected future dividends equal to today’s market price), and (3) a consensus estimate. The easiest but not necessarily most accurate way to estimate the expected market risk premium is to extrapolate from historical data. These data are readily available from numerous sources. Most introductory finance textbooks contain some version of Table 11.1, which presents historical data on stock and bond returns. The table shows arithmetic average returns since 1928 and since 1960 and the respective standard deviations of returns, and also shows the geometric average returns over the same periods. Geometric averages are calculated based on the beginning and ending values from 1928 through 2017 and 1960 through 2017. Tab le 11 .1 Historical stock and bond returns
1928–2017 Series
S&P 5001, 2 U.S. Treasury Bonds (LT)2 U.S. Treasury Bills (ST)2 Inflation3 Series S&P 5001, 2 U.S. Treasury Bonds (LT)2 U.S. Treasury Bills (ST)2 Inflation3
1960–2017
Arithmetic Mean
Standard Deviation
Arithmetic Mean
Standard Deviation
11.53% 5.15% 3.44% 3.05%
19.62% 7.72% 3.05% 3.84%
11.27% 6.64% 4.64% 3.78%
16.31% 9.05% 3.11% 2.81%
Geometric Mean
Geometric Mean
10.15% 6.18% 4.62% 3.78%
9.32% 4.93% 3.40% 3.04%
Composite Total Return Index (includes dividend reinvestment). Sources: Stock, T-bond, and T-bill annual returns data downloaded from http://pages.stern.nyu.edu/~ adamodar/New_ Home_ Page/d atafile/ histretSP.html. 3 Annual CPI (inflation) data from U.S. Department of Labor, Bureau of Labor Statistics. 1 2
416
Chapter Eleven
To estimate the expected market risk premium with these data, we would calculate the market risk premium as the difference between the historical average return on the S&P 500 (the market) and the historical average risk-free rate. If we are valuing a near-term cash flow (such as a year or two), we would use the historical short-term risk-free rate in the calculation. If we were valuing a longer-term cash flow, such as a return in five years, we would use the historical long-term rate, represented by the yield on U.S. Treasury bonds. Using historical data to estimate the expected market risk premium, while simple, still requires some choices for implementation. For example, over what period should the average returns be measured, and is it better to use arithmetic or geometric averages? In using historical data to develop a forward-looking estimate of the market risk premium, we need to balance the relevance of older data with the statistical unreliability that comes with using fewer observations. Table 11.1 includes historical average data for two time periods, but other windows could be used. In principle, the historical returns averages will be more reliable when measured over a long period, as long as the fundamentals that drive the risk premia are consistent over time. In this case, the period since 1960 represents a time when modern portfolio theory was generally accepted so that investors would mainly have been concerned about compensation for bearing systematic risk. Arguably, the earlier years are ones where the premium could be affected by total risk. As to the choice between arithmetic and geometric average returns, it is common practice among academics and practitioners to use the arithmetic averages.6 On balance, evidence favors using arithmetic averages, though geometric averages may be more appropriate for long-term projects. In Figure 11.3, we explain why arithmetic averages work well. Table 11.1 also reports historical inflation rates. Consistent with our discussion of the need to adjust the risk-free rate for inflation to derive an estimate of the real risk-free rate, we could use this information to infer the historical real rates for short- and long-term riskless debt. Based on the period since 1960, Treasury bill rates have averaged 0.86% higher than the average rate of inflation, so the apparent short-term real risk-free rate is 0.86% and the long-term Treasury rate averaged 2.86% above the inflation rate. Similarly, the risk premium of the S&P 500 averaged 6.63% above the short-term risk-free rate and 4.63% above the long-term Treasury rate. For a variety of reasons, the historical average may not be the best measure of the market risk premium. Recent forward-looking estimates of the risk premium (derived by discounting forecasts of future dividends) suggest that the long-term historical average overstates the market risk premium somewhat. For example, Fama and French (2002) use dividend and earnings growth rates to
New Venture Valuation in Practice 417 Fi g u r e 11 . 3 Using arithmetic or geometric average returns
Suppose an asset that is correctly priced is equally likely to return either 50% or 0% each year. If the cost of capital for an investment is inferred by calculating the average annual return, the resulting estimate is 25% (i.e., (50% + 0%)/2). The geometric return, in contrast, is 22.5% [i.e., ((1 + 50%) × (1 + 0%))0.5 – 1]. Which one is right? Consider a one-period investment of $1.00. At the end of the period, it will be worth either $1.50 or $1.00. The average is $1.25. If you discount $1.25 at the arithmetic average rate of 25%, you find that the present value is $1.00. If you discount it at the geometric average rate of 22.5%, you find that the present value is $1.02. Apparently, since we know the correct value is $1.00, the arithmetic average is right. However, consider a two-year investment. Suppose you get a 50% return the first year and 0% the second (or the reverse). The cumulative value is $1.50 for two years. If you discount the $1.50 to present value at 25%, you get $.96, whereas if you discount it at 22.5%, you get $1.00. Now, it appears that the geometric average is right if you are valuing long-term cash flows. This is the common argument for using the geometric average. However, look again. The previous calculation does not consider all possible outcomes. If the probabilities remain stable over time, there is a 25% chance that the investment will be worth $2.25 in two years (i.e., $1.00 × (1 + 50%) × (1 + 50%)), a 50% chance that it will be worth $1.50 (i.e., $1.00 × (1 + 50%) × (1 + 0%)), and a 25% chance that it will be worth $1.00 (i.e., $1.00 × (1 + 0%) × (1 + 0%)). The expected value is $1.56 (25% × $2.25 + 50% × $1.50 + 25% × $1.00 = $1.56), not $1.50 as in the above calculation. Discounting $1.56% at 25% for two periods yields $1.00, the correct present value. Thus, even for a longer holding period, the arithmetic average yields the correct value. So both short-term and
long-term projects are correctly valued by the arithmetic average.
estimate the equity premium over the period 1951–2000 and find equity premia of 2.6–4.3%. Welch (2000) surveys a large number of finance and economics professors and reports that the consensus forecast of the long-term equity risk premium is 5–5.5%. There is also a significant difference between the opinions of U.S. academics and practitioners. Graham and Harvey (2007) survey 465 CFOs and find an average expected risk premium of 3.8%, materially lower than that suggested by most academics. Recent estimates of the market risk premium are generally in the 2.5 to 5.5% range. In light of the historical and recent empirical evidence, we would normally assume a market risk premium of between 5% and 7%. When possible, it makes
418
Chapter Eleven
sense to consult managers of asset allocation funds and similar investment vehicles to learn their current forecasts of the market risk premium. Those estimates are usually based on finding the discount rate that equates their forecast of future dividends to the current stock index value (i.e., dividend discount models of value). One should understand the range of possible estimates and remain alert to the prospect of better methods of estimating the market risk premium. As any rate is going to be an estimate of the true rate, it can also be helpful to assess the sensitivity of the decision to reasonable variations in the assumed market risk premium. If the venture looks good under both “low” and “high” assumptions, then the choice of whether to proceed is not sensitive to the assumed rate.
Estimating the New Venture Beta The final variable we need if we are using the RADR form of the CAPM is the beta of the venture. Equation (11.4), which we saw first in Chapter 10, is the formula for computing the beta of an asset based on its returns and the returns of the market: CovrA , rM ρrA , rM σ A (11.4) βA = = σM σ 2M The expression implies several different approaches that can be used to estimate βA. We describe each and discuss implementation in the following subsections. Using the betas of comparable firms. For established public companies, beta is often estimated by regressing a time series of historical stock returns on a time series of historical returns on a market index. The regression coefficient is the beta term in Eq. (11.4). For new ventures, however, there is no publicly traded stock. One common solution is to use betas estimated from comparable publicly traded firms. Beta estimates for the common stock of public companies can be calculated easily and are available from several sources.7 The case for using data on comparable firms is strongest when there are public companies with market risk comparable to that of the new venture. For example, if the new venture is a grocery store, it could be reasonable to use market information on public grocery store companies to infer beta. Thus, Safeway or another grocery chain might use the company beta as an estimate of the beta for analyzing a new store. Table 11.2 contains data on equity betas for a number of different sector (industry) groups. Ranges of equity betas reported in the table are based on aver-
New Venture Valuation in Practice 419
ages for firms in the sector and are estimated based on two years of weekly stock returns. It is apparent that beta values vary in a systematic fashion across sectors. The sectors with low-equity betas include banks and public utilities, whereas the high-equity beta sectors include cyclical industries. As beta measures the Tab le 11 . 2 Average beta estimates for selected sectors
Industry Name
Utility (general) Bank (money center) Utility (water) Power Banks (regional) Telecom (wireless) Green and renewable energy Environmental and waste services Investments and asset management Healthcare support services Information services Electronics (general) Healthcare information and technology Software (entertainment) Computer services Computers/peripherals Telecom equipment Entertainment Semiconductor equipment Business and consumer services Software (systems and applications) Electronics (consumer and office) Semiconductors Drugs (pharmaceutical) Retail (online) Software (internet) Oil/gas (integrated) Drugs (biotechnology) Food wholesalers Steel Chemical (diversified) Total market Total market (without financials)
Number of firms
Equity beta
D/E Ratio
Unlevered beta
18 11 23 61 612 18 22 87 165 115 61 167 112 13 111 58 104 90 45 169 255 24 72 185 61 305 5 459 15 37 7 7247 6057
0.293 0.638 0.342 0.505 0.502 1.302 1.202 0.877 0.987 0.898 0.882 0.938 0.979 0.891 1.101 1.010 1.034 1.152 0.982 1.169 1.088 1.092 1.172 1.209 1.182 1.203 1.372 1.440 1.786 1.817 2.034 1.000 1.069
0.672 1.573 0.381 0.764 0.587 1.200 0.982 0.349 0.421 0.248 0.157 0.150 0.193 0.065 0.308 0.182 0.207 0.337 0.115 0.274 0.141 0.069 0.131 0.146 0.114 0.033 0.153 0.158 0.375 0.362 0.272 0.590 0.307
0.175 0.248 0.248 0.286 0.316 0.592 0.606 0.650 0.695 0.719 0.762 0.816 0.821 0.837 0.842 0.854 0.857 0.861 0.880 0.917 0.953 1.021 1.037 1.055 1.061 1.165 1.190 1.243 1.299 1.334 1.599 0.629 0.817
Equity beta estimates and leverage ratios are estimated by Damodaran and are updated regularly. Asset betas are computed as discussed in this section of the text. Sectors reported are the five with the highest and lowest asset betas as well as a number of sectors that are the focus of high levels of entrepreneurial activity. source : http://p ages.stern .nyu .edu/~ adamodar/New_ Home_ Page/d atafile/B etas.html
420
Chapter Eleven
market component of total risk, these values make sense given the underlying systematic risks of the industries. As can be seen in Table 11.2, for public companies, at the sector level, equity beta values of less than 0.5 or greater than 2.0 are rare. Keep in mind that the betas typically available from published sources are equity betas and are derived by regressing stock returns on market returns. They are affected by the financial leverage choices of the public companies. Implementation of Eq. (11.4) for a new venture, however, requires an estimate of the asset beta. A firm’s asset beta removes the effect of financial leverage and reflects only the market risk of the business; consequently, it is always less than or equal to its equity beta. To see this, consider two operationally identical businesses. One is financed entirely with equity, while the second uses considerable debt. Since equity is a residual claim, a shareholder in the leveraged company will be behind the bondholders in the line to claim the cash flows generated by the firm’s assets. The riskiness of asset cash flows is the same for both companies and is represented by βA. The riskiness of the cash flows to shareholders, represented by the equity beta, βE , is greater for the leveraged firm. If debt is riskless or has no market risk, equity betas can be converted to asset betas using Eq. (11.5): E β A = β E (11.5) V In the expression, βE is usually estimated for a comparable firm by regression analysis (or possibly taken from a published source), E is the market value of the comparable firm’s equity, and V is the market value of the comparable firm’s total capital (debt plus equity). The computation in Eq. (11.5) yields only an approximation of the asset beta because it assumes that the comparable firm’s debt has no market risk (i.e., the debt beta is zero). If the comparable is highly leveraged, the expression will underestimate the asset beta.8 In practice, because market values of debt are usually not available, it is common to estimate V as the sum of the market value of equity plus the book value of debt. Sometimes, asset beta estimates are also adjusted for the tax deductibility of debt interest and for cash balances (since the cash may be unnecessary for the enterprise and has a zero beta).9 Assuming comparable-firm data are available for more than one firm, we would take the following steps to estimate the new venture asset beta and to calculate the appropriate discount rate: Step 1. Calculate or collect equity betas and data on E and V for the comparable firms. Step 2. Use Eq. (11.5) to convert each equity beta to an asset beta.
New Venture Valuation in Practice 421
Step 3. Use the comparable firm asset betas to compute a weighted average asset beta for the new venture. The weightings should be based on your judgment about comparability. Step 4. With the definitions from Chapter 9, if we are using the cash flows to all investors to value the venture, use the weighted average βA in the CAPM to estimate rA, the discount rate.10 One significant concern with using market comparables to estimate beta is that many ventures are highly specialized and involve products and market opportunities that are difficult to analogize to existing public companies. Even if a prospective comparable firm is in the same industry, it has survived long enough and grown large enough to have gone public. While good public-firm comparables for early-stage ventures are uncommon, the late 1990s constitutes a rare exception. During that period of several years, a large number of “new economy” ventures went public at early stages of development. Based on sample data of over 2,600 firm-years from the last half of the 1990s, Kerins, Smith, and Smith (2004) estimate the equity betas of newly public firms, including many observations of very small firms, firms that had yet to achieve profitability, and firms that had yet to begin revenue generation. Table 11.3 reports some of the findings from that study. Based on all of the observations, the average equity beta is very close to 1.0, similar to the total risk of the S&P 500 (the “market” risk). Because these firms tend not to use debt financing, their equity betas are approximately equivalent to asset betas. The highest reported beta in the table is 1.27, and the evidence also suggests that younger and smaller firms have lower betas. These results suggest that reasonable estimates of beta for nonpublic ventures will generally be in the range of about 0.6 to 1.25.
Estimating the Components of Beta Separately The term on the right-hand side of Eq. (11.4) describes the beta for use in the RADR form as a function of the standard deviation of asset holding-period returns, the standard deviation of market returns, and the correlation coefficient between holding-period returns and market returns. Equation (10.8) describes the cash flow beta for use in the CEQ form as a function of the standard deviation of asset cash flows, the standard deviation of market returns, and the correlation between asset cash flows and market returns. Thus, another approach to valuation is to derive the estimate of the returns beta or the cash flow beta from separate estimates of these three elements.
422
Chapter Eleven
Tab le 11 . 3 Beta estimates and market correlations for newly public, VC-backed firms
Mean β
Correlation with the Market
Standard Deviation of Returns
2,623
0.99
0.195
1.20
501 105 247 407 130 440 39 754
0.75 0.80 1.16 1.02 1.02 0.81 1.24 1.20
0.149 0.237 0.215 0.241 0.208 0.172 0.217 0.200
1.04 0.87 1.20 1.04 0.93 1.44 1.06 1.37
1,263 957 403
0.93 0.96 1.27
0.162 0.212 0.259
1.35 1.04 1.14
102 1,475 1,033
0.82 1.14 0.82
0.165 0.197 0.200
1.19 1.35 1.00
187 496 1,661
0.59 0.86 1.14
0.117 0.153 0.231
1.26 1.28 1.13
# of Observations
All Observations Industry Biotechnology Broadcast and cable TV Communication equipment Communication services Computer networks Computer services Catalog/mail order (internet) Software Age (Years After IPO) 0–1 years 2–3 years >3 years Financial Condition No revenue Revenue but negative income Positive income Employees 0–25 26–100 Over 100
source : Kerins, Smith, and Smith (2004). Beta estimates of recently public, VC-backed firms that went public during the 1995–2000 period. Betas and correlations are computed using the S&P 500 index as the “market.”
Estimating the standard deviation of cash flows or holding-period returns. If expected cash flows are estimated by probability-weighted scenario analysis or simulation, the information needed to estimate the standard deviation of cash flows is generated at the same time and is easily used. An illustration appears later in the chapter. However, as stated, it is not easy to go from the standard deviation of cash flows to the standard deviation of holdingperiod returns. This may not be a serious problem. If the objective is simply to determine whether the expected cash flow is high enough to justify investment of a certain amount that is known in advance, then an approximation of the standard deviation of holding-period returns (and beta) can be computed on the basis of the
New Venture Valuation in Practice 423
required investment. This is how we computed the variance of holding-period returns when we used scenario data to estimate beta. However, if the objective is to determine how much a particular risky asset is actually worth, as we explained in Chapter 10, it is not possible to determine the correct standard deviation of holding-period returns without simultaneously determining the value of the cash flows, which involves a trial-and-error process that can be tedious. If you cannot derive a reliable estimate of beta from data for comparable firms, it can be better and easier to circumvent the problem of estimating beta by using the CEQ form of the CAPM. The advantage of the CEQ approach is that it is not necessary to determine the standard deviation of holding-period returns and the PV of the cash flow simultaneously. Look back at Eq. (10.8). Instead of estimating the cash flow beta from scenarios, you can estimate the elements of the cash flow beta. Estimating the standard deviation of market returns. We suggest using historical data on market returns to estimate the market standard deviation. In Table 11.4, we report the standard deviation of a long-run historical average of annual holding-period returns for investing in the S&P 500. Our estimate is an annual standard deviation of 16.7%. To compute the standard deviation for a holding period other than one year, we assume that holding-period returns from one year to the next are independent of each other and the risk distributions are identical over time.11 Under this assumption, we can compute the standard deviation for a different holding period either by Tab le 11 . 4 Standard deviation of market returns (S&P 500)
Holding Period Length
1 year 2 years 3 years 4 years 5 years 6 years 7 years 8 years 9 years 10 years
Standard Deviation
Variance
16.69% 23.60% 28.91% 33.38% 37.32% 40.88% 44.16% 47.21% 50.07% 52.78%
2.79% 5.57% 8.36% 11.14% 13.93% 16.71% 19.50% 22.28% 25.07% 27.86%
The 1-year standard deviation is computed over the 50 years ending with 2017, based on the annual holding-period returns of the S&P 500 index. Returns for longer holding periods are calculated assuming time-series independence of annual returns.
424
Chapter Eleven
ultiplying the standard deviation by the square root of the time interval in m years or by multiplying the annual variance by the time interval and then taking the square root. For example, the two-year standard deviation is
σ 2 years = 16.69% × 2 = 23.60% or
σ M = (16.69%) 2 x 2 = 23.60% For convenience, in Table 11.4 we report market standard deviations calculated for holding periods of up to 10 years. Estimating the correlation between project cash flows and market returns. Although Eq. (10.8) circumvents the need to determine the risk of the holding-period return for a cash flow, it does raise another difficult question: How can the correlation between venture cash flows and market returns be estimated? One approach is to base the estimate on judgment, in light of the nature of the risks that the venture will face. Alternatively, it may be possible to use stock returns data for public companies to gain perspective on the range of appropriate assumptions. Although the prospect of estimating a correlation coefficient between a new venture cash flow and the market portfolio may seem daunting, a realistic range of values can be narrowed easily. Assuming you can do a good job of estimating total risk, you can use data for comparable firms to generate a reasonable estimate of the correlation coefficient. Large and highly diversified corporations have returns that are more highly correlated with the market than do small and undiversified corporations. Nonetheless, even diversified corporations seldom have correlations with the market that are in excess of 0.7; a more typical level would be 0.4. For example, General Electric, widely considered one of the most diversified public companies, had a correlation coefficient of 0.50 with the S&P 500 using monthly returns data for the five-year period ended January 1, 2018. Over the same period Apple’s correlation was 0.54, and Amazon’s was also 0.54. In contrast, GoPro, a focused recent IPO firm, had an estimated correlation of 0.01 (over a shorter period since the company has not been public for five years). Low correlations with the market are not limited to young firms. Genentech, a focused biotechnology firm that was founded in 1976, had a recent correlation with the market of 0.07, and Newmont Mining, a focused gold mining company established in 1916, had a recent correlation of 0.06. Thus, for a new venture with a high degree of idiosyncratic risk and little diversification, it is unlikely that the correlation of returns with the market will exceed 0.3.
New Venture Valuation in Practice 425
This is supported by the data in Table 11.3, where the highest reported correlation is 0.26 and most correlations are in the 0.15–0.25 range. Correlations with the market increase with firm financial maturity (measured by age and financial condition) and size (measured by the number of employees). Given the narrow range of realistic values, attempting to develop a precise estimate of the correlation between project returns and market returns is probably not worth the effort. A number near 0.5 is appropriate for a market-sensitive venture such as a financial institution or public utility. A number around 0.1 is appropriate for early-stage ventures that have characteristics more like lotteries. As with all variables, a range of estimates should be assessed for their impact on the valuation.
Shortcuts for Estimating Opportunity Cost of Capital Sometimes it is possible to develop an estimate of the cost of capital without basing it directly on the CAPM. Suppose we can find mature public companies that are comparable to the venture that is being valued and that are expected to pay dividends at a constant rate. If we know the price of the stock, the dividend level, and the expected growth rate of dividends, we can estimate the cost of capital using the dividend discount model: D D P0 = 1 or r = 1 + g (11.6) r−g P0 where D1 is the dividend expected next year, g is the expected rate of dividend growth, r is the unknown opportunity cost of capital, and P0 is the current stock price. For example, if the expected dividend is $1, expected growth of dividends is 3% per year, and the current price is $12, then r=
D1 $1 +g= + 0.03 = 0.083 + 0.03 = 0.113 = 11.3% P0 $12
The resulting estimate of cost of capital is 11.3%. Another possible shortcut is to recognize that if a public company has no opportunity to invest retained cash flows at a rate different from its cost of capital, then the earnings/price ratio is an estimate of cost of capital. With no growth, an investor is basically purchasing a perpetual stream of risky earnings. For example, if earnings are expected to be $1 and the current price is $9, the estimated cost of capital is 11.1%. If the company has attractive investment opportunities, then the ratio and cost of capital will be lower, depending on the value of the growth opportunities. Both of these shortcuts are ways to estimate the cost of equity. To generate a cost of capital for assets, they must be unlevered or incorporated into a weighted average cost-of-capital calculation. For most new ventures, the approaches
426
Chapter Eleven
are unlikely to be of much value as stand-alone methods. Companies that are candidates for using these approaches are very different from new ventures, where the entire value usually depends on prospective growth. Nonetheless, for some private businesses, such as retailing or service businesses, the shortcut approaches may work well. In any case, they can serve as a reality check on the cost-of-capital estimate you derive using a CAPM-based approach.
Shortcuts in the Application of DCF The cash flows of a venture are not all the same. Some are low risk and some are high risk. Differences in risk can arise over time as well as across activities within a single period. Ideally, we would discount each cash flow at a rate that takes account of its timing and risk characteristics; in practice, this is rarely done. Short of dealing with the unique risk characteristics of each cash flow, we might aggregate the cash flows of different risks that occur in a given period and find a weighted average discount rate that is appropriate for the aggregate. This also is rarely done, as it would imply using different discount rates for different periods. What usually happens in the RADR approach is that a single discount rate is used to discount all of the project cash flows. In part this is because the data from comparable public firms, on which the cost-of-capital estimate is based, are for the entire firm and do not distinguish among separate cash flows based on risk or timing. If this shortcut is used in the RADR approach, it makes sense to select parameters for the risk-free rate and market risk premium that are of approximately the same duration as the average duration of project cash flows. Thus, if continuing value is used and almost all of the value of the project is captured in the continuing value, it would make sense to use a long-term risk-free rate and a correspondingly lower market risk premium. In the CEQ approach, it is easier to reflect risk differences in cash flows over time. It could still make sense to simplify the analysis by using a single risk-free rate and market risk premium for all periods, but this does not automatically result in a single beta. Nor does it need to in the RADR approach, if the risk attributes of the project are determined by simulation rather than simply assumed to be similar to those of comparable firms.
Testing the Consistency of Assumptions If market data for comparable firms are available, those data can be evaluated to see if the assumptions are internally consistent and reasonable. Compared
New Venture Valuation in Practice 427
to public companies, which are better established than new ventures and normally are more diversified, a new venture is likely to have higher total risk but more of the risk is likely to be idiosyncratic. Because these deviations are offsetting, Table 11.3 suggests that the asset betas of new ventures are similar to the asset betas of comparable public companies but that new ventures have higher total risk (standard deviations) and lower correlations with the market. If the factors are offsetting, then you can use your assumptions about project risk to make an ex post check. Do the assumptions result in an implied beta that is reasonable in light of what you know about the betas of comparable public companies? If not, you may want to reconsider your assumptions. If comparable firm data are available, why not just estimate the asset beta from the comparables, such as those reported in Table 11.3? The answer is twofold. First, for the following reasons, it is important to assess the total risk of the venture, not only the market risk represented by beta: • It is unlikely that you can generate a reasonable estimate of the expected cash flows without considering total risk. • A reasonable measure of total risk is important for valuing the project from the perspective of the underdiversified entrepreneur. • Total risk is also critical to the valuation of the complex financial claims and real options that are frequently features in new venture financing. Second, coming at the question of beta risk from both directions—that is, by using market comparables for beta and by inferring a beta estimate from the assumptions used in the CEQ valuation—is a good way to check the reasonableness of underlying assumptions. For an unlevered comparable firm, the estimate of correlation is given by Eq. (11.7): σM ρ (11.7) C r = βComp σr j M
Comp
In the preceding expression, the subscript, Comp, stands for comparable firm assets. Because the comparable will generate cash flows in many years, the estimate in Eq. (11.7) is best interpreted as a weighted average correlation over the life of the venture.
A Caveat The CEQ form of the CAPM works well as long as project risk (cash flow standard deviation) is not too large compared to the expected cash flow and/or as long as the correlation between project cash flows and the market is not too great. For very risky cash flows, especially if correlation with the market is
428
Chapter Eleven
high or if value is based on total risk instead of market risk only, CAPM-based valuation models tend to undervalue risky cash flows. Because this problem is more likely to arise in determining value to the entrepreneur, we defer a more complete examination of the problem to Appendix 11A. Sometimes, however, the problem can arise even for a well-diversified investor; if it does, the valuation should address it.
11.4 New Venture Valuation: An Illustration To illustrate implementation, we introduce the example of Micro Components, Inc. (MCI), a hypothetical start-up that will provide components to manufacturers of lithium-ion batteries. The MCI entrepreneur seeks funds to increase production capacity to commercial scale. The scale-up will take one year, with revenues beginning in Year 2. The entrepreneur has prepared financial projections and developed three scenarios that yield the cash flow forecast shown in the following table. She has also assigned a probability to each scenario, as shown. M icro Co m p o n ent s , I n c . (M CI) ca s h flow fo reca st ($ t h ou s a n d s)
Year Scenario
Probability
0
1
2
3
4
5
Success Likely Failure Expected
0.25 0.50 0.25
($3,000) ($3,000) ($3,000) ($3,000)
($1,500) ($1,500) ($1,500) ($1,500)
$1,000 $500 $0 $500
$3,000 $500 $0 $1,000
$5,000 $500 $0 $1,500
$9,000 $500 $0 $2,500
Continuing value
$108,000 $4,000 $0 $29,000
The expected cash flow in each period is simply a probability-weighted average of each scenario’s cash flow. To simplify, we assume that MCI is entirely equity financed and that the cash flows shown are the cash flows to shareholders. This means that the appropriate risk-adjusted discount rate is the cost of equity.12 The explicit value period is five years, at the end of which the entrepreneur anticipates two possible exits for investors. If the “success” scenario is realized, MCI would go public. You have collected data on comparable IPO transactions and have estimated that if the “success” scenario is realized, the appropriate continuing value multiple will be 12.0 times MCI’s Year 5 cash flow. Under the “likely” scenario, MCI will be sold by acquisition to a customer or supplier interested in vertical integration. You have collected information on comparable trade sale transactions and have estimated a continuing value multiple of 8.0
New Venture Valuation in Practice 429
times the Year 5 cash flow. The “failure” scenario results in a full loss of the investment, so that the continuing value in Year 5 is zero.
Using the RADR Form of the CAPM Equation (11.8) is the CAPM valuation model stated in RADR form. C jt C jt PV = Σ Tt = 0 = Σ Tt = 0 (11.8) t t (1 + rjt ) 1 + rFt + β jt ( RPMt ) In this expression, PV is the present value of all cash flows to investors in the venture. You can interpret the periods from Time 0 to T as the explicit value period, where the Time T cash flow includes continuing value. To allow for the possibility that different cash flows in the same period can vary in risk, Cjt represents a particular expected cash flow, j, at time t. For example, except for uncertainties about the tax status of a venture, depreciation cash flows (i.e., the tax savings in each period) are determined at the time the investment is made. Consequently, they are likely to be much less risky than net income or operating cash flow. Each cash flow in Eq. (11.8) is discounted by a factor that is determined by the CAPM. The elements of the discount factor are the risk-free rate, rF, the market risk premium, RPM, and beta risk, βjt. Equation (11.8) is very general. The first two variables are specific to the time period t, while βjt, which reflects the riskiness of the cash flow, is specific to both time period t and cash flow j. Thus, within a given period we may have cash flows of different risk, while across periods the risk-free rate and market risk premium may also vary. This implies that the discount rate can differ from one period to the next and for different cash flows in the same period. The cash flows in Eq. (11.8) are expected cash flows in the statistical sense, including the risk of failure. They are not the cash flows anticipated if the venture is a success. Nor are they the “likely” cash flows, which can be seen clearly in the MCI data above. The estimate of continuing value is also an expected value estimate that is conditional on the various assumptions about the venture’s exit strategy. When the RADR form is used, uncertainty is addressed in the discount rate, so it is not necessary to make a direct estimate of the uncertainty of cash flows. In practice, this may not be much of an advantage: it is hard to envision good ways of estimating expected cash flows that do not involve assessing uncertainty at the same time. When inflows and outflows within a period differ in risk, it is unlikely that you can find a single, theoretically correct discount rate that can be used to value cash flows in different periods. It is, however, possible to estimate a single (weighted average) rate that can be applied to the net cash flows in a given period.
430
Chapter Eleven
This would seem to imply that a changing (time-varying) weighted average discount rate for valuing net cash flows might be appropriate. Equation (11.8) suggests segregating the streams by riskiness and using different discount rates, both within and across time periods. In practice, it is common to simplify the RADR process by using a single rate for all cash flows. Suppose we wish to calculate a schedule of discount rates that reflect each expected cash flow’s risk and timing. We could begin by asking whether a single discount rate is appropriate. To answer, consider the venture’s cash flows at Time 0 and in Year 1. Most of the cash is being used to expand capacity, and the business has not yet begun to generate revenue. Because the costs of building capacity are assumed to be certain, there is no variance in the cash flows across the three scenarios for the first two years. Given this, these cash flows should be valued at the risk-free rate. The important risks of the venture start in Year 2, when product sales begin. The revenue-based cash flows have much more risk than the early investment cash flows. These risks can be seen in the considerable variation across the cash flows by scenario in each of Years 2–5. Because these cash flows are risky, the discount rates that are applied in Years 2–5 should not be the same as those applied to the Time 0 and Year 1 cash flows. This is intuitive, yet as noted, it is common practice when using the RADR approach to apply a single discount rate to all cash flows. Doing so in this case would result in underestimating the PV of investment outflows, which then results in an overestimation of project NPV. Nevertheless, to be consistent with standard practice (and to illustrate an advantage of the CEQ approach), we will first follow the common approach. Suppose that using the data on risk-free rates and returns, we have estimated the following CAPM parameters: Current rate on long-term Treasury bonds 4.0% Market risk premium 6.5% The final variable we need before we can implement Eq. (11.8) is the estimate of beta. MCI has identified the following three firms as being comparable, and we have collected data on their equity betas and capital structures: Comparable
Genric, Inc. Preces Systems Visania Co. note :
Equity β
Market value of equity
Debt value
1.9 1.5 1.2
$12.0 $24.0 $7.0
$4.0 $3.0 $0.0
All monetary values are in $ millions.
New Venture Valuation in Practice 431
Because these are equity betas, the first step is to determine the asset betas using Eq. (11.5):
Comparable
Equity β
MV equity
Debt
Asset value
Equity to asset value
Asset β
1.9 1.5 1.2
$12.0 $24.0 $7.0
$4.0 $3.0 $0.0
$16.0 $27.0 $7.0
0.75 0.89 1.00
1.43 1.33 1.20
Genric, Inc. Preces Systems Visania Co. note :
All monetary values are in $ millions.
Based on a simple average of the asset betas, we arrive at an estimated βA of 1.32 for MCI. Using our estimates of a 4.0% risk-free rate and a 6.5% market risk premium, we can now calculate the required return on assets, rA: rA = rF + β A (rM − rF ) = 4% + 1.32(6.5%) = 12.58% The cash flows, discount rate calculations, and PV computation are summarized in Table 11.5, “Valuation Template 1,” which is available on the companion website as a downloadable spreadsheet. The “Project Information” section is the same data we saw earlier for MCI and includes each scenario’s cash flow Tab le 11 . 5 Valuation Template 1: Valuation by the RADR method based on discrete scenario cash flow forecast
Year
Project Information Cash Flows ($000s) Success scenario
Probability
0
1
2
3
4
5
0.25
–$3,000
–$1,500
$1,000
$3,000
$5,000
$117,000
Expected scenario
0.50
–$3,000
–$1,500
$500
$500
$500
$4,500
Failure scenario
0.25
–$3,000
–$1,500
$0
$0
$0
$0
–$3,000
–$1,500
$500
$1,000
$1,500
$31,500
Expected Cash Flow Market Information Risk-free rate
4.00%
8.16%
12.49%
16.99%
21.67%
10.50%
22.10%
34.92%
49.09%
64.74%
Market risk premium
6.50%
13.94%
22.44%
32.10%
43.08%
Comparable firm beta
1.32
1.32
1.32
1.32
1.32
12.58%
26.56%
42.10%
59.36%
78.53%
–$1,332
$395
$704
$941
$17,644
Market rate
Estimated Cost of Capital Market Value Estimate Present Value of Expected CF Sum of PVs
–$3,000 $15,352
Value is estimated using the RADR form of the CAPM. Shaded cells are inputs. All dollar values are in thousands.
432
Chapter Eleven
forecast and the expected cash flow in each year. The operating cash flows in Year 5 and the continuing values have been combined in this table and shown as Year 5 cash flows. In the “Market Information” section of the template, we provide the data and calculations used to estimate the RADR each year. In this table, we use the same annual discount rate each period, ignoring the differences in risk we noted earlier. As a result, the beta is assumed to be constant, consistent with the most common use of the RADR approach. The starting values for the relevant assumptions are shown in Year 1 for a one-year holding period. The risk-free rate is 4%, and the market rate is 10.5% (based on our assumed market risk premium of 6.5%). For longer holding periods, the cumulative risk-free rates and cumulative market rates shown are found by compounding the one-year rates; the market risk premium is just the difference between the two. For example, a two-year investment in the riskless asset is expected to yield a total return of 8.16%, and the two-year market return is estimated to be 22.10%. This results in the following estimated discount rate for the Year 2 cash flow: rYear 2 = 8.16% + 1.32(22.10% − 8.16%) = 26.56% Summing the PVs of each year’s expected cash flow produces an estimate of $15,352 as the value of MCI.
Using the CEQ Form of the CAPM We now use the CEQ approach to estimate the value of MCI. We have already estimated RPM and rF (6.5% and 4.0%, respectively) and have also calculated the expected cash flows, Ct. Recall that Eq. (10.8) values a single cash flow at time t, meaning that we must apply it to each period’s cash flow. We have estimates of the standard deviation of the market from Table 11.4, so the only variables we still need are the standard deviation of MCI’s cash flows and their correlation with the market. For the latter, we decided to use the overall mean correlation value of 0.195 from Table 11.3 as a reasonable estimate. The cash flows, discount rate calculations, and CEQ PV computation are summarized in Table 11.6, “Valuation Template 2,” which is also available as a downloadable spreadsheet on the companion website. First, we note the main differences between Table 11.6 and Table 11.5. Under “Project Information,” the cash flows and scenarios are the same, but Table 11.6 adds the “Standard deviation of CFs” line. These cash flow standard deviations are computed separately for each year based on the three cash flow scenarios. Consistent with our earlier discussion, for Time 0 and Year 1, the standard
New Venture Valuation in Practice 433 Tab le 11 . 6 Valuation Template 2: Valuation by the CEQ method based on discrete scenario cash flow forecast
Year Project Information
Cash Flows ($000s)
Probability
0
1
2
3
4
5
Success scenario
0.25
–$3,000
–$1,500
$1,000
$3,000
$5,000
$117,000
Expected scenario
0.50
–$3,000
–$1,500
$500
$500
$500
$4,500
Failure scenario
0.25
Expected Cash Flow Standard Deviation of CFs
–$3,000
–$1,500
$0
$0
$0
$0
–$3,000
–$1,500
$500
$1,000
$1,500
$31,500
$0
$0
$354
$1,173
$2,031
$49,398
Market Information Risk-free rate Market rate
4.00%
8.16%
12.49%
16.99%
21.67%
10.50%
22.10%
34.92%
49.09%
64.74%
6.50%
13.94%
22.44%
32.10%
43.08%
14.50%
20.51%
25.11%
29.00%
32.42%
0.195
0.195
0.195
0.195
0.195
($1,442)
$419
$707
$907
$15,371
Market risk premium Market variance Market standard deviation Correlation Market Value Estimate Present Value of Expected CF
($3,000)
Sum of PVs
$12,963
Diagnostic Information Annualized required return
4.0%
9.2%
12.2%
13.4%
15.4%
Standard deviation of returns
0.00%
0.00%
84.39%
165.76%
223.82%
321.36%
Covariance with market
0.00%
0.00%
3.37%
8.12%
12.66%
20.32%
Beta
0.00
0.00
0.80
1.29
1.51
1.93
note:
Value is estimated using the CEQ form of the CAPM. Shaded cells are inputs. All dollar values are in thousands.
deviation is zero since the cash flows are invariant across scenarios. For Year 2, the standard deviation of $354 is calculated as follows:
($1, 000 − $500)
σ CF
=
σ CF
= $354
Year 2
Year 2
2
(
)
2
(
)
2
× 0.25 + $500 − $500 × 0.5 + $00 − $500 × 0.25
In the “Market Information” section of the template, we show the values for all of the relevant assumptions in Year 1. In subsequent years, the risk-free rate and market rate, both expressed in annual terms, are simply being compounded. Values for market standard deviation are taken directly from Table 11.4, and the correlation value of 0.195 is an assumption based on Table 11.3.
434
Chapter Eleven
In the “Market Value Estimate” panel, each period’s value in the template reflects the application of Eq. (11.6) to the data for that year. In Year 3, for example, the calculation is as follows: PVt =
Ct −
ρC , r σ C
M t × ( rM − rF ) $1, 000 − 0.195 × $1,173 × 0.2244 σM $796 0.2511 = = 1 + 0.1249 1.1249 1 + rF
t
PVt = $707
The numerator in the next-to-last term of the equation, $796, is the certainty equivalent CF, which is being discounted by the three-year risk-free rate of 12.49%. Summing each PV gives a total (net) present value for MCI of $12,963. Finally, the “Diagnostic Information” panel in Table 11.6 contains information that can be used to help understand the valuation and to assess the reasonableness and internal consistency of the assumptions. The first line in the panel shows the annualized required rate of return for each annual cash flow. As shown, the required rates are highly variable. Each is calculated in the spreadsheet by comparing the expected cash flow in each year to its present value and converting to an annual rate. For example, the correct risk-adjusted holding-period return for the Year 2 cash flow is Ct −1 = PVt
$500 − 1 = 0.092 or 9.2% $419
The “Diagnostic Information” panel also shows the beta estimate for each period’s cash flow. The estimate for each year is calculated based on the data in the template and the formula for the cash flow beta. For example, the Year 4 beta estimate is computed as follows: σCt ρCt , rM × PVYear 4 βt = σM
βYear 4
σC ρCYear 4 , rM Year 4 0.195 $2, 031 PVYear 4 $907 = 1.51 = = 0.2900 σ M Year 4
The estimated betas vary substantially from year to year and are also different from the assumed beta of 1.32 used in the RADR approach.
Comparing the CEQ and RADR Approaches We now can examine the differences between our CEQ and RADR results. The CEQ approach produced a value estimate of $12,963, while the RADR re-
New Venture Valuation in Practice 435
sult is $15,352. The most important factor in the difference is that the RADR approach as applied in Table 11.5, by assuming a single beta, uses the wrong discount rate to value each annual cash flow. This can be seen by comparing the PV estimates for the annual cash flows computed by the RADR approach with the estimates computed by the CEQ approach. The Year 1 cash flow is a good place to start. MCI will spend the first year building capacity, so there is little uncertainty about the required investment. In fact, the standard deviation of the cash flows is zero, which implies no risk. But in Table 11.5, the riskless Year 1 expected cash flow of negative $1,500 is being discounted at the RADR of 12.58%. In Table 11.6, the expected cash flow and the CEQ cash flow are identical, and the discount rate applied is the riskless rate of 4%. Rather than the correct PV of negative $1,442, the RADR PV of the Year 1 cash flow is negative $1,332, which contributes to overvaluation of the venture. In fact, in every year the RADR discount rate is incorrect. The biggest discrepancy is in Year 5, when the CEQ model estimates the required risk-adjusted return at 108%, but the rate used in the RADR calculation is only 78.5%. By using a discount rate that is too low, the RADR estimate is overvalued by $2,507 in Year 5 and by $2,635 in aggregate across all years. With the RADR approach, the discount rate each period is based on the assumption that the beta stays constant at 1.32. However, when we compare the two templates, we can see that an assumption of constant risk is not correct. The betas implicit in Table 11.6, the CEQ model, are different every year. Our simple example exposes the limitations of the RADR approach when compared to the CEQ approach and highlights some of the challenges of implementing both. We do not, however, want you to lose track of its advantages. The principal strengths of the RADR approach are as follows: • Valuation is based on expected cash flows. • The discount rate is intended to be opportunity cost of capital. • Market data can be used to estimate cost of capital. • It is unnecessary to estimate the total risk or the correlation with the market. The main disadvantages of the RADR approach are as follows: • If separate information is unavailable on cost of capital, then holdingperiod returns and cost of capital must be determined simultaneously. • Truly comparable firms are unlikely to be available for most new ventures.
436
Chapter Eleven
• The model works best if it is applied to cash flows that are segregated according to market risk; but in that case, the appropriate discount rate for valuing a single cash flow cannot normally be determined based on data from comparable firms. • If information on total risk is not generated, it is difficult to value complex financial claims on the underlying asset. The benefits of the CEQ form of the CAPM are as follows: • Valuation is based on expected cash flows. • Cost of capital is used to value each annual cash flow. • Cash flows that differ in terms of total risk are handled easily. • Cash flows at different times can easily be valued separately. • Any financial claim can be valued, as long as the CAPM assumptions hold. • A measure of the total risk of cash flows is generated and will be useful for valuation by an underdiversified entrepreneur. The main disadvantages of the CEQ approach are as follows: • An estimate of the full distribution of cash flow possibilities is required, instead of a single estimate. • The correlation coefficient between venture cash flows and the market can be difficult to estimate, even if data on the betas of comparable firms are available.
Using the Relative Value Method Because MCI is an early-stage venture with no revenue, it would be very difficult to find comparable public companies to apply relative valuation to the venture at an early stage. However, we can use the relative value approach to estimate the company’s continuing value at the end of the explicit value period, when it is expected to look more like established firms. Under the “success” scenario, we assumed that MCI will go public at the end of Year 5 and under the “likely” scenario that the venture will be sold to a customer in a trade sale. To estimate value at the end of Year 5 under the “success” scenario, we have collected the data shown in the following table on recent IPO transactions. Like MCI, all of the comparable firms are financed entirely with equity, so we do not need to adjust for capital structure.
New Venture Valuation in Practice 437 R ece nt I P O s
Financial data (last 12 months) Company
Market cap at IPO
Revenue
EBIT
$134,125 $10,500 $97,350
$89,420 $6,180 $44,250
$14,420 $1,460 $7,490
Greyport Networks Spectrial Labs Indeve Inc. note :
Earnings
Cash flow to all investors
$5,590 $330 $5,120
$8,770 $880 $11,095
All data are in $ thousands.
Market capitalization in the preceding table is the total value of each firm calculated as the IPO issue price times total post-IPO shares outstanding, including IPO shares. It represents the value the entrepreneur or pre-IPO investors would receive by selling their shares as part of the IPO. MCI’s forecast is of cash flow to all investors, so we focus our comparables analysis on the last column of data. In practice, we would also use MCI’s forecast data on revenue, EBIT, and earnings, as well as multiples of the additional data for the comparables; for this example, for illustration, we use only cash flows to all investors. As discussed earlier, we would also want to consider the current market conditions and determine whether today’s multiples represent reasonable estimates for transactions 5 years in the future. For each IPO transaction, we calculate the ratio of market capitalization to CF to all investors. This tells us approximately how much IPO investors were willing to pay per dollar of each comparable firm’s cash flow. R ece nt I P O s
Company
Greyport Networks Spectrial Labs Indeve Inc. Average note :
Market cap at IPO
$134,125 $10,500 $97,350
Cash flow to all investors
$8,770 $880 $11,095
Market cap/cash flow
15.3 11.9 8.8 12.0
All monetary values are in $ thousands.
Because we consider the comparables to be equally informative about the value of MCI, we average the market cap ratio across the three comparables. The resulting market capitalization/cash flow ratio is 12.0. Based on this, we infer that 12.0 times cash flow to all investors is a reasonable estimate for the continuing value of MCI at the end of 5 years, if the venture achieves the “success” scenario. If MCI develops according to the “likely” scenario, the exit at the end of Year 5 will be by trade sale to a strategic partner (i.e., acquisition). Relative valuation
438
Chapter Eleven
can be used to estimate the continuing value at this point, and we have collected the data below on comparable M&A transactions. R ece nt M & A t ra n s ac t io n s
Financial data (last 12 months) Target
Acquirer
Price paid
Biros Inc. Viage Ent. Mecent Labs Protoscan Inc.
Kinerion Inc. Bantic Networks Mercuron Co. Neurovage, L.V.
$75,650 $32,500 $145,950 $88,275
note :
Revenue
$58,190 $32,500 $153,630 $73,560
EBIT
Earnings
Cash flow to investors
$9,460 $5,000 $12,160 $9,100
$3,960 $2,560 $5,770 $4,700
$7,200 $4,710 $17,388 $14,240
All monetary values are in $ thousands.
In these transactions, price paid is what the purchaser paid for the target firm’s assets and their associated cash flows. For each comparable, we compute the ratio of price paid to CF to all investors, which is shown in the right-hand column in the following table. Considering the transactions to be equally informative, we then average the four ratios to get 8.0. R ece nt M & A t ra n s ac t io n s
Target
Acquirer
Biros Inc. Viage Ent. Mecent Labs Protoscan Inc. Average
Kinerion Inc. Bantic Networks Mercuron Co. Neurovage, L.V.
note :
Price paid
$75,650 $32,500 $145,950 $88,275
CF to investors Price/cash flow
$7,200 4,710 $17,388 $14,240
10.5 6.9 8.4 6.2 8.0
All monetary values are in $ thousands.
Based on this analysis, we use an estimate of 8.0 times CF to all investors for the continuing value of MCI if it develops according to the likely scenario. For consistency, we have used the multipliers derived by application of the RV method as continuing value throughout this comparison of valuation methods.
Using the Venture Capital Method We now consider valuing MCI by the VC Method. In our example, the final year is Year 5, and under the “success” scenario we assume that MCI will go public with a continuing value of 12.0 times cash flow to shareholders. We know from the CEQ model that the correct PV is $12,963. The VC Method includes the following steps:
New Venture Valuation in Practice 439
Step 1. Select a year for the start of the continuing value period of the valuation by determining a point where, if the venture is successful, harvesting by acquisition or IPO or other means would be feasible. Estimate net income or other cash flow in that year based on the “success” scenario. Step 2. Use the appropriate P/E ratio or other multiple and the harvest-date earnings or cash flow projection to compute continuing value. The multiple should reflect the expected capitalization of earnings or cash flow for a company that has achieved the level of success reflected in the scenario. Step 3. Convert the continuing value estimate to PV by discounting at a hurdle rate that you believe is high enough to compensate for time value, risk, and the probability that the “success” scenario will not be achieved. Step 4. Based on estimated PV, compute the minimum fraction of ownership an investor would require in exchange for contributing a given amount of capital. In the VC Method, we focus only on the cash flows of the “success” scenario, valuing them at a hurdle rate appropriate for the stage of development of the project. Unfortunately, this approach does not provide much guidance. If the project is considered to be in first-stage development, then the table of hurdle rates presented in Chapter 10 suggests a discount rate of 40–60%. In Table 11.7, we compute the present values of MCI’s “success” scenario cash flows at hurdle rates of both 40% and 60%. At 40%, the value is $23,588, and at 60%, it is $12,106, slightly lower than the CEQ valuation of $12,963. Such a broad range is not very helpful, especially if the valuation is to be used as the basis for an investment decision, such as to assess whether it would be a good
Tab le 11 .7 Valuation at various discount rates by the Venture Capital Method
Cash Flows
Total
Success scenario Discount Rate = Present value Discount Rate = Present value
0
1
2
3
4
5
($3,000)
($1,500)
$1,000
$3,000
$5,000
$117,000
($3,000)
($1,071)
$510
$1,093
$1,302
$21,754
($3,000)
($938)
$391
$732
$763
$11,158
($3,000)
($950)
$401
$763
$806
$11,943
40% $23,588 60% $12,106
Implied Single Rate Rate
57.84%
Present value
$12,963
440
Chapter Eleven
idea to invest $18,000 today to acquire the venture. In Table 11.7, we also solve for the single hurdle rate that generates the true PV, which works out to be 57.84%. One problem with the VC Method should now be obvious: there is little basis for selecting the hurdle rate. Furthermore, small errors in the rate can dramatically affect the estimated PV. A second problem can be detected by comparing the PVs of the individual annual cash flows that are shown in Table 11.7 with the PVs of the same cash flows in Table 11.6. Using a 57.84% discount rate produces the same overall project value as the CEQ approach, but the value of each annual cash flow provided by the VC Method is wrong, usually by an amount larger than when the RADR approach is used. There is no good way to generalize about the nature of these errors except to say that they are certain to occur and that the magnitudes of the errors will increase with increases in the variability of cash flows in different years. Perhaps with enough experience using hurdle rates that eventually proved to be too high or too low, an entrepreneur or investor could find the VC Method to be a useful heuristic approach. It is easy to criticize the VC Method on quantitative grounds. However, the more qualitative advantages are also worth noting: • The valuation can be driven by the financial projections for a “success” scenario that may be reported in the business plan. • The negotiation process may be facilitated by centering the negotiations on the entrepreneur’s projections. • The investor’s experience may be easiest to apply without formal analysis when comparisons of ventures are made on the basis of “success” scenarios. • The method is easy to use and may be adequate for simple investment decisions. The disadvantages include the following: • Lack of precision due to reliance on unnecessarily limited information and rules of thumb. • Biases resulting from discounting optimistic cash flow projections at a hurdle rate that is above cost of capital. • Lack of information about uncertainty, which would be useful for valuing complex financial claims. Before you decide not to use the VC Method, you should give some thought to the negotiation process. Consider the following scenario: An entrepreneur who is enthusiastic about an idea comes to you with a business plan. You, as an investor, like the concept and believe the team is the right one to undertake the
New Venture Valuation in Practice 441
project. Like many entrepreneurs, however, this one seems overly optimistic; in particular, the entrepreneur’s financial projections are based on a scenario in which the venture develops according to plan, with no major obstacles or setbacks. You would like to go forward, but you believe that the entrepreneur’s projections seriously overstate the likely success. What do you do? Do you engage in a debate with the entrepreneur about the projections, or do you simply try to counter optimism with pessimism by applying a discount rate that is biased upward to compensate for the entrepreneur’s inflated projections? In the past, at least, the answer has often been the latter. It sometimes is easier to build the valuation on a structure of compensating errors than to work toward unbiased value projections on which both parties agree. Although the negotiation might be centered on the VC Method valuation, you can protect yourself from mistakes by backing up your internal analysis with a method that does not suffer from the biases.
Using the First Chicago Method In the way we applied the RADR approach to value expected cash flows generated from discrete scenarios, we already illustrated the First Chicago Method, with the specific added assumption that the CAPM is the correct asset pricing model. As such, there is no reason to repeat. Although the First Chicago Method proposes to discount expected cash flows at their opportunity cost of capital, it does not provide guidance on how to determine cost of capital. A natural approach would be to use comparable ventures, but the method provides no guidance on how to assess comparability. With simple modifications to the “Market Information” panel of Table 11.5, you can, if you wish, apply the First Chicago Method without the constraint that the discount rate be determined by the CAPM. The strengths of the First Chicago Method include the following: • Discrete scenarios provide a simple and easy method of determining both risk and expected return. • The intent is to value expected cash flows. • The intent is to discount the cash flows at an estimate of opportunity cost of capital. • Because information about total risk is derived, the method provides a basis for valuing complex financial claims. The disadvantages are as follows:
442
Chapter Eleven
• Discrete scenarios discard information about the risk that could be useful, especially for valuing complex claims. • No basis is provided for assigning probabilities to the different scenarios used in the valuation. Although any valuation model has limitations, it generally makes sense to use the best one you can. The best model is one that values expected cash flows based on the opportunity cost of alternative investments, giving reasonable consideration to the cost of conducting the valuation and the importance of making a decision quickly. Generally, this points to either the First Chicago Method or the CEQ model. The latter has the advantage of building on the discipline of the CAPM as a basis for selecting the discount rate and can deal with a comprehensive description of venture uncertainty.
11.5 The Cost of Capital for Non-U.S. Investors Perhaps the investor lives in a country other than the U.S., possibly a developed country such as the U.K. or Japan or maybe an emerging economy such as Mexico or China. If so, you may be wondering about the relevance of diversification and the CAPM as drivers of opportunity cost of capital. While there are differences, the differences are not as great as they appear. Whether the investor is in a developed country or an emerging economy, opportunity cost is always the guiding principle of new venture investing, and the ability to invest in a diversified portfolio that includes an investment in the venture is a determinant of opportunity cost. In all developed and many other countries, investors have the opportunity to invest in a diversified market portfolio. However, because of currency exchange rates, doing so could subject the investor to somewhat different risks. In countries such as the U.K. and Japan, an investor can also invest in highly diversified domestic portfolios. So while the market index may be different, the economic principle that only nondiversifiable risk matters still applies, although estimating opportunity cost may be more difficult because of data limitations. Except for differences in expected inflation, however, the opportunity cost of capital implied by the CAPM for U.S. investors is likely to be similar to the cost of capital implied by well-diversified portfolios throughout the world. The practical challenge of estimating opportunity cost may be greater in emerging economies. Prohibitions may exist against investments in foreign diversified portfolios; exchange rates may be subject to dramatic swings or may
New Venture Valuation in Practice 443
be artificially constrained; investors may have limited opportunities to diversify domestically; and they may face other risks of long-term investing, such as a potential for expropriation. Nevertheless, the investment principles are the same. Projections of expected cash flows should include scenarios that incorporate such factors as the risk of expropriation, and the opportunity cost of capital should reflect the ability to diversify. An estimate of the opportunity cost of capital should therefore consider how the risk of investing in the venture would limit the investor’s ability to diversify or would contribute to diversification. As a first approximation, the investor could still use the CAPM model but with different assumptions for the risk-free rate, the risk premium on a portfolio that was diversified as much as possible, the total risk of the venture, and the correlation between the venture and the diversified portfolio.
11.6 Some Practical Caveats on Implementation Be cautious in relying on comparable-firm valuations that are based on private transactions. As discussed in Chapter 4, VC funding rounds often include sweeteners such as ratchets and preferences that make the shares issued in those rounds more valuable than plain vanilla common stock. Yet the share prices from these later rounds are often used as if there were no sweeteners to compute the post-money value of the venture. As a result, the private valuation can overstate the public market value. How serious is this potential for overvaluation? To address this question, Gornall and Strebulaev (2017) compare the reported private valuations of unicorns to their subsequent IPO values or valuations based on or fundamental valuations of all share classes. The term “unicorn” is applied to private firms with reported valuations of at least $1 billion. They include such firms as Airbnb, Box, CloudFlare, Dropbox, GoPro, Lyft, Pinterest, Snap, Square, Uber, and Zynga. The authors estimate that post-money valuations based on the last private funding round overstated firm value by an average of 50%. Do not forget about stock-based compensation. VC-backed private firms often have established large stock-based option pools that are used to compensate early-stage employees in lieu of higher cash compensation. Such pools have the potential to materially dilute company earnings relative to comparable public companies. Moreover, if valuations or market values decline, to
444
Chapter Eleven
retain their employees, firms are sometimes compelled to sweeten stock-based compensation arrangements. Watch for bias in the selection and use of comparable firms. In a recent study of data from countries where underwriters are required to disclose the comparable firms used in their valuations, Paleari, Signori, and Vismara (2014) find that underwriters bias the selection of comparables by systematically excluding firms that would make a prospective issuer appear to be overvalued. While they price the offering at a discount relative to selected comparables, the price is still above what would be implied by an unbiased selection of comparables.
11.7 Summary Given the growth in the VC industry and the intensity of competition among investors, methods of private firm valuation reflect increasing sophistication. The (possibly indirect) focus of the valuation effort is on the PV of future cash flows the investor would receive. Much of this chapter is devoted to methods for determining the information needed to carry out a valuation. Although we are careful and systematic in coverage of these methods, we also identify shortcuts that can sometimes be used. The valuation approaches we apply in this chapter differ in several dimensions. The more traditional approaches are easy to use, in a computational sense, and potentially can facilitate negotiation between investor and entrepreneur in simple transactions. On the other hand, they are more prone to result in valuation errors than are approaches based on financial economic theory. Although all valuation models have limitations, it generally makes sense to use as many approaches as the data allow. The best model from a theoretical perspective is one that values expected cash flows based on the opportunity cost of alternative investments, giving reasonable consideration to the cost of the valuation and the importance of making a decision quickly. Generally, this points to either the First Chicago Method or the CEQ model. The latter has the advantage of building on the discipline of the CAPM as a basis for discount rate selection and can deal with a more comprehensive description of venture risk. That said, relative valuation and multiples that incorporate current market data on comparable ventures and transactions can be used to assess continuing value and to test the DCF results.
New Venture Valuation in Practice 445
Review Questions
1. What are some of the key criteria for choosing a valuation method? 2. Explain the theory behind continuing value. What are the important factors to consider when calculating continuing value? 3. What are some of the ways to estimate the beta for a new venture? Describe some of the challenges to developing an accurate estimate. How does financial leverage impact your beta estimate? 4. Is the dividend discount model a good shortcut for estimating the cost of equity for a new venture? Why or why not? 5. Describe the process of estimating the appropriate risk-free rate for estimating a DCF discount rate. Where can data be found for estimating the risk-free rate? 6. Describe the process of estimating the appropriate market risk premium. Where can data be found for estimating the market risk premium? 7. Describe the various methods of computing beta in the CEQ approach. 8. Is relative valuation a useful approach for estimating the value of new ventures? Why or why not? Under what circumstances would you be confident using relative valuation? 9. What are the quantitative criticisms of the VC Method? In spite of its shortcomings, why is this valuation approach still popular? 10. For social networking ventures, what kinds of indicators do you think might be useful at early stages to help assess the value of an opportunity?
Notes
1. Application of multiplier-based approaches to valuation is more defensible when the venture is well established than when it is at an early stage of development. At later stages, inference of value from income or balance sheet information is more reasonable and accurate. 2. Sometimes continuing value is referred to as “terminal value.” The rationale is that it is a valuation at a point where existing investments could reasonably be “terminated” by sale to others. It does not mean that the venture is expected to be terminated or even that the financial claims are expected to be sold, only that they could be sold without unusual difficulty. We prefer “continuing value” because it suggests that the venture is ongoing.
446
Chapter Eleven
3. See Damodaran (2006) and McKinsey and Co., Koller, Goedhart, and Wessels (2015, chap. 12). 4. Historical information on risk-free yields and real rates is provided by the U.S. Treasury at https://w ww.treasury.gov/resource- center/data- chart -center/interest-rates/Pages/TextView.aspx?data=yield. 5. The preferred rate is from a zero-coupon security. However, only the three-month, six-month, and one-year Treasury bills are sold as zero-coupon securities. For longer-term risk-free rates, Treasury “STRIPS” (Separate Trading of Registered Interest and Principal of Securities) are used to proxy for zero-coupon bonds. Information on current U.S. government bond interest rates is readily available from free services such as Yahoo! Finance, the Federal Reserve Bank of St. Louis (http://research.stlouisfed.org/fred2/), WSJ .com, and Bloomberg.com. Yields on STRIPS are available at WSJ.com. 6. See, for example, Brealey, Myers, and Allen (2017); Titman (2016); and Ross, Westerfield, Jaffe, and Jordan (2016). 7. These include Bloomberg, Yahoo! Finance, and Value Line, among others. 8. For a more detailed exposition, including the possibility of risky debt, see Koller, Goedhart, and Wessels (2005, appendix D). 9. See http://pages.stern.nyu.edu/~adamodar/New_Home_ Page/datafile/ Betas.html for more recent estimates of asset betas that are adjusted for taxes and cash balances. 10. We could also value the equity directly using the cash flows to stockholders. The correct discount rate is the cost of equity, rE , which can be computed using the CAPM and βE for the new venture. We can estimate βE using the weighted average βA and values for E and V in Eq. (11.5) specific to the new venture. 11. There is both empirical and theoretical support for assuming independence of the time series. 12. Under these assumptions, the cost of equity will be the same as the required return on assets, since without leverage, βE and βA are the same.
References and Additional Reading
Black, E. 2003. “Usefulness of Financial Statement Components in Valuation: An Examination of Start-Up and Growth Firms.” Venture Capital: An International Journal of Entrepreneurial Finance 5 (1): 47–69. Brealey, R. A., S. C. Myers, and F. Allen. 2017. Principles of Corporate Finance, 12th ed. New York: McGraw-Hill. Damodaran, A. 2006. Damodaran on Valuation: Security Analysis for Investment and Corporate Finance, 2nd ed. New York: Wiley.
New Venture Valuation in Practice 447
DeAngelo, H., and R. Masulis. 1980. “Optimal Capital Structure Under Corporate Taxation.” Journal of Financial Economics 8: 5–29. Fama, E., and K. French. 2002. “The Equity Premium.” Journal of Finance 47: 637–59. Fernandez, P. 2009. “Market Risk Premium Used in 88 Countries in 2014: A Survey with 8,228 Answers.” SSRN working paper. http:// papers .ssrn .com/sol3/papers.cfm?abstract _ id=2450452 . Gornall, W., and I. A. Strebulaev. 2017. “Squaring Venture Capital Valuations with Reality.” Stanford University Graduate School of Business research paper. SSRN: https://ssrn.com/abstract=2955455. Graham, J. R., and C. R. Harvey. 2007. “The Equity Risk Premium in January 2007: Evidence from the Global CFO Outlook Survey.” ICFAI—Journal of Financial Risk Management 9: 46–61. Kaplan, P. D., and J. D. Peterson. 1998. “Full Information Industry Betas.” Financial Management 27: 85–93. Kerins, F., J. K. Smith, and R. L. Smith. 2004. “Opportunity Cost of Capital for Venture Capital Investors and Entrepreneurs.” Journal of Financial and Quantitative Analysis 39: 385–405. Koller, T., M. Goedhart, and D. Wessels. 2005. Valuation: Measuring and Managing the Value of Companies. New York: Wiley. McKinsey and Co., T. Koller, M. Goedhart, and D. Wessels. 2015. Valuation: Measuring and Managing the Value of Companies, 6th ed. New York: Wiley. Miller, M. H. 1977. “Debt and Taxes.” Journal of Finance 32: 261–75. Paleari, S., A. Signori, and S. Vismara. 2014. “How Do Underwriters Select Peers When Valuing IPOs?” Financial Management 43: 731–55. Robichek, A. A., and S. C. Myers. 1966. “Conceptual Problems in the Use of Risk-Adjusted Discount Rates.” Journal of Finance 21: 727–30. Ross, S. A., R. Westerfield, J. Jaffe, and B. Jordan. 2016. Corporate Finance, 11th ed. New York: McGraw-Hill Irwin. Scherlis, D., and W. Sahlman. 1987. “Method of Valuing High-Risk, LongTerm Investments.” Harvard Business School Note 9-288-006. Boston: Harvard Business Publishing. Sick, G. 1986. “A Certainty-Equivalent Approach to Capital Budgeting.” Financial Management 15: 23–32. Titman, S. 2016. Valuation: The Art and Science of Corporate Investment Decisions, 3rd ed. London: Pearson. Welch, I. 2000. “Views of Financial Economists on the Equity Premium and Other Issues.” Journal of Business 73: 501–37, with 2009 update at http:// www.ivo-welch.info/academics/equpdate-results2009.html).
C h a p t e r Ele ven Appen dix
Th e E ntr e pr e n e u r ’ s Pe rs pec tive o n Valu e
E n t r e p r e n e u r s h i p i s r i s k y— e s p ec i a l ly since the entrepreneur
is necessarily underdiversified. In Chapter 10, we showed how underdiversification increases the entrepreneur’s opportunity cost of capital relative to that of a well-diversified investor. We demonstrated how to extend the CAPM to estimate the effect of underdiversification on the entrepreneur’s opportunity cost. In this appendix, we continue the analysis by illustrating how the value of the entrepreneur’s claim depends on how much of the entrepreneur’s wealth is committed to the venture, and we use simulation to show the impact of underdiversification on value. This analytical approach, which is based on opportunity cost, is of practical use whenever an entrepreneur would be compelled to invest a material fraction of personal wealth and human capital in a venture in order to pursue it. In such cases, it is also useful for evaluating alternative financial structures of ownership between an underdiversified entrepreneur and welldiversified investors.
11A.1 Estimating the Entrepreneur’s Commitment to a Venture To estimate the entrepreneur’s opportunity cost, we need an estimate of the entrepreneur’s wealth that includes the value of human capital and an estimate of the fraction of that wealth that is committed to the venture. Much of the entrepreneur’s investment is likely to be in the form of effort (i.e., human capital) committed to the venture. Effort has an opportunity cost; the entrepreneur could have devoted the effort to a different occupation than the venture. Accordingly, the entrepreneur’s wealth includes both present financial 448
The Entrepreneur’s Perspective on Value 449
capital and the PV of the entrepreneur’s human capital in its highest-valued alternative use. Two issues are of primary concern: 1. How can we estimate the value of the entrepreneur’s human capital? 2. How can we estimate the value of the human capital committed to the venture? As to the first, we use a growing annuity approach to estimating the PV of expected future earnings in the highest-valued alternative use. t C 1+ g PVcomp = 1 1 − (11A.1) r − g 1+ r where PVcomp is the PV of future compensation over the entrepreneur’s work life if the venture is not undertaken, C1 is first-year annual compensation in the best alternative employment, g is the expected growth rate of compensation in that employment over the entrepreneur’s work life, r is an estimate of the cost of capital for alternative compensation, and t is the expected years of work life remaining. Suppose, for example, that a prospective entrepreneur whose expected salary in alternative employment next year is $150,000. If the individual does not pursue the venture, her salary is expected to grow at a rate of 5% per year for her remaining 15-year career. The correct rate depends on the risk of the individual’s employment. For our purposes, it is sufficient to use a low rate if employment is very secure and a higher rate if employment is very risky. For this example, we assume that employment is about as risky as investing in the market and therefore use a 12% rate. The resulting estimated PV of human capital is $1.329 million. PVcomp =
t t C1 1 + g $150, 000 1.05 = − 1 − 1 = $1, 328, 973 r − g 1 + r 0.12 − 0.05 1.12
The value of the entrepreneur’s investment in the venture can be approached in a similar way. For example, if the individual commits one year to the venture and works without compensation, then the entrepreneur loses the first year of income in the preceding calculation. Its PV is an investment of human capital in the venture. The commitment may be partly offset by a salary paid by the venture, which would reduce the effective investment. In addition, forsaking one’s current career for a new venture might reduce prospects for starting salary and subsequent raises upon returning to the workforce. In our example, the commitment to the new venture is for only one year, which is likely shorter than the commitment most entrepreneurs make. Given that total wealth includes both financial and human capital, we define the entrepreneur’s commitment to a venture in terms of the fraction of total wealth committed. A full commitment would mean that the entrepreneur devotes
450
Chapter Eleven Appendix
all financial and human capital to the venture. In practice, no entrepreneur can make a full commitment. We are more concerned with investment decisions involving substantial partial commitments (i.e., neither a full commitment nor the commitment of a well-diversified investor). Scenario 1. Consider the commitment of the “typical” entrepreneur. Numerous studies and surveys based on self-employment data have shown the typical entrepreneur to be a married male in his 40s with a college degree or some college experience. With this picture in mind, we can make some reasonable assumptions about the potential entrepreneur and his level of commitment. Assume he is 45 years old with a current salary (including benefits) of $150,000 and net worth of $2.0 million, including $800,000 of liquid assets and $1.2 million in retirement savings. He is planning to retire at age 60, which means that he has 15 years left to work. He is considering resigning to pursue a new venture that would require a five-year commitment at two thirds of his current salary ($100,000) plus a financial investment of $400,000. To measure the entrepreneur’s commitment, we need to estimate percentage of his total wealth that will be invested in the venture. This investment is composed of two parts: the direct financial investment of $400,000 plus the opportunity cost of accepting a reduced salary for 5 years. Earlier we estimated the human capital wealth of this individual if he continues in his current job to be $1.329 million. Adding this to his financial net worth of $2.0 million yields estimated total current wealth of $3.3 million. Undertaking the new venture means a $100,000 salary for the next 5 years. We will assume that after 5 years, if the venture fails or if he exits, he can return to employment in his prior field but at a starting salary equal to his current pay of $150,000. Further, after he reenters the workforce, his salary will only grow at 4% annually for the remaining 10 years. Thus, if he decides to pursue the venture and it fails, the PV of his future earnings is:
PVcomp
$150, 000 (1 + 0.04)10 × 1 − 10 5 1 + 0.0) 0.12 − 0.04 (1 + 0.012) ( $100, 000 = 1 − + 0.12 − 0.0 (1 + 0.12)5 (1 + 0.12)5
PVcomp = $917, 337 The first term represents the PV of the five-year, $100,000 new venture salary, and the second is the PV of $150,000 starting in Year 6 and then growing 4% annually for 10 years discounted to today. The difference of $411,636 ($1,328,973 − $917,337) represents the implicit investment in the form of reduced compensation if the venture is undertaken. We add to this the $400,000 cash investment,
The Entrepreneur’s Perspective on Value 451
making the potential entrepreneur’s total investment $811,636, or approximately 24% of his total wealth of $3.3 million. With more information about the risk and return of the venture, we could estimate the entrepreneur’s required return based on his 24% commitment. Scenario 2. An entrepreneur has just completed college and has an opportunity to begin working in a corporate position with a starting salary (including benefits) of $50,000. Her salary is expected to grow at a rate of 5% per year on average, and her expected work life is 40 years. The rate of future salary growth is subject to the same degree of risk as in the previous example. The entrepreneur has no savings or other financial assets of consequence. The venture would require a five-year commitment during which her expected salary would be only $25,000 per year. Should the venture fail, the entrepreneur could take on other employment, with a starting salary of $50,000 in Year 6, with the same average annual growth of 5% for the remaining 35 years of her work life. The entrepreneur’s commitment is measured against the opportunity cost of accepting employment immediately. Using Eq. (11A.1), we can find the PV of her employment opportunity, using $50,000 as the Time 1 salary, 12% as the required rate of return, 5% as the growth rate, and 40 years as the horizon: 40 $50, 000 1 + 0.05 1 PVcomp = − = $660, 245 0.12 − 0.05 1 + 0.12 This is also her total wealth, since all she has is human capital. Should the entrepreneur decide to undertake the venture, the calculation becomes
PVcomp =
5 $25, 000 1 1 − + 0.12 1.12
35 $50, 000 1.05 1 × − 0.12 − 0.05 1.12
(1.12)5
PVcomp = $453, 082 The PV of the entrepreneur’s current commitment to the venture is the difference of $207,163 ($660,245 − $453,082), or 31% of the total PV of the entrepreneur’s human capital. The assumptions in this example are very different from those in Scenario 1, where the salary and tangible wealth were higher, the work life was shorter, and a significant cash investment was required of the entrepreneur. Yet the commitments in both cases yield similar percentages of investment. Generalizations. The previous scenarios represent two very different situations but yield similar results in terms of the fraction of present-valued wealth invested by the entrepreneur. Given the specifics, they also represent reasonable
452
Chapter Eleven Appendix
upper bounds on how high the entrepreneur’s fractional commitment of wealth is likely to be, and they shed light on the factors that affect the required rate of return. Generally, the fraction of risk capital required decreases as time to retirement increases; it decreases as the shortfall in compensation for involvement with the venture decreases compared to compensation in alternative employment; it increases with increases in the duration of commitment; and it decreases with the entrepreneur’s long-run ability to maintain his or her level of earnings in alternative employment. The fraction of risk capital committed increases with commitment of financial capital to the venture and decreases as more financial capital is allocated to other investments.
A Shortcut for Estimating Wealth and Investment Because the fraction of wealth committed to the venture is important to the entrepreneur’s valuation and because calculating the human capital commitment can be tedious, we provide a template, Table 11A.1, to streamline the calculation. In the table, “Valuation Template 3,” we assume that an entrepreneur with a 40-year remaining work life is considering giving up current employment that pays a salary of $100,000 per year to commit 3 years to a venture that will pay only $40,000 each year. His salary, if he forgoes the venture, is expected to grow at a rate of 5% per year, but the position is moderately risky, so that expected future earnings in that employment are discounted at 12% per year (similar to the prevailing market rate). If the entrepreneur undertakes the venture and it fails, the entrepreneur anticipates returning to his alternative employment but at a starting salary of $90,000, 10% lower than his current salary. He expects to get 5% annual increases after returning to the workforce. The entrepreneur has also estimated the PVs of his other assets to be $2.5 million and liabilities to be $1.0 million. His estimate of the value of assets other than human capital that he will commit to the venture if it is undertaken is $200,000. Based on the inputs, the table shows the entrepreneur’s human capital to be worth $1.32 million and his total current wealth to be $2.82 million. Undertaking the venture would involve a human capital investment of about $393,000. Combining this with his financial investment of $200,000, the entrepreneur would be committing about $593,000 to the venture, or about 21% of his total wealth. It is noteworthy that the estimate is not sensitive to the discount rate assumption. Rates from 6% to 20% all yield about the same estimate of the entrepreneur’s percentage commitment.
The Entrepreneur’s Perspective on Value 453 Tab le 11A .1 Valuation Template 3: Present value of the entrepreneur’s wealth and commitment
Present Value of Total Wealth
Inputs
Value
40
Years until expected retirement
$100,000
Current annual compensation
5.00%
Expected growth rate of compensation
12.00%
Cost of capital for compensation (based on risk) Present Value of Earnings in Alternative Employment
$1,320,490
Other assets (savings, retirement savings, house, etc.)
$2,500,000
Other liabilities (mortgage, etc.)
$1,000,000
Net Other Assets
$1,500,000
Total Wealth
$2,820,490
Commitment to the Venture Years committed to the venture Starting annual compensation in the venture Growth in annual compensation in the venture
3 $40,000 0% $96,073
Present value of annual compensation while in the venture Starting compensation if the venture is abandoned Growth in compensation if the venture is abandoned
$90,000 5%
Present value of compensation if the venture is abandoned
$831,117
Present Value of Human Capital if the Venture Is Pursued and Abandoned
$927,191
Present Value of Human Capital Investment in the Venture
$393,299
Present value of other assets invested in the venture Entrepreneur’s Total Investment in the Venture
$200,000 $593,299
Total Investment as a Percentage of Total Wealth note:
21.0%
Shaded cells are inputs. See the website version of this table for guidance on using this template.
11A.2 Valuing Partial-Commitment Investments The entrepreneur’s valuation problem is significantly more difficult than the investor’s. However, we have already done much of the spadework in Chapters 10 and 11 to show you how to address it. Here we take up the remaining difficulties and demonstrate how to use the RADR and CEQ approaches to value partial-commitment investments by entrepreneurs. Every DCF approach requires an estimate of expected future cash flows. The CEQ approach also requires an estimate of the standard deviation of cash flows. The discrete scenario First Chicago Method and simulation approaches can be used to estimate both the expected cash flow and cash flow uncertainty.
454
Chapter Eleven Appendix
Using the RADR Approach to Estimate Value To simplify the entrepreneur’s valuation problem, we assume that a partially diversified entrepreneur can allocate wealth between only two investment opportunities: the new venture and a market index. Of course, the restriction to two risky assets does not square with reality, but the additional complexity of treating human capital as a third risky asset does not qualitatively affect any of our conclusions and is unlikely to have an important effect on estimated value. Hence, we limit the presentation to the two-asset assumption. To summarize, the required return for a partial commitment can be estimated by the RADR approach as follows: Step 1. Using the percentage of total wealth allocated to the venture, estimate the standard deviation of holding-period returns of the entrepreneur’s total portfolio of risky assets (i.e., the venture and the market portfolio) (Eq. 10.9). Step 2. Use the CAPM to estimate the required return on the portfolio (Eq. 10.10). Step 3. Set the portfolio required return equal to the weighted average of the required returns on the market and the venture (Eq. 10.11) and solve for the venture required return. The RADR approach works well as long as it is possible to specify ex ante assumptions about correlation and the equilibrium standard deviation of holdingperiod returns for a well-diversified investor. The standard finance approach for ex ante specification is to rely on public market data from comparable firms for this type of information.
Using the CEQ Approach to Estimate Value For new ventures, there often is no good public market analogy. In such cases, the CEQ approach is easier and likely to yield more reliable results. Valuation of the venture by the CEQ approach involves two steps: Step 1. Use the CEQ approach to value the entrepreneur’s portfolio. Step 2. Find the value of the venture by deducting the investment in the market portfolio from total portfolio value. The first step is accomplished by using Eq. (11A.2), the CEQ form of the DCF valuation model, applied to valuing the entrepreneur’s portfolio of risky investments.
The Entrepreneur’s Perspective on Value 455
PV port =
C port −
σC port
σM 1 + rF
RPM (11A.2)
In Eq. (11A.2), for convenience, we drop the time subscripts. The expression requires an estimate of the expected harvest-date cash return of the portfolio, Cport, and the standard deviation of the portfolio cash flow, σCport. Given that you already have developed an estimate of the expected harvest-date cash flow and cash flow standard deviation of the venture, Eq. (11A.3) gives the expected harvest-date cash flow of the portfolio, and Eq. (11A.4) gives the standard deviation of the portfolio’s cash flow.
C port = CP + wM (1 + rM ) (11A.3)
σC port = σC2 P + ( σ M wM ) + 2ρP , M σCP ( σ M wM ) (11A.4) 2
The expected cash return of the entrepreneur’s portfolio is the sum of the expected cash return from investing in the venture and the expected cash return from the investment in the market. In Eq. (11A.3), we compute the expected cash return for investing in the market index as 1 plus the expected rate of return, rM, times the dollar value of the investment, wM. The dollar investment in the market is the PV of the entrepreneur’s total wealth (including human capital) minus the PV of financial wealth and human capital investment in the venture. The standard deviation of the entrepreneur’s portfolio cash flows is computed in Eq. (11A.4) but based on dollar-valued investments instead of percentages of invested wealth. For the second step, we use the CEQ value of the portfolio from Eq. (11A.2). In Eq. (11A.5), project value is found by deducting the investment in the market.
PVP = PV port − wM (11A.5)
11A.3 Implementation: Partial Commitment Forecasting Expected Cash Flow and Risk To focus on the essential issues of valuation, Table 11A.2 portrays a simple venture in which the entrepreneur makes an initial investment and receives an uncertain cash return in Year 6, when the investment is assumed to be sold. The spreadsheet is constructed so that the Year 6 cash flow includes continuing value as well as any accumulated free cash from the first six years of operation.
456
Chapter Eleven Appendix
The table sets out the simulation model of pro forma cash flows of the venture. The values actually shown are from a single random iteration of a simulation model. The relatedness of cash flows over time is built into the model by making potential sales in each year depend on prior-year potential sales, plus a random growth component. Other assumptions are detailed in the table. We further simplify by assuming that operating cash flow and profit are equivalent. In the table, the entrepreneur makes an initial investment of $2 million. Free cash is a balance that is computed each year as the balance from the previous year, plus interest computed at 4% (the risk-free rate), plus profit, and minus the year-to-year increase in cash needed. We simplify the valuation problem by assuming that cash is not distributed until Year 6. The six-year investment horizon is assumed to be long enough to enable the entrepreneur to harvest by selling or liquidating the venture. We estimate value to the entrepreneur by assuming a hypothetical sale at the end of the explicit value period. At that point, the entrepreneur withdraws any free cash and sells the venture for an estimate of the Year 6 value based on its ability to generate cash flows after Year 6. The assumption that cash cannot be distributed early approximates reality, as early distributions by new ventures are usually limited, and many ventures do not generate much free cash flow during the early stages. We assume that the entrepreneur sells the venture for a multiple of Year 6 profit; the specific multiple reflects the level of sales in Year 6 and the growth rate in the prior year. Thus, we have built into the multiplier the realistic assumption that ventures with high expected growth rates sell for higher multiples. The potential for failure and early liquidation is also incorporated in Table 11A.2. This is done by assuming that if cash is not sufficient to meet interim financing needs, the residual cash balance is invested at the riskless rate until Year 6. Retaining the cash until Year 6 even in the event of failure is a convenient way to treat the venture as a one-period valuation problem (cash in at Time 0 and cash out at Year 6). By simulating the model, we were able to estimate that, for the $2 million investment at Time 0, the expected cash flow in Year 6 is $10,743, with a standard deviation of $5,527.
Valuing the Venture as a Partial Commitment Suppose that undertaking the venture described in Table 11A.2 does not require the entrepreneur’s entire wealth. Specifically, to undertake the venture the entrepreneur would have to invest $2 million of financial capital from $6 million of total wealth. Doing so would leave $4 million of financial and human capital to invest in a market index and result in 33% of her total wealth
The Entrepreneur’s Perspective on Value 457 Tab le 11A . 2 New venture simulation model ($000)
Forecast of Market Demand
Year
0
1
2
3
4
5
6
Potential sales
–
928.9
984.9
1,486.6
1,838.5
2,706.9
3,833.5
Forecast of Revenue, Profit, Free Cash Flow, and Cash Flow to Investor/Entrepreneur at Harvest 1
Sales Profit Cash needed Initial investment Free cash Out of cash test Invest Ending value Ending cash Total to Investor/Entrepreneur
$0.0 $100.0 $2,000.0 $1,900.0
2
3
4
5
6
$928.9 –$74.9 $464.4
$984.9 –$55.3 $492.5
$1,486.6 $120.3 $743.3
$1,838.5 $243.5 $919.2
$2,706.9 $547.4 $1,353.5
$3,833.5 $941.7 $1,916.8
$1,536.7 0
$1,514.8 0
$1,444.9 0
$1,570.2 0
$1,746.2 0
$2,194.5 0
$2,000.0 $8,600.8 $2,194.5 $10,795.3
Assumptions Potential sales is a simulated series including a random starting value in year 1 and a random growth rate to year 2. Profit is 35% of sales, less a fixed cost of $400. Cash needed is 50% of sales. Free cash is prior year free cash, plus interest at 4%, plus prior cash needed, plus profit minus current cash needed. If free cash is negative, the venture is terminated the next year. Ending value is a multiple of year 6 profit. The multiple is an increasing function of sales growth in the last 2 years. Ending cash is year 6 free cash and is assumed to be distributed to the entrepreneur. Total to investor/entrepreneur is ending value plus ending cash.
in the venture, and the remainder is invested in a market portfolio. To estimate the value of the venture as a partial commitment, we use the approach described in the previous section. The analysis is summarized in Table 11A.3, “Valuation Template 4,” which can be modified to evaluate other choices of input variables. The “Market Information” panel contains details on market returns and risk. The Year 1 risk-free rate is 4%. Rates for longer holding periods are derived by compounding, so that the cumulative six-year risk-free rate is 26.53%. The annual market rate of 12.0% is compounded in the same manner. The market risk premium for the holding period is computed as the difference between the market rate and the risk-free rate. Market variance and market standard deviation are computed from the assumption that the market has a 15% standard deviation of annual returns and that market returns are uncorrelated over time. We also assume that the correlation between the venture and the market is 0.2.
458
Chapter Eleven Appendix Tab le 11A . 3 Valuation template 4: Partial commitment of the entrepreneur Annual
Market Information
Holding Period
4.00%
26.53%
12.00%
97.38%
Market risk premium
8.00%
70.85%
Market variance
2.25%
13.50%
15.00%
36.74%
Invest Date
Harvest Date
Risk-free rate Market rate
Market standard deviation
0.2
Correlation Cash Flows (000s)
6
Years until harvest $2,000
Investment in project
$10,743
Expected project cash flow
$5,527
Project standard deviation of cash flows Investment in market
$4,000
Percentage of entrepreneur’s total wealth invested in project
33.3%
Expected market cash flow
$7,895
Market standard deviation of cash flows Expected portfolio cash flow
$1,470 $6,000
Portfolio standard deviation
$18,638 $5,996
CAPM Market Value Estimate (Diversified Investor) Portfolio present value
$10,806
Market investment
$4,000
Present Value—Project
$6,806
Net Present Value—Project Project required return for diversified investor
$4,806 7.90%
Project equilibrium standard deviation of returns
57.85% 81.21%
Project implied beta
0.44
CAPM Private Value Estimate—Partial Commitment Entrepreneur (000s) Present value—portfolio
$5,592
Market investment
$4,000
Present value—project
$1,592
Net present value—project Project required return for underdiversified investor
($408) 37.47%
Equilibrium standard deviation of returns note:
574.90% 347.22%
Shaded cells are inputs. See the website version of this table for guidance on using this template.
The “Cash flows” panel of Table 11A.3 includes not only the venture-specific information but also the expected cash flow and standard deviation of cash flows for investing in the market and for the portfolio that combines the two investments. The expected market cash flow is computed by applying the market rate of return to the $4 million investment in the market. The standard deviation of cash flows for the investment in the market is computed by multiplying the
The Entrepreneur’s Perspective on Value 459
market standard deviation in percentage terms by the $4 million investment in the market. The project expected cash flow and standard deviation are determined by simulating the model in Table 11A.2. The expected portfolio cash flow and portfolio standard deviation values are computed using Eqs. (11A.3) and (11A.4). In the “CAPM Market Value Estimate” panel, we show the PV of the project cash flows as $6.806 million, which yields an NPV of $4.806 million after subtracting the $2 million investment. For a diversified investor in the venture, because of a low beta, the annual required return is 7.9%. In the “CAPM Private Value Estimate—Partial-Commitment Entrepreneur” panel, we value the entrepreneur’s investment in the new venture using the CEQ form of the CAPM-based model, Eq. (11A.2). Under that approach, the entrepreneur’s total portfolio is worth $5.592 million. To find the value of the venture, we subtract the $4 million value of the market index. The resulting estimate of venture PV is $1.592 million, or an NPV of negative $0.408 million based on the entrepreneur’s $2 million investment. With this level of partial diversification, the entrepreneur’s cost of capital is 37.5%, and the project is worth much less to the entrepreneur than to a well-diversified investor. Calculations aside, the entrepreneur’s required return declines as the fraction of risk capital invested in the venture declines. If the entrepreneur were to invest a trivial fraction of total wealth in the venture, the required return would drop to 7.9%, the same as for a welldiversified investor. Using the spreadsheet in Table 11.3, with some trial and error you could determine that the venture would have a zero NPV at a 32.3% discount rate, which corresponds to 27.2% of wealth being invested, somewhat less than the current level. If the cost of the venture remains $2 million, then the entrepreneur needs to be a bit wealthier and to have $5.355 million to invest in the market index to yield a zero NPV for the venture.
11A.4 Shortcuts and Extensions Shortcuts to Valuing Partial-Commitment Investments Valuation Template 4 (Table 11A.3) requires only a few inputs: 1. The annualized risk-free rate. It makes sense for an investor in a U.S. venture to choose a currently available U.S. Treasury borrowing rate for a term that is approximately the same as the anticipated holding period. 2. The annualized expected return on the market portfolio. In Chapter 11, we saw that the expected market return and market risk premium are
460
Chapter Eleven Appendix
ifficult to estimate with precision. Values in the 5–8% range for the mard ket risk premium can all be justified. Our calculations in this appendix are based on 8%. 3. The annualized standard deviation of market returns. The long-term historical average return on the S&P 500 provides a reasonable estimate of expected forward-looking volatility. We use an estimate of 15% in our calculations. 4. The estimated correlation between the venture and the market. While it might be best to generate this estimate from a financial model, in Chapter 11 we present estimates of correlation generated by Kerins, Smith, and Smith (2004). The overall average correlation from that study is 0.195; the table also shows estimates by industry sector, stage of development of the venture, and venture size. 5. The expected holding period for the venture. For a venture that does not have staged investments, this is normally the time until the entrepreneur believes exit would be possible, meaning that the venture will have established a track record that is sufficient to enable a third party to make a reasonable assessment of value. 6. The entrepreneur’s investment in the venture. This includes the financial investment and the PV of time invested, based on the opportunity cost of forgone employment. 7. Expected cash return. This is an estimate of the expected cash flow to the entrepreneur when the venture is harvested. 8. Standard deviation of harvest cash return. This is an estimate of the standard deviation of the cash return. The estimate can be derived from scenario analysis, from simulation, or by other means.
Using Data for Public Firms to Estimate the Entrepreneur’s Cost of Capital It is possible to use market data to infer Cport and σCport, the two parameters we need for the CEQ approach in Eq. (11A.2), to determine the PV of an underdiversified entrepreneur’s portfolio. We can then solve for the entrepreneur’s cost of capital. Suppose an entrepreneur who has total wealth of $1.0 million is considering investing $150,000 of financial and human capital in a venture. The investment is expected to return $350,000 in 2 years. The entrepreneur would like to know if the return is high enough to justify the investment. One difficulty with trying to use public data is that not much data are available for early-stage ventures. Using post-IPO data for early-stage high-
The Entrepreneur’s Perspective on Value 461
technology ventures, however, Kerins, Smith, and Smith (2004) report an average beta of 0.99, which, combined with the average correlation of 0.195, suggests average that total risk is a little more than 5 times as high as market risk (because correlation is about one fifth of market risk). Their evidence can be used to estimate cost of capital for entrepreneurial investments in a representative early-stage venture. First, we assume (counterfactually) that the entrepreneur uses the required rate for a well-diversified investor to find the zero-NPV cash flow for a portfolio that is invested partly in the market and partly in the venture. We use the beta of the venture to compute the zero-NPV portfolio cash return for any given arbitrary allocation of wealth between the market and the venture. Because the entrepreneur and the diversified investor value the venture differently, they will not have the same value weights. To find the entrepreneur’s required rate of return, we will need to experiment with these weights later on, but for now, we can start with the entrepreneur’s weights of 15% in the venture and 85% in the market. For an investor who values the opportunity using the required returns for diversified investors, we can use Eq. (11A.6) to compute the expected cash flow: Cport = wM(1 + rM) + wIP[1 + rF + βP(rM − rF)] (11A.6) where wIP is based on the diversified investor’s valuation. For convenience, we assume that Eq. (11A.6) is standardized so that wM and wIP sum to $1. The result is that Cport is the required cash return per dollar invested in the portfolio by a diversified investor. With a risk-free rate of 4% per year, a market return of 12% per year, a beta of 0.99, and 15% of wealth allocated to the venture, the required portfolio cash flow per dollar invested is calculated as follows: Cport = wM(1 + rM) + wIP[1 + rF + βP(rM − rF)] Cport = $0.85(1 + .254) + $0.15[1 + .082 + .99(.254 − .082)] = $1.254 The market and risk-free rates have been compounded to reflect the two-year investment. We then can use the same weights to compute the equilibrium portfolio cash flow risk, σCport, for an investor who uses only market risk to value the project:
σC port =
( wM σ M )2 + ( wPj σ P )
2
+ 2ρP , M wM wPj σ M σ P (11A.7)
With the already stated assumptions, annualized market risk of 15.0%, and a correlation of 0.195, the portfolio cash flow standard deviation for an investor who considers only market risk is $0.265 per dollar invested in the portfolio.
462
Chapter Eleven Appendix
Given the portfolio expected cash flow and standard deviation, we can use Eq. (11A.2) to solve for the value of the portfolio to an entrepreneur who takes account of underdiversification. Also, because we know that in the framework we are using, the investment in the market is valued the same by the entrepreneur as by the investor, we can determine the value of the project to the entrepreneur by subtracting wM from the entrepreneur’s value of the portfolio. Because the entrepreneur cares about diversification, her valuation will be less than that of an investor who is diversified. In the example, the entrepreneur’s valuation of the portfolio turns out to be $0.96 (relative to the $1 value determined by the investor), which means that the entrepreneur’s realized weight of investment in the venture would be 11.5% [i.e., ($0.96 − $0.85)/$0.96]. Recall that we are searching for the required rate of return for an entrepreneur who invests 15% of wealth in the venture. Therefore, the 15% weight we have been using for an investor who ignores nonmarket risk is too low. We need to find an allocation to that investor that would result in our target weights for the underdiversified entrepreneur. To find this, we need to adjust the weights for the investor who only cares about market risk until we find the point where they produce a valuation where the entrepreneur would have 15% in the venture. This is the point where the entrepreneur realizes a zero NPV on the investment. Once we find this zero-NPV point, we can use Eq. (11A.8) to solve for the entrepreneur’s required rate of return on the project.
rPE =
CP −1 (11A.8) PVPE
In this example, increasing the weight in the project for the investor who ignores underdiversification to 21.1% results in a value of the entrepreneur’s portfolio ($0.928) that yields a 15% weight for the entrepreneur [i.e., ($0.928 − $0.789)/$0.928]. Using Eq. (11A.8), we find that the entrepreneur’s required return is 38.0% per year. Discounting the expected $350,000 payoff by 38.0% per year yields a present value of $183,785, or a positive NPV of $33,785.
Shortcuts for Estimating the Entrepreneur’s Cost of Capital from Market Data The process for valuing the entrepreneur’s investment, as described above, is complicated. To simplify the calculations, Table 11A.4, “Valuation Template 5,” can be used to find the entrepreneur’s cost of capital by using risk and expected return data for a comparable public firm. Because it is based on the CEQ approach, required-rate-of-return estimates derived from the template correctly reflect the effect of underdiversification and yield a cost of capital that can be used to value project cash flows.
The Entrepreneur’s Perspective on Value 463
In the table, we illustrate the solution discussed earlier. The highlighted cells are assumptions you will need to supply. One of them, the weight invested in the market by the diversified investor, is labeled “Adjustment” because it is the cell that you would need to adjust until the cell at the bottom of the figure labeled “Target” reaches the entrepreneur’s planned weight of investment in the venture. Other than the assumptions, most of the information in the table is diagnostic. The important result is the entrepreneur’s cost of capital. We used the Goal Seek feature of Excel to find the value of the diversified investor’s investment in the
Tab le 11A . 4 Valuation template 5: Entrepreneur’s cost of capital
Per Year
Partial Commitment
Inputs Years until expected harvest Hypothetical fraction of wealth invested in market portfolio by diversified investor Hypothetical fraction of wealth invested in venture by diversified investor Market Data Risk-free rate of interest Expected return on market Standard deviation of market returns Comparable Public Firm Data Correlation of comparable public firm with market Beta of comparable public firm Standard deviation of comparable public firm returns
(Adjustment)
2.0 78.8% 21.2%
4.0% 12.0% 15.0%
8.2% 25.4% 21.2%
0.195 0.990
0.195 0.990 107.7%
Portfolio Cash Flow Results Expected harvest cash flow of entrepreneur’s portfolio Cash flow standard deviation of portfolio Value of entrepreneur’s portfolio
$1.254 $0.308 $0.928
Venture Valuation Results Value of entrepreneur’s investment in venture Expected cash flow from investment in venture
$0.139 $0.265
Venture Cost of Capital Estimates 90.4% 38.0% 11.9% 26.1% 163.7%
Entrepreneur’s holding-period cost of capital for venture Entrepreneur’s annualized cost of capital for venture Diversified investor’s annualized cost of capital for venture Entrepreneur’s annualized underdiversification premium Standard deviation of entrepreneurs return from venture Entrepreneur’s Wealth Allocation Fraction of entrepreneur’ s wealth invested in venture Fraction of entrepreneur’s wealth invested in market Estimation is based on data from comparable public firms.
(Target)
15.0% 85.0%
464
Chapter Eleven Appendix
market that would make the entrepreneur’s fraction of present-valued wealth in the venture equal to 15%. Another way to think of this approach to estimating the entrepreneur’s required rate of return is to focus on the underdiversification premium to the standard CAPM: rEP = rIP + UPEP = rF + βP(rM − rF) + UPEP (11A.9) where UPEP is the underdiversification premium that is specific to the risk properties of the venture and the fraction of wealth invested by the entrepreneur. In the example in Table 11A.4, a well-diversified investor would have an annualized cost of capital for this venture of 11.9%; the entrepreneur, who requires 38.0%, therefore expects an annualized underdiversification premium of 26.1%. Recall that we introduced the idea of an underdiversification premium early in the chapter. The approach we illustrated at that point produced negatively biased estimates of the entrepreneur’s required rate of return. This example is a bit more complex but yields estimates that are unbiased in the CAPM context. Another shortcut approach is simply to look up an estimate of the underdiversification premium in a table. Table 11A.5 shows various estimates of the entrepreneur’s underdiversification premium as a function of the percentage of the diversified investor’s wealth in the venture and the holding period. As in Table 11A.4, calculations are based on the CEQ approach, with an assumed beta of 1.0, a correlation of 0.20, a risk-free rate of 4% per year, an annual market return rate of 12%, and standard deviation of the market rate of 15% per year. Reasonable changes in assumptions about the risk-free rate and market risk premium have little effect on the estimated underdiversification premium. The assumed correlation and standard deviation are comparable to those presented in Chapter 11. Table 11A.5 shows that the underdiversification premium is higher the larger the fraction of the entrepreneur’s wealth that is invested and the longer the holding period. This makes sense because investing a large fraction of wealth for a very short period does not expose the entrepreneur to much nonmarket risk; neither does investing a small fraction of wealth for a long period. Table 11A.5 provides reasonable estimates of the underdiversification premium for most ventures.
11A.5 Summary In this appendix, we develop a framework for evaluating the investments of entrepreneurs. This framework recognizes that entrepreneurs, as undiversified investors, should value projects at discount rates that can be substantially
The Entrepreneur’s Perspective on Value 465 Tab le 11A . 5 Underdiversification premium to required rate of return
Percentage of Diversified Investor’s Wealth in the Venture
Holding Period
1 Year 2 Years 3 Years 4 Years 5 Years
40.0%
35.0%
30.0%
25.0%
20.0%
15.0%
10.0%
5.0%
0.0%
33.6% 43.2% 65.1% NM NM
30.9% 39.0% 55.4% 159.4% NM
27.9% 34.3% 46.2% 84.5% NM
24.4% 29.3% 37.4% 55.9% NM
20.4% 23.8% 28.9% 38.1% 64.5%
15.8% 17.9% 20.8% 25.1% 32.5%
10.7% 11.8% 13.1% 14.8% 17.1%
5.3% 5.7% 6.1% 6.6% 7.1%
0.0% 0.0% 0.0% 0.0% 0.0%
NM = not meaningful.
higher than the rates appropriate for well-diversified investors. Because the entrepreneur’s required rate depends on total risk and not just market risk, a significant opportunity exists to select value-enhancing strategies for undertaking new ventures and to design value-enhancing contracts between entrepreneurs and outside investors. This appendix explores the benefits that can derive from regarding entrepreneurial investment in this way. Among other factors, the entrepreneur’s required return declines as the fraction of wealth that must be invested in the venture declines. This gives rise to a significant opportunity to create value by selecting the size of the investment or bringing in an outside investor to share in the risk of the venture. Doing so can reduce the overall required rate of return. It is intuitive that underdiversification should motivate the entrepreneur to seek a higher return than a well-diversified investor would require. However, the appropriate magnitude of the return premium that should be required to compensate for underdiversification is not intuitive. In fact, it is common for advisors to suggest arbitrary “haircuts” to value in order to compensate for such things as illiquidity and underdiversification. The approaches presented in this chapter are more systematic and are solidly grounded in financial economic theory. We devote significant attention to presentation and discussion of the analytics to help make clear that (1) the premiums to required rates of return that arise from underdiversification can be substantial, and (2) there is no single, arbitrary haircut that can do a good job of helping you to make good decisions in light of opportunity cost. Beyond the analytics, the appendix provides several shortcut approaches to estimating discount rates that take account of underdiversification.
References and Additional Reading
Chen, H., J. Miao, and N. Wang. 2010. “Entrepreneurial Finance and NonDiversifiable Risk.” Review of Financial Studies 23 (12): 4348–88.
466
Chapter Eleven Appendix
Evans, D., and B. Javanovic. 1989. “An Estimated Model of Entrepreneurial Choice under Liquidity Constraints.” Journal of Political Economy 97: 808–27. Gentry, W. M., and G. Hubbard. 2004. “Entrepreneurship and Household Saving.” Advances in Economic Analysis and Policy 4. http://www.bepress. com/bejeap/advances/vol4/iss1/art8/. Hall, R. E., and S. E. Woodward. 2010. “The Burden of Nondiversifiable Risk of Entrepreneurship.” American Economic Review 100: 1163–94. Heaton, J., and D. Lucas. 2000. “Portfolio Choices and Asset Prices: The Importance of Entrepreneurial Risk.” Journal of Finance 55: 1163–98. ———. 2004. “Capital Structure, Hurdle Rates, and Portfolio Choice: Interactions in an Entrepreneurial Firm.” University of Chicago working paper. http://fisher.osu.edu/fin/dice/seminars/Lucas.pdf. Hurst, E., and A. Lusardi. 2004. “Liquidity Constraints, Household Wealth, and Entrepreneurship.” Journal of Political Economy 112: 319–47. Kartashova, Katya. 2014. “Private Equity Premium Puzzle Revisited.” American Economic Review 104: 3297–334. Kerins, F., J. K. Smith, and R. L. Smith. 2004. “Opportunity Cost of Capital for Venture Capital Investors and Entrepreneurs.” Journal of Financial and Quantitative Analysis 39: 385–405. Moskowitz, T. J., and A. Vissing-Jorgensen. 2002. “The Returns to Entrepreneurial Investment: A Private Equity Premium Puzzle.” American Economic Review 92: 745–78. Wasserman, N. 2006. “Stewards, Agents, and the Founder Discount: Executive Compensation in New Ventures.” Academy of Management Journal 49: 960–76.
C h a p t e r T welve
D e s ig n i n g an d Valu i n g Stag e d I nve stm e nt with R e al O p tio n s
S tag i n g i n v e s t m e n t i n new ventures can enhance value by enabling
the parties to test hypotheses, reduce uncertainty, and generally exploit the structure of real options that comprise the venture. Because the effects are complex, comparing deal structures can be challenging. In this chapter, we use the valuation methodology developed in earlier chapters to examine how staged investment can enhance value. Suppose you are an angel investor and that an entrepreneur has approached you for financing. His venture will need substantial capital before harvesting is feasible. You have several concerns. First, the entrepreneur’s projections seem too optimistic. Second, the entrepreneur may not be an effective manager. Third, competing ventures are under way, and one of them may capture the opportunity sooner. Finally, the entrepreneur will not be motivated unless he can expect to retain significant control and a large fraction of ownership. In this chapter, we focus on designing staged investment structures that deal with these complex yet common issues. Among other considerations, staging has great potential to help mitigate the full range of information and incentive concerns. A well-planned staging strategy can create tremendous value, especially for venture opportunities that succeed, and can help the entrepreneur realize a larger share of the value. As usual, unless noted, we adopt the convention of maximizing value for the entrepreneur.
12.1 Staged Investment: The Venture Capital Method As a first look into the benefits of staging investment, we use the VC Method. Recall that under the VC Method, the cash flows of a success scenario are 467
468
Chapter Twelve
discounted at a hurdle rate that is greater than cost of capital. Because the probability of failure declines as a venture achieves its milestones, as shown in Section 10.2, hurdle rates are lower for investments made at later stages. Consider a venture that needs $700,000 of capital per year for five years, beginning at Time 0. Table 12.1, “Valuation Template 6,” summarizes the essential information about the venture. The entrepreneur’s business plan includes a projection of negative earnings during the first few years, followed by rapid growth. The venture is projected to reinvest any free cash flows during the first five years. By the end of Year 5, the plan shows earnings before interest and after tax (EBIAT) of $2.5 million. The entrepreneur projects a public offering of equity at the end of Year 5. Because (we have determined) the typical earnings multiple of comparable public companies is 15, we project that in the success scenario continuing value as of Year 5 will be $37.5 million. The Investor Valuation and Ownership Allocation panel of the table shows hurdle rates for investments made each year. The rates are consistent with the VC Method rates an investor might use. They begin at 50% for Time 0 investment and decline by 5% per year. A hurdle rate of 50% means that for an investment at Time 0, success-scenario cash flows must be expected to return 50% annuTab le 12 .1 Valuation template 6: Single-stage investment—Venture Capital Method
Year
0
1
2
3
4
5
Income Statement Information
Earnings before interest and after tax
($500,000) ($200,000)
$400,000 $1,400,000 $2,500,000
Cash Flow Information External funds required to support operations
$700,000
$700,000
$0
Equity capital raised
$3,240,927
Beginning cash balance
$3,240,927 $2,642,564 $2,020,267 $1,373,077
$700,000
$0
Uses of cash Cash invested in marketable securities Return on invested cash Ending cash balance
$700,000
$700,000
$700,000
$700,000
$700,000
$700,000
$0
$2,540,927 $1,942,564 $1,320,267
$673,077
$0
$0
$52,811
$26,923
$0
$0
$2,642,564 $2,020,267 $1,373,077
$700,000
$0
$0
35.00%
30.00%
25.00%
$101,637
$700,000
$700,000
$77,703
Investor Valuation and Ownership Allocation Investor hurdle rate Continuing value earnings multiplier
50.00%
45.00%
40.00%
15
Continuing value of venture
$37,500,000
Required future value of investment
$24,610,789
Ownership Share Required
65.63%
Designing and Valuing Staged Investment with Real Options 469
ally; for an investment at Year 3, success-scenario cash flows must be expected to return 35% annually. We use this example to see how, based on the VC Method, staging affects the allocation of ownership. To establish a baseline, we first determine the allocation under single-stage investment.
Single-Stage Investment Suppose that the entrepreneur would like to raise enough cash at Time 0 to cover projected cash needs for the entire five years. In Table 12.1, capital raised at Time 0 but not needed until later is invested at a risk-free rate of 4% until it is needed. On that basis, the entrepreneur seeks $3.241 million of equity at Time 0. The cash balance portion of the “Cash Flow Information” panel demonstrates that this is sufficient to yield the required $700,000 of cash per year. Our objective is to determine the fraction of equity an investor would require in exchange for an investment of $3.241 million at Time 0. Two equivalent approaches are possible. The normal approach from corporate finance is to discount projected continuing value by 50% per year and compare this (postmoney) PV to the amount invested. The easier approach when we use the VC Method to consider staging, however, is to find the Year 5 future value of the Time 0 investment and compare it to projected continuing value at Year 5. We use the second approach in Table 12.1. The $3.241 million investment earning 50% annually must generate a Year 5 cash return of $24.611 million in the success scenario. As shown in Table 12.1, because the continuing value of the entire venture is $37.5 million, the investor would require 65.63% of the equity. If the entrepreneur cares about control and about holding a majority stake of the equity, he will not find this financing arrangement attractive. However, because we do not know the entrepreneur’s cost of capital, we cannot use Table 12.1 to value his financial claim.
Multi-Stage Investment Suppose that instead of investing only at Time 0, the investor proposes to contribute the capital in stages. Table 12.2 illustrates a staging alternative in which capital is invested at Time 0 and in Years 2 and 4. Because some of the capital contributed at Time 0 can be invested at the risk-free rate until it is needed in Year 1, the amount required for the first two years is $1.373 million. A like amount is needed at the end of Year 2, and an investment of $700,000 is required at the end of Year 4. In aggregate, the investor’s three infusions total $3.446 million. However, because the investor can also earn the risk-free rate,
470
Chapter Twelve
Tab le 12 . 2 Valuation template 7: Multistage investment—Venture Capital Method
Year
0
1
2
3
4
5
Income Statement Information
Earnings before interest and after tax
($500,000) ($200,000) $400,000 $1,400,000 $2,500,000
Cash Flow Information External funds required to support operations
$700,000
$700,000
$700,000 $700,000
$700,000
$0
Equity capital raised
$1,373,077
$1,373,077
$700,000
Beginning cash balance
$1,373,077
$700,000 $1,373,077 $700,000
$700,000
$0
Uses of cash
$700,000
$700,000
Cash invested in marketable securities
$673,077
$0
$26,923 $700,000
Return on invested cash Ending cash balance
$700,000 $700,000
$700,000
$0
$673,077
$0
$0
$0
$0
$26,923
$0
$0
$0
$0
$700,000
$0
$0
$0
Investor Valuation and Ownership Allocation Investor hurdle rate
50.00%
45.00%
40.00%
35.00%
30.00%
25.00%
Continuing value earnings multiplier
15
Continuing value of venture Investor’s Required Future Value and Equity Share Third stage Second stage First stage Ownership Required
$37,500,000 Required Beginning Share
Required Ending Share
Value
2.43% 10.30% 31.77%
2.43% 10.05% 27.80%
$910,000 $3,767,723 $10,426,803
40.28%
$15,104,527
the Time 0 equivalent is the same $3.241 million as in the nonstaged scenario (if the venture is successful). To find the investor’s required ownership interest under the staging scenario, we must evaluate each stage separately. As with the decision tree analysis in Chapter 5, we approach the valuation and ownership requirements by reverse induction—beginning with the final stage before proceeding to the earlier one. This is because, except for the first-stage investment, each subsequent investment is evaluated conditional on the venture actually reaching that stage. Thus, working from the last to the first, we begin with the $700,000 third-stage investment in Year 4. According to the hurdle rates in the table, in the third stage the investor seeks a return of 30%. Thus, the Year 5 future value of the third-stage investment that
Designing and Valuing Staged Investment with Real Options 471
is made in Year 4 must be $910,000. Given the total continuing value of $37.5 million, to achieve a zero NPV at Year 4 the investor would require 2.43% of the equity in exchange for the $700,000 third-stage investment. We determine the required percentage of ending ownership for the secondstage investment in a similar manner. The zero-NPV success-scenario return on the $1.373 million investment is 40% per year for three years, resulting in a Year 5 value of $3.768 million. This corresponds to 10.05% of continuing value. By the same reasoning, the required ownership percentage for the first stage is 27.80%. Collectively, if the venture proceeds through all three financing stages, the investor would require a 40.28% interest. The maximum required ownership percentage of the outside investor is considerably lower with staging than without. In fact, with staging, the entrepreneur can retain control and majority ownership.
Why Does Staging Reduce the Required Outside Ownership Share? Comparing Tables 12.1 and 12.2, according to the VC Method, staging reduces the investor’s ownership and apparently benefits the entrepreneur. Staging clearly benefits the investor by reducing the investor’s risk and reducing the initial investment amount. To understand why this happens, you must know why the hurdle rates decline at later stages. How is the investor’s commitment different in Table 12.2 from that in Table 12.1? For a single-stage investment, the commitment is absolute. Even if, during the five years, it becomes clear that the project is destined to fail, the investor’s capital is irrevocably committed. In the event of liquidation, because (we have assumed) the investment is in the form of common stock, the investor would recover only a 65.63% share of liquidation value. The balance would go to the entrepreneur. There is no economic difference between the Table 12.1 scenario and an alternative in which investment is staged as in Table 12.2 except that the commitment to invest is made irrevocably at Time 0. Because the commitment is made at Time 0, the investment must be evaluated as if it were made at Time 0 even though the funds are expended at a rate of $700,000 per year over the five years. Because the commitment is made at Time 0, the 50% hurdle rate would apply to all investments. It follows that the lower hurdle rates in Table 12.2 arise because some of the uncertainty about venture performance is resolved before the decision to invest in later stages is made. The later-stage investments occur only if the venture continues to be attractive. That is, in Table 12.2, the investor has the options to abandon the venture at the ends of Years 2 and 4.
472
Chapter Twelve
Because staging shifts additional risk to the entrepreneur, it is not clear which of the two scenarios is a better deal for him. In Table 12.2, the entrepreneur can expect to retain control if the “success” scenario is realized. However, the entrepreneur also bears more risk that either financing will not be available or the investor will require a larger fraction of equity in a scenario that is not as good. The VC Method provides no clear guidance to the entrepreneur for comparing these alternatives. The entrepreneur cannot choose between the scenarios in Tables 12.1 and 12.2 simply by comparing the fraction of equity retained in the success scenario.
Providing for Dilution When Follow-On Investment Rounds Are Expected Investors who invest during one round often anticipate that the venture will seek additional financing in subsequent rounds. Because additional rounds increase the number of shares outstanding, an investor who bases her decision on continuing value must take future financing rounds into account. Table 12.2 demonstrates how, under the VC Method, anticipated future financing affects the fraction of equity an investor requires at each stage. Beginning with the third stage, because no subsequent rounds are anticipated, the required ownership share is 2.43%. A second-round investor who estimates continuing value of $37.5 million anticipates a third round of financing in exchange for 2.43% of the equity. Before the third round, the combined shares of the entrepreneur and investor represent 97.57% of ending (Year 5) equity. Because they own 100% of the outstanding equity at the time of the investment, a second-round investor who wishes to end up with 10.05% of equity at harvest must seek a larger fraction at the time of investment. Equation (12.1) shows how you can determine the required percentage of equity when future rounds are anticipated. Ending % required % required in current round = (12.1) (1 − Sum of % s required in future round ) Using this equation, in Table 12.2 the investor’s required ownership percentage in the second round is 10.30% 10.05% % required in current round = = 10.30% 1 − 2.43% Similarly, the investor’s required share in the first round is 31.77%: 27.80% % required in current round = = 31.77% 1 − 2.43% − 10.05%
Designing and Valuing Staged Investment with Real Options 473
This method for determining the investor’s required ownership fraction when additional rounds of financing are expected is recursive, the same for other valuation methods discussed later in the chapter.
How the Capitalization of a Venture Relates to the Stage of Financing Normally, investors expect the capitalization of a venture to increase at each round. If it does not, or even if the increase is small, they may become concerned about the venture’s prospects. Table 12.2 demonstrates the reason for such concerns. In the first stage, the investors contribute $1.373 million in exchange for 31.77% of the equity. This implies that they are capitalizing the venture at $4.322 million (i.e., the post-money valuation). $4.322 million = $1.373 million/31.77% Because, under the success scenario, the second-stage investor contributes $1.373 million for 10.30% of outstanding equity, the second-stage capitalization is $13.330 million, a substantial gain over the first stage. Similarly, the third-stage investor is expected to contribute $700,000 for 2.43% of equity, an implied capitalization of $28.807 million. These increases occur because the venture is progressing according to the success scenario. The gain is due partly to reduced uncertainty and partly to increasing proximity of exit. Even if the uncertainty were not reduced, capitalization would increase merely due to proximity of a harvest opportunity. If implied capitalization decreases from one round to the next or does not increase very much, it follows that the venture is not performing according to the success scenario.
Limitations of the Venture Capital Method The VC Method is easy to use but has two major limitations. First, as we have already discussed, it provides no analytical way to determine the appropriate hurdle rates. The rates are based only on rules of thumb and experience. Second, it does little to help an entrepreneur decide whether to proceed with a venture or compare financing alternatives. The first limitation can be addressed by incorporating risk into the analysis and focusing on expected cash flows. The second requires explicit valuation of the entrepreneur’s financial claims in light of necessary underdiversification.
474
Chapter Twelve
12.2 Staged Investment: CAPM Valuation with Discrete Scenarios This section and the next include core material on structuring contracts that include real options. Because the analysis of staged financing quickly becomes involved, we present only a basic example. The entrepreneur proposes to develop a new avionics system add-in that will be attractive to private aircraft owners. The entrepreneur plans to market the system first to small aircraft manufacturers (OEMs), and if that proves to be successful, to market directly to the owners of small aircraft. The entrepreneur expects that the venture can be harvested in 5 years. According to the entrepreneur’s business plan, the venture requires cash infusions from an investor totaling $3.5 million. The entrepreneur needs $2 million now and an additional $1.5 in 2 years. At the current risk-free rate of 4.0%, the total infusions have a PV of $3.387 million. The parties do not expect the venture to generate any free cash flow before harvest.
Single-Stage Venture Capital Method Valuation For simplicity, we assume that the entrepreneur and the VC investor agree about the projections for the venture. The investment and payoff projections in Table 12.3 reflect their belief that if the venture is fully successful, it can genTab le 12 . 3 Investment and payoff assumptions
Conditional Cash Flows
Scenario Probabilities
Round One
Round Two
$14,000,000 $3,750,000 $0
$21,538,462 $5,769,231
Expected Cash Flow at Harvest
$17,750,000
$27,307,692
Standard Deviation of Cash Flows
$15,204,851
$9,730,085
$1,000,000 2
$500,000 3
Harvest Value Forecast Success with both groups Success with first group Failure
$35,000,000 $15,000,000 $0
0.4 0.25 0.35
Investment Burn rate per year Years Total investment Present value of unconditional total investment Present value of expected investment
$3,500,000 $3,386,834 $2,901,442
The table shows the conditional payoffs and the single-stage investment requirements. It also shows round information relevant to the staged investment valuation.
Designing and Valuing Staged Investment with Real Options 475
erate sales from two different customer groups: OEM aircraft avionics suppliers and aircraft owners. The parties estimate that if it is successful with both groups, the venture will have a total harvest value of $35 million in Year 5. The VC Method success scenario assumes that both groups adopt the new product. The entrepreneur’s VC Method analysis is summarized in Panel A of Table 12.4. Believing that the historical average annual return to limited partners of VC funds has been around 18%, the entrepreneur proposes to offer a VC investor 25% of the equity in exchange for the $3.387 million investment. Discounting the projected success-scenario harvest value by 18% for five years yields a value of about $15,299 million. The offer of 25% seems to the entrepreneur to be more than sufficient to compensate the VC for the $3.387 million of initial financing. Although the VC investor agrees with the entrepreneur’s projections, the VC disagrees that a 25% interest is sufficient to justify the investment. Recognizing that the business plan reflects only complete success and considering the early stage of development, the investor believes that her success-scenario projection Tab le 12 . 4 Venture Capital Method single-stage valuation
Panel A
Venture Capital Method (Single-Round Financing) Entrepreneur’s Assumptions
5 $35,000,000 18.00%
Number of years until projected harvest date Harvest-date value of the venture assuming success Annual hurdle rate: Customary discount rates for seed to start-up ventures are 50%–100% per year. Discount rates are lower for more established ventures, declining gradually to rates of 20%–30% for expansion-stage financing. The discount rate for turnaround financing is comparable to that of a start-up.
$15,298,823
Post-money valuation of the venture
$3,386,834 22.14%
Amount of financing required Fraction of harvest-date equity needed to justify the financing
Panel B
Venture Capital Method (Single-Round Financing) Investor’s Assumptions
5 $35,000,000 40.00%
Number of years until projected harvest date Harvest-date value of the venture assuming success Annual hurdle rate: Customary discount rates for seed to start-up ventures are 50%–100% per year. Discount rates are lower for more established ventures, declining gradually to rates of 20%–30% for expansion-stage financing. The discount rate for turnaround financing is comparable to that of a start-up.
$6,507,705
Post-money valuation of the venture
$3,386,834
Amount of financing required
52.04%
Fraction of harvest-date equity needed to justify the financing
Panel A shows valuation under the entrepreneur’s VC discount rate assumption and Panel B shows valuation under the VC’s discount rate assumption.
476
Chapter Twelve
of $35 million should be discounted at a VC Method hurdle rate of 40%. The resulting value is $6.508 million. Consequently, as shown in Panel B of Table 12.4, the investor would require more than 52% of the equity in exchange for committing the $3.386 million.
Staged Investment Given the disparity between their projections, it may seem that the parties should give up on reaching a deal. However, by staging the investment they may be able to reach agreement. Rather than using the VC Method (with its ambiguity about appropriate hurdle rates), we use the CAPM-based valuation approach. Identifying the real options. Suppose the parties recognize that three outcomes reasonably describe the range of plausible scenarios. In addition to complete success, the venture might succeed only in marketing the add-in to aircraft OEMs, or it might fail entirely. Development and marketing of the avionics system can be staged easily. The add-in is unlikely to be successful with aircraft owners unless it can first be sold to OEMs. If marketing to OEMs is successful, the venture can initiate marketing to aircraft owners. Figure 12.1 illustrates the interplay between the investor’s financial options and the real options of the staged venture. The investor’s initial financial commitment (Round 1) enables the entrepreneur to pursue the OEM market (Stage 1). If those efforts are sufficiently successful, the investor will fund the next stage (Round 2), marketing to aircraft owners, but if Stage 1 is not successful, the venture will be abandoned. If Round 2 financing is provided, the harvest value depends on whether the venture is a complete success or succeeds only with OEMs. Of course, Figure 12.1 simplifies the actual decision and range of outcomes. At each stage in the figure, there are only two outcomes, success and failure, and the next round of investment depends only on which of the two occurs. The discrete scenario valuation approach is simplified here by using the knife-edged assumption that failure at Stage 1 means that no subsequent investment occurs and that success at Stage 1 results in the Stage 2 investment being made. The figure is a representation of the parties’ understanding. For example, the investor expects that it will be valuable to invest in Round 2 if the entrepreneur is successful in Stage 1. However, there is no specific definition of success in the figure that the entrepreneur could use as a milestone for triggering the next financing round.
Designing and Valuing Staged Investment with Real Options 477 Fi g u r e 12 .1 Staged investment decision model Success
Success with both groups
Failure
Success with one group
Fund Stage 2: Aircraft owners
Fund Stage 1: Avionics OEM Failure
Success
Abandon
Failure
Round 2
Exit
Financing: Round 1
Specifying key assumptions. The staging structure in Figure 12.1 requires that the parties estimate the expected harvest value associated with each scenario, the probability of each scenario, and the investment assumption for each financing round. In Table 12.3, we had set out the shared assumptions. Based on those assumptions, we computed the expected harvest cash flow and standard deviation of cash flows at each investment round. The expected cash flow and standard deviation by round will become inputs to the valuation. Panel A of Table 12.3 also shows the “burn-rate” assumption for each stage and the expected duration of the stage. CAPM-based valuation with irrevocable commitment to invest. To establish a baseline we can use to assess how staging affects value, it is useful to begin with a valuation where the investor’s entire commitment is made at the beginning. Thus, under the assumptions from Table 12.3, the investor would commit all $3.387 million at Time 0 (the PV of the $3.5 million commitment). The entrepreneur would use the money to achieve as much success as possible. This unconditional valuation is the same as if the venture were not staged and can be calculated using approaches developed earlier. In Table 12.5, we compute the value from the perspective of the VC as a welldiversified investor who also seeks a return on effort. The table combines parts of other valuation templates we have seen in earlier chapters. The Market and Contract Data Assumptions panel (numbered “1”) shows assumptions about the market and the VC’s contract with limited partners. We assume that the VC seeks an annual management fee of 2.5% based on the value of the investment and also seeks a carried interest of 20% of the return to the fund. The SingleStage Investment and Timing panel (numbered “2”) shows the present valued investment and expected time to harvest. The Investor Valuation of Harvest Cash Flows panel (numbered “3”) shows the mean and standard deviation of
478
Chapter Twelve
harvest date cash flows as computed in Table 12.3. In this panel, we use the CEQ approach to estimate the present value of the venture. In this calculation, we implement the VC fund contract provisions in a simplified way, by adjusting the harvest-date cash flow down by 20% (i.e., deducting the carried interest) and then adding the annual fee (as a percentage) to the discount rate that is used to compute PV. Based on the harvest cash flow assumptions, the venture has a PV of $8.279 million to a well-diversified investor, as shown in the third panel. In the Investor’s Required Rate of Return panel (numbered “4”), we use this valuation to impute cost of capital that would be used in the RADR approach. The assumptions yield a cost of capital of 16.48% for the VC investor. Because we imputed the cost of capital based on the gross harvest cash flow, the management fee and carried interest are built into the imputed cost of capital. In other words, a discount rate of 16.48% would discount the expected harvest date cash flow of $17.5 million to a present value of $8.279 million. Accordingly, this is
Tab le 12 . 5 Valuation of unstaged investment: VC investor
VC Limited Partners’ Valuation
1. Market and Contract Data Assumptions Annual risk-free rate Annual market rate Standard deviation of market General partner’s annual fee General partner’s carried interest Correlation of venture with market 2. Single-Stage Investment and Timing Total investment committed Years to harvest 3. Investor Valuation of Harvest Cash Flows Expected harvest value Standard deviation of harvest cash flows Value of Venture at Time of Investment 4. Investor’s Required Rate of Return Annualized VC cost of capital
4.0% 10.0% 15.0% 2.5% 20.0% 0.20 $3,386,834 5.00 $17,750,000 $15,204,851 $8,279,258 16.48%
Investor Ownership Requirement
5. Ownership Requirement Round 1 investment Value at Round 1 Share of Incremental Value Required
$3,386,834 $8,279,258 40.91%
The table shows the valuation for limited partners of the VC investor, taking account of the general partner’s annual management fee and carried interest.
Designing and Valuing Staged Investment with Real Options 479
the relevant cost of capital for the entrepreneur to consider in seeking capital from the VC fund. As shown in the Ownership Requirement panel (numbered “5”), based on the entrepreneur’s valuation, the investor would need 40.91% of the equity in exchange for the $3.387 million commitment. While the entrepreneur would retain the remaining 59.09% of the equity, the value of that ownership fraction could be considerably less, depending on the extent of the entrepreneur’s underdiversification. For example, based on the approach outlined in the Appendix to Chapter 11, if we assume that the entrepreneur has total wealth of $2.0 million and will commit $100,000 per year (present valued) in forgone salary to the venture. For a five-year commitment, this leaves $1.5 million invested in the market. Under these assumptions, based on the entrepreneur’s expected portfolio harvest cash flows and risk, the entrepreneur’s share of the venture is expected to yield $10.489 million, for a total expected harvest-date portfolio value of $12.905 million. In PV terms, the entrepreneur’s portfolio is worth $1.825 million. Subtracting the market investment of $1.5 million means that the entrepreneur’s interest in the venture would be worth only $325,000, a negative NPV of $175,000 on the entrepreneur’s human capital investment of $500,000.1 Thus, to make the venture attractive to the entrepreneur, it would be necessary to find a way to reduce the VC investor’s required ownership and the entrepreneur’s commitment. Staging the investment provides a way to accomplish both.
Valuing the Staged Venture at Each Investment Round When investment is staged, the investor’s claims and those of the entrepreneur can be adjusted at each investment round. The value of staging derives primarily from reductions of uncertainty about the likelihood of subsequent success. The changes in value that occur as the venture evolves affect the compensation the investor must receive for investing. In Table 12.6, we compute values that correspond to each round. The values reflect the assumptions in Table 12.3. While actual results will certainly be different, the valuations in Table 12.6 are the best estimates the parties can make at the time of the initial investment decision. The first two panels summarize previously discussed assumptions from Tables 12.3 and 12.5. Valuations at each round are computed in the third panel in a recursive fashion, beginning with the Round 2 valuation and then turning to Round 1. These valuations are based on the parties’ beliefs as to the expected harvest cash flow and the standard deviation of harvest cash flows at each round from Table 12.3. As in Table 12.5, the values are for a well-diversified limited
480
Chapter Twelve Tab le 12 . 6 Valuation of staged investment: VC investor
Valuation of staged investment: VC investor
1. Market and Contract Data Assumptions Annual Risk-free Rate Annual Market Rate Standard Deviation of Market General Partner’s Annual Fee General Partner’s Carried Interest Correlation of Venture with Market Time from first investment to harvest (years) 2. Staged Investment Timing and Burn Rates Round 1 - Aircraft OEM Burn Rate per Year Years to Complete Round 2 - Aircraft Owners Probability of Stage 1 Success Burn Rate per Year Years to Complete
4.0% 10.0% 15.0% 2.5% 20.0% 0.20 5.00
$1,000,000 2.00 65% $500,000 3.00
3. Valuation of Harvest Cash Flows Round 2 Expected Harvest Value Standard Deviation of Harvest Cash Flows Value to Investor at Time of Investment
$27,307,692 $9,730,085 $17,062,756
Round 1 Expected Harvest Value Standard Deviation of Harvest Cash Flows Value of Round 1 at Time of Investment
$17,750,000 $15,204,851 $8,279,258
4. Investor’s Implicit Cost of Capital at Each Round Round 1 Round 2
16.48% 16.97%
partner investor in a VC fund, after providing for the VC general partner’s management fee and carried interest. The Round 1 valuation of $8.279 million in Panel 3 (and implicit cost-of-capital estimate, in Panel 4) are the same as for the unstaged valuation in Table 12.5. Comparing the Round 1 and Round 2 results, you can see the benefit of staging. Based on the entrepreneur’s beliefs, the Round 2 value (shown in Table 12.7) is $17.063 million, a 106% increase over the Round 1 value. Most of the increase arises because of the option to make the Round 2 investment only if Stage 1 is successful. Thus, staging does two things: it reduces the expected investment and it results in additional (Round 2) investment only when prior success warrants
Designing and Valuing Staged Investment with Real Options 481
it. In this example, the implicit cost of capital in the second round is slightly higher than the cost of capital in the first round. Thus, conditional on investing in Round 2, although the likelihood of ultimate success increases, the risk measured statistically does not go down.
Determining the Investor’s Required Ownership Share at Each Stage The valuation results from Table 12.6 can be used to determine the minimum equity the investor would need at each stage. We focus on minimums to maintain consistency with our general approach of assigning positive NPV to the entrepreneur, who is the residual claimant. Thus, in Table 12.7, we compute the zero-NPV (after fees and carried interest) ownership percentages of the VC limited partner investor.2 We use backward induction to establish the minimum equity allocations that will trigger exercise of the financing option. Beginning with Round 2, assuming the venture has been successful in the first stage, the venture will require $1.5 million to fund Stage 2. Because the venture is worth $17.063 million at the time of the Round 2 decision, in exchange for investing $1.5 million the VC investor would need to realize 8.79% of the expected harvest value at that time to achieve an NPV of zero. As we saw with the VC Method example earlier, the Round 2 investment reduces the percentage ownership held by prior investors and the entrepreneur. In the current example, to maintain a zero NPV for the investor at each round, we employ a structure where the early-round investor’s claims are not diluted by later-round financings. The “Investor’s Required Ownership After I nvesting”
Tab le 12 .7 Valuation-based contracting model of VC ownership shares
Venture Capital Ownership Requirement
Investor’s Round 2 investment Venture value at Round 2 Incremental Equity Required Investor’s value from Round 1 Investor’s value required with investing Investor’s Required Ownership after Investing
$1,500,000 $17,062,756 8.79% $4,121,808 $5,621,808 32.95%
Investor’s Round 1 investment Venture value at Round 1 Investor’s Required Ownership after Investing
$2,000,000 $8,279,258 24.16%
The VC limited partners’ required ownership percentages are computed by backward induction to achieve zero-NPV returns first on the conditional Stage 2 investment and second on the unconditional investment in Stage 1.
482
Chapter Twelve
percentages in Table 12.7 are the percentages that result in no dilution of the investor’s existing equity claims. Thus, assuming that the venture is successful in Stage 1, the parties expect the venture to be worth $17.063 million at the time of the Round 2 financing. At this point, the investor’s 24.16% of harvest value from Round 1 is worth $4.122 million. This would represent a positive NPV for the investor. However, the NPV is offset by what the entrepreneur believes is a 35% probability that the venture will fail in Stage 1 and be worth zero. The table provides a way to find the investor’s required zero-NPV ownership stake in Round 2, without diluting the Round 1 investment. The required ownership stake is computed by adding the new investment to the $4.122 million value of the original 24.16% interest. To avoid diluting the gain realized on the initial investment, the investor would require a total of 32.95% of harvest value in exchange for investing in both rounds. If the venture is successful in the first stage, so that the Round 2 investment is made, the investor would end up with 32.95% of the equity after her Round 2 investment. This compares with the 40.91% we found the investor would need if the investment were not staged. With staging, the entrepreneur’s ownership percentage is higher in both the first and second rounds. Moreover, his initial commitment is for only two years instead of five. The Round 2 commitment is made only if the venture proves to be successful in Round 1. As a result, the value of the entrepreneur’s commitment to the venture is increased. The relevant question to the entrepreneur is whether to invest two years of time in exchange for 75.84% of the first-round equity and the option to invest three more years to pursue the second-round real option. The value to the entrepreneur is much higher than if the venture were pursued without staging.
12.3 Using Simulation to Design Financial Contracts Analyzing discrete scenarios is not a very convenient or accurate way to evaluate deal structures. Furthermore, limiting the analysis to a small number of scenarios is likely to undervalue embedded real options. Options are most valuable when results are either very good or very bad, but extreme outcomes are usually ignored in discrete scenario analysis. Simulation affords an expedient and insightful way of comparing alternatives. The value of any venture depends on how the deal is structured, and specific contract alternatives can be assessed only in the context of the venture. The process parallels our discussion from Chapter 5 but with additional specificity as to the distribution of possible payoffs.
Designing and Valuing Staged Investment with Real Options 483
Step 1: Model the venture and the deal. Build a financial model that reflects the structure of the venture and any important real options. A model is simply a set of mathematical expressions that converts operating and financial assumptions into a forecast of project cash flows. When a financial contract allows for abandonment if revenue is low or expansion if revenue is high, then the model must incorporate those options. To be used for valuation, the model must be structured to yield the cash flows that would be committed or received by investors and the entrepreneur. Step 2: Specify the assumptions and uncertainties, including terms for exercising the options. The assumptions include such factors as how rapidly sales are expected to grow and how profitable sales are expected to be, the structure of ownership claims that controls the cash flow allocation, and so on. In addition to specifying expected performance, it is necessary to make assumptions about the uncertainty of key financial variables, such as sales and profitability and to make specific assumptions about the conditions under which each real option would be exercised. If each party can have different expectations about cash flows, different sets of assumptions can be used to reflect the different expectations. Specific assumptions may depend on the signaling and bonding aspects of the proposed financial contract. Step 3: Run the simulation. Based on the assumptions, you can run the simulation model to generate expected cash flows and standard deviations for investors and the entrepreneur. Estimates of incremental cash flows and uncertainty, conditional on the exercise of any embedded options can then be derived. Step 4: Value the expected cash flows. Use the valuation models to evaluate the financial claims and conditional investment decisions. If the investor is well diversified, evaluate the investment based on the CAPM. If the entrepreneur is poorly diversified, evaluate the financial claim based on the total risk of the entrepreneur’s portfolio. Step 5: Search for the option-exercise criteria that result in the highest value. Consider the problem of contract design from the perspective of the entrepreneur. For options where the investor has the control rights, the entrepreneur would like to predict the investor’s exercise decision and understand the implications for the value of the entrepreneur’s claim. In addition, the entrepreneur would like to determine optimal exercise decisions for options where he has control rights. Evaluate the exercise decisions by experimenting with a range of possible exercise choices.
484
Chapter Twelve
Step 6: Search for the allocation of claims that results in the highest value. By changing assumptions in the model, you can search for the allocation of ownership claims that yields the highest value to the entrepreneur or to the investor. In the remainder of this section, we illustrate the approach and discuss its use to evaluate contract alternatives.
Developing the Financial Model of the Venture and Specifying the Assumptions To illustrate the use of simulation for designing a financial contract, we construct a financial model based on a decision tree. The model provides for much of the uncertainty about the venture’s probable success to be resolved by the second year. It includes a real option where the investor can decide, at that point, whether to make a second-stage investment. The investment would expand the capacity of the venture, which could be warranted if market response to the venture during its first two years is high enough. The model we use for this illustration is simple. An initial investment of $2.5 million is made at Time 0 and a conditional second round investment of $5 million is made at Time 2. We use simulation partly to determine the demand conditions under which it would be valuable to make the second-round investment. Table 12.8 shows the results from one simulated trial where the secondround investment is made. The model is designed to generate forecasts of the cash flows that the investor and entrepreneur will commit and receive. Because we addressed financial modeling in detail in earlier chapters, we abbreviate description of the financial aspects of the model here. The model is driven (in the first panel) by a forecast of market potential that is subject to uncertainty. Much of the uncertainty is resolved by the end of the second year, so that if potential revenue is high at that point, it probably will continue to be high.3 The venture’s capacity to supply the market depends on the total amount invested (in the second panel). Time 0 investment is $2.5 million, which includes $500,000 of the entrepreneur’s human capital. The investor has an option to expand capacity at the end of Year 2 by investing an additional $5 million. The investor will want to make the second-round investment only if market potential is expected to be high. The precise conditions under which the investor would want to exercise this expansion option depend on the venture’s profitability, its other cash needs, and the financial claim that the investor would gain by exercising the option. In the table, we must evaluate the venture in light of the terms of a specific deal. Accordingly, we assume that in the first round, the investor commits
Tab le 12 . 8 New venture simulation with conditional second-stage investment
Year
0
1
2
3
4
5
101.16 $100 $10,116
204.80 $100 $20,480
435.20 $100 $43,520
606.92 $100 $60,692
833.88 $100 $83,388
$2,500
$5,000 $7,500
$7,500
$7,500
$7,500
$10,116 $8,092 $2,023 $1,809 $72 $0 $214
$20,480 $16,384 $4,096 $2,638 $479 $0 $1,458
$43,520 $34,816 $8,704 $4,482 $1,471 $0 $4,222
$60,692 $48,553 $12,138 $5,855 $2,335 $0 $6,283
$83,388 $66,710 $16,678 $7,671 $2,753 $0 $9,007
$0 $214 $1,012 ($798) $798 $0 $0
$0 $1,458 $5,182 ($3,725) $3,725 $0 $0
$0 $4,222 $11,520 ($7,298) $7,298 $0 $0
$0 $6,283 $8,586 ($2,303) $2,303 $0 $0
$0 $9,007 $11,348 ($2,341) $2,341 $0 $0
$798
$4,522
$11,820
$14,123
$16,464
Market Conditions
Market potential Price (dollars) Potential revenue Total Investment Investment Cumulative investment
$2,500 $2,500
Total Income Sales revenue Cost of sales Gross profit Operating expenses Interest expense Interest income Net profit Total Cash Flow Beginning cash Operating cash flow NWC required Free cash Borrowing Repay loan Ending free cash Loan balance
$0
Total Ending Value Continuing value Plus: Free cash Less: Loan balance Value
$72,052 $0 $16,464 $55,588
Total Investment $2,000 $5,000 $500
Investor’s first-stage investment Investor’s second-stage investment Entrepreneur’s first-stage investment Total Equity Share and Return Investor’s share in first stage Investor’s share in second stage Investor’s Total Share Entrepreneur’s Total Share
10.00% 30.00% 37.00% 63.00%
$20,568 $35,020
The table shows one randomly generated trial from a simulation model with conditional second-stage investment. Dollar figures except unit price are in thousands.
486
Chapter Twelve
$2 million in exchange for 10% of the equity and the entrepreneur commits $500,000 as the value of human capital. In the second round, if the option is exercised, the investor receives new shares sufficient to increase total equity by 30% (i.e., the investor’s new shares equal 30% of existing shares). Because of dilution, this would make the investor’s total stake 37%. We assume that the investor exercises the option if Year 2 revenue is at least $15 million. We selected this exercise-decision criterion after we used the simulation to evaluate several alternative exercise decision rules to see what would work best for the investor. In addition, for evaluating alternative contracts, we assume arbitrarily that the investor will not participate unless the investor’s NPV is at least $1 million (rather than setting it to zero as we usually do). That is, the market for VC funding is not perfectly competitive and the VC can command a positive NPV in exchange for investing. Thus, the entrepreneur searches for deal terms that maximize value subject to a constraint that the NPV of the investor’s claim must be worth at least $1 million. The deal terms and investment levels are advantageous for the entrepreneur. We settled on them only after some experimentation. The assumptions, including investment levels and exercise decision, are reflected in the investment line of the table. You can evaluate alternative terms and exercise decisions by modifying the formulas in these cells. You can design allocations that are more complex by making other simple modifications, for example, to allow for the possibility that the bonding and monitoring aspects of contract terms affect revenue and profitability expectations. In the model, the cash flows available to the parties depend on net profit and annual free cash flow. We compute net profit as operating profit, plus interest income on surplus cash, minus interest expense on any borrowing needed to fund the operation. We assume that a liquidity event occurs at the end of Year 5 and compute value at that point by capitalizing Year 5 earnings, adding the free cash balance and deducting any outstanding debt. The parties allocate total ending value according to their relative shares. For simplicity, we assume that the venture pays no taxes.
Using Simulation to Estimate Expected Cash Flows and Risk Simulating Table 12.8 generates forecasts of expected cash flows and cash flow risk for both parties. The investor is interested in two questions. First, assuming that the investor makes the initial investment, under what conditions would it make sense to make the second-stage investment? Second, is the NPV of the investor’s interest high enough to justify making the first-stage invest-
Designing and Valuing Staged Investment with Real Options 487
ment? The entrepreneur would like to know the condition under which the investor would make the second-stage investment. Given that condition, the entrepreneur would like to know the value of his financial claim. Because the entrepreneur’s commitment is irrevocable, the entrepreneur has no decision to make at the end of the second year and needs to try to design an agreement in which the investor’s decision would also work to the benefit of the entrepreneur. As with the evaluation of any real option, we must approach this recursively by first developing the condition for exercise and evaluating the final (second) stage investment. To assess the investor’s option-exercise decision, Table 12.9 is a version of Table 12.8 that is modified to isolate the incremental cash flows from exercising the option by making the second-round investment. Most of the rows used to determine value in Year 5 are hidden. The table includes (1) a projection where the second-stage investment is assumed to be zero regardless of potential revenue and earnings and (2) the incremental revenue and earnings. Together with cash flow returns to the investor and entrepreneur, these are measured as the difference between the projection from the model in Table 12.8 and the model with no second-stage investment. This difference is used to compute the incremental Year 5 value of investing in the second round. The table shows the results of one random trial, an example in which the VC investor exercised the option to make the second-stage investment. When the condition for exercise is satisfied, the investor commits an additional $5 million at the end of Year 2 and receives an increased interest in the venture’s cash flow. Table 12.9 presents a summary version of the three parts of the spreadsheet for one trial.4 The “Total Income” panel shows the cash returns and valuations to the parties if the investor makes the second-stage investment whenever the exercise condition is satisfied. The “Income—No Second Stage” panel shows cash returns and valuations for the same trial, assuming no second-stage investment is made. The “Incremental Income—Second Stage” panel shows the incremental cash returns and valuation differences due to the option-exercise decision. We use the results from these three panels to evaluate the conditions under which it would be rational to make the investment, to test whether the investor’s incentive is aligned with the interest of the entrepreneur, and to evaluate the financial interests of both parties. To assess the deal structure, we ran 1,000 iterations of the model in Table 12.9 and used the results to compile statistics for only the iterations that satisfy a stated condition. In this case, we collected results for five variables: For all 1,000 iterations, we collected data on the investor’s total share of cash flows, the entrepreneur’s total share, and the investor’s second-stage investment. Then, looking only at the iterations where the second-stage investment occurs, we collected
Tab le 12 . 9 New venture simulation: incremental effects of conditional second-stage investment
Year
0
1
2
3
4
5
$9,365
$19,531
$41,088
$54,199
$64,717
$41,088 $3,931
$54,199 $5,504
$64,717 $6,766
Market Conditions
Potential revenue Total Investment Investment
$2,500
$5,000
Total Income Sales revenue Net profit
$9,365 $124
$19,531 $1,344
Value
$43,184
Total Equity Share and Return Investor’s total share Entrepreneur’s total share
37.00% 63.00%
$15,978 $27,206
Investment—No Second Stage Investment
$2,500
$0
Income—No Second Stage Sales revenue Net profit
$9,365 $124
$19,531 $1,344
$30,000 $2,600
$30,000 $2,600
$30,000 $2,600
Ending Value—No Second Stage Value
$22,285
Equity Share and Return—No Second Stage Investor’s total share Entrepreneur’s share
10.00% 90.00%
$2,229 $20,057
Incremental Investment—Second Stage Investment
$0
$5,000
Incremental Income—Second Stage Sales revenue Net profit
$0 $0
$0 $0
$11,088 $1,331
$24,199 $2,904
$34,717 $4,166
Incremental Ending Value—Second Stage Value
$20,899
Incremental Equity Share and Return— Second Stage Investor’s incremental share Entrepreneur’s incremental share
27.00% –27.00%
$13,749 $7,149
The table is used to generate information on the simulated value of the venture with and without second-stage investment and the incremental value of the second-stage investment. The information is used to develop the investor’s decision rule for exercising the investment option in the second round.
Designing and Valuing Staged Investment with Real Options 489
data on the investor’s incremental share and the entrepreneur’s incremental share. We will use these results to value the option. Table 12.10 summarizes the simulation results. The table shows the expected second-stage investment of the investor together with the means and standard deviations of cash flows for the parties from all 1,000 iterations of the financial model. In addition, it shows the expected incremental cash flows and standard deviations computed over the 495 iterations where the condition for making the second-stage investment (Year 2 revenue of at least $15 million) is satisfied.
The Value of the Second-Stage Investment to the Investor We now can evaluate the various decisions to invest. We follow the conventional approach of evaluating the conditional (second-stage) decision first. Accordingly, Table 12.11 concerns the investor’s decision to invest in the second stage. A minimum condition for investing is that the investor’s NPV for exercising be positive at the time of the decision. To assess this, we use the “Investor Interest and Value—Conditional Second Stage Investment Only” panel of the spreadsheet to compute the value of the investor’s incremental cash flow as of the end of the second year. We compute discount rates and market risk over the three-year holding period from Year 2 to Year 5. For the investor, the Year 2 value of the expected $14.312 million incremental cash flow in Year 5 is $12.121 million. This value is based on the CAPM and on the assumption, shown in the “Market Information” panel of Table 12.11, that the correlation between incremental venture cash flows and the market is 0.2. The resulting (Year 2) NPV of the $5 million investment is $7.121 million. We can see from the table that exercising the expansion option is beneficial for the investor whenever Year 2 market potential is at least $15 million. It may be that the trigger value of $15 million is too high, however, so that additional opportunities to add value are overlooked. Alternatively, it may be too low, so that the investment is sometimes made even though the expected value is negative. To determine the optimal critical value, we would need to run the Tab le 12 .1 0 Summary of simulation investment and return results
Variable
Investor’s second-stage investment Investor’s total share Entrepreneur’s total share Investor’s incremental share Entrepreneur’s incremental share
Mean
Standard Deviation
Number of Cases Where Condition Is Satisfied
($2,475) $9,112 $19,505 $14,312 $7,824
$2,501 $8,483 $10,842 $4,187 $5,429
1,000 1,000 1,000 495 495
490
Chapter Twelve
Tab le 12 .11 Valuing the investor’s second-stage real option
Year 2
Year 5
Market Information
Risk-free rate of interest (compounded) Market rate of return (compounded) Market risk premium Market standard deviation Correlation between venture and market
12.49% 40.49% 28.01% 34.64% 0.2
Investor Interest and Value—Conditional Second-Stage Investment Only
Cash Flows Expected cash flow Standard deviation
$ $
– –
Present value of option cash flows (Year 2) NPV of option to invest in second stage (Year 2) NPV of Interest in Venture (Year 2)
$ $
7,121 –
$ $
(5,000) –
$ $
14,312 4,187
$
(5,000)
$
12,121
The table is used to compute the present value of the investor’s second-stage investment based on the simulation results in Table 12.9.
s imulation model several more times, trying different critical values for exercising the second-stage investment option.
How the Investor’s Decision Affects Value to the Entrepreneur Ideally, the entrepreneur would like to design the contract so that the investor would make the second-stage investment only when doing so is beneficial to the entrepreneur as well. To help design deal terms, the entrepreneur would need to value his share of the return on the conditional investment. Because total risk matters, this value depends on the entrepreneur’s other assets. For example, if the entrepreneur expects to have $1.7 million invested in a market index at the Year 2 decision point and to commit $100,000 per year of human capital (present valued), the value of the committed human capital investment would be $300,000. Based on this, and using the approach from the appendix to Chapter 11, we can find that the NPV of the entrepreneur’s incremental human capital investment is $3.071 million. Thus, the entrepreneur would regard the assumed deal terms as positive. To determine whether this is the best choice, we would need to test other assumptions.5
Designing and Valuing Staged Investment with Real Options 491
The Value of the Investor’s Interest in the Venture Now that we have valued the option for both parties, we need to back up one step and determine whether the overall deal is sufficiently attractive for the investor. Then we need to determine whether undertaking the venture is worth the entrepreneur’s commitment. The “Investor Interest and Value” panel of Table 12.12 contains the evaluation of the investor’s total financial claims. The calculations assume that whenever the trigger value is reached, the secondstage investment is made. For the investor, the Time 0 NPV of the deal is $1.5 million; this includes the PV of the $9.112 million expected cash flow in Year 5 (from Table 12.10) and the PV of the negative $2.475 million expected secondstage investment in Year 2. By investing $2.0 million and an expected $2.475 million in Year 2, the investor gets an expected Year 5 return of $9.112 million. The PV of the expected return in Year 5 is $5.788 million and the PV of the expected Year 2 investment is negative $2.288 million, resulting in a Time 0 PV of $3.5 million. Deducting the Time 0 investment results in a $1.5 million NPV.
Tab le 12 .12 Valuing the investor’s financial claim by discounting all expected cash flows
Year 2
Year 5
8.16% 25.44% 17.28% 28.28% 0
21.67% 76.23% 54.57% 44.72% 0.2
Market Information
Risk-free rate of interest Market rate of return Market risk premium Market standard deviation Correlation between venture and market Investor Interest and Value Cash Flows Expected cash flow Standard deviation Present value of first-stage interest NPV of option to invest in second stage NPV of Interest in Venture
$ $
(2,000) –
$ $
(2,475) 2,501
$ $
$9,112 $8,483
$
3,500
$
(2,288)
$
$5,788
$
1,500
The table is used to compute the NPV of the venture from the perspective of the investor, taking account of the decision rule concerning the conditions under which the option to make the second-stage investment is exercised.
492
Chapter Twelve
The Value of the Entrepreneur’s Interest in the Venture Based on the expected Year 5 value of the entrepreneur’s expected share of the venture of $7.824 million, as shown in Table 12.10, and taking account of the entrepreneur’s underdiversification, the NPV entrepreneur’s interest in the venture is $5.177 million. Thus, the venture is acceptable to both. This is determined using the methodology from the appendix to Chapter 11.6 The valuation in Table 12.12 is based on one set of assumptions about when the second-stage investment option would be exercised and one allocation of ownership claims. To determine whether the exercise decision is the highestvalued one for the investor or whether some other assignment of ownership claims would be better for the entrepreneur, we must repeat the simulation and valuation steps using different assumptions for option exercise and/or ownership shares.
Simplifying the Valuation Process In the preceding example, we assumed that we knew the marketwide parameters, the expected cash flows and standard deviation of cash flows of the venture, and the correlation between venture cash flows and the market. Since we did not assume we knew the required return on the venture, we used the CEQ form of the CAPM to value the opportunity, including the real option. As it turns out, even though the total risk of the ultimate cash flows is high (a standard deviation estimated to be 93% of the expected cash flow), the low correlation with the market results in an effective cost of capital that is low. The $14.3 million Year 5 expected cash flow is valued at $12.1 million at Year 2, indicating a RADR of 5.7%. Risk and required return in the second stage are low because the investment is made only if stage 1 is successful. Given the success in the first stage, the likelihood of success in the second stage is high. The ultimate expected return over all trials of $9.1 million in Year 5 has a Time 0 value of $5.8 million, indicating a 9.3% required rate of return. The valuation problem for multi-stage investments can be simplified if we can assume that the RADR is known. Some of the evidence presented in Chapter 11 suggests that a discount rate around the expected return on the market would be appropriate for many ventures. In the current example, we assumed a required return on the market of 12%. If we were to discount the second-stage investment, the implied Year 2 value would be $10.2 million instead of $12.1 million, and the ultimate value of the expected $9.1 million would be $5.2 million instead of $5.8 million.
Designing and Valuing Staged Investment with Real Options 493
The point here is that if we are willing to assume that the required rate of return is known, the venture can be valued based entirely on the expected cash flow that is determined by simulating the investor’s payoffs.
12.4 Valuing Different Types of Financial Claims In this section, we discuss ways of applying financial modeling and simulation to other contracting issues. Our treatment here is conceptual but can be integrated within the structural simulation model of the venture. In the preceding example, the portfolios of claims of the parties were different. In exchange for the first-stage investment, the investor received common stock plus a call option on additional shares, with an exercise price of $5 million. For someone to buy a call, someone else must write (sell) it. In that example, because the second-stage investment would increase shares outstanding, both parties effectively were involved in writing the call. In that analysis, we did not attempt to value the equity of the venture or to determine value per share; instead, we valued the complete distribution of cash flows each party would expect to receive, given their deal. This approach, of valuing the entire distribution of cash flows, can be applied to any contracting structure the parties may consider. The following are examples. Debt financing. Straight debt can be incorporated easily in a financial simulation model. If competition generates a competitive interest rate, then the entrepreneur does not need to value the financial interest of a creditor. All that is necessary is to model and value the residual cash flows of the entrepreneur in a way similar to that shown in the previous illustration. On the other hand, if the entrepreneur has to negotiate debt financing terms bilaterally with a prospective lender, there may be good reason to estimate the NPV of the creditor’s financial claim—for example, to design terms that would work to the parties’ mutual benefit. The model of the venture can be structured to incorporate default states and generate the corresponding cash flows to the creditor. Thus, simulation of the debt contract can be used to estimate the series of periodic expected cash flows and standard deviations of cash flows, which are inputs to the valuation. Convertible securities. Often, the investor’s financial claim is a hybrid, such as convertible preferred stock or convertible debt. Valuing convertible securities is not fundamentally different from what we have already discussed.
494
Chapter Twelve
The simulation analysis must be based on a financial model that anticipates the conditions under which conversion is expected to occur. This is similar to simulating when a real option is likely to be exercised. Preferred shares and debt give the investor priority in the event of liquidation. They may also generate dividend or interest payments or accrue rights to dividends or interest that are to be paid at exit. The model must incorporate the dividend or interest provisions in the cash flow allocations. With these modifications, the valuation methodology is the same as in the previous example. Each claim is valued on the basis of expected cash flows, standard deviation, and the estimated correlation with the market. Ratchets or antidilution rights. A ratchet is a right to receive future shares in the event that value per share is lower in a subsequent round of financing. If value declines, the investor’s average cost per share is reduced by issuing additional shares for a nominal price. One difficulty in evaluating a ratchet is that the modeling must include an estimate of the variable that can precipitate exercise. If exercise depends on share price in a subsequent round, exercising the ratchet would increase shares outstanding without adding to value. Consequently, the share price in the round depends on whether the ratchet would be exercised. Share price and exercise of the ratchet provision are simultaneous outcomes; the model needs to incorporate the simultaneity.
12.5 Summary The primary focus of this chapter is on demonstrating how staging a venture can create value and providing the tools for valuing staged investments. We first demonstrate use of the VC Method in a staged investment context. While the method is helpful, focus on the success scenario is more problematic when an investment is staged than when it is not. We turn then to using the CAPM and value expected cash flows in a few discrete scenarios. Finally, we use simulation to show how real option and ownership structures can be used to enhance value and illustrate the use of simulation to design staged contract structures that are beneficial to both parties. The financial modeling techniques presented here enable the entrepreneur to assess how the required ownership percentage changes when follow-on investment is expected. The financial modeling approach can be applied to a myriad of
Designing and Valuing Staged Investment with Real Options 495
related contracting issues, such as the impact of various types of financial claims on value and how debt financing and similar provisions can increase value. Review Questions
1. In the context of the VC Method, describe how to address dilution to determine the amount of ownership an investor would require for a staged investment. Refer to Tables 12.1 and 12.2. 2. How does staging of new venture investments create value for the entrepreneur? What about for the investor? 3. What is the likely impact of staging on the level of risk facing the entrepreneur? 4. Why might an investor in an early financing round want to anticipate the need for additional financing rounds even in cases where the entrepreneur and investor agree on the terms for investing? 5. Explain how an entrepreneur can use staging of investment to both signal the entrepreneur’s beliefs and better align the entrepreneur’s incentives with the interest of investors. 6. Why is simulation preferred over scenario analysis for valuing real options embedded in new ventures? 7. Describe the ideal “trigger” for modeling the value of an embedded real option. 8. How can you model the impact on value of the personal guarantee given by entrepreneurs when they raise new venture debt? 9. How can the entrepreneur use debt terms to signal confidence in the venture? 10. Describe the process of valuing the entrepreneur’s ownership interest in a two-stage investment. Refer to Tables 12.8 through 12.11. Notes
1. Because these calculations are based on the Chapter 11 Appendix, these calculations are not shown in Figure 12.5. Interested readers can download the Excel version of the figure and unhide the rows that contain the details of the entrepreneur’s valuation. 2. The zero-NPV increase in ownership is not necessarily the same as the minimum ownership change necessary to cause the investor to exercise the option to invest. To determine the necessary ownership change requires
496
Chapter Twelve
c omparing the value of the ownership claim with exercise to an explicit assumption of the value without exercise. That analysis depends on what would happen if the option were not exercised. Not exercising could mean, for example, that the real option to go to the next stage could not be exercised. In this example, if the venture is successful in the prior stage but the investor does not invest in the next round, we effectively assume that another investor provides the funds on the same terms but in a way that does not dilute the existing investor’s claim. The assumption assigns all positive NPV to the entrepreneur. 3. To see the random process that generates the forecast of potential revenue, you would need to download the @Risk or CrystalBall version of Table 12.8. 4. The hidden rows and cell formulas for @Risk and CrystalBall versions can be studied in the Excel files, which are available on the companion website. 5. To see the entrepreneur’s valuation, Table 12.11 can be downloaded and the relevant rows unhidden. 6. The Excel file for Table 12.12 can be downloaded and the rows showing the entrepreneur’s valuation can be unhidden.
References and Additional Reading
Berger, P. G., E. Ofek, and I. Swary. 1996. “Investor Valuation of the Abandonment Option.” Journal of Financial Economics 42: 257–87. Bernardo, A., and I. Welch. 2001. “On the Evolution of Overconfidence and Entrepreneurs.” Journal of Economics and Management Strategy 10: 301–30. Bhide, A., and H. H. Stevenson. 1990. “Why Be Honest If Honesty Doesn’t Pay.” Harvard Business Review 68 (September–October): 121–9. Brennan, M., and A. Kraus. 1987. “Efficient Financing Under Asymmetric Information.” Journal of Finance 42: 1225–43. Chua, J. H., and R. S. Woodward. 1993. “Splitting the Firm Between the Entrepreneur and the Venture Capitalist with the Help of Stock Options.” Journal of Business Venturing 8: 43–58. Cornelli, F., and O. Yosha. 2003. “Stage Financing and the Role of Convertible Securities.” Review of Financial Studies 70: 1–32. Cumming, D. 2005. “Capital Structure in Venture Finance.” Journal of Corporate Finance 11: 550–85. Gompers, P. A. 1995. “Optimal Investment, Monitoring, and the Staging of Venture Capital.” Journal of Finance 50: 1461–89.
Designing and Valuing Staged Investment with Real Options 497
Kaplan, S. N., and P. Stromberg. 2002. “Characteristics, Contracts, and Actions: Evidence from Venture Capitalist Analyses.” Journal of Finance 59: 2177–210. ———. 2003. “Financial Contracting Theory Meets the Real World: An Empirical Analysis of Venture Capital Contracts.” Review of Economic Studies 70: 281–316. Landier, A., and D. Thesmar. 2009. “Financial Contracting with Optimistic Entrepreneurs.” Review of Financial Studies 22: 117–50. Ravid, S. A., and M. Spiegel. 1997. “Optimal Financial Contracts for a StartUp with Unlimited Operating Discretion.” Journal of Financial and Quantitative Analysis 32: 269–85. Sahlman, W. 1987. “Note on Financial Contracting: Deals.” Harvard Business School Press Note 9-288-014. Boston: Harvard Business School Press. Stevenson, H. 1988. “Deal Structure.” Harvard Business School Press Note 9-384-186. Boston: Harvard Business School Press.
This page intentionally left blank
C h a p t e r Th i rteen
Harve sti n g
H a rv e s t i n g i s t h e final stage of the entrepreneurial investment process.
It also is a critical component of initial investment decisions. Investors evaluate opportunities based on the expectation of a liquidity event that will enable them to realize a return and allow them to shift attention to other projects. To estimate exit value when they make the investment decision, investors must make assumptions about how and when the investment will be harvested and about the return they will realize. Harvesting is also of direct importance to entrepreneurs, but the issues may be different. Some entrepreneurs are like investors in their desire to harvest and move on to other endeavors. Others view the venture as a lifetime commitment and seek to earn returns through the continuing free cash flow of the venture or through partial liquidation of their holdings. Both types should be seeking to realize a return that compensates for the opportunity cost of their capital and effort. In this chapter, we examine the more important harvesting alternatives. As with financing, there is a menu, and it includes some of the same possibilities. For any given venture, the harvesting alternatives realistically available are limited. Accordingly, much of the chapter concerns the factors that bear on the decision of how and when to harvest. Going public and private sale of the venture to another firm (i.e., acquisition) are the harvesting alternatives that normally receive the most attention in business plans. Other important alternatives include a management buyout (MBO), sale of the business to employees or other members of the team, and continuing to operate the venture. We begin by surveying the alternatives and conclude by examining the factors that affect the harvesting choice. 501
502
Chapter Thirteen
13.1 Going Public When a venture goes public, an ongoing public market is established for its common stock. IPOs in the U.S. are usually “primary market” transactions. The issuing firm raises capital by selling new shares to public market investors. The IPO is only the first step in the process of harvesting by going public. By placing freely tradable shares in the hands of investors, the IPO enables investors in the stock market to establish a public market value for the equity. The public market itself is predominantly a “secondary market” where market participants transact with each other. Thus, secondary market transactions transfer ownership among investors; they do not generate new funds for the firm that originally issued the shares. A venture that has already gone public can raise additional capital by issuing new shares in a subsequent primary offering. Such an offering is called a “seasoned offering,” meaning that the issuer’s shares have already gone through a “seasoning” period of trading in the public market. A seasoned offering is not the same as a secondary offering unless the shares that are sold in the offering are not newly issued. In a secondary offering, the shares that are sold were issued previously and are being resold by existing investors. Sellers may include VCs and other early-stage investors, the entrepreneur, and others who acquired shares from the firm. Market value (or market capitalization) is the share price times the number of shares outstanding. It is determined by the price that equates secondary market demand with secondary market supply. Except for primary market transactions and first-time public sales of shares by entrepreneurs and others who invested before the IPO, secondary market supply (the “float”) is fixed. With supply fixed, fluctuations in market value result from changes in demand. Such value changes reflect changes either in expectations about future free cash flows or in the rate of return required by investors. Equity investors can harvest their investments in public companies in four ways: 1. Secondary sale as part of the IPO. Existing investors, including the entrepreneur and the VC, can sometimes harvest a portion of their investment by selling some of the shares they own in the IPO.1 2. Sale in the secondary market. Once the public market is established, investors who acquired their shares before the company became publicly traded can sell small amounts of shares from time to time directly into the public market.2
Harvesting 503
3. Seasoned offerings of secondary shares. Investors who acquired their shares before the company was publicly traded can sell larger quantities of shares in a formal secondary public offering after the IPO. 4. Private placement of secondary shares. Investors can use the public market value to benchmark the price and make a private sale to another investor, the entrepreneur, or the company.3
The IPO Process A firm-commitment general cash offering is the process for going public that is used most often in the U.S. In some countries a rights offering process, which will be discussed later, is more common. Selecting the underwriter. A management team that is considering a public offering normally begins by selecting an investment banker or team (syndicate) of investment bankers to underwrite the offering.4 The selection process, sometimes referred to as a “bake-off” or “beauty contest,” can include multiple contestants. The selected underwriter is responsible for forming the underwriting syndicate, advising the issuer, distributing the shares, and underwriting the risk of market price fluctuations during the offering. A broader notion of the term “underwrite” is that the underwriter also makes explicit and implicit representations about share value. The choice of investment banker involves matching underwriter capabilities with the issuer’s specific interests and objectives. Investment bankers have different capabilities and different approaches for carrying out the offerings. Some, for example, focus on institutional brokerage clients, whereas others focus more on retail. This focus affects the kinds of investors with whom the underwriter is likely to place the stock. Some investment bankers have established reputations and expertise for handling offerings in particular industries. Some have more limited capabilities but may be able to underwrite small issues at lower cost. These different capabilities explain why the issuer may select a lead and co-lead underwriter and may go further by using a syndicate of underwriters to help market the offering. Other considerations can affect the issuer’s choice of underwriter. If the issue is to be large, is the underwriter capable of organizing the syndicate needed to complete a large offering quickly? If secondary market liquidity is important, how strong is the market-making operation of the investment banker? In some cases, one or more investment banker may already have identified the venture as
504
Chapter Thirteen
a prospect for public offering and may have initiated a relationship. Sometimes VC investors draw upon their relationships with investment bankers as prospective underwriters. The investment banker may even have a VC operation that has invested in the venture at an earlier stage.5 It may seem that an issuer could use competitive bidding to select an underwriter; that is, the issuer could select the investment banker who bids the lowest fee (i.e., spread) for underwriting the issue. However, since many factors besides fee are important to issuers and bidding on the IPO of an unknown firm would be subject to great uncertainty, current practice is for the issuer to negotiate fees and other terms bilaterally with the prospective underwriter. Providing multiple underwriters with information sufficient to enable them to bid effectively is not practical, and new information and market fluctuations during the process of preparing the offering make negotiating specific terms infeasible. Due diligence and issue pricing. The underwriter is an intermediary between the issuer and public market investors. As a well-informed seller of shares, the issuer may be motivated by the desire to raise capital for investment opportunities or by a perception that the market will overvalue the shares.6 The seller may also want to establish a public market value for the venture. The value can be used as a basis for future transactions, or the public market can enable existing investors to harvest.7 Public market investors, by comparison, are uninformed about the issuer. They run the risk of being exploited by issuers who expect that the market will overvalue their shares. Serving as an intermediary in such transactions is challenging. Investors who are uninformed and concerned about paying too much are likely to be conservative in their valuations. Their concerns about overpaying discourage prospective issuers. One of the underwriter’s responsibilities is to mitigate the effects of the information disparity. In part, the underwriter does this by preparing a prospectus that describes some of the issuer’s operating history, the market, the management team, and how offer proceeds will be used. The underwriter verifies the information in the prospectus through an elaborate due-diligence process. The objective of due diligence is to ensure that the underwriter discovers any material adverse information about the issuer.8 The underwriter, however, cannot disclose all relevant information in the prospectus. For example, some information is strategically sensitive. Still, the information must factor into the underwriter’s decision about pricing. As a result, due diligence is sometimes sufficient to discourage private firms with negative information from issuing via public offering. Such firms, if they need equity capital, are more likely to negotiate private transactions.
Harvesting 505
The other way in which the underwriter addresses information disparity is by establishing the offer price. Information contained in the prospectus is never sufficient to enable investors to arrive at an accurate valuation. Some relevant information is too commercially sensitive to publish.9 Moreover, issuers and underwriters uniformly refrain from making forecasts during the offering process.10 Based on information acquired in due diligence, the underwriter (in consultation with the issuer) establishes an offer price. The actual process of pricing an issue is more protracted. Figure 13.1 illustrates the process in the U.S.11 The underwriter first establishes a “filing range” of prices at which the offering is likely to be made and publishes the filing range and other information in a preliminary prospectus. At that point, the underwriter and the issuer normally begin a series of “road show” presentations to brokers and institutional investors. During the road show, the underwriter collects information on demand, in the form of nonbinding indications of interest. This process is referred to as “bookbuilding” (building an order book). The final issue price is determined the day before the offering in light of demand information from due diligence, indications of interest, and new information that arises between the time of the preliminary prospectus and the offer date.12 The offer price represents an opinion about the value of the issue relative to the contemporaneous market values of other companies. Even though prospectus information is incomplete, many investors rely on the issue price as a conservative estimate of what the shares are worth. On average, their reliance is justified, because, as documented later, the immediate aftermarket prices of most IPOs in the U.S. are above the issue prices.13
Fi g u r e 13 .1 IPO issue pricing process for a firm commitment underwriting
New information from market
Comparable firm values
Comparable transactions and IPOs
Discounted cash flow valuation
Information from issuer
New information from market
Filing range reported in preliminary prospectus
Preliminary estimate of value
New information from due diligence
Issue price reported in final prospectus Indications of interest from roadshow “Take down” due diligence
506
Chapter Thirteen
Underwriter and Venture Capitalist Certification and Issue Pricing Why are investors able to rely on the offer price as a representation of value? The financial economics literature indicates that one function of the underwriter is to certify the offer price.14 An underwriter with an established reputation uses reputational capital to certify that the offer price is consistent with information the underwriter could reasonably be expected to discover through due diligence and with prevailing market values. The underwriter’s fee includes an (implicit) premium for using its reputational capital to certify the price. If the issue is overpriced, the underwriter’s reputation is harmed and its ability to charge a premium fee for underwriting future offerings is reduced.15 A similar argument is that if an issue is overpriced, the underwriter may bear the cost of defending itself in litigation brought by investors who purchased shares in the IPO. This is a version of the certification hypothesis, as underwriters must charge premiums to offset the expected cost of litigation.16 VC firms also appear to contribute to the certification and monitoring of companies that go public. VCs and other early investors are usually precluded from selling their positions in or shortly after the IPO. “Lockup” agreements with the underwriter typically prevent selling for 180 days.17 VC investors normally retain ownership in their portfolio companies for at least one year after the companies go public.18 During the period while the VC firm is still invested, it has an incentive to continue to monitor the company. Thus, the VC firm can provide an important monitoring function that works to the benefit of public market investors. In addition, there is evidence that in the U.S., companies with VC backing have been less underpriced than have other companies.19 The finding suggests that VC firms contribute to certification, enabling companies to issue shares at prices closer to market value. However, in more recent periods VCbacked IPOs have tended to have higher initial returns. Lee and Wahal (2004) compare underpricing of VC-backed IPOs to that of matched non-VC-backed IPOs, and find underpricing to be 5–10% higher in the VC-backed sample.
Examples: Underwriter Selection and the Challenges of Issue Pricing and Distribution To illustrate the mechanics underlying the IPO process, consider an actual IPO by a company we will call Alias. The issuer entertained proposals from 15 leading investment banks, each of which prepared a preliminary estimate of value. The offer prices inferred from those presentations ranged from $12 to
Harvesting 507
$38 per share. At the end of the process, two investment banks were selected as co-lead underwriters. One had indicated a preliminary valuation of $31 to $36 per share. The other had indicated $16 to $24. Two months later, a preliminary prospectus was circulated, indicating a filing range of $20 to $22. In the interim, the overall market had declined, and the overall level of IPO activity had fallen. During the road show, it became apparent that demand for the shares was not as high as had been hoped. The IPO took place about one month after the preliminary prospectus was circulated. The shares were offered at $20, the bottom of the range, and the total size of the offering was reduced in response to information from the road show. Immediately following the IPO, the closing price in the aftermarket was $21, up 5% from the issue price. Subsequently, the market for the shares languished, and the stock price fell to below the issue price. The May 1999 IPO of eToys, which went public during the Internet bubble, provides a counterexample with substantially more underpricing. Goldman Sachs was selected as lead underwriter. The preliminary prospectus indicated a filing range of $10 to $12 per share. In response to indications of strong demand and overall market gains for Internet stocks, the offering was priced at $20. The stock closed on the first day of trading at $80 per share, a gain of 300%, which gave eToys a market capitalization of $7.8 billion. However, the company went bankrupt in 2001 and KB Toys purchased its remaining assets for $5 million. Shareholders from before the IPO filed a lawsuit, alleging that Goldman Sachs underpriced the offering for its own purposes. These stakeholders argued that if the issue had been priced closer to market, eToys would have had enough cash to avoid bankruptcy.
The Cost of Initial Public Offerings Going public is expensive. The costs associated with an IPO include the fee or spread charged by the underwriter, the out-of-pocket costs of the issuer, and the effective cost of underpricing. Underpricing is normally measured as the initial return ((first aftermarket closing price – issue price)/issue price). Historically, in the U.S., underpricing since 1980 has averaged about 18% of the issue price. However, this includes the tech bubble in the late 1990s and 2000, when average underpricing was very high. From 1995 through 2000, the annual average was 33.7%, whereas the overall annual average was much lower, at 14.7%.20 Figure 13.2 shows average initial returns by year for IPOs with issue prices of at least $5. Underpricing tends to be greater for smaller and riskier issues. Underpricing is usually the most important component of the total cost of an IPO.21 The second most important is the spread. In a general cash offer, by far
Chapter Thirteen
Fi g u r e 13 . 2
80%
Average IPO underpricing by year (measured as initial return)
70% 60%
source: Data are from https://site.warrington .ufl .edu/r itter/ipo - data/.
50% Percent
The figure shows equalweighted and proceedsweighted annual averages of IPO underpricing where underpricing is measured as the initial return from the offer price to the first aftermarket closing price.
Annual average Average weighted by proceeds
40% 30% 20% 10% 0%
1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017
508
Year
the most common type of offering in the U.S., the underwriter buys the shares from the issuer for a negotiated net price and seeks to resell to the public at the offer price, which is higher. The difference between the offer price and the net price is the underwriter spread. Typically, in the U.S., the spread is around 7% of the issue price. As shown in Figure 13.3, the annual average underwriter spread has declined over time.22 The lower average when the data are weighted by proceeds indicates that the spread is lower for larger issues. In addition, the issuing firm will incur other direct costs in preparation of the IPO. As shown in Table 13.1, evidence from a 1996 study suggests that direct costs and the underwriter spread ranged from 6% to about 17% of gross proceeds and that the percentage is inversely related to issue size. Evidence from Lowry, Michaely, and Volkova (2017) that is more recent and the evidence in Figure 13.3 indicate that spreads have declined, especially for small and intermediatesized offerings. Why are IPOs so expensive, and why do both fees and underpricing tend to be larger percentages of proceeds for smaller issues? One reason is that there are significant fixed costs associated with underwriting a new issue. Much of the cost of preparing the prospectus, for example, is fixed. Fixed costs contribute to the negative relationship between issue size and the fee percentage. Furthermore, due-diligence costs depend on how fully the issuer hopes to price the offering. The closer to expected market value the price is set, the harder the underwriter must work to discover information that could adversely affect value and the more likely is a price decline after the offering to result in litigation.23 Also, the
Harvesting 509 Fi g u r e 13 . 3
9%
Average IPO underwriter spread by year
The figure shows equalweighted and proceedsweighted annual averages of IPO underwriter spreads where the spread is measured as the difference between gross and net proceeds, expressed as a percentage of gross proceeds.
7%
Percentage of gross proceeds
source: Data are from https://site.warrington.ufl .edu/r itter/ipo - data/.
8%
6% 5% 4% 3% 2%
Equal-weighted mean Proceeds-weighted mean
1% 0%
1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 2017
Year
Tab le 13 .1 IPO direct costs and underwriter spread by issue size
Issue Size
Total Direct Cost
(proceeds in $ millions) 2.0–9.99 10.0–19.99 20.0–39.99 40.0–59.99 60.0–79.99 80.0–99.99 100.0–199.99 200.0–499.99 500.0 and up
(% of proceeds) 17.0 11.6 9.7 8.7 8.2 7.9 7.1 6.5 5.7
The cost of a public offering includes three components: the spread between the offer price and net proceeds to the issuer, issue costs borne directly by the issuing firm, and underpricing. This table provides estimates of the first two components (excluding underpricing) for different sizes of issues. source : I. Lee, S. Lochhead, J. Ritter, and Q. Zhao, “The Cost of Raising Capital,” Journal of Financial Research 19 (1996): 59–74, Table 1.
more fully the offering is priced, the more direct costs the underwriter will have to incur to market the issue. For a small offering, it may be less expensive to underprice more and save some of the costs of due diligence and marketing.24 Another way to think about why IPOs are underpriced is to recognize that prospective investors include both those investors who are informed about
510
Chapter Thirteen
emand for the issue and those who are uninformed.25 If, on average but not in d individual cases, issues were priced correctly relative to market demand after the offering, then informed investors (who are better able to assess value) would tend not to participate in overvalued offerings. So informed investors would buy only the undervalued shares, and uninformed investors would get 100% of the overvalued shares but a smaller fraction of the undervalued shares. Thus, when issue pricing is correct on average but too high or too low in each individual case, the result is losses for uninformed investors. These expected losses would cause uninformed investors to withdraw from the IPO market. If it is important to keep them in, then an underwriter must set offer prices low enough, on average, so that uninformed investors end up with fair rates of return. The winner’s curse problem described in the preceding paragraph seems to suggest that issuers would fare better by using an auction process to sell their shares. That way, the private knowledge of informed investors could influence the offer price. However, auctions still reward investors for concealing their private knowledge about value, do not eliminate underpricing, and diminish the ability of the market to rely on representations of value by the underwriter. As an alternative to auction, underwriters can allocate IPO shares in a way that rewards investors for revealing their private knowledge or even encourages investors to produce information about value.26 A continuing puzzle that deserves recognition is that the average underwriter spread is materially larger in the U.S. than in many other countries. Chen and Ritter (2000) refer to “The Seven Percent Solution” and find that over 90% of the IPOs over $20 million in the late 1990s had spreads of exactly 7%. Citing an earlier paper, they note that this is roughly twice as high as IPO spreads in a number of other developed countries. In a more recent study, Abrahamson, Jenkinson, and Jones (2011) compare fees and underpricing for IPOs in the U.S. and Europe. While average fees have declined, they find that 7% continues to be the norm in the U.S. Large IPOs have lower percentage spreads, so that the proceeds-weighted-average spread is 5.64%. This compares to an average in their sample of European IPOs of 2.68%. Average proceeds are similar in the U.S. and Europe, so the difference is not explained by issue size. The authors consider the hypothesis that the higher fees in the U.S. may be offset by a more intensive selling effort and less underpricing. Their evidence does not support this explanation. They also consider several other possible explanations for the difference but are unable to find a compelling explanation. Banerjee, Dai, and Shrestha (2011) study underpricing in a number of countries and find that underpricing is greater when country-level information asymmetry is high and access to legal recourse is limited. It is lower in countries with welldefined property rights. The difference in fees remains a puzzle.
Harvesting 511
Methods of Valuation and Pricing Underwriters develop their own procedures when establishing offer prices. As a common first step, the underwriter estimates enterprise value on a pre-money basis (before the IPO proceeds but reflecting the value of opportunities). That valuation may be based on the market values of comparables, the prices of other recent IPOs or other transactions, or the present value of projected cash flows. Normally, an underwriter uses several approaches and, based on the results, makes an assessment of enterprise value. Next, the underwriter deducts the value of debt financing. The residual equity value is divided by the existing number of shares to arrive at an estimate of equity market value per share. The underwriter determines the number of shares to be offered and the offer price per share based on this estimate of market value per share. To illustrate, suppose an issuer hopes to raise $20 million in net proceeds. The underwriter estimates that the market value of equity is $10 per share. From this, the underwriter might deduct $1.50 of desired underpricing, and the underwriter spread might be estimated at 10% of gross proceeds. Thus, the venture would issue shares priced at $8.50 and would net $7.65 per share. Given the target of $20 million in net proceeds, the venture would need to issue 2.614 million shares. Using this information, the underwriter might specify a filing range of $7.50 to $9.50 and report the filing range in the preliminary prospectus, along with the fact that the company plans to issue about 2.614 million shares, to net $20 million. If, at the time of the offering, the market has improved, the underwriter is likely to select an offering price near the top of the range and may issue more shares, depending on how the issuer would like to respond to the change in the market. In many cases, it can make sense for the issuer to agree to an overallotment option. An overallotment option allows the underwriter to offer additional shares, anticipating that some investors may not honor their commitments to buy or may “flip” their shares to make a quick profit, which would depress the price. The option, sometimes referred to as a Greenshoe option (named after the first company to use them), allows the underwriter to issue additional shares if demand turns out to be higher than anticipated. In the U.S., the option can vary in size up to an SEC regulated maximum of 15% of the original number of shares offered. The feature allows the underwriter to affect the supply of shares, thereby providing more latitude to affect the aftermarket price.
The Cost of Harvesting by Going Public The high percentage cost of an IPO does not translate to a high percentage cost of harvesting. To see why, consider a venture that seeks to raise $30 m illion via
512
Chapter Thirteen
an IPO. Prior to the offering, the venture’s investment banker estimates the pre-money valuation to be $120 million. To assess the cost of the IPO, suppose the offering is structured to achieve a target aftermarket value per share of $30.00. This means that after the IPO, there should be 5 million shares outstanding. Consistent with normal practice, the underwriter builds in expected underpricing of 15%, compared to the estimated aftermarket value. Thus, the IPO is priced at $25.50 per share. In addition, the underwriter’s fee is 7% of gross proceeds. So the issuer nets $23.715 per share (0.93 × $25.50). To achieve net proceeds of $30 million, the firm plans to issue 1.265 million shares ($30 million/$23.715). Existing shareholders (the entrepreneur, VC investors, and others) retain the balance of 3.735 million shares. There are several ways to measure the cost of the IPO. Total cost per share issued is $6.285 ($30.00 − $23.715). This is 24.65% of gross proceeds, or 20.95% of expected aftermarket value of the issued shares. In dollars, the cost is $7.95 million ($6.285 per share times the number of shares issued). Alternatively, cost can be measured as the difference in the value of shares held by the existing investors compared to if the offering were costless. If the issue could have been done at $30, with no issue cost, the firm would only have needed to issue 1 million shares; existing shareholders could have retained 4 million. The 265,000 fewer shares they retain, valued at $30, yields a total cost of $7.95 million. Although the percentage issue cost is substantial, the market value of the venture after the IPO is expected to be $150 million. If the $7.95 million is viewed as the cost of harvesting the pre-money value of $120 million, it amounts to 6.625%.27 The percentage cost would be even lower if the venture’s pre-money value were higher compared to the size of the offering. Therefore, do not be misled into thinking that the high percentage issue cost is the same as the percentage cost of harvesting. The dollar cost is the same, but the percentage cost is much lower. Instead of focusing on the percentage cost of the IPO, it can be more useful to think about the cost to existing investors of creating a public secondary market for firm shares. If raising new financing is a minor concern, it could make sense to select the offer size to minimize the expected dollar cost of the offering, which, even with scale economies of the IPO, would probably indicate a small offering.
Other Approaches to Public Offering The approach we have described is a firm commitment general cash offering. The term “firm commitment” means that the underwriter commits to buying all of the issued shares at the net price in the prospectus. The underwriter’s compensation is realized by reselling at a higher price. The difference between
Harvesting 513
the gross price and the net price is the underwriter’s fee or spread.28 There are other ways to issue public equity. Best-efforts offering. An offering can be done on a best-efforts basis. In a best-efforts offering, the investment banker acts more like an agent of the issuer and does not guarantee net proceeds. You might expect that issuers would select best-efforts offerings when they wish to raise capital but do not have specific concerns about the amount that is actually raised. However, the bestefforts approach also suggests that the issuer is concerned about achieving a high price and that the issuer and the underwriter may disagree about the value of the shares.29 In a best-efforts offering, the underwriter provides distribution services but does not certify the price. The evidence on best-efforts offerings is not positive; generally, it appears that issuers could have achieved more success with a general cash offering.30 Rights offering. The term “general cash offering” signifies that the shares are sold to the investing public at large in exchange for cash. In contrast, in a “rights offering,” the firm offers shares to existing shareholders, in exchange for cash and warrants. Rights offerings are more common outside the U.S.31 In a rights offering, an issuer that is already publicly traded raises equity by issuing warrants to existing shareholders. The warrants are separately tradable call options that enable the shareholders to purchase new shares at a fixed price for a specified period. Shareholders who do not wish to exercise their rights can sell the warrants to others. Assuming that each new share requires one warrant, the effective price of a new share is the exercise price of the warrant plus the market value of the warrant. The issuing firm receives the exercise price, and existing shareholders receive the market value of the warrants. If the issuer is not yet public, a public market for the shares can be created by first allowing the warrants to trade and then allowing the shares to trade. For example, suppose that the stock of a company is trading for $20 per share and the company announces a rights offer that will increase the number of outstanding shares by 20%. The firm issues warrants to existing shareholders, one warrant for each outstanding share. The warrants give the shareholders the ability to buy new shares at a price of $15 plus 5 warrants per share. The 5-to-1 ratio will result in a 20% increase in shares outstanding. Because the shares are being sold at a discount from the market price, other things equal, the issue will cause the market price per share to decline. We can easily calculate the expected effect on the share price: Before the offering, based on the market price, 5 shares would have a total value of $100. After the
514
Chapter Thirteen
ffering, 6 shares would have a value of $115 (the $100 plus an additional $15). o So the new value per share is $115/6, or $19.167 per share. Since an investor can either exercise the rights by paying $15 and 5 warrants, or sell the warrants, and since the new share value will be $19.167, we can determine that each warrant should be worth $0.833/5, or $0.167. The strongest argument for the rights offering approach is, perhaps, that the issuer and the underwriter do not need to predict the market price after the offering or establish an offer price for the issue that is closely tied to aftermarket value. Because the effective price of newly issued shares is determined by trading in the market, existing shareholders do not lose if the exercise price of the warrants is below market. So there is no wealth transfer to new investors as there is when a general cash offer is underpriced. Similarly, new investors do not need to worry that the issue price is set above market value.32 An argument against rights offerings is that the approach limits the certification function of the underwriter to warranting the accuracy and completeness of the prospectus. Because the underwriter does not establish an issue price, information in the prospectus gains importance. Another concern is that rights offerings can be difficult to administer and some existing investors may fail to either exercise or sell their warrants. This concern is elevated because many investors are not holders of record. Instead, their securities are held in “street name” and the issuing company does not know that they are holders of the shares. In the U.S., issuing firms are free to select either a general cash offering or a rights offering. The fact that rights offerings are rare is evidence that for most U.S. issuers, the price certification function contributes more to expected proceeds than concern with mispricing takes away.33 Auction method offering. General cash offerings involve intentional underpricing, such that the aftermarket price is expected to be above the issue price. This indicates that at the issue price, there is more demand for the shares than there are shares available. Underwriters respond to excess demand by rationing the shares in ways that appear to favor preferred customers, such as those who are willing to buy even when an issue seems overpriced, and who refrain from “flipping” the shares to make a quick profit, or who may be sources of future business (“spinning”). On the one hand, underpricing may be an efficient aspect of the offering process that contributes to the certification efficacy of the underwriter and reduces costs of due diligence and marketing. On the other, it may be an unnecessarily good deal for the underwriter’s preferred customers. As an alternative, a few companies in the U.S. have gone public using an auction method.
Harvesting 515
The rationale for the auction method is that it can reduce or eliminate underpricing and the practices of flipping and spinning. In the U.S., W. R. Hambrecht and Company originated the OpenIPO, which represents a test of the efficiency of general cash offerings. Under this IPO process, bidders submit their bid prices and the quantities they are willing to buy to the underwriter via an Internet site. When the offering goes effective, demand information is matched with the quantity of shares available and the market-clearing price is found. Orders may be filled at or below the market-clearing price, but in the U.S., because of SEC regulations, all bidders whose orders are accepted pay the same price, even if they bid higher. The first OpenIPO offering was Ravenswood Winery in 1999. The auction price for that IPO was determined to be $10.50, and in the subsequent month the stock continued to trade in the $10.50 to $11 range. It appears that the process eliminated underpricing for this issue. In contrast, Google’s controversial auction-method IPO in August 2004 did not eliminate underpricing. In late July 2004, Google publicized its target price range of $108 to $135. Just before the auction, it lowered the range to $85 to $95. The IPO price was $85 and the first public price after the auction was $100, almost 18% above the issue price. What are some possible explanations for the underpricing in the auction process? One is that individual investors, who are supposed to be the primary participants in this auction, may lack sufficient information to appropriately price the security. Another is that because the offer price is determined by auction, the underwriter is less able to certify. A third is that, while the general principles of the auction are followed, the underwriter can (and often does) still decide to fill orders at a price below the market-clearing price in order to include some preferred investors who have submitted low bids (a “dirty auction”). In this case, there is still rationing, and underpricing is likely. Auction method offerings have been tried in many other countries, most often in small and emerging capital markets.34 In a study of Japanese IPOs during a period when both an auction method and bookbuilding (general cash offering) were available to issuers, bookbuilding was the clear winner. The concern is that the auction method gives rise to an adverse selection or “lemons” problem and that bookbuilding solves the problem. Availability of bookbuilding can reduce reliance on auctions or eliminate them altogether, which is what happened in Japan and other places.35 It remains to be seen whether the auction process is as effective in raising capital as the general cash offering approach. So far, the market test suggests that while this approach has appealed to some issuing firms, the number is limited and the concept does not have much momentum. U.S. interest in the OpenIPO has largely died out.36
516
Chapter Thirteen
Reverse merger. A reverse merger (also called a reverse IPO) is an alternative way for a private company to become public. In a reverse merger, an existing public company nominally acquires the shares of the private company and possibly then changes its name to that of the private company. The public company is usually nothing more than a shell—often a remnant of an unsuccessful public company. Its only remaining value is that its shares are publicly registered and tradeable, making it a target for a firm that wants to establish a public market but does not need to use an IPO to raise capital and does not want to incur the expense of a public offering. The notional acquirer (the public company shell) does not need to have any resources to accomplish the acquisition. Rather, the owners of the private company simply exchange their shares for shares of the public company. Normally, at the same time, they will change the name of the acquiring public company shell to that of the acquired formerly private company. Reverse merger does not require an underwriter, avoids the cost of an IPO, and can be accomplished relatively quickly. As there is no certification of value, the aftermarket capitalization may be less than if the company were to go public via IPO. Because such companies tend to have little liquidity, lack of analyst following, and few active market makers, the valuations may remain low for sustained periods. Whether, in individual cases, reverse merger makes sense depends on the firm’s circumstances. By creating a public market for company shares, reverse merger enables existing investors to eventually harvest their investments in a public market, creates a medium of exchange that can be used in other M&A (merger and acquisition) transactions, and may enable the formerly private firm to use market-based incentive compensation more effectively. However, because it does not generate immediate capital and may result in a lower valuation, reverse merger is not suitable for near-term capital raising and may not be ideal for harvesting when the firm is large enough and mature enough for an IPO or a true acquisition. In a recent study, Lee, Li, and Zhang (2014) examine the performance of reverse mergers that became active on U.S. stock markets between 2001 and 2010. They report that, as a group, reverse mergers are early-stage companies that typically trade over-the-counter but that Chinese reverse merger firms trading on U.S. exchanges tend to be more mature and less speculative. They find that these Chinese firms outperformed their matched peers and conclude that their results shed light on the balance between credibility through underwriting and access to well-functioning markets. Direct public listing. Direct listing is a rarely used alternative to a traditional IPO that offers benefits similar to reverse merger. Direct listing involves
Harvesting 517
registering existing private company shares for public market trading on an exchange such as NASDAQ or the New York Stock Exchange (NYSE). In a direct listing, the firm issues no new shares and does not normally engage the services of an underwriter. The main functions of direct listing are to enable existing shareholders to harvest their investments through aftermarket public sale and to establish a public market value for the shares. Direct listing in the U.S. requires financial disclosures similar to those required in an IPO, and the firm is subject to the SEC’s ongoing reporting requirements. Spotify is the most significant recent example of direct listing. Historically, most direct listings have been by small companies. Spotify is an exception and may pioneer this approach for other firms that are interested in public market visibility but are not in need of the capital that would be generated through an IPO. Direct public offering. Direct public offering (DPO) is not the same as direct listing. The most common type of DPO is a Small Corporate Offering Registration, or SCOR. The SEC provided this option to small businesses through Regulation D, Rule 504. SCOR gives an offering exemption to private companies that raise no more than $1 million in any 12-month period through the sale of stock, without restriction on the type of investor. A related type of DPO is permitted under Rule 505 of Regulation D. This option enables a small business to sell up to $5 million in stock during a 12-month period to an unlimited number of accredited investors. Partly in an effort to encourage smaller firms to go public, the U.S. Congress, through the JOBS Act, sought to relax some of the prior restrictions and to provide a lower-cost way for small issuers to go public. Regulation A+ permits issues of up to $50 million within one year to register with simplified filings and to reduce ongoing reporting requirements. There are other restrictions related to sale of the securities to the general public. The regulation and the other DPO provisions are described more fully in Chapter 2. Over a period of about 2 years following passage of the JOBS Act, Dambra, Field, and Gustafson (2015) estimate that the Act led to about 21 additional IPOs per year, predominantly in industries with high disclosure costs. However, Lowry et al. (2017) note that more recent data indicate that this effect of the 2012 JOBS Act may have been transitory.
Harvesting in the IPO For early investors in a company that later goes public, the IPO is only the beginning of the harvesting process. The entrepreneur, VCs, and other early
518
Chapter Thirteen
investors do not normally sell shares during the IPO. Historically, the lead VC investors in the U.S., for example, sell in only about 27% of the IPOs where they hold shares.37 When they do sell, they typically sell only about 20% of their holdings. There are three reasons for an investor, such as a VC, to refrain from selling during the IPO of a portfolio company. First, potential public market investors are likely to perceive insider selling as a negative signal (though not when the VC is facing expiration of the fund). A second reason is that the investor expects to continue to add significant value to the venture. The investor would want to hold the shares until the ability to add value falls below what the investor could achieve by focusing on other ventures. The third reason is that shares sold in the IPO are likely to be underpriced. By waiting, the investor can usually obtain a higher price.38
Harvesting After the IPO Typically, even if an early investor does not sell in the IPO, the investor’s shares will be registered along with those sold by the issuing firm. Registered shares are freely tradable in the aftermarket and can be sold. The main limitation is that existing shareholders normally have agreed with the underwriter to refrain from trading, usually for 180 days after the IPO. This “lockup period” serves at least two functions: First, it removes the need for an underwriter, who normally supports the market during an offering, to purchase shares that are being sold into the market at the same time by existing shareholders. Second, it allows enough time to pass so that the sellers cannot easily be accused of trading on the basis of information that should have been disclosed in the prospectus. In the U.S., there are several ways that early investors can harvest after the IPO. Rule 144 sales. In the U.S., shares of a public company that are not registered can still be sold into the market, but SEC Rule 144 limits the rate at which the sales can occur. An investor who owns unregistered shares but is not a control person (such as an officer, director, or large block holder) can periodically sell small amounts of unregistered shares but may not be able to liquidate his entire position. After one year of ownership, any remaining restricted shares become freely tradable. If the investor is a control person, the shares remain restricted and subject to the gradual liquidation process. One consideration behind Rule 144 is a concern that rapid sale into the public market could depress share prices.39 Gradual sale gives the market time to absorb the shares and reduces concerns about opportunistic selling.
Harvesting 519
Seasoned offerings that include secondary sales. Sometimes existing investors are eager to liquidate; in such cases, gradual sale may not be attractive. The alternatives are to arrange a large private transaction or to participate in a seasoned offering. Companies sometimes use seasoned offerings to lower the overall cost of raising equity or to enable investors to harvest. The issuer may undertake a small IPO to establish a market price and then follow a few months later with a seasoned equity offering (SEO). It is more common for existing shareholders to participate by selling in the SEO than in the IPO. Normally, the underwriter’s fee for an SEO is less than the fee for an IPO. In addition, SEOs are priced very close to the prevailing market price at the time of the offering. When an SEO is announced, the market price of the stock drops by about 3% on average, far less than the 17% average underpricing of IPOs.40
13.2 Acquisition While many entrepreneurs aspire to going public, by far the more common successful exit, especially after 2000, has been acquisition or merger. When investment in a venture is harvested by acquisition or merger with another company, the transaction can take any of several forms. The acquirer can purchase the equity of the venture for cash or exchange it for shares of the acquiring company. Alternatively, the acquirer can purchase all or some of the venture’s assets for cash or stock. The choice is affected by a variety of considerations.
Purchasing Venture Equity for Cash By purchasing the equity, the buyer acquires ownership of the assets of the venture and assumes the liabilities.41 The parties enter a stock purchase agreement that sets out the terms and describes the assets and liabilities. The acquirer normally undertakes some level of due diligence to verify its assessment of value. In addition, sellers are normally expected to provide representations and warranties that go to issues of value. For example, the seller may warrant that the accounts receivable balance the purchaser will acquire was prepared in a manner consistent with GAAP; that the venture is the legal owner of any intellectual property being transferred; and that the sellers are unaware of any existing or prospective litigation. To an extent, due-diligence efforts and representations and warranties are substitutes. A seller may be hesitant to open its records to the review of
520
Chapter Thirteen
a prospective buyer. This is understandable because if the transaction is not consummated, the buyer could have gained access to confidential information about the seller. In addition, if the venture is small, the cost of extensive due diligence can exceed the expected benefit. From the perspective of an investor, such as a VC firm, sale of equity for cash is likely to be the preferred form of transaction. However, it may be the most difficult form on which to agree. It may also be unattractive to the entrepreneur. For the investor, selling for cash facilitates harvesting because cash is easily distributed to limited partners. In addition, because many institutional limited partners are not taxed on their earnings, receiving cash does not give rise to adverse tax consequences. An entrepreneur who is selling equity in a venture may have a different perspective. If the sale involves a gain, the entrepreneur and members of the management team may be exposed to a significant immediate tax liability. On a more qualitative level, selling for cash has an aspect of finality that may not be to the entrepreneur’s liking. For the purchaser, acquiring equity for cash involves the most risk. By acquiring shares and paying cash, the acquirer is giving up a safe asset and acquiring one that is risky. Furthermore, if the acquirer must borrow the cash, then leverage increases. The increase in risk is not particularly important in itself, but adding leverage can deprive the acquirer of financial slack that could be used for other opportunities.
Purchasing Venture Assets for Cash Transferring only the essential assets of the venture to the buyer can reduce the buyer’s risk. In an asset purchase, the seller continues to be responsible for the venture’s liabilities. Because the buyer does not assume direct liability, the cash price will be higher than for an equity purchase. The seller can use the extra cash to pay off the liabilities. Automobile dealerships, for example, are often transferred using asset purchase agreements. The buyer acquires the facility, the vehicle inventory, and other essential business assets but does not assume liability for the loan the seller used to finance the inventory. Vehicle inventories are normally financed with “flooring” loans secured through specific claims against individual vehicles in the dealership inventory. When a vehicle is sold, the dealer is supposed to repay the loan amount associated with that vehicle. When new inventory arrives, the dealer can borrow more, based on the specific vehicles received. If a dealer sells vehicles without repaying the loan, a purchaser who assumes the obligation can be liable for the repayment. Instead of verifying that the inventory securing the loan is still on the dealer’s lot, the purchaser simply buys the inventory (possibly arranging new
Harvesting 521
financing) and leaves the seller to repay the loan. For similar reasons, a purchaser may decide to acquire only those assets that are central to future operations. The accounts receivable balance is an example. Rather than trying to verify the quality and status of receivables, it could be easier not to purchase them.
Exchanging Venture Equity or Assets for Equity of the Acquirer Some transactions are structured as exchanges of the acquirer’s stock for either the stock or assets of the venture. Exchanges of stock for assets are difficult because the seller is left with the liabilities and may not have the liquidity to service the debt. Accordingly, we focus on exchanges of stock for stock. In such exchanges, uncertainty about value exists on both sides. Not only must the buyer conduct due diligence, but sellers must also value the shares they will receive. The problem is particularly difficult if the acquirer is not a public company. In such cases, there are no public stock prices and no SEC reporting requirements or analyst reports to facilitate agreement. Instead, both parties must rely on due diligence and the representations and warranties of the other party. A further issue, if the acquirer is not public, is that the exchange may not directly accomplish the likely objective of investors in the venture, which is to harvest their investments. For VCs and similar investors, exchange of nonpublic shares for other nonpublic shares is helpful only if the resulting entity is better able to undertake a public offering in the future, or if the acquirer is willing to purchase the holdings of the investor for cash, even though the entrepreneur and members of the management team may receive shares. Exchanges of stock for stock have some advantages over cash transactions. An acquirer who wants to expand rapidly may be able to consummate acquisitions more easily by using stock. Another advantage is that, to the extent the share values move together, exchanging shares can address the moving target problem that arises when using cash for shares. Negotiating and completing an acquisition can take several months, during which the values can change. If the value changes are driven by a common factor such as an overall increase in stock prices, the exchange ratio between target and acquirer shares may be stable. Also, in some cases exchanging shares enables the seller to postpone recognizing a capital gain, which adds value to the exchange.
After the Acquisition In some transactions, the acquirer would like to retain members of the venture’s management team. Employment agreements protect the acquirer from
522
Chapter Thirteen
loss of value resulting from resignations by team members. The acquirer may also seek to align the incentives of team members with the interests of the acquirer. Approaches include asking team members to accept restricted shares of the acquirer (so their ability to trade and defeat incentive alignment is limited), compensating team members with equity claims or options on the acquiring firm, or offering incentive pay. The acquirer may also be concerned that once the venture is sold, team members will decide to reenter the market as competitors. Accordingly, sellers are sometimes asked to accept noncompete agreements and incentive compensation that increases the return to selling founders for the success of the enterprise that results from their continuing involvement.
Agreeing to Disagree When sellers are more optimistic about the venture than is the acquirer, reaching agreement on fixed terms of exchange is difficult. The parties can try to address the difference by devoting resources to due diligence or by strengthening representations and warranties. The parties also can use an earn-out provision to make the acquisition price contingent on its future performance. Compensation to sellers can be adjusted whether or not they remain with the venture. If a seller is no longer with the venture, contractual compensation can be adjusted based on performance. If the seller has stayed on as an employee, then the employee’s bonuses can be tied to firm performance.
13.3 Valuing Private Transactions In recent years, many ventures have elected to remain private or have been acquired by nonpublic companies. As of late 2014, Forbes reports that there were 49 private companies valued at more than $1 billion (the so-called “unicorns”) and that 35 of those had raised venture capital during 2014 at valuations of $1 billion or more. Airbnb and Dropbox both raised capital at valuations over $10 billion.42 As of this writing, such well-known companies as Airbnb and Uber remain private. Increasingly, private markets are finding ways to provide liquidity to large firms without the costs and regulatory concerns of being public. Late-stage private tech companies are providing liquidity to early investors by privately placing equity with accredited investors and institutions and using the proceeds of that placement to buy back shares from existing shareholders. Hedge funds
Harvesting 523
and institutions are investing in private firms. In October 2017, Google’s parent (Alphabet) led a $1 billion private funding round.43 One caveat about these high private company valuations, as discussed in Chapter 4, is that post-money valuations from private transactions are based on the prices paid for the most recently issued shares, which are often preferred shares with important sweeteners relative to founder shares and earlier-round shares. Thus, valuations inferred in this way can substantially overstate fundamental value. Many of the companies that raised capital with unicorn valuations did so by issuing such preferred shares.44 A study by Gornall and Strebulaev (2017) of 135 unicorns finds that, on average, valuations based on pricing of the last round are more than 50% higher than estimated fair value. They estimate that pricing based on the last round of preferred shares results in overvaluation of common shares by 58% on average. Even adjusting for this distortion, however, these companies would still have valuations that could easily support an IPO but they elect to remain private. How much is a venture worth in a private transaction? The answer depends on a myriad of factors, but the overriding considerations are its value in the next best alternative use (e.g. acquisition by a public company) and costs of the transaction. We consider the transactions costs first, followed by opportunity cost. Costs of the transaction. As discussed earlier, the cost of raising capital in the public market can be substantial. The costs of private transactions are of a similar nature but less directly observable. For a private transaction, we usually know only the net proceeds the sellers receive; hence, there is no convenient measure of transactions costs. We can gain perspective on the transactions costs of private sales by examining the pricing of private equity sold by companies that have public equity outstanding. A study of equity private placements by Hertzel and Smith (1993) that includes data for a sample of small public companies found that, on average, shares of private transactions were priced 20.1% below the contemporaneous public market value. Discounts were larger for smaller firms, when small amounts of proceeds were raised, and when a placement was large relative to firm size. Discounts were also larger when uncertainty about value was high, such as when a firm was involved in development of a speculative product or was threatened with financial distress. With the growth of the private market, it is unclear whether such large discounts will continue to be found. Assuming that private value is similar to public market value, the discounts estimated by Hertzel and Smith are estimates of the purchaser’s due-diligence cost. Private transaction prices may also reflect a discount for illiquidity;
524
Chapter Thirteen
however, illiquidity discounts generally are small.45 Comparing these results with the total issue cost for IPOs, including underpricing, shows that the average transactions costs for private sales are somewhat lower. Opportunity cost. The value of a venture to an acquirer depends on its stand-alone value, plus the value of any synergistic gains, less transactions costs. For a private venture that is acquired by a public firm, some value can possibly be created by transfer of ownership from underdiversified investors to the capital market, though recent evidence indicates that the net value of being public is often small. Whether sellers can capture any of the synergistic gains or the gains from movement to a public market depends on comparison of opportunities. If the venture is unique and multiple public acquirers are competing to acquire it, then sellers should realize most of the full public market value, including synergies, less the buyer’s transactions costs. Conversely, if there is only one prospective buyer and there are no good private market alternatives, and if other ventures would be similarly attractive to the buyer and would substitute for each other, the buyer should be able to capture most of the value that is created. These are the extremes. In other cases, the gains are likely to be shared on the basis of availability of substitute buyers and sellers.46 It follows that a reasonable approach to valuation is to estimate public market value assuming gains are captured by the seller and estimate stand-alone private market value. These are the bounds on value before subtracting the buyer’s transaction costs. The value can be estimated by considering how the parties are likely to share the gains. Noncash transactions. How should the entrepreneur and other sellers value their claims if the venture is sold for claims other than cash? Clearly, if freely tradable stock is received, then except for tax considerations there is little difference between receiving stock and receiving cash. However, the stock is often restricted from trading for a significant period. Sometimes, the restriction is designed to align the entrepreneur’s incentives with the acquirer’s interests. Restricting resale is also a way of making sure the entrepreneur’s gain is related to the acquirer’s subsequent performance (including the acquirer’s interest in the venture). Stock that is not freely tradable leaves the entrepreneur with a financial position that is still underdiversified. This form of compensation is less valuable than if the entrepreneur were to receive cash or tradable shares. The methods
Harvesting 525
developed in earlier chapters can be used to assess the impact of trading restrictions on value. For a diversified investor, such as a VC firm, it is reasonable to value the shares as if they were tradable.
Illustration: Estimating the Cost of Private Transactions It is possible, using comparable public firms or other data, to estimate the value of a venture as if it were public. Assuming that the acquirer is public and values the acquisition on the basis of comparables, the acquisition price can be estimated by deducting the expected discount to reflect the cost of the private transaction: Acquisition Price = Public Company Market Value − Private Transaction Discount Table 13.2 summarizes some of the statistical evidence of private transaction discounts. Consider a biotechnology venture that has begun to generate revenues but is not yet profitable and for which there is significant potential that the firm will Tab le 13 . 2 Private placement discounts compared to equity market value
Variable
Mean Discount Proceeds (Millions)
< $1.0 $1.0–$5.0 $5.0–$10.0 $10.0–$20.0 > $20.0
43.7% 33.1% 15.1% 10.1% 0.2%
Variable
Mean Discount
Market Value of Equity (Millions)
< $10.0 $10.0–$25.0 $25.0–$75.0 > $75.0
34.60% 35.60% 17.20% 7.60%
Single Investor Book-to-Market Equity Ratio