Handbook of Financial Decision Making (Research Handbooks in Money and Finance series) 1802204164, 9781802204162

Citation preview

HANDBOOK OF FINANCIAL DECISION MAKING

M10794_HILARY&MCLEAN_v1.indd 1

18/07/23 5:24 PM

RESEARCH HANDBOOKS IN MONEY AND FINANCE The Research Handbooks in Money and Finance series presents a thorough analysis of recent scholarly developments in monetary and financial economics, forming an essential, authoritative and comprehensive reference guide to the field. Edited by esteemed international scholars, these Handbooks contain a wide range of specially commissioned chapters covering the latest advances and research, and aim to be prestigious, high-quality works of lasting significance. Each Handbook consists of original contributions by an international team of scholars, and contributes to both the expansion of current debates and the development of future research. For a full list of Edward Elgar published titles, including the titles in this series, visit our website at www.e-elgar.com.

M10794_HILARY&MCLEAN_v1.indd 2

18/07/23 5:24 PM

Handbook of Financial Decision Making

Edited by

Gilles Hilary Houston Professor of Accounting, McDonough School of Business, Georgetown University, USA

David McLean William G. Droms Professor of Finance, McDonough School of Business, Georgetown University, USA

RESEARCH HANDBOOKS IN MONEY AND FINANCE

Cheltenham, UK • Northampton, MA, USA

M10794_HILARY&MCLEAN_v1.indd 3

18/07/23 5:24 PM

© Gilles Hilary and David McLean 2023 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical or photocopying, recording, or otherwise without the prior permission of the publisher. Published by Edward Elgar Publishing Limited The Lypiatts 15 Lansdown Road Cheltenham Glos GL50 2JA UK Edward Elgar Publishing, Inc. William Pratt House 9 Dewey Court Northampton Massachusetts 01060 USA A catalogue record for this book is available from the British Library Library of Congress Control Number: 2023939588 This book is available electronically in the Economics subject collection http://dx.doi.org/10.4337/9781802204179

ISBN 978 1 80220 416 2 (cased) ISBN 978 1 80220 417 9 (eBook) Typeset by Cheshire Typesetting Ltd, Cuddington, Cheshire

EEP BoX M10794_HILARY&MCLEAN_v1.indd 4

18/07/23 5:24 PM

Contents vii xii

List of contributors Acknowledgements PART I  INTRODUCTION Financial decision making: an overview Gilles Hilary and David McLean

2

PART II NATURAL AND ENVIRONMENTAL FACTORS THAT IMPACT FINANCIAL DECISION MAKING Part II.1  Natural Factors   1 Limited attention and financial decision-making Alexander Nekrasov, Siew Hong Teoh, and Shijia Wu

17

  2 Seasonality in stock returns and government bond returns Mark J. Kamstra and Lisa A. Kramer

36

  3 Preference for lottery-like securities Turan G. Bali and Quan Wen

63

 4 Neurofinance Elise Payzan-LeNestour

91

Part II.2  Environmental, Social, and Cultural Factors   5 Corporate culture: a review and directions for future research Jillian Grennan and Kai Li

112

  6 Geography and financial decision making Qinghai Wang

133

  7 Language in financial disclosures Natasha Bernhardt, Mandy T. Ellison, Kristina M. Rennekamp, and Brian J. White

154

  8 The impact of word-of-mouth communication on investors’ decisions and asset prices Byoung-Hyoun Hwang   9 The role of media in financial decision-making Kenneth R. Ahern and Joel Peress

171 192

v

M10794_HILARY&MCLEAN_v1.indd 5

18/07/23 5:24 PM

vi  Handbook of financial decision making

PART III  INSTITUTIONS, FRAMEWORKS, AND TOOLS Part III.1  Institutions 10 Disclosure regulation: past, present, and future S.P. Kothari, Liandong Zhang, and Luo Zuo

215

11 The audit in a modern economy W. Robert Knechel and Eddie Thomas

235

Part III.2  Frameworks 12 Accounting and prices Steven J. Monahan

256

13 Managerial accounting and decision-making Satish Joshi and Ranjani Krishnan

277

14 Social responsibility in business and finance Hao Liang and Tran Bao Phuong Nguyen

296

Part III.3  Tools: Computer-Based Advising 15 Artificial intelligence in financial decision-making Allen H. Huang and Haifeng You

315

16 IT meets finance: financial decision-making in the digital era Francesco D’Acunto and Alberto G. Rossi

336

PART IV  SETTINGS: ADVISORS AND DECISION MAKING 17 Financial analysts Daniel Bradley

356

18 Household financial decision making Sumit Agarwal and Nithin Mannil

375

19 Behavioral finance and retirement planning in defined contribution plans Julie Agnew

411

Index

432

M10794_HILARY&MCLEAN_v1.indd 6

18/07/23 5:24 PM

Contributors Sumit Agarwal is Low Tuck Kwong Distinguished Professor of Finance at the Business School and Professor of Economics and Real Estate at the National University of Singapore. Julie Agnew is the Richard C. Kraemer Professor of Business at the Raymond A. Mason School of Business at William & Mary, Williamsburg, USA. Kenneth R. Ahern is Associate Professor of Finance and Business Economics at the University of Southern California and Research Associate at the National Bureau of Economic Research (NBER). His research is distinguished by the examination of networks to study how economic outcomes spread through interactions, including among industries, individuals, and information sharing. His research has been published in leading finance and economics journals, and has been cited in both the popular press and legislative hearings around the world. Turan G. Bali is the Robert S. Parker Chair Professor of Business Administration at the McDonough School of Business at Georgetown University, Washington, DC, specializing in asset pricing, risk management, fixed income securities, and financial derivatives. Professor Bali has published three books and more than 50 articles in economics and finance journals, and also serves as an Associate Editor of several prestigious journals in his field. He has won several awards, including the Q Group’s Jack Treynor Prize in Finance. Natasha Bernhardt is an accounting PhD student in the Samuel Curtis Johnson Graduate School of Management, SC Johnson College of Business at Cornell University, New York. Daniel Bradley is the Lykes Professor of Finance and Sustainability at the Kate Tiedemann School of Business and Finance at the University of South Florida, USA. Francesco D’Acunto is James A. Clark Chair and Associate Professor of Finance at Georgetown University, Washington, DC. His research interests include beliefs and decision-making, inequalities, and FinTech. His work has been published in top ­academic  journals such as the Journal of Political Economy, the Review of Economic Studies, the Review of Financial Studies, the Journal of Financial Economics, Proceedings of the National Academy of Sciences (PNAS), and the Journal of Economic Perspectives. Mandy T. Ellison is an accounting PhD student in the McCombs School of Business at the University of Texas at Austin, and a Certified Public Accountant. vii

M10794_HILARY&MCLEAN_v1.indd 7

18/07/23 5:24 PM

viii  Handbook of financial decision making

Jillian Grennan is an Associate Professor of Finance in the Leavey School of Business, an Associate Professor of Law, and a Faculty Scholar in the Markkula Center for Applied Ethics at Santa Clara University (USA). She is also a Visiting Professor in the Institute for Business and Social Impact at the University of California Berkeley’s Haas School of Business. Gilles Hilary is a Chaired Professor at Georgetown University, Washington, DC, and a former Chaired Professor at INSEAD. He is a founding member of Cercle-K2, a French think-tank on risk management, a Senior Fellow at the Asian Bureau of Finance and Economic Research, and a Research Fellow at the French Military Police Academy (CREOGN). Allen H. Huang is an Associate Professor of Accounting at the School of Business and Management at Hong Kong University of Science and Technology. He is also the Associate Dean of the School of Business and Management, the Associate Director of the Center of Business and Social Analytics, and the Faculty Associate of the Institute of Emerging Market Studies at HKUST. Byoung-Hyoun Hwang is the Provost’s Chair in Finance and an Associate Professor in the Nanyang Business School at Nanyang Technological University (NTU), Singapore. Prior to joining NTU, he was an Associate Professor of Finance at the SC Johnson College of Business at Cornell University, USA. His main research areas are in empirical asset pricing, behavioral finance, and social finance. Satish Joshi is a Professor in the Department of Agricultural, Food and Resource Economics at Michigan State University, USA. Mark J. Kamstra is Professor of Finance at York University’s Schulich School of Business in Toronto, Canada. His research revolves around topics in behavioral finance and empirical asset pricing, including time-varying risk premia related to links between human sentiment and financial risk tolerance, rooted in the medical, psychology, and economics literatures. Other interests include investor attention and social interaction, statistical modeling, and machine learning. W. Robert Knechel, PhD, is Distinguished Professor and the Frederick E. Fisher Eminent Scholar in Accounting at the University of Florida, Gainesville, USA. He is currently the Director of the International Accounting and Auditing Center (IAAC) located within the Fisher School of Accounting. Robert also holds an appointment as a Professor of Accounting Research at the University of Auckland, New Zealand. S.P. Kothari is Gordon Y Billard Professor of Accounting and Finance at MIT’s Sloan School of Management, having previously served there as Deputy Dean. Most recently, he was Chief Economist and Director of the Division of Economic and Risk Analysis at the US Securities and Exchange Commission (SEC). Kothari has been global head of equity research for Barclays Global Investors and director of the Bombay Stock Exchange. He is currently a director of Velan Studios and EbixCash. From 1997 to 2019 he edited

M10794_HILARY&MCLEAN_v1.indd 8

18/07/23 5:24 PM

Contributors  ­ix

the world-renowned Journal of Accounting and Economics. Kothari has received India’s Padma Shri civilian award, and holds Honorary Doctorates from the London Business School, the University of Cyprus, and the University of Technology Sydney. Lisa A. Kramer is Professor of Finance at the University of Toronto, Canada, including affiliations with its Department of Economics and Rotman School of Management (where she is a Research Fellow of Behavioral Economics in Action), as well as with the University of Toronto Mississauga. She conducts interdisciplinary work in the field of behavioral finance, blending tools from psychology and economics (such as surveys, experiments, and statistical analysis) to better understand markets and financial decisions. Ranjani Krishnan is Ernest W. and Robert W. Schaberg Chair in Accounting, and a professor in the Department of Accounting and Information Systems at Michigan State University, USA. She was also a visiting professor at Harvard Business School and Georgetown University, Washington, DC. Kai Li holds the Canada Research Chair in Corporate Governance and the W. Maurice Young Endowed Chair in Finance at the Sauder School of Business, University of British Columbia (UBC). She is also a Fellow of the Royal Society of Canada, and was Senior Associate Dean, Equity and Diversity at UBC Sauder between 2015 and 2021. She received her PhD in Economics from the University of Toronto. Hao Liang is an Associate Professor of Finance and the Co-Director of the Singapore Green Finance Centre at Singapore Management University (SMU), where he also held the BNP Paribas Fellowship, the DBS Sustainability Fellowship, and the Lee Kong Chian Fellowship. He received his PhD in Finance from Tilburg University (Netherlands). He is the Section Editor (Finance and Business Ethics) of the Journal of Business Ethics, and is on the editorial boards of Asia-Pacific Journal of Financial Studies, British Accounting Review, and Strategic Management Journal. Nithin Mannil is a finance doctoral student at the London School of Economics and Political Science (LSE). He was previously a research associate at the Indian School of Business. David McLean is the William G. Droms Professor of Finance at Georgetown University. His teaching and research interests are in capital market imperfections and their ramifications for both asset prices and corporate finance. His research has won several awards, including the Amundi Smith Breeden Award for the best paper in the Journal of Finance and the Jensen Prize for the best paper in the Journal of Financial Economics. He serves on the editorial boards of several academic journals, including the Journal of Financial and Quantitative Analysis and Management Science. Steven J. Monahan is an Associate Professor of Accounting at the University of Utah, David Eccles School of Business, USA. His research and teaching focuses on two related issues: the role of accounting and non-accounting information in the fundamental ­analysis process; and the economic causes and consequences of alternative disclosure

M10794_HILARY&MCLEAN_v1.indd 9

18/07/23 5:24 PM

x  Handbook of financial decision making

policies/regimes. He has published articles in leading academic journals. He is serving or has served on numerous editorial boards and is a former Associate Editor of the European Accounting Review and a current Associate Editor of the Journal of Accounting Auditing and Finance. Alexander Nekrasov is an Associate Professor of Accounting at the University of Illinois, Chicago, USA. His broad research interests include limited attention, capital markets, financial reporting and disclosure, financial analysts, and labor. His work has been published in leading academic journals, has featured in the Wall Street Journal, and won the 2010 Review of Accounting Studies Conference Best Paper Award. Tran Bao Phuong Nguyen is a Postdoctoral Research Fellow at Lee Kong Chian School of Business, Singapore Management University (SMU), from where she obtained her PhD in Economics. Her research interests include international trade, international economics, corporate finance, and sustainability. Elise Payzan-LeNestour is a Scientia Associate Professor in Finance at the University of New South Wales Business School in Sydney, Australia. Joel Peress is Professor of Finance and the Claude Janssen Chair in Business at the European Institute of Business Administration (INSEAD) in France. His theoretical and empirical research focuses on the generation and diffusion of information in financial markets, with applications to asset management, financial disclosures, the media, and economic growth. His work has been published in leading journals and has received several awards. Joel served as Editor of The Review of Finance. Kristina M. Rennekamp is a Professor of Accounting in the Samuel Curtis Johnson Graduate School of Management, SC Johnson College of Business at Cornell University, New York, and a Chartered Financial Analyst. Alberto G. Rossi is a Professor of Finance at the McDonough School of Business, Georgetown University, Washington, DC, where he is also Director of the AI, Analytics, and Future of Work Initiative. His research interests include FinTech, household finance, machine learning, and asset pricing. Professor Rossi’s work has been published in leading academic journals such as the Journal of Finance, the Review of Financial Studies, and the Journal of Financial Economics and Management Science. Siew Hong Teoh is the Lee and Seymour Graff Endowed Professor of Accounting at the University of California, Los Angeles (UCLA) Anderson School of Management, where she studies earnings management, accounting-based anomalies, and behavioral issues about information disclosure in the capital markets. Her more than 50 articles published  in leading journals in accounting, finance and economics are widely cited and profiled in the media. Her recent research examines visual information for attention effects on the capital markets, and personality trait impressions from the faces of market participants.

M10794_HILARY&MCLEAN_v1.indd 10

18/07/23 5:24 PM

Contributors  ­xi

Eddie Thomas, PhD, is an Associate Professor of Accounting at Georgia College & State University, Milledgeville, USA. He was previously an auditor at Butler, Williams & Wyche, LLP, a Certified Public Accountancy firm in Macon, Georgia, and a tenured Professor of Philosophy at Mercer University, Macon. Qinghai Wang is Professor of Finance and the Richard T. Crotty Endowed Chair in Finance in the College of Business Administration at the University of Central Florida, USA. Quan Wen is the Associate Professor of Finance at the McDonough School of Business at Georgetown University, Washington, DC. He received his PhD in Finance from Emory University’s Goizueta School of Business, USA. His research interests are empirical asset pricing (with a particular focus in understanding what drives asset returns), risk-return trade-off, institutional investors, and frictions in financial markets and their implications for market efficiency. Brian J. White is an Associate Professor of Accounting in the Samuel Curtis Johnson Graduate School of Management, SC Johnson College of Business at Cornell University, New York, and a Certified Public Accountant. Shijia Wu is an Assistant Professor in Accounting at the Chinese University of Hong Kong, Shenzhen. Her primary research interest is understanding the fundamental changes in information acquisition and dissemination in the digital age. Specifically, she studies the visual presentation of the information, social media as the information intermediary, and earnings management. She earned her PhD in Accounting from the University of California Irvine, and her research has been published in the Review of Accounting Studies and the Journal of Financial Reporting. Haifeng You is a Professor of Accounting at the School of Business and Management at Hong Kong University of Science and Technology (HKUST), where he is also a Co-Director of the Center for Securities Analysis with Financial Technology. Liandong Zhang is Lee Kong Chian Professor of Accounting and Associate Dean of Research at the School of Accountancy of Singapore Management University. He currently serves as an Associate Editor of Asia-Pacific Journal of Accounting and Economics and an editor of Corporate Governance: An International Review. Luo Zuo is a Professor of Accounting at Cornell University’s Samuel Curtis Johnson Graduate School of Management, New York, where he currently serves as the Faculty Director of the Cornell–Tsinghua Finance MBA Program and the Faculty Lead of the Cornell–HKUST Partnership. Professor Zuo is the President of the Chinese Accounting Professors’ Association of North America, an Editor of The Accounting Review, an Associate Editor of Management Science and the Journal of Accounting and Economics, and an editorial board member of Review of Accounting Studies.

M10794_HILARY&MCLEAN_v1.indd 11

18/07/23 5:24 PM

Acknowledgements We acknowledge and thank Daniel Mather, Senior Editor at Edward Elgar Publishing, for his continuous assistance in the production of this Handbook. We also thank the numerous authors who contributed their time and knowledge to this project.

xii

M10794_HILARY&MCLEAN_v1.indd 12

18/07/23 5:24 PM

PART I INTRODUCTION

M10794_HILARY&MCLEAN_v1.indd 1

18/07/23 5:24 PM

Financial decision making: an overview Gilles Hilary and David McLean

1 INTRODUCTION Financial decision making is an increasingly complex subject. Academic research on this topic has been expanding rapidly over the last half-century. For this Handbook, we invite leading researchers in accounting and finance to summarize 19 topics that are at the forefront of financial decision-making research. We hope that these chapters can serve both as entry points for persons interested in learning about a topic and as points of reference for researchers working on these and related topics. In this introductory chapter, we begin by offering a brief chronology of how the last 50 years of academic research have led us to where we are today. We then summarize the various sections and chapters in this Handbook, which we organize into three sections based on three broad themes. The first theme deals with the natural and environmental factors that impact financial decision making. Several chapters consider behavioral factors that are grounded in psychology and neuroscience, while other chapters focus on the social and cultural elements that affect financial decision making. The second theme is concerned with institutions, frameworks, and tools created for the purpose of aiding financial decision making. The chapters in this section cover topics such as regulation, mental models for financial decision making, and new data-driven tools such as artificial intelligence. The third theme is concerned with enabling financial decision making. This last section offers chapters on financial advisors and household finance. We close this introductory chapter by offering our opinions on where opportunities lie for future research on financial decision making.

2  A BRIEF HISTORY OF FINANCIAL DECISION MAKING In the first part of the twentieth century, financial decision making was largely done on an ad hoc basis that combined the benefit of experience with some institutional accounting and law knowledge. Financial markets were largely unregulated, at least until the Great Depression. A large change occurred when people started to incorporate the insights of formal economics into financial management theory. For example, in 1952, Harry Markowitz built on the work of von Neumann and Morgenstern (1947) and Savage (1954) on utility functions. In doing so, he became a pioneer of what would be called “modern finance”. In his seminal publication Markowitz postulated a world in which a representative investor optimizes a portfolio of risky assets to obtain the highest possible expected return for a given level of risk. Other scholars followed his work and deepened the formalization of financial decision making by using formal equilibrium conditions. William Sharpe’s capital asset pricing model (CAPM) is a prominent example of this (Sharpe (1964)). 2

M10794_HILARY&MCLEAN_v1.indd 2

18/07/23 5:24 PM

Financial decision making: an overview  ­3

Eugene Fama developed and tested the notion of market efficiency, the idea that information should be incorporated into asset prices both quickly and accurately (Fama (1965a, b)). Market efficiency is consistent with the rational expectation framework developed by John Muth (1961). In equilibrium, forecast outcomes do not differ systematically from actual results. The rational expectations framework assumes that representative agents can costlessly acquire and process all relevant information in an unbiased fashion. The efficient market hypothesis was initially very successful at matching empirical data, slowly but profoundly affecting practice. A key prediction from this line of research is that passive investment should dominate active investment. Many studies show that very few (if any) funds can outperform the market on a risk-adjusted basis (Jensen (1968, 1969)). Over time, index funds become a primary investment vehicle for individual investors and institutions. The neoclassical view of the world was then extended to financial decision making in corporate finance. On the investment side, this approach characterized the optimal capital level by considering the relative prices of production factors (Jorgenson (1963)). New investment is determined by the comparison between the marginal product of capital and its marginal cost. Aside from installation costs, these optimal results are achieved without much friction. In contrast to this traditional neoclassical theory, James Tobin (1969) (and Nicholas Kaldor before him) proposed a popular alternative view that links corporate decisions and financial markets: q-investment theory explains investment using the difference between the stock market valuation of firms’ real assets and their replacement costs. In this context, investment rates fluctuate with changes in stock market valuation. While the initial neoclassical approach focuses on the optimal level of capital, q-theory focuses on the optimal adjustment rate. Hayashi (1982) justified the measurement of marginal q with a valuation ratio, average q, suggesting that a simple regression of investment on Tobin’s q should have a strong fit. On the financing side, Modigliani and Miller (1958) showed the irrelevance of financing decisions in a world without friction. The value of a firm is independent of how it is financed; and, conversely, the cost of capital is also independent of financing sources (a second paper, Miller and Modigliani (1961), argued the irrelevance of dividend policies). These conclusions had to be caveated by the existence of taxes and bankruptcy costs but were still striking at the time. The trade-off theory of capital structure directly emerged from these features by asserting that corporate leverage is determined by balancing the tax-saving benefits of debt against the dead-weight costs of bankruptcy (see Ai et al. (2021) for a recent review of this literature). Although the neoclassical framework has been criticized for its extreme stylization of reality, this paradigm yielded useful insights into financial decision making. However, the story did not end there. In the 1970s, researchers became interested in the role of information constraints, an element ignored by earlier research. For example, a key element in Miller and Modigliani’s proposition is the inexistence of information asymmetry, which occurs when one party to an economic transaction (typically but not always the seller) possesses greater material knowledge than the other party. A series of developments in economics brought information asymmetry to the core of financial decision making by showing the existence of two frictions generated by this asymmetry. The first, adverse selection, was introduced by Akerlof (1970). Adverse selection involves a situation that occurs in markets where buyers cannot ascertain the quality of the products sold before acting (that is, buying). When knowing the presence of

M10794_HILARY&MCLEAN_v1.indd 3

18/07/23 5:24 PM

4  Handbook of financial decision making

low-quality items, buyers refuse to pay the full price for high-quality products, chasing them away from the market and leaving only low-quality goods. The second friction is the existence of moral hazard, a situation where one party cannot ascertain the actions of a second party after contracting.1 One example is the principal–agent problem, where employee efforts may remain hidden from the employer. Information asymmetry in general, and these two mechanisms in particular, ushered in several streams of research.2 In asset pricing, scholars began to study models in which some investors had better information than others. A common result in these models is an equilibrium in which markets are not fully efficient. The knowledge of informed investors does not fully impound into prices. An important example of this effect is provided by Grossman and Stiglitz (1980), who argue that informationally efficient markets are not possible, pointing out that if prices reflect all available information, then it is not profitable to gather information. In this case, people should not spend resources gathering information, creating a paradox for market efficiency. Investors need to gather information to make prices efficient; but if prices are already efficient, then investors will not gather information. Other early papers on asset pricing and information asymmetries include Hellwig (1980) and Admati (1985). A more recent paper by Kelly and Ljungvist (2012) uses brokerage firm closures and restructurings as shocks to the information environment, finding evidence that is consistent with the common predictions in these models, which is that prices and demand from uninformed investors both fall when information asymmetries among investors increase. Derrien and Kecskés (2013) show the real consequences of these shocks, in that firms raise less capital and invest less. In corporate finance, the literature stressed the importance of conflicts of interest between shareholders and managers. Jensen and Meckling (1976) show how agency problems can impact firm value, but that managerial ownership can mitigate such problems and increase firm value. Jensen (1986) suggests that managers may overinvest (compared to the optimal point for shareholders) to increase their compensation. Jensen (1986) also shows that increasing the amount of debt in the capital structure mitigates this issue. Myers and Majluf (1984) develop the “Pecking Order Theory”, which posits that managers adopt a financial policy that minimizes the costs associated with adverse selection. In the pecking order framework, managers prefer internal financing over external financing, and issuing debt over issuing equity. Fazzari et al. (1988) use the idea of costly external finance to challenge classical q-­theory, showing that not only does q fail to explain investment, but also that investment is sensitive to the firm’s internally generated cash flow. The importance of cash flows in explaining investment suggests a wedge between the costs of internal and external funds, where external funds are more costly. This finding is consistent with Myers and Majluf (1984) and conflicts with classical q-theory, which assumes a similar cost for internal and external funds. Many subsequent papers debate this finding, with several pointing out that q is measured with error and that if that error is corrected, then q exhibits greater explanatory power for investment, and the cash flow effect is muted (e.g., Erickson and Whited (2000), Peters and Taylor (2017)). The neoclassical-rational expectation type of framework does not leave much room for research in accounting, which is basically assumed away. Somehow, everyone has full and complete information. How information is created and subsequently disseminated, and why anyone trusts it, is not given much thought or attention, even though the answering

M10794_HILARY&MCLEAN_v1.indd 4

18/07/23 5:24 PM

Financial decision making: an overview  ­5

of these questions is not straightforward. Yet Ball and Brown (1968) had shown that accounting is important for equity prices, and that financial markets may not process accounting information effectively (Bernard and Thomas (1989) subsequently identified the “Post Earnings Announcement Drift” that characterizes a systematic delay in accounting information). Further research shows that disclosure quality impacts, among other things, the cost of capital (see Botosan (2006) for a review), corporate investment (Biddle et al. (2009)), and capital structure choices (Chang et al. (2009)). In the late 1990s, organizational governance (see Shleifer and Vishny (1997) for an initial review and Di Vito and Trottier (2022) for a more recent review), financial disclosure regulations (particularly accounting frameworks), and national institutions (see Zattoni et al. (2020) for a review) are shown to have an impact on financial decision making. In addition to considering informational issues, the literature has extended the neoclassical framework to incorporate behavioral elements; but, to some extent, this is old news. For example, Keynes (1936) used psychological insights in general theory (e.g., animal spirits), while Nobel Prize winner Herbert Simon introduced the term “bounded rationality” in the 1950s (Simon (1957)). However, the work of Daniel Kahneman and Amos Tversky moved this approach to the center of financial theory (see Kahneman (2011) for a review). Essentially, this literature suggests that cognitive biases (“heuristics”) influence people’s decisions and are important in aggregate markets. In contrast, other scholars conclude that even if people make mistakes in judgment, these errors are likely to be random and cancel out. To the extent these errors are nonrandom, they should be arbitraged away by rational traders in financial markets (e.g., Friedman (1953, pp. 3–43) and Fama (1965a)). As such, the study of such errors is of limited interest to financial economists. In the 1990s, research began to show that financial markets could be affected by errors if (i) errors are correlated, and (ii) arbitrage is risky and costly. In classical finance, arbitrage is assumed to be a riskless and costless transaction. In the real world, virtually nothing is riskless and without cost. Researchers began to incorporate these facts into their models, showing that they are important for equilibrium outcomes (see Gromb and Vayanos (2010) for a review). De Long et al. (1990a) showed that noise traders, irrational investors who trade on noise as if it were relevant information, made arbitrage risky and therefore limited. The presence of noise traders can thus mean that markets are not fully efficient. De Long et al. (1990b) showed that, in some settings, arbitrageurs find it optimal to trade with, instead of against, irrational investors, making mispricing greater, not smaller, and markets less efficient. These studies led to a large body of behavioral finance studies in which empirical irregularities are explained by a combination of investor irrationality and limited arbitrage (e.g., Hirshleifer (2001) and Barberis and Thaler (2003)). One early paper in this genre, Pontiff (1996), showed that idiosyncratic risk deters arbitrage and is important for explaining the puzzle of closed-end fund mispricing, building on Treynor and Black’s (1973) overlooked model, which solved for arbitrageurs’ optimal portfolio weight in a mean-variance setting. An arbitrageur who is chasing abnormal returns may decide not to diversify, and instead make a single asset a large part of his or her portfolio. In this case, the idiosyncratic volatility of the asset, which cannot be hedged, impacts the volatility of the portfolio. Rational arbitrageurs realize this situation and trade off the benefit of a higher alpha with the cost of higher volatility, which results in assets with high idiosyncratic risk receiving lower portfolio weights and maintaining larger mispricing. A large stream of empirical literature on capital market anomalies

M10794_HILARY&MCLEAN_v1.indd 5

18/07/23 5:24 PM

6  Handbook of financial decision making

shows that idiosyncratic risk plays a central role in explaining many anomalies (see Pontiff (2006) for a review). Researchers do, however, continue to go back and forth on how best to explain a long list of empirical anomalies, such as stock return momentum (Jegadeesh and Titman (1993)) and the accrual anomaly (Sloan (1996)). The competing frameworks include behavioral finance, rational expectations combined with frictions such as liquidity, and statistical artifacts resulting from data mining and empirical design. These competing frameworks are not mutually exclusive, and there is evidence that each may play some role (see Fama (1998), Chordia et al. (2014), Harvey et al. (2016), and McLean and Pontiff (2016)). Corporate finance scholars also began to incorporate behavioral finance and the idea of irrationality on the part of either managers or investors to impact how firms make decisions. Minsky (1977) and Kindleberger (1978) provided early insights into this idea, concluding that excess speculation can cause rapid credit expansion, bubbles in asset prices, and overinvestment. Since then, a large body of literature on corporate finance has built on the idea that markets are not totally efficient and that investors and managers are not fully rational. Baker and Wurgler (2011) reviewed the behavioral corporate finance literature and broke it into two segments. The first segment assumes that the managers of firms are rational but that investors are not. In this case, managers may cater to investors’ beliefs or time markets, and choose to issue or repurchase securities and invest based on mispricing. Such behavioral studies also weigh in on the debates surrounding q-theory of investment. Behavioral studies reason that investor sentiment can impact stock prices and thus alter the cost of external finance, and that this should be reflected in both q and investment (e.g., Morck et al. (1990), Baker et al. (2003), and McLean and Zhao (2014)). The second segment of behavioral corporate finance studies the effects of irrational managers under the assumption that the markets in which they operate are efficient. One consistent theme in this research is that entrepreneurs and managers have optimism bias (e.g., Cooper et al. (1998)), a tendency to be overconfident (see Malmendier and Tate (2015) for a review), or both (e.g., Hilary et al. (2016)). Among others, Heaton (2002) developed models that study the actions of managers who are too optimistic about the firm’s assets and investment opportunities. These studies show that, in such cases, both underinvestment and overinvestment can arise. Similarly, accounting researchers have revived a long tradition of using experimental work to better understand individual decision making. By doing so, they have been able to obtain granular knowledge of what type of information is relevant to agents and the issues they face when making economic decisions (see Swieringa and Weick (1982) and Kachelmeier (2020) for reviews). At the same time, an archival stream of literature that considered, among other things, the effect of agent characteristics (e.g., gender and life experience) on decision making,3 the actual behavior of individual decision makers, or the effects of agents’ broader environment on their decisions (e.g., religiosity and trust) also started to emerge in the 2000s.4 Theories of economic regulation have also been affected. While the 1970s and 1980s saw a wave of deregulation broadly supported by a neoclassical view of markets, it was succeeded by the view that regulation, although not costless, could help with market failures (for instance, those caused by informational issues) and that costs and benefits should be balanced. For example, in 1993 President Clinton issued the Regulatory Planning and

M10794_HILARY&MCLEAN_v1.indd 6

18/07/23 5:24 PM

Financial decision making: an overview  ­7

Review executive order, which stated that “Federal agencies should promulgate only such regulations […] made necessary by compelling public need, such as material failures of private markets […]. In deciding whether and how to regulate, agencies should assess all costs and benefits” (Clinton, 1993). The development of behavioral economics also had an effect on regulation (e.g., Sunstein et al. (1998)). In particular, Thaler and Sunstein (2008) suggested that regulators use insights from behavioral research to “nudge” people to make prosocial choices.

3  THEMES AND CHAPTERS OF THE HANDBOOK Many of the themes discussed above are now mature and have been the subjects of welldeveloped reviews. For this Handbook, we solicited chapters that focus on more contemporary themes. We organize the chapters into three parts based on three broad topics: (i) the natural and environmental factors that impact financial decision making; (ii) the institutions, frameworks, and tools created to aid financial decision making; and (iii) ­advisors and household finance. 3.1  Natural and Environmental Factors 3.1.1  Natural factors Several chapters consider behavioral factors that are grounded in psychology and neuroscience. Alex Nekrasov, Siew Hong Teoh, and Sijia Wu contribute a chapter on limited attention. This chapter reviews both the theoretical and empirical literature. Investor attention is a limited resource. The central tenet of this research is that, when attention is limited, markets are less efficient. Prices do not immediately respond to new information and can drift post-announcement. Inattention is stronger among retail investors than among other types of investors; however, more sophisticated participants, such as sell-side analysts and institutional investors, can exhibit inattention as well. Firms are aware of and capitalize on investor inattention and choose disclosure strategies that increase attention to good news and reduce attention to bad news. In sum, investor inattention has important effects on financial markets. Mark Kamstra and Lisa Kramer contribute a chapter on seasonal affective disorder (SAD) and its impact on financial markets. The amount of daylight people receive can vary predictably across seasons. For some people, less daylight can result in depression, which is referred to as SAD. The finance research on SAD shows that investors demand greater risk premiums during seasons when there is less daylight. This effect is documented both with stock market index returns from multiple countries and with US Treasury bonds. Kamstra and Kramer also provide some new evidence of the SAD effect, showing that SAD has a stronger effect on small stocks than on large stocks. The authors also create a new SAD proxy based on Google searches for the term “seasonal affective disorder” and show that it strengthens the SAD effect in international stock markets. Turan Bali and Quan Wen study the preference for lottery-like securities or assets with a small probability of a large payoff. This literature shows that some investors show a strong preference for lottery-like stocks, which are typically defined as low-priced stocks, stocks with high idiosyncratic volatility or high idiosyncratic skewness, or stocks with very

M10794_HILARY&MCLEAN_v1.indd 7

18/07/23 5:24 PM

8  Handbook of financial decision making

high daily returns over the previous month. Stocks with lottery-like characteristics have lower returns, suggesting that investors’ preference for this trait is strong enough to lead to the overvaluation of these assets. Bali and Wen also contribute some novel results in their chapter, such as showing that the underperformance of lottery stocks is greater when institutions are selling and retail investors are buying lottery-like stocks. The emerging field of neurofinance takes what we know about decision making at the neurobiological level and uses that information to better understand how individuals make financial decisions. Elise Payzan-LeNestour contributes a chapter on this topic. One interesting fact is that the brain is a “Bayesian sampler”, evolved to perform natural sampling tasks that promote survival, such as where to forage for food. For this reason, the brain operates poorly in terms of elementary probability tasks, but can solve more complex probability tasks through sampling probability distributions. Practical implications emerge from this research. As an example, an investor who makes poor choices when presented with explicit statistics may make better choices if given information that is more easily processed by the brain. 3.1.2  Environmental factors A number of the authors offer chapters focused on social and cultural elements that affect financial decision making. Jillian Grennan and Kai Li define corporate culture as “an informal institution typified by patterns of behavior and reinforced by people, systems, and events”. As a motivation for the need to better understand corporate culture, Grennan and Li point out that there is a good deal of heterogeneity across firms in terms of productivity, and that culture can help explain this. Their chapter describes how advances in measurement techniques, such as surveys, proxies such as executives’ cultural heritage, computational linguistic models, and experiments are enabling researchers to better measure and study the effects of corporate culture. The chapter concludes by offering a rich set of topics that can be addressed by future research on corporate culture. One current topic is understanding how technology and remote working impact corporate culture. Qinghai Wang explains how geographic preference is an important factor in financial decision making, with a focus on the equity market. One pervasive finding is the “home bias”, in that investors overweight their portfolios toward domestic stocks and away from international stocks. However, even within domestic holdings, both retail and institutional investors overweight local stocks. This is puzzling, as domestic investing does not face the same costs and barriers as international investing. One explanation for this effect is that local investors may have advantages at acquiring information. The chapter also discusses stylized facts regarding the geographic distributions of firms, individual investors, and institutional money managers, and how differences among these actors can lead to various financial market frictions. The language that a firm uses to communicate its financial disclosures can also impact the recipient’s judgment and decisions. This aspect is the central theme of a chapter on language in financial disclosures, contributed by Natasha Bernhardt, Mandy Ellison, Kristina Rennekamp, and Brian White. There is a good deal of research focused on how investors are affected by linguistic choices regarding casual language, tone, and jargon. Looking ahead, however, the effect of stories, which reflect the way words are used to form a narrative, may play a greater role in financial disclosure research. The role of stories has been extensively studied in the fields of communications, marketing, and psychology, but

M10794_HILARY&MCLEAN_v1.indd 8

18/07/23 5:24 PM

Financial decision making: an overview  ­9

not in that of financial disclosure. The authors discuss how this could be a direction for future research on financial disclosure. A person’s surroundings can also impact financial decisions. Byoung-Hyoun Hwang develops this idea in a chapter on “social” or “word-of-mouth” finance. The notion that social interactions can impact investment decisions and asset prices has a long history, dating at least back to the South Sea Bubble of 1720. Social interactions may be even more important for investment today, as social media widens investors’ social networks and makes it easier for them to share information. The impact of word-of-mouth investing on market efficiency is a topic where more research is needed. One conclusion from Hwang’s chapter is that, on balance, word of mouth does not help investors make better decisions. Kenneth Ahern and Joel Peress study the role that financial media plays in financial markets. Their chapter identifies three media activities: circulating unedited information, selecting information for publication, and creating new information. A common message from studies on the effect of media is that it improves financial decision making, even if it can lead to herding and overreaction. Ahern and Peress explain how traditional media makes markets more efficient, can enable investors to earn higher returns, and allows firms to have lower costs. Interestingly, Hwang’s preceding chapter concludes that “word of mouth” or social network information does not lead to better financial decision making. 3.2  Institutions, Frameworks, and Tools 3.2.1 Institutions This Handbook contains two chapters concerned with regulation and how it is implemented. S.P. Kothari, Liandong Zhang, and Luo Zuo contribute a chapter on disclosure regulation. Disclosure is important because the accuracy and timeliness of financial information disclosure are critical for an efficient financial market, which in turn enables efficient resource allocation. The chapter focuses on three questions: Why do we need disclosure? Which theories best explain the current state of disclosure regulation? What are the economic consequences of disclosure regulation? The chapter also provides an overview of the current debate surrounding environmental, social, and governance (ESG) disclosure and regulation. Robert Knechel and Edward Thomas offer a chapter on the role of auditors. Without enforcement, regulation is meaningless. In this context, auditors represent a second line of defense for the integrity of financial markets between the first, the preparers of financial statements, and the third, the regulators. The authors discuss fundamental operational issues such as how to maintain a balance between independence and the need to collaborate with clients to understand increasingly complex environments. They also discuss the role of auditors in a world where the information is not merely contained in standardized financial statements. This discussion relates to the role played by the press in financial markets or to the issues related to ESG reporting. The authors link their ­discussion to the auditing-regulatory framework. 3.2.2 Frameworks Steven Monahan offers a framework to understand the role that the financial reporting system plays in optimizing the pricing of financial assets. The view taken in the chapter is

M10794_HILARY&MCLEAN_v1.indd 9

18/07/23 5:24 PM

10  Handbook of financial decision making

that the primary objective of financial reporting is to provide investors with information that they can use when valuing companies and the securities that they issue. Monahan discusses various pricing models and relates them to financial information. The goal here is to improve investment decisions. However, he points out that the links between financial reporting, financial decisions, and economic efficiency are not fully understood. In contrast to Monahan, who primarily focuses on external users, Satish Joshi and Ranjani Krishnan focus on the internal users of financial information. They focus on the context of management control systems, using the standard principal–agent framework as a starting point, and also discussing extensions that incorporate behavioral theories. Aside from taking stock of where we are currently, Joshi and Krishnan cover emerging issues; for example, they discuss what control systems may look like in organizations that focus on stakeholders and are concerned with socially responsible investment (SRI). This last point, SRI, is the focus of the chapter by Hao Liang and Tran Bao Phuong Nguyen, who note that, despite its contemporary importance, the term SRI is not well defined – or at least not subject to a widely accepted common definition. Although Liang and Nguyen review different frameworks that identify a multiplicity of pecuniary and nonpecuniary incentives for corporations and investors to engage in SRI, they also note that the existence of financial payoffs for SRI remains disputed. 3.2.3  Technological tools Two chapters focus on new and technologically advanced tools that enable financial decision making. Allen Huang and Haifeng You offer a broad review of how artificial intelligence (AI) offers great potential in improving financial decision making. They review increasingly common machine learning (ML) algorithms that extract and aggregate information from structured and unstructured data to facilitate financial decisions. The authors discuss how AI and humans can complement each other to improve financial decision making. Francesco D’Acunto and Alberto Rossi focus on how information technology has been transforming consumers’ financial decision making through robo-advising. After characterizing the unique features of this approach, the authors describe four pioneering commercial applications as case studies. They conclude by discussing opportunities for research in various fields. 3.3  Advisors and Household Finance After reviewing the increasing role of ML advising in previous chapters, the Handbook concentrates on human advising. Daniel Bradley offers a chapter on sell-side analyst research companies. Analysts have received a good deal of attention from academics in both finance and accounting. Bradley shows that there are more than 2,500 articles in the top four accounting journals and the top four finance journals that mention analysts in their abstracts. Although analysts have historically played a significant role in capital markets, Bradley notes that their survival is jeopardized by several threats, and discusses how analysts have responded. We conclude the Handbook by exploring two settings in which individuals make financial decisions. The field of household finance has become more prominent in recent academic research, while the prior literature has focused more on institutional and corporate

M10794_HILARY&MCLEAN_v1.indd 10

18/07/23 5:24 PM

Financial decision making: an overview  ­11

actors. Sumit Agarwal and Nithin Mannil offer a chapter exploring traditional topics such as consumption, savings, borrowing, and investment, as well as new ones such as fintech. They focus on the point of view of individuals or households, linking these decisions to different factors explored elsewhere in this Handbook, such as information constraints, behavioral aspects, and social dimensions. Julie Agnew focuses on one specific but very important financial dimension, individual retirement decision making. Many countries have aging populations and longevity is increasing, so prudent decisions in this dimension are critical. Agnew offers a summary of financial decisions in both the preretirement accumulation phase and the postretirement “decumulation” phase. The chapter highlights how behavioral biases and targeted interventions can affect the quality of individual choices.

4  LOOKING AHEAD Although we are mindful of the (alleged) quote by Niels Bohr that “Prediction is very difficult, especially if it’s about the future”, we nonetheless try, here, to point out possible areas of interest for researchers in the future. At a broader level, we note that the explosion of data has allowed people to examine topics that were difficult to analyze in the past. We believe that the rise in household finance and social finance can be traced to this development. Machine learning has also been enabled by this, leading the way to its analysis as a topic of interest and yielding new empirical tools to study other topics. While this trend is likely to continue, it paradoxically increases the need for models that prevent us from drowning in oceans of data. Furthermore, using new or unconventional data to revisit findings that are well established through more traditional channels has some merit, but its full potential is likely to be fulfilled if it is used to answer new questions. For example, the within-household dynamics of decision making are poorly understood at the moment, but an analysis of communication between domestic partners could shed some light on this topic. The same can be said of communication among employees within their organization. We also note that modern research in economics and business studies draws increasingly from other fields, such as psychology or sociology. However, since one cannot be an expert at everything, streams of research often focus on a very select part of noneconomic fields. For example, the extraordinary success of Kahneman and Tversky has led to a field of behavioral economics, which nearly ignores findings outside this paradigm. Many theories that are well accepted in psychology (e.g., construal level theory) are often ignored by researchers in accounting and finance. Broadening our collective horizons may yield new insights. Finally, we note that some populations have been intensively studied, while others have been studied much less. For example, past research has focused extensively on the US and, more recently, on Chinese populations. In contrast, our understanding of African and, to some extent, Indian populations is much more limited. These new settings (from the point of view of published academic research) may yield powerful findings for key topics such as the effect of global warming, financial innovations, or ­cultural traits on financial decision making. Overall, we hope that this Handbook offers a review of contemporary issues in financial decision making. Before leaving readers with the chapters, we would like to thank all the contributors to this volume. We hope that everyone enjoys their work as much as we did.

M10794_HILARY&MCLEAN_v1.indd 11

18/07/23 5:24 PM

12  Handbook of financial decision making

NOTES 1. The notion “moral hazard” had existed for a long time (e.g., Rowell and Connelly (2012)) but was more commonly formalized using mathematical tools in the 1970s (e.g., Jensen and Meckling (1976)). 2. Signaling (Spence (1973)) is another important mechanism that was delineated in the 1970s. Signaling is the notion that an economic agent may engage in costly observable actions to show his or her greater hidden ability. 3. See Reddy and Jadhav (2019) for a review of the literature on gender diversity in boardrooms or Teodósio et al. (2021) for a review of the literature on gender diversity and corporate risk taking. 4. For example, Hilary and Hui (2009) and Hilary and Huang (2023). See Hanlon et al. (2022) for a review.

REFERENCES Admati, A. R. 1985. A noisy rational expectations equilibrium for multi-asset securities markets. Econometrica 53, 629–657. Ai, H., M. Frank and A. Sanati. 2021. The trade-off theory of corporate capital structure. Oxford Research Encyclopedia of Economics and Finance. Retrieved 30 July 2022, from https://oxfordre. com/economics/view/10.1093/acrefore/9780190625979.001.0001/acrefore-9780190625979-e-602. Akerlof, G. A. 1970. The market for lemons: quality uncertainty and the market mechanism. Quarterly Journal of Economics 84, 488–500. Baker, M. and J. Wurgler. 2011 Behavioral corporate finance: An update survey, National Bureau of Economic Research Working Paper 17333. Baker, M., J. C. Stein, and J. Wurgler. 2003. When does the market matter? Stock prices and the investment of equity-dependent firms. Quarterly Journal of Economics 118, 969–1005. Ball R. and P. Brown. 1968. An empirical evaluation of accounting income numbers. Journal of Accounting Research 6, 159–178. Barberis, N. and R. Thaler. 2003. A survey of behavioral finance. In: Constantinides, G. M., M.  Harris and R. Stulz (eds.), Handbook of the Economics of Finance, Vol. 1B: Financial Markets and Asset Pricing, Elsevier, Amsterdam, 1053–1128. Bernard V. L. and J. K. Thomas. 1989. Post-earnings-announcement drift: Delayed price response or risk premium? Journal of Accounting Research 27, 1–36. Biddle, G. C., G. Hilary, and R. S. Verdi. 2009. How does financial reporting quality relate to investment efficiency? Journal of Accounting and Economics 48, 112–131. Botosan, C. A. 2006. Disclosure and the cost of capital: what do we know? Accounting and Business Research 36, 31–40. Chang, X., S. Dasgupta, and G. Hilary. 2009. The effect of auditor quality on financing decisions. The Accounting Review 84, 1085–1117. Chordia, T., A. Subrahmanyam, and Q. Tong. 2014. Have capital market anomalies attenuated in the recent era of high liquidity and trading activity? Journal of Accounting and Economics 58, 41–58. Clinton, W. 1993. Executive Order 12866 of September 30, 1993 Regulatory Planning and Review, Federal Register Presidential Documents Vol. 58, No. 190, Monday, October 4. Cooper, A. C., C. Y. Woo, and W. C. Dunkelberg. 1988. Entrepreneurs’ perceived chances for success. Journal of Business Venturing 3, 97–108. De Long, J. B., A. Shleifer, L. Summers, and R. Waldmann. 1990a. Noise trader risk in financial markets. Journal of Political Economy 98, 703–38. De Long, J. B., A. Shleifer, L. Summers, and R. Waldmann. 1990b. Positive feedback investment strategies and destabilizing rational speculation. Journal of Finance 45, 379–395. Derrien, F. and A. Kecskés. 2013. The real effects of financial shocks: Evidence from exogenous changes in analyst coverage. Journal of Finance 68, 1407–1440.

M10794_HILARY&MCLEAN_v1.indd 12

18/07/23 5:24 PM

Financial decision making: an overview  ­13

Di Vito, J. and K. Trottier. 2022. A literature review on corporate governance mechanisms: Past, present, and future. Accounting Perspectives 21, 207–235. Erickson, T. and T.M. Whited. 2000. Measurement error and the relationship between investment and q. Journal of Political Economy 108, 1027–1057. Fama, E. F. 1965a. The behavior of stock-market prices. Journal of Business 38, 34–105. Fama, E. F. 1965b. Random walks in stock market prices. Financial Analysts Journal 21, 55–59. Fama, E. F. 1998. Market efficiency, long-term returns, and behavioral finance. Journal of Financial Economics 49, 283–306. Fazzari, S., R. G. Hubbard, and B. Petersen. 1988. Financing constraints and corporate investment. Brookings Papers on Economic Activity 19, 141–206. Friedman, M. 1953. Essays in Positive Economics. University of Chicago Press, Chicago. Gromb, D. and D. Vayanos. 2010. Limits of arbitrage. Annual Review of Financial Economics 2, 251–275. Grossman S. J. and J. E. Stiglitz. 1980. On the impossibility of informationally efficient markets American Economic Review 70, 393–408. Hanlon, M., K. Yeung, and L. Zuo. 2022. Behavioral economics of accounting: A review of archival research on individual decision makers. Contemporary Accounting Research 39, 1150–1214. Harvey, C. R., Y. Liu, and H. Zhu. 2016. … and the cross-section of expected returns. Review of Financial Studies 29, 5–68. Hayashi, F. 1982. Tobin’s marginal q and average q: A neoclassical interpretation. Econometrica 50, 213–224. Heaton, J. B. 2002. Managerial optimism and corporate finance. Financial Management 31, 33–45. Hellwig, M. F. 1980. On the aggregation of information in competitive markets. Journal of Economic Theory 22, 477–498. Hilary, G. and S. Huang. 2023. Trust and contracting: Evidence from church sex scandals. Journal of Business Ethics 182, 421–442. Hilary, G., and K. W. Hui. 2009. Does religion matter in corporate decision making in America? Journal of Financial Economics 93, 455–473. Hilary, G., C. Hsu, B. Segal, and R. Wang. 2016. The bright side of managerial over-optimism. Journal of Accounting and Economics 62, 46–64. Hirshleifer, D. 2001. Investor psychology and asset pricing. Journal of Finance 56, 1533–1597. Jegadeesh, N. and S. Titman. 1993. Returns to buying winners and selling losers: Implications for stock market efficiency. Journal of Finance 48, 65–91. Jensen M. C. 1968. The performance of mutual funds in the period 1945–1964. Journal of Finance 23, 389–416. Jensen, M. C. 1969. Risk, the pricing of capital assets, and the evaluation of investment portfolios. Journal of Business 42, 167–247. Jensen, M. C. 1986. Agency costs of free cash flow, corporate finance, and takeovers. American Economic Review 76, 323–329. Jensen, M. C. and W. H. Meckling. 1976. Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics 3, 305–360. Jorgenson, D. 1963. Capital theory and investment behavior. American Economic Review 53, 247–259. Kachelmeier, S. J. 2020. Financial accounting and auditing experiments in the Journal of Accounting Research: Historical background and recent advances, Journal of Accounting Research virtual issue. https://dx.doi.org/10.2139/ssrn.3678868. Kahneman, D. 2011. Thinking, Fast and Slow. Farrar, Straus and Giroux, New York. Kelly, B. and A. Ljungqvist. 2012. Testing asymmetric-information asset pricing models. Review of Financial Studies 25, 1366–1413. Keynes, J. M. 1936. The General Theory of Employment, Interest and Money. Macmillan, London. Kindleberger. C. P. 1978. Manias, Panics, and Crashes: A History of Financial Crises. Basic Books, New York. Malmendier, U. and G. Tate. 2015. Behavioral CEOs: The role of managerial overconfidence. Journal of Economic Perspectives 29, 37–60.

M10794_HILARY&MCLEAN_v1.indd 13

18/07/23 5:24 PM

14  Handbook of financial decision making

Markowitz, H. 1952. Portfolio selection. Journal of Finance 7, 77–91. McLean, R. D. and J. Pontiff. 2016. Does academic research destroy stock return predictability? Journal of Finance 71, 5–32. McLean, R. D. and M. Zhao. 2014. The business cycle, investor sentiment, and costly external finance. Journal of Finance 69, 1377–1409. Miller, M. and F. Modigliani. 1961. Dividend policy, growth, and the valuation of shares. Journal of Business 34, 411–433. Minsky, H. 1977. The Financial Instability Hypothesis: An interpretation of Keynes and an alternative to “standard” theory. Challenge 20, 20–27. Modigliani, F. and M. Miller. 1958. The cost of capital, corporation finance and the theory of investment. American Economic Review 48, 261–297. Morck, R., A. Shleifer, and R. W. Vishny. 1990. Do managerial objectives drive bad acquisitions? Journal of Finance 45, 31–48. Muth, J. F. 1961. Rational expectations and the theory of price movements. Econometrica 29, 315–335. Myers, N. and C. S. Majluf. 1984. Corporate financing and investment decisions when firms have information that investors do not have. Journal of Financial Economics 13, 187–221. Peters, R. and L.A. Taylor, 2017. Intangible capital and the investment-q relation. Journal of Financial Economics 123, 251–272. Pontiff, J. 1996. Costly arbitrage: evidence from closed-end funds. Quarterly Journal of Economics 111, 1135–1151. Pontiff, J. 2006. Costly arbitrage and the myth of idiosyncratic risk. Journal of Accounting and Economics 42, 35–52. Reddy, S. and A. M. Jadhav. 2019. Gender diversity in boardrooms: A literature review, Cogent Economics and Finance 7. doi: 10.1080/23322039.2019.1644703. Rowell, D. and L. B. Connelly. 2012. A history of the term “moral hazard”. Journal of Risk and Insurance 79, 1051–1075. Savage, L. J. 1954. The Foundations of Statistics. John Wiley & Sons, New York. Sharpe, W. F. 1964. Capital asset prices: A theory of market equilibrium under conditions of risk. Journal of Finance 19, 425–442. Shleifer, A. and R. W. Vishny. 1997. A survey of corporate governance. Journal of Finance 52, 737–783. Simon, H. A. 1957. Models of Man: Social and Rational; Mathematical Essays on Rational Human Behavior in Society Setting. Wiley, New York. Sloan, R. G. 1996. Do stock prices fully reflect information in accruals and cash flows about future earnings? The Accounting Review 71, 289–315. Spence, M. 1973. Job market signaling. Quarterly Journal of Economics 87, 355–374. Sunstein, C. R., C. Jolls, and R. H. Thaler. 1998. A behavioral approach to law and economics. Stanford Law Review 50, 1471–1548. Swieringa, R. J. and K. E. Weick. 1982. An assessment of laboratory experiments in accounting. Journal of Accounting Research 20, 56–101. Teodósio, J., E. Vieira, and M. Madaleno. 2021. Gender diversity and corporate risk-taking: A literature review. Managerial Finance 47, 1038–1073. Thaler, R. H. and C. R. Sunstein. 2008, Nudge: Improving Decisions about Health, Wealth, and Happiness. Yale University Press, New Haven, CT. Tobin, J. 1969. A general equilibrium approach to monetary theory. Journal of Money, Credit and Banking 1, 15–29. Treynor, J. L. and F. Black. 1973. How to use security analysis to improve portfolio selection. Journal of Business 46, 66–86. von Neumann, J. and O. Morgenstern. 1947. Theory of Games and Economic Behavior (2nd edn). Princeton University Press, Princeton, NJ. Zattoni, A., E. Dedoulis, S. Leventis, and H. Van Ees. 2020. Corporate governance and institutions: A review and research agenda. Corporate Governance: An International Review 28, 465–487.

M10794_HILARY&MCLEAN_v1.indd 14

18/07/23 5:24 PM

PART II NATURAL AND ENVIRONMENTAL FACTORS THAT IMPACT FINANCIAL DECISION MAKING

M10794_HILARY&MCLEAN_v1.indd 15

18/07/23 5:24 PM

Part II.1 Natural Factors

M10794_HILARY&MCLEAN_v1.indd 16

18/07/23 5:24 PM

1.  Limited attention and financial decision-making Alexander Nekrasov, Siew Hong Teoh, and Shijia Wu

1 INTRODUCTION In the current world of the Big Data revolution, where a vast amount of information is available to investors, attention is a scarce resource. The time, effort, and skill required to identify, acquire, and process all relevant information for decision-making can be substantial. Consider, for example: the large volume of information available about a single firm from mandated financial statements and voluntary disclosures produced by the firm itself; reports generated by intermediaries such as analysts; news articles by business journalists; postings on social media by investors and other stakeholders; and data publicly available on the Internet or curated from proprietary sources. Even reading a single document, such as this chapter on limited attention, requires time and attention. Moreover, investors must evaluate and compare information for many other firms before they can make informed investment decisions. On top of this, firms compete for their attention. During a busy earnings announcement season, hundreds of firms will announce their earnings on the same day. This deluge of information and the limits of attention create a bottleneck. To understand how capital markets function when this bottleneck occurs, we need to know how investors, managers, and other stakeholders make decisions when faced with limited attention. The traditional theory assumes that information processing costs are negligible, and that all publicly available information is incorporated into stock prices immediately and fully. In contrast, limited attention theory posits that investors have finite attention and processing power, and that this constraint can prevent them from making optimal decisions. Investors cannot fully attend to all public news. Instead, only a subset of ­investors heeds any specific piece of information at any given time, while the attention of others is consumed by other news. In this market with limited investor attention, the  key prediction is that the stock price cannot fully incorporate all available ­information. In this chapter, we first provide a simple example of a theory model to illustrate testable predictions for the consequences of limited attention for capital markets. Then we discuss empirical studies examining a wide set of predictions from limited attention theory. Limited attention theory can explain a broad set of empirical findings, including misreaction to public information, post-earnings announcement drift (PEAD), effects of news salience, and managerial choices regarding the format and timing of disclosures. In the next section, we discuss a simple limited attention model and its predictions about investor responses to firms’ earnings announcements. The model illustrates how investor reaction to a firm’s announcement depends on the degree of investor attention and the amount of supplemental information disclosed by the firm in the announcement. We then discuss empirical evidence on the factors that influence investors’ attention and responses to earnings news. We discuss the effects of distraction due to competing stimuli, 17

M10794_HILARY&MCLEAN_v1.indd 17

18/07/23 5:24 PM

18  Handbook of financial decision making

investor sophistication, news salience and information processing ease, allocation of attention to competing tasks, information intermediaries, and disclosure timing. We turn next to examining investor limited attention on managerial decisions, including disclosure timing and qualitative disclosure attributes. The final section concludes our discussion.

2 ATTENTION THEORY, MARKET REACTION TO NEWS, AND RETURN PREDICTABILITY When some investors have limited attention, they attend only to a subset of publicly available information. The equilibrium price then reflects the weighted average of the expectations of attentive and inattentive investors. As a result, the price underreacts to the public news at the announcement date and continues to drift in the direction of the news when inattentive investors catch up with the news. In the context of an earnings news announcement, this price pattern is the post-earnings-announcement drift. Bernard and Thomas (1989, 1990) document that stock prices continue to drift upward (downward) following earnings announcements when the quarterly earnings news exceeds (falls below) expectations. This price pattern is one of the most famous and robust anomalies. Fink (2021) provides a survey review of the various characteristics of PEAD documented in the literature. PEAD is a global phenomenon; it is observed in both developed and emerging financial markets. It is stronger for small firms, firms with lower analyst following, and firms with lower institutional ownership. PEAD is not subsumed by other anomalies, such as price momentum, accruals anomaly, or value-growth anomaly. In this section, we present a simple limited attention model adapted from Hirshleifer and Teoh (2003), Hirshleifer et al. (2011), and Li et al. (2020). Assume that a fraction 1 − γ of investors are fully attentive to public information and that the other fraction γ of investors are inattentive. Attentive investors acquire and fully process the disclosed information, whereas inattentive investors are either unaware of the disclosure or do not process it. In other words, attentive investors update as rational Bayesians using the public information, whereas inattentive investors do not update at all. We can think of attention in the investor population, 1 − γ, as increasing with information salience, processing ease, and investor sophistication and decreasing with the distraction due to competing events. There is a single risky security (stock) and a risk-free asset (cash) in the market. Investors can trade assets at each of dates 0, 1, 2, and 3 and can consume at terminal date 3. Date 0 is before the earnings announcement, and the stock price at that date is denoted as ​​P0​  ​​. At date 1, the firm announces earnings, which provides investors with public information about the terminal value of the stock. At date 2, the firm’s filing date of financial statements with the authorities, investors may receive further information about the terminal value of the stock. At date 3, they receive the terminal payoff of the stock, ​P3​  ​, and consume. Following Hirshleifer and Teoh (2003), we assume that the stock is in zero net supply, which implies that there is no risk premium. At date 1, solving for optimal investor trading positions in the stock as a function of price ​P1​  ​and then imposing the market clearing condition that the sum of the trading positions is zero, we can solve for the equilibrium price. This is a weighted average of the expectations of attentive and inattentive investors,

M10794_HILARY&MCLEAN_v1.indd 18

18/07/23 5:24 PM

Limited attention and financial decision-making  ­19

​ ​P1​  ​  =  f ​E​ I​[​P3​  ​ │ Φ]​ + ​(​1 − f​)​ ​E​ A​[​P3​  ​ │Φ ]​ ​​,

(1.1)

where ​Φ​denotes the information available at date 1 and superscripts A and I represent the beliefs of attentive and inattentive investors, respectively. The parameter ​f​is an increasing function of the fraction of inattentive investors, γ: ​γ _

​    ​  ​var​​ I​(​ ​P3​  ​)​ ​ f  =  ____________ ​       ​,​ _ ​ ​var​​ Iγ​( ​ P​ ​  ​)​ ​  + ​_  ​var​1−γ   ​  ​ A​(​ P ​ 3​  ​)​ 3

(1.2)

where v​​ ar​​ I​(​ ​P3​  ​)​​and ​​var​​ A​(​P3​  ​)​​are the variances of future firm value unconditionally or conditional upon public signals, respectively. Now consider two alternative information disclosure regimes. In the timely disclosure (TD) regime, at date 1, the firm discloses both earnings ​e​and a financial statement item ​λ​, which together comprise the date 1 information set ​Φ​. In the delayed disclosure (DD) regime, the firm delays the disclosure of the financial statement item to date 2, and ​Φ​ consists of only earnings ​e​. Attentive investors incorporate the additional information ​λ​ disclosed at date 1 into their expectations, whereas inattentive investors do not. The price at date 1 in the TD regime is: ​ P ​ 1​  ​(e, λ)​  =  ​f​ e,λ​ E​[​P3​  ​]​ + ​(1 − ​f​ e,λ​)​E[​ ​P3​  ​ │ e, λ]​,​

(1.3)

where the expectation of inattentive investors is the prior expectation, ​E​ I​[​P3​  ​ │ e, λ]​ = E​[​P3​  ​]​​, and similarly inattentive variances are equal to prior variances ​v​ ar​​ I​(​P3​  ​)​  =  var​(​P3​  ​)​​. In contrast, the expectation of attentive investors is the fully rational Bayesian update conditional on all available information, E ​ ​ A[​ ​P3​  ​│e, λ]​ = E​[​P3​  ​│e, λ​]​, and variances are conditional upon all available information as well. By (1.3), the effective weight for the beliefs of inattentive investors in the TD regime is: ​γ   ​  ​ _ var​(​ ​P3​  ​)​ ​ ​f​ e,λ​  =  ______________ ​       ​.​ γ​ _ ​    ​  + ​_   1−γ    ​ var​(​ ​P​  ​)​ ​var​​ A​(​ ​P​  ​|​e, λ​)​ 3

(1.4)

3

In the DD regime, where only earnings ​e​is disclosed, the price at date 1 is: ​ P ​ 1​  ​(​e​    ​​ )​  =  ​f​ e​ E​[​P3​    ​​ ]​ + ​(1 − ​f​ e​)​E[​ ​P3​  ​ │ e]​.​

(1.5)

Similar to the TD regime, the expectation of inattentive investors is the prior expectation,​ E​​ I[​ ​P3​  ​ │ e]​  =  E​[​ ​P3​  ​]​; inattentive variances are equal to prior variances; and the expectation of attentive investors is the fully rational Bayesian update, conditional upon the available earning information, ​E​ A[​ ​P3​  ​ │ e]​  =  E​[​ ​P3​  ​│e​]​, and variances are conditional upon e as well. The effective weight for the beliefs of inattentive investors in the DD regime is: ​γ _

​ var​(​  ​P​  ​)​​  3 ​ f​ ​  ​  =  ______________ ​       ​​. e

_ ​ var​(γ​​  ​P​  ​)​ ​ + ​_  ​var​​ 1−γ   ​  A ​(​ P ​ 3​  ​|​e)​ ​ 3

(1.6)

Since variance ​​var​​ A​(​P3​  ​ │ e)​  >  ​var​​ A​(​P3​  ​│e, λ​)​​, the effective weight ​f​ ​ e,λ​  ​on the beliefs of ­inattentive investors is lower in the TD regime than the weight f​ ​ e​in the DD regime.

M10794_HILARY&MCLEAN_v1.indd 19

18/07/23 5:24 PM

20  Handbook of financial decision making

Using (1.3) or (1.5), the immediate reaction to the earnings news, as reflected in the earnings response coefficient (ERC), is obtained by differentiating the change in the stock _ price from date 0 to date 1 with respect to the earnings news e​  − ​ e ​ at date 1. The ERC in the two cases is as follows. TD regime: ​​ERC​TD   ​  =  ​(1 − ​f​ e,λ​)​ β​ ​ ​P​  ​,e​.​

(1.7)

DD regime: ​​ERC​DD   ​  =  ​(1 − ​f​ e​)​ ​β​ ​P​  ​,e​.​

(1.8)

3

3

The term ​β​P​  ​  ​,e​is the ERC if the fraction of inattentive investors were zero (that is, γ​, ​f​ e,λ​​, 3 and f​​ e​ were all zero). However, when inattentive investors are present, the fractions of inattentive investors ​f​ e,λ​and ​f​ e​in the TD and DD regimes, respectively, are positive. Thus, Equations (1.7) and (1.8) show that, in both regimes, the ERC decreases with the proportion of inattentive investors (​​f​ e,λ​or ​f​ e​)​. Thus we have the following observation. Observation 1a: The immediate market reaction to earnings news is lower when the proportion of inattentive investors is high. Furthermore, as explained above, ​f​ e,λ​​ < ​f​ e​; so, comparing Equations (1.7) and (1.8), the ERC is lower in the DD regime when the financial statement item is not disclosed at the same time that the earnings news is disclosed, date 1. Thus we have the following ­observation. Observation 1b: The immediate market reaction to earnings news is lower when financial statement disclosure is delayed (DD regime) than when it is not (TD regime). Turning to PEAD, the delayed market reaction to earnings news is obtained in a similar way by differentiating the change in the stock price from date 1 to date 3 with respect to _ the earnings news ​e − ​e    ​ at date 1. The PEAD coefficients in the two cases are as follows. TD regime: ​​PEAD​TD   ​  =  ​f​ e,λ​ β​ ​ ​P​  ​,e.​​

(1.9)

DD regime: ​​PEAD​DD   ​  =  ​f​ e​ ​β​ ​P​  ​,e​.​

(1.10)

3

3

Equations (1.9) and (1.10) show that (1) the PEAD increases with the proportion of inattentive investors (that is, ​f​ e,λ​or ​f​ e​) in both regimes, and (2) the PEAD is higher in the DD regime than in the TD regime (since ​f​ e​ > ​f​ e,λ​). Thus we have the following observations. Observation 2a: The post-earnings announcement drift is greater when the proportion of inattentive investors is high. Observation 2b: The post-earnings announcement drift is greater when financial statement disclosure is delayed (DD regime) than when it is not (TD regime). Finally, we consider the price reaction at the filing date, date 2. For brevity, we limit our discussion to a summary of the key elements of this analysis. Because there is a fixed cost

M10794_HILARY&MCLEAN_v1.indd 20

18/07/23 5:24 PM

Limited attention and financial decision-making  ­21

of attending to information and because the filing is later than the earnings announcement, the proportion of inattentive investors at the filing date, ​f  ′​, is lower than that at the earnings announcement date, f​ ​. Therefore the effective weight in the market price at date 2 on investors who pay attention to both the earnings news and the financial statement item is lower in the DD regime, where the financial statement is disclosed at a time when investor attention is low. As a result, the total price reaction to earnings news from date 0 to date 2 is lower in the DD regime than in the TD regime. The price correction occurs only at a later date when investors realize the terminal payoff. Thus we have the following observations. Observation 3a: The sum of the earnings response coefficients at the earnings announcement date and at the filing date is lower when financial statement disclosure is delayed (DD regime) than when it is not (TD regime). Observation 3b: The price drift after the filing date is higher when financial statement disclosure is delayed (DD regime) than when it is not (TD regime). The above framework comparing TD and DD regimes can be reinterpreted to analyze the situation where observant investors notice both pieces of information, earnings e and additional information λ, versus inattentive investors who notice only earnings e. The framework can also be adapted to obtain testable predictions about the effects of limited attention on stock market reactions to news in a wide range of other situations. For example, as mentioned earlier, the attention variable (1 − γ) can be used to represent the salience of the information disclosure or the ease of processing of the information item to study capital market effects of different presentation formats of disclosures. Higher salience or easier processing translates to more attentive investors, and the limited attention model predicts stronger immediate price response to the more salient presentation of news and a consequent more muted, longer-window price response or drift. The parameter γ can also be used to proxy for investor characteristics, such as financial sophistication, or it can proxy for situational characteristics, such as the degree of distraction due to contemporaneous events. A key advantage of this limited attention framework is its adaptability to accommodate a wide set of realistic disclosure characteristics and provide a rich set of testable predictions to study many disclosure-related issues.

3  EMPIRICAL EVIDENCE 3.1  Investor Inattention Due to Competing Stimuli Attention is a limited cognitive resource. When multiple stimuli compete for investor attention, attention to one task requires a substitution of attention from other tasks (Kahneman and Tversky, 1973). Therefore, on days with higher investor distraction, the limited attention model predicts a more muted immediate price response, and consequently a stronger price drift in the subsequent period. This prediction has been tested in various settings or contexts. Hirshleifer et al. (2009) find that, when many earnings announcements occur on the same day, investor attention to a focal firm’s earnings

M10794_HILARY&MCLEAN_v1.indd 21

18/07/23 5:24 PM

22  Handbook of financial decision making

announcement can be distracted by other firms’ announcements and irrelevant stimuli, such as industry-unrelated news. They find evidence that the distraction leads to a weaker investor reaction to the earnings news and a stronger PEAD. Interestingly, they find that, when the competing firms announcing on the same day belong to unrelated industries, investor distraction effects are stronger. This is intuitive, as a concentration of peer firms announcing on the same day can actually draw attention to the industry and therefore lead to more attention on the announcing firms within the same industry. DellaVigna and Pollet (2009) propose that investor attention is lower on Friday, as the upcoming weekend distracts investors from the task of stock valuation in response to earnings news made on Fridays. They find muted market response to Friday earnings announcements and a larger drift in the subsequent days. Because the choice of the day in the week to announce earnings is endogenous, they expect the total price reaction over the quarter to the earnings news may vary. Therefore they test the limited attention prediction for the price reaction as the ratio of the immediate price response to the drift response (or the total quarter return response). In the context of merger announcements, Louis and Sun (2010) find muted market reaction to Friday stock swap announcements, suggesting investor inattention occurs even in the context of one of the largest corporate events. Israeli et al. (2022) use a daily news pressure variable to measure the availability of newsworthy material on a given day to proxy for potential investor distraction. They find that investor attention to earnings announcements is weaker on days with high levels of unexpected distractions as measured by the daily news pressure. Investor attention can also be influenced by non-information events. For instance, Drake et al. (2016) find that the National Collegiate Athletic Association (NCAA) basketball tournament during March each year diverts millions of investors’ attention away from earnings news, and therefore the price reaction to earnings news released during NCAA basketball tournaments is muted. Brown et al. (2022) use the exogenous outages of the Blackberry Internet Service (BIS) and study whether the mobile Internet distracts investors from participating in financial markets. Consistent with the distraction hypothesis, they find a significant increase in trading volume and trading frequency when BIS unexpectedly goes offline. On the other hand, Madsen and Niessner (2019) study how investors respond to attention-grabbing events with little public information, specifically firms’ print advertisements. They find that print ads, especially in business publications, trigger temporary spikes in trading volumes. The evidence is consistent with the notion that, in the presence of limited attention, advertisements remind potential investors about the company and result in increased search and trading for its stock. 3.2  Limited Attention of Analysts, Institutional Investors, and Loan Officers Past evidence also shows that limited attention effects extend to sophisticated financial analysts. Several accounting studies examine analysts’ use of financial information assuming analysts’ limited attention. For example, Koester et al. (2016) examine whether firms use the announcement of extreme positive earnings surprises to attract analysts’ attention. They indeed find an intuitive result: that extreme positive earnings surprises are more salient to analysts, and therefore draw more of their attention. Choi and Gupta-Mukherjee (2022) find that analysts with larger workloads and fewer resources are more likely to rely

M10794_HILARY&MCLEAN_v1.indd 22

18/07/23 5:24 PM

Limited attention and financial decision-making  ­23

on industry rather than firm information. They find that this reliance is associated with lower forecast accuracy. Recent studies support a distraction effect even among sell-side analysts. For example, deHaan et al. (2017) find that analysts are slower to respond to an earnings announcement when they experience lower attention due to inclement weather. Driskill et al. (2020) provide compelling evidence that even information specialists such as financial analysts are subject to limited attention. Specifically, they find that analysts are less likely to: (1) issue timely earnings forecasts; (2) ask questions during an earnings conference call; and (3) provide prompt stock recommendations for a firm when there is another firm in their coverage portfolio that announces earnings on the same day. Du (2022) and Li and Wang (2021) study the influence of childcare responsibilities on analyst forecast outcomes, especially for female analysts. Using the shocks of school closures caused by the COVID-19 pandemic, these authors find that female analysts are less likely to issue timely forecasts, and that their forecasts become less accurate compared to those of their male counterparts, consistent with the notion that female analysts are distracted by their childcare responsibilities. Large institutional investors likewise have limited attention. Based on the findings from a large-scale survey, the Investor Responsibility Research Center (IRRC, 2011) expresses concern about the influence of limited institutional investor attention on their monitoring: “Three-fourths of institutions report that time is the most common impediment to engagement [with corporations], while staffing considerations rank second.” Kempf et al. (2017) identify distracted shareholders by exploiting exogenous shock to unrelated parts of the institutional shareholders’ portfolios. They find that firms with distracted shareholders are more likely to engage in diversifying and value-destroying acquisitions. The result is consistent with the concern that managers take advantage of the looser monitoring when institutional shareholders are distracted. We will discuss managerial decisions when investors have limited attention in a later section. Campbell et al. (2019) study the influence of attention on decision-making for loan officers. They find that lending decisions based on soft information (e.g., qualitative and hard-to-verify information) lead to worse loans when loan officers are distracted and fail to accurately interpret and reflect on soft information. Investor attention may also be affected by geographic location. Dyer (2021) studies the demand for public information by local versus nonlocal investors. He finds that the same investor chooses to acquire more public information for local firms than for nonlocal ones. But the local preference in information acquisition tends to decrease as proxies for information processing capacity increases. These findings are consistent with investors being more attentive to local investments, which diverts attention from nonlocal investments. 3.3  Investor Sophistication Sophisticated investors—such as sell-side analysts (Driskill et al., 2020; Chiu et al., 2021) and institutional shareholders (Kempf et al., 2017)—exhibit limited attention. However, retail investors are more subject to attention and time constraint in processing financial information. The accounting and finance literatures provide a large body of evidence that the effects of limited attention are more pronounced for less sophisticated or less experienced investors and analysts. For example, Bartov et al. (2000) use institutional ownership

M10794_HILARY&MCLEAN_v1.indd 23

18/07/23 5:24 PM

24  Handbook of financial decision making

as a proxy for investor sophistication and find that PEAD is concentrated among firms with a high percentage of unsophisticated investors. In responding to calls from the US Securities and Exchange Commission (SEC, 2007) and the Financial Accounting Standards Board (FASB, 2010) for clear and concise disclosures and for improving the readability of financial disclosures to individual investors, Lawrence (2013) studies whether retail investors are more subject to attention and time constraints when processing complicated financial disclosures. He finds that retail investors have lower relative information disadvantages and invest more in firms with clear and concise financial disclosures. The evidence highlights the importance of disclosure clarity, which directly impacts the amount of information that can be absorbed by investors with limited attention. In recent years, companies and the capital market have started to adopt new technologies to help investors (especially retail investors) suffer less from their limited attention to news events. For example, Blankespoor et al. (2018) study the implementation of “robojournalism” technology by The Associated Press (AP) and how it influences investor attention to earnings news. Robo-journalism produces articles about a firm’s earnings press release by synthesizing information from the press release, analyst reports, and stock information, and then distributes the articles over national and local outlets. These articles only contain public information. More importantly, they concisely synthesize salient information likely to attract investor attention. As a result, the increasing awareness of earnings news due to robo-journalism is associated with increasing investor trading, especially by retail investors. Moss (2022) examines how retail brokerages’ use of push notifications impact retail investor attention and trading. He finds that push notifications increase the number of retail trades by approximately 25 percent in the minutes following a notification. The evidence is consistent with the idea that investors are more attentive to the stocks that get pushed to the front of their minds, and so they are more likely to act by trading these stocks. Early studies in accounting and finance rely on indirect measures of attention such as extreme returns and trading volume (Barber and Odean, 2008; Hou et al., 2009). However, limited attention encompasses not just mere awareness of the information but extends also to the acquisition and processing costs that can prevent investors from incorporating information into trading decisions (Blankespoor et al., 2019). Recent studies construct novel measures that can directly capture investor attention. Da et al. (2011) use Google searches as a proxy for retail investor attention. They argue that Internet users commonly use search engines to collect information, and that Google accounts for most search queries in the United States. More importantly, if someone searches for a stock in Google, then that person is undoubtedly paying attention to the stock. The authors find that firms with an increase in Google searches experience more retail trading as well as higher shortterm stock price and long-term price reversal. Relatedly, Ben-Rephael et al. (2017) measure institutional investor attention using searching and reading on Bloomberg terminals. They find that their institutional investor attention measure differs significantly from other investor attention measures and is highly correlated with institutional trading. They also compare retail attention proxied by Google searches and institutional investor attention from Bloomberg searches, and find that institutional attention responds in a more timely manner to major news events.

M10794_HILARY&MCLEAN_v1.indd 24

18/07/23 5:24 PM

Limited attention and financial decision-making  ­25

Lu et al. (2016) examine limited attention among another type of sophisticated i­nvestors—hedge fund managers. They find that hedge fund managers’ marriages and divorces are associated with lower fund alpha in both the short and long term, consistent with a distraction effect. 3.4  News Salience and Information Processing Ease People are more likely to process information that is more salient and easier to process (Fiske and Taylor, 2016). Early studies use the placement, categorization, and labeling of information in financial statements to measure its salience and to examine how salience affects investor incorporation of the information into stock prices. One important finding is that investors put higher weights on information presented in the financial statements compared to information disclosed in the footnotes (Aboody, 1996; Ahmed et al., 2006; Amir, 1993). Similarly, Files et al. (2009) show that the market reaction to restatement announcements is stronger when the restatement is disclosed in the headline of the press release, a location of high salience, than when it is disclosed in the body or footnote of the press release. Studies in the finance and accounting literatures provide evidence that the salience of earnings news increases with broader dissemination via information intermediaries such as the business press and social media. Klibanoff et al. (1998) find that the price reaction to closed-end country funds is stronger when the country-specific news appears on the front page of the New York Times, consistent with the notion that media coverage increases the salience of country-related information and then triggers investor trading. Kimbrough (2005) finds that the initiation of conference calls is associated with reduction in analyst forecast errors, consistent with the notion that conference calls provide managers with the opportunity to direct investor attention to the key earnings metrics that otherwise may be ignored. Drake et al. (2014) provide evidence that disseminating accounting information via the business press increases the number of investors who are aware of the news and reduces the mispricing of accounting information. Fang and Peress (2009) find that firms with no media coverage earn higher stock returns than firms with high media coverage, which is consistent with the notion that media coverage increases investor awareness about stocks. In a field experiment, Lawrence et al. (2018) promote earnings announcements to a subset of Yahoo Finance users for randomly selected firms, and find that promoted firms experience stronger market reaction to earnings news. This finding provides direct evidence that investors tend to trade attention-grabbing stocks (Barber and Odean, 2008). Recent years have seen a sharp increase in the use of social media for disseminating financial information. The SEC issued a report on April 2, 2013, that makes it clear that companies can use social media outlets, such as Facebook and Twitter, to announce key information in compliance with Regulation Fair Disclosure (Regulation FD). Blankespoor et al. (2014) study the role of social media in disseminating earnings news, and find that firms disseminating news via Twitter are associated with lower information asymmetry. The evidence is consistent with the notion that disseminating news on Twitter increases the visibility of earnings news by reaching a broad set of investors. Psychology research shows that individuals put more weight on cues with higher processing fluency in their decision-making (Song and Schwarz, 2008). An experimental

M10794_HILARY&MCLEAN_v1.indd 25

18/07/23 5:24 PM

26  Handbook of financial decision making

study by Rennekamp (2012) shows that a more readable disclosure with higher processing fluency increases investors’ confidence that they can incorporate the disclosed information into valuation decisions. Umar (2022) finds that the textual complexity of article titles on Seeking Alpha is associated with lower investor attention to the news contained in the articles and less trading, consistent with the notion that investors are complexity averse, especially less sophisticated investors. Specifically, in a field experiment setting by holding the article content constant, Umar finds that an article with a complex title receives fewer views from Seeking Alpha users compared to the same article with a less complicated title. Miao et al. (2016) study the effect of limited attention on investor valuation of accruals by comparing market reactions to earnings announcements between two subsamples. In one subsample, firms disclose only the balance sheet in the earnings press release, while in the other firms disclose both the balance sheet and the statement of cash flows (SCF). The availability of the SCF makes accruals more salient and easier to process for investors with limited attention, whereas the accruals information needs to be inferred from comparative balance sheets when the SCF is unavailable at the announcement date. The authors find strong evidence that SCF disclosure reduces the accruals anomaly, especially in firms with more retail investors. Cronqvist et al. (2022) examine limited attention effects in the context of FAS 123-R, which requires firms to report the cost of stock option compensation on the income statement instead of disclosing it in financial statement footnotes. The authors find that, after FAS 123-R’s adoption, firms with high option compensation are more likely to miss analyst earnings forecasts and experience stock recommendation downgrades and market valuation declines. The findings suggest that investors and analysts ignored stock option cost when it was less salient and harder to process. Cardinaels et al. (2019) study how investor judgment of earnings announcements is influenced by the automatic summarization of earnings press releases, compared to summaries written by managers who have incentives to strategically choose the tone and content of the summary. They find that investors are less subject to limited attention and value information in the earnings releases more conservatively if the earnings release is accompanied by an automatic summary, which is less biased, compared to a managergenerated summary. A substantial body of literature finds that, owing to limited attention, the stock market sometimes underweights relevant non-accounting information signals as well. For example, measures of innovative efficiency and originality based on patents and patent citations positively predict profits and seem to be underweighted by the stock market. As a result, such measures predict positive abnormal stock returns (Hirshleifer et al., 2013, 2018). Although investors are likely to benefit when the increased salience of information matches its importance to the relevant decision, recent studies show that salience can be a double-edged sword. If salience excites investors to overweight salient but transitory earnings news, price overreaction at the time of the announcement may occur. Huang et al. (2018) use a new salience measure—defined as the number of quantitative items in an earnings press release headline—and find that high salience is associated with a stronger immediate market reaction, followed by a subsequent return reversal. Notably, they find that managers take advantage of the headline salience by highlighting good but less persistent financial performance when they plan to sell their shares after earnings ­announcements.

M10794_HILARY&MCLEAN_v1.indd 26

18/07/23 5:24 PM

Limited attention and financial decision-making  ­27

More recently, accounting and finance studies have started to investigate the influence of nontextual information on salience and investor attention. Nekrasov et al. (2021) use visuals as a novel proxy for salience, and find that visuals in firms’ Twitter earnings announcements are associated with more retweets, suggesting greater attention to and engagement with those announcements. Consistent with managerial opportunism, they find that managers use visuals to attract investor attention when the quarterly earnings performance is good but less persistent. Visuals are also associated with a stronger immediate reaction to earnings news and a subsequent return reversal. Liaukonytė and Žaldokas (2022) study how retail investor behavior responds to television advertising. They find that television advertising is associated with an increase in Electronic Data Gathering, Analysis, and Retrieval (EDGAR) and Google searches for financial information within 15 minutes of the airing of the ad, suggesting that advertising increases the attention not only of consumers but also of investors. Gu et al. (2022) study how investor sentiment reacts to the use of dynamic visuals— graphics interchange format (GIF), a novel attention-grabbing communication tool that is increasingly used on social media. They find that GIFs are associated with an increase in net bullish sentiment. Moreover, firms discussed with GIFs experience stronger immediate stock returns that are followed by long-term reversals, consistent with the notion that investors overreact to information presented with GIFs. 3.5  Allocation of Attention to Competing Tasks When investors and analysts face competing tasks, how do they allocate their attention? Gibbons et al. (2021) find that analysts access firms’ financial filings on EDGAR more frequently for companies with more volatile returns or recent mergers and acquisitions. These findings are consistent with the notion that analysts’ attention and information acquisition are driven by the demand for information from their clients. Financial analysts are well known for their extremely long working hours (Bradshaw et al., 2017). Their job entails producing research reports; conducting calls, meetings, and on-site visits with clients; and meeting with the sales and trading departments within their brokerages. During the earnings announcement seasons, when multiple earnings announcements are issued on the same day, an analyst must decide which announcement to cover first. Chiu et al. (2021) study whether analysts prioritize firms that are more important to their institutional clients when facing competing tasks during the earnings announcement days. Consistent with expectations, they find a positive association between institutional attention and the order in which an analyst produces research for multiple firms that announce earnings on the same day. In addition, they find that analysts’ timely forecasts are rewarded by better career outcomes. Specifically, analysts who issue more timely forecasts in response to institutional investor attention are more likely to be named all-star analysts by Institutional Investor magazine in the following year, and they are less likely to be demoted to a smaller brokerage. Similarly, Harford et al. (2019) find that analysts make more accurate, frequent, and informative earnings forecasts and recommendations for firms ranked higher within their portfolio based on proxies for the importance of firms to institutional investors. Relatedly, Driskill et al. (2020) provide evidence that, when analysts face concurrent earnings

M10794_HILARY&MCLEAN_v1.indd 27

18/07/23 5:24 PM

28  Handbook of financial decision making

announcements, they tend to allocate their limited attention to firms that are more important to their careers. Han et al. (2020) investigate the allocation of attention by analysts who experience major climatic disasters. They find that analysts in disaster zones are more likely to allocate their attention to firms of higher importance or salience. Chakrabarty and Moulton (2012) investigate the allocation of attention by market makers. They find that, when some stocks handled by a designated market maker have earnings announcements, liquidity is lower for non-announcement stocks handled by the same market maker. Furthermore, they find that the effect of this attention constraint is reduced after the NYSE introduces the hybrid market, which increases the automation and speed of trading. 3.6  The Role of Information Intermediaries Recently, researchers have been interested in whether information intermediaries, such as business press and equity analysts, can mitigate the effect of limited attention by drawing more attention to the news or facilitating processing of the information contained in announcements. Zhang (2008) studies the impact of analyst forecast timeliness on market reactions to earnings announcements. She finds that the earnings response coefficient is higher for firm-quarters with timely analyst forecast revisions and the corresponding post-earnings announcement drift is lower, suggesting that prompt analyst revisions help market participants process and respond promptly to information disclosed in earnings announcements. With the revolution in financial technology (fintech), many research firms have started to adopt robo-analysts to provide investment recommendations. Robo-analysts use stateof-the-art technology, such as natural language processing (NLP) and machine learning, to produce investment recommendations along with research reports. Coleman et al. (2022) conduct a comprehensive analysis comparing the recommendations generated by robo-analysts versus human analysts. They find that automation allows robo-analysts to revise their recommendations more frequently and incorporate information from complex periodic filings. The evidence overall suggests that robo-analysts suffer less from limited attention and can be used by the sell-side research industry to produce high-quality outputs. Finance and accounting research also provides ample evidence that highlights the business media as a key player in financial markets. For example, Peress (2014) exploits newspaper strikes to assess the causal impact of the media on market reactions to firm news. He finds that trading volume falls 12 percent on strike days. This evidence suggests that disseminating information via the business press helps investors become aware of and incorporate firm news into stock prices. Drake et al. (2014) show that press coverage of annual earnings announcements mitigates cash flow mispricing. Using high-frequency intraday data, Rogers et al. (2016) find that media dissemination affects how the market responds to insider trading news in the minutes after its release. Twedt (2016) finds that newswire dissemination of management earnings guidance is associated with more ­efficient incorporation of guidance information into price.

M10794_HILARY&MCLEAN_v1.indd 28

18/07/23 5:24 PM

Limited attention and financial decision-making  ­29

3.7  The Effect of Timing of Disclosures on Investor Attention to News Since the 1930s, both the frequency and length of firm disclosures have considerably increased due to the expanded mandatory disclosure rules and investor demand for information (Paredes, 2003; Radin, 2007). Regulators and practitioners have expressed concerns that investors are overloaded with disclosures, which may reduce investor ability to adequately incorporate firm disclosures into decision-making (White, 2013; Higgins, 2014). Furthermore, the growing number of concurrent announcements impedes the prompt processing of information. Arif et al. (2019) document that a growing percentage of firms disclose earnings announcements concurrently with their 10-K filings (that is, approximately 9 percent in 2002 compared to 43 percent by 2016). These concurrent filings increase the amount of information at the announcement, which may increase distraction. As a result, investor difficulty in instantaneously processing the greater amount of information leads to muted reactions. Chapman et al. (2019) study managers’ response to the potential information overload due to increasing amounts of disclosed information. They find that managers combat information overload by adjusting the timing of mandatory disclosures. Specifically, managers spread the disclosures out over several days when there are multiple disclosures for the same event date. In addition, managers are more likely to delay a disclosure when a disclosure has been made in the three days before the event date. However, disclosure of supporting financial information in earnings announcements may help investors process earnings news. A recent study by Li et al. (2020) investigates the effect of releasing information in installments rather than all at once. The authors find that, when firms delay disclosure of financial statement items in earnings announcements, investors and analysts underreact to earnings news. The underreaction continues even when the delayed items are disclosed in 10-Q filings. This finding is consistent with the maxim that “opportunity knocks but once.” If a firm mismatches the timing of disclosure to when investors and analysts are most attentive (e.g., at the earnings announcement date), then the disclosed earnings information will not be fully impounded into valuations. 3.8  Investor Limited Attention and Firms’ Strategic Disclosure Choices The evidence reviewed so far suggests that: (1) retail and institutional investors, sell-side analysts, and other capital market participants (such as loan officers and hedge fund managers) are subject to limited attention; and (2) market participants’ limited attention affects firms’ stock prices. These findings suggest that managers will consider information users’ limited attention when making disclosure decisions. Several studies examine whether managers are being strategic in the choice of disclosure timing by exploiting variation in investor attention. For instance, DellaVigna and Pollet (2009) provide evidence that managers with short-term objectives tend to release bad earnings news on Fridays, consistent with managers attempting to reduce investor reaction to the negative news. DeHaan et al. (2015) provide evidence that the lulls and peaks in investor attention are predictable by the managers. Specially, they find that investor attention is lower after trading hours, on busy days, and with less advance notice of the forthcoming announcements. More importantly, they show that managers exploit investor limited attention by announcing bad earnings news during periods of low attention.

M10794_HILARY&MCLEAN_v1.indd 29

18/07/23 5:24 PM

30  Handbook of financial decision making

Managers can also divert investor attention away from negative news by obfuscating the information and making it more difficult to process. Bloomfield (2002) proposes a management obfuscation hypothesis whereby managers have more incentives to obfuscate information when earnings news is bad, under the assumption that investors are less likely to fully incorporate disclosed information into stock prices when the information is hard to digest. Consistent with this hypothesis, Li (2008) provides initial evidence that the readability of annual reports is lower for firms with poor earnings performance. However, the positive relation between obfuscation and bad news could be driven by (1) managers’ incentive to obfuscate or (2) the inherent difficulty of communicating bad news (Bloomfield, 2008). Lo et al. (2017) attempt to disentangle these two explanations by studying the readability of management discussion and analysis (MD&A) sections of annual reports. They find that the MD&A section is more complicated when firms have managed earnings by beating prior year’s earnings. This finding contradicts the explanation that good news is inherently easier to communicate, and supports the management obfuscation hypothesis. Bushee et al. (2018) examine obfuscation in the setting of quarterly earnings conference calls. They argue that obfuscation could prevent analysts from asking follow-up questions about bad news, and therefore delay market reaction to bad news. Using a novel approach to disaggregate the linguistic complexity in firm disclosures into obfuscation and information components, they find a positive association between obfuscation and information asymmetry. In the context of mutual funds, deHaan et al. (2021) provide evidence that funds use unnecessarily complex disclosures to obfuscate high fees. Huang et al. (2014) study managers’ choice of the tone of words in earnings press releases and the implications for financial performance. They estimate abnormal positive tone as a measure of tone management, and find that the abnormal positive tone relates to lower future earnings and cash flows, suggesting managers’ strategic use of tone when disclosing quarterly earnings news. However, investors with limited attention are misled by the abnormal positive tone and do not discount for the negative information about future performance, resulting in an overvaluation of the stock at the time of the earnings announcements. Relatedly, Huang et al. (2018) find that managers opportunistically headline positive financial information in earnings press releases. Nekrasov et al. (2021) find that managers strategically present earnings news with visuals to attract investor attention when the quarterly earnings performance is good but less persistent. Both studies find that investors overreact to the salient good earnings news. Jung et al. (2018) study whether firms are being strategic when disseminating earnings news on social media, and find that firms are less likely to disseminate news on Twitter when the quarterly earnings news is bad, especially when firms have high litigation risk. Additionally, they show that disseminating bad news on social media attracts more attention from the traditional media, as evidenced by the increasing number of negative news articles following the earnings announcements. The recent surge in inflation across the globe has prompted renewed interest in research about inflation. One emerging research strand documents inadequate managerial attention to inflation dynamics or inflation risk. Coibion et al.’s (2018) survey of New Zealand firms finds that firms tend to ignore inflation and devote only limited resources to learning about it. Coibion et al. (2020) find that firms do not react to inflation until publicly available information about recent inflation is made salient to them. Konchitchki

M10794_HILARY&MCLEAN_v1.indd 30

18/07/23 5:24 PM

Limited attention and financial decision-making  ­31

and Xie (2022) consider inattention to inflation risk, and find that managers of firms with high exposure to inflation risk fail to disclose that risk, despite the SEC’s Regulation S-K requiring the disclosure of inflation risk factors. These inattention effects are costly as they render firms unprepared to handle problems that arise in an inflationary ­environment.

4 CONCLUSION Limited attention theory posits that investor attention is a scarce resource. The growing stream of research reviewed in this chapter finds that the effects of limited attention on capital markets are significant and pervasive. Theoretical and empirical studies show that, when some investors are inattentive to public information, in general the immediate stock price reaction to the information is incomplete and future stock prices exhibit postearnings announcement drift (PEAD). The degree of investor inattention explains the magnitude of the stock price underreaction. In some cases, however, when relevant information is ignored because of inattention so there is inadequate discounting of disclosed information (such as the lower persistence of accruals than cash flows), investor overreaction can result, followed by a post-announcement reversal. The literature has identified several factors affecting investor attention to news. These include: news salience; distraction by competing stimuli; information processing ease; investor sophistication; the economic importance of the firm relative to other firms; the timing of disclosure; and the processing and dissemination of the news by information intermediaries. Retail investors are more prone to limited attention. However, the effects of limited attention are nontrivial even for professional market participants such as equity analysts, institutional investors, financial data providers, and market makers. Limited investor attention influences investor trading, stock returns, and trading volume. The limited attention of professional equity analysts influences the timing of their forecasts and research reports, their underreactions to public information, and the allocation of their effort across firms they follow. The effects of limited attention can also be observed in the patterns of investor information acquisition through different channels, such as performing Internet searches, accessing of firm filings on EDGAR, and the searching and reading of professional investors on Bloomberg terminals. Finally, we discussed how investor limited attention affects firms’ choices regarding disclosure timing and format. Studies find that firms disclose bad news during periods of low investor attention and use various qualitative disclosure attributes to increase the salience of favorable information. These managerial choices can decrease stock market efficiency in processing financial information. Mandated disclosures can be double-edged swords and can have negative unintended consequences. Regulations on disclosures to increase investor limited attention can be beneficial, but they need to balance the trade­ anagerial offs of benefits from increased investor attention with the potential harms from m responses to limited attention.

M10794_HILARY&MCLEAN_v1.indd 31

18/07/23 5:24 PM

32  Handbook of financial decision making

REFERENCES Aboody, D., 1996. Recognition versus disclosure in the oil and gas industry. Journal of Accounting Research 34, 21–32. Ahmed, A.S., Kilic, E., Lobo, G.J., 2006. Does recognition versus disclosure matter? Evidence from value-relevance of banks’ recognized and disclosed derivative financial instruments. The Accounting Review 81(3), 567–588. Amir, E., 1993. The market valuation of accounting information: the case of postretirement benefits other than pensions. The Accounting Review 68(4), 703–724. Arif, S., Marshall, N.T., Schroeder, J.H., Yohn, T.L., 2019. A growing disparity in earnings disclosure mechanisms: the rise of concurrently released earnings announcements and 10-Ks. Journal of Accounting and Economics 68(1), 101221. Barber, B.M., Odean, T., 2008. All that glitters: the effect of attention and news on the buying behavior of individual and institutional investors, Review of Financial Studies 21, 785–818. Bartov, E., Radhakrishnan, S., Krinsky, I., 2000. Investor sophistication and patterns in stock returns after earnings announcements. The Accounting Review 75(1), 43–63. Ben-Rephael, A., Da, Z., Israelsen, R.D., 2017. It depends on where you search: institutional investor attention and underreaction to news. Review of Financial Studies 30(9), 3009–3047. Bernard, V.L., Thomas, J.K., 1989. Post-earnings-announcement drift: delayed price response or risk premium? Journal of Accounting Research 27, 1–36. Bernard, V.L., Thomas, J.K., 1990. Evidence that stock prices do not fully reflect the implications of current earnings for future earnings. Journal of Accounting and Economics 13(4), 305–340. Blankespoor, E., deHaan, E., Wertz, J., Zhu, C., 2019. Why do individual investors disregard accounting information? The roles of information awareness and acquisition costs. Journal of Accounting Research 57(1), 53–84. Blankespoor, E., deHaan, E., Zhu, C., 2018. Capital market effects of media synthesis and dissemination: evidence from robo-journalism. Review of Accounting Studies 23(1), 1–36. Blankespoor, E., Miller, G.S., White, H.D., 2014. The role of dissemination in market liquidity: evidence from firms’ use of Twitter™. The Accounting Review 89(1), 79–112. Bloomfield, R., 2008. Discussion of “Annual report readability, current earnings, and earnings persistence”. Journal of Accounting and Economics 45(2–3), 248–252. Bloomfield, R.J., 2002. The “incomplete revelation hypothesis” and financial reporting. Accounting Horizons 16, 233–243. Bradshaw, M.T., Ertimur, Y., O’Brien, P.C., 2017. Financial analysts and their contribution to wellfunctioning capital markets. Foundations and Trends® in Accounting 11(3), 119–191. Brown, N.C., Elliott, W.B., Wermers, R., White, R.M., 2022. News or noise: mobile internet technology and stock market activity. Available at SSRN 3585128. Bushee, B.J., Gow, I.D., Taylor, D.J., 2018. Linguistic complexity in firm disclosures: obfuscation or information? Journal of Accounting Research 56(1), 85–121. Campbell, D., Loumioti, M., Wittenberg-Moerman, R., 2019. Making sense of soft information: interpretation bias and loan quality. Journal of Accounting and Economics 68(2–3), 101240. Cardinaels, E., Hollander, S., White, B.J., 2019. Automatic summarization of earnings releases: attributes and effects on investors’ judgments. Review of Accounting Studies 24(3), 860–890. Chakrabarty, B., Moulton, P.C., 2012. Earnings announcements and attention constraints: the role of market design. Journal of Accounting and Economics 53(3), 612–634. Chapman, K.L., Reiter, N., White, H.D., Williams, C.D., 2019. Information overload and disclosure smoothing. Review of Accounting Studies 24(4), 1486–1522. Chiu, P.C., Lourie, B., Nekrasov, A., Teoh, S.H., 2021. Cater to thy client: analyst responsiveness to institutional investor attention. Management Science 67(12), 7455–7471. Choi, H.M., Gupta-Mukherjee, S., 2022. Analysts’ use of industry-level and firm-specific information: implications for information production. Available at SSRN 2803324. Coibion, O., Gorodnichenko, Y., Kumar, S., 2018. How do firms form their expectations? New survey evidence. American Economic Review 108(9), 2671–2713. Coibion, O., Gorodnichenko, Y., Ropele, T., 2020. Inflation expectations and firm decisions: new causal evidence. Quarterly Journal of Economics 135(1), 165–219.

M10794_HILARY&MCLEAN_v1.indd 32

18/07/23 5:24 PM

Limited attention and financial decision-making  ­33

Coleman, B., Merkley, K.J., Pacelli, J., 2022. Human versus machine: a comparison of robo-analyst and traditional research analyst investment recommendations. The Accounting Review 97(5), 221–244. Cronqvist, H., Ladika, T., Sautner, Z., 2022. Limited attention to detail in financial markets. Working Paper, University of Miami. Da, Z., Engelberg, J., Gao, P., 2011. In search of attention. Journal of Finance 66(5), 1461–1499. DeHaan, E., Madsen, J., Piotroski, J.D., 2017. Do weather-induced moods affect the processing of earnings news? Journal of Accounting Research 55(3), 509–550. DeHaan, E., Shevlin, T., Thornock, J., 2015. Market (in) attention and the strategic scheduling and timing of earnings announcements. Journal of Accounting and Economics 60(1), 36–55. DeHaan, E., Song, Y., Xie, C., Zhu, C., 2021. Obfuscation in mutual funds. Journal of Accounting and Economics 72(2–3), 101429. DellaVigna, S., Pollet, J.M., 2009. Investor inattention and Friday earnings announcements. Journal of Finance 64(2), 709–749. Drake, M.S., Gee, K.H., Thornock, J.R., 2016. March market madness: the impact of value-­ irrelevant events on the market pricing of earnings news. Contemporary Accounting Research 33(1), 172–203. Drake, M.S., Guest, N.M., Twedt, B.J., 2014. The media and mispricing: the role of the business press in the pricing of accounting information. The Accounting Review 89(5), 1673–1701. Driskill, M., Kirk, M.P., Tucker, J.W., 2020. Concurrent earnings announcements and analysts’ information production. The Accounting Review 95(1), 165–189. Du, M., 2022. Locked-in at home: female analysts’ attention at work during the COVID-19 pandemic. Available at SSRN 3741395. Dyer, T.A., 2021. The demand for public information by local and nonlocal investors: evidence from investor-level data. Journal of Accounting and Economics 72(1), 101417. Fang, L., Peress, J., 2009. Media coverage and the cross-section of stock returns. Journal of Finance 64(5), 2023–2052. Files, R., Swanson, E.P., Tse, S., 2009. Stealth disclosure of accounting restatements. The Accounting Review 84(5), 1495–1520. Financial Accounting Standards Board (FASB), 2010. Conceptual framework for financial ­reporting—chapter1, the objective of general purpose financial reporting, and chapter3, qualitative characteristics of useful financial information. FASB, Norwalk, CT (Statement of Financial Accounting Concepts No. 8). Fink, J., 2021. A review of the post-earnings-announcement drift. Journal of Behavioral and Experimental Finance 29, 100446. Fiske, S.T., Taylor, S.E., 2016. Social Cognition: From Brains to Culture. McGraw-Hill. Gibbons, B., Iliev, P., Kalodimos, J., 2021. Analyst information acquisition via EDGAR. Management Science 67(2), 769–793. Gu, M., Teoh, S.H., Wu, S., 2022. Contagion of investor sentiment in online investment communities: evidence from dynamic visuals on StockTwits. Available at SSRN 4110191. Han, Y., Mao, C.X., Tan, H., Zhang, C., 2020. Distracted analysts: evidence from climatic disasters. Available at SSRN 3625803. Harford, J., Jiang, F., Wang, R., Xie, F., 2019. Analyst career concerns, effort allocation, and firms’ information environment. Review of Financial Studies 32(6), 2179–2224. Higgins, K.F., 2014. Disclosure effectiveness: remarks before the American Bar Association Business Law Section Spring Meeting. Securities and Exchange Commission. Available at https:// www.sec.gov/News/Speech/Detail/Speech/1370541479332. Hirshleifer, D., Hsu, P.H., Li, D., 2013, Innovative efficiency and stock returns. Journal of Financial Economics 107(3), 632–654. Hirshleifer, D., Hsu, P.H., Li, D., 2018, Innovative originality, profitability, and stock returns. Journal of Financial Economics 31(7), 2553–2605. Hirshleifer, D., Lim, S.S., Teoh, S.H., 2009. Driven to distraction: extraneous events and underreaction to earnings news. Journal of Finance 64(5), 2289–2325. Hirshleifer, D., Lim, S.S., Teoh, S.H., 2011. Limited investor attention and stock market ­misreactions to accounting information. Review of Asset Pricing Studies 1(1), 35–73.

M10794_HILARY&MCLEAN_v1.indd 33

18/07/23 5:24 PM

34  Handbook of financial decision making

Hirshleifer, D., Teoh, S.H., 2003. Limited attention, information disclosure, and financial reporting. Journal of Accounting and Economics 36(1–3), 337–386. Hou, K., Xiong, W., Peng, L., 2009. A tale of two anomalies: the implications of investor attention for price and earnings momentum. Available at SSRN 976394. Huang, X., Nekrasov, A., Teoh, S.H., 2018. Headline salience, managerial opportunism, and overand underreactions to earnings. The Accounting Review 93(6), 231–255. Huang, X., Teoh, S.H., Zhang, Y., 2014. Tone management. The Accounting Review 89(3), 1083–1113. Investor Responsibility Research Center (IRRC), 2011. The state of engagement between U.S. corporations and shareholders. Available at https://cpb-us-w2.wpmucdn.com/sites.udel.edu/ dist/8/12944/files/2022/08/IRRC-ISS_EngagementStudy1.pdf. Israeli, D., Kasznik, R., Sridharan, S.A., 2022. Unexpected distractions and investor attention to corporate announcements. Review of Accounting Studies 27(2), 477–518. Jung, M.J., Naughton, J.P., Tahoun, A., Wang, C., 2018. Do firms strategically disseminate? Evidence from corporate use of social media. The Accounting Review 93(4), 225–252. Kahneman, D., Tversky, A., 1973. On the psychology of prediction. Psychological Review 80(4), 237–251. Kempf, E., Manconi, A., Spalt, O., 2017. Distracted shareholders and corporate actions. Review of Financial Studies 30(5), 1660–1695. Kimbrough, M.D., 2005. The effect of conference calls on analyst and market underreaction to earnings announcements. The Accounting Review 80(1), 189–219. Klibanoff, P., Lamont, O., Wizman, T.A., 1998. Investor reaction to salient news in closed-end country funds. Journal of Finance, 53 673–699. Koester, A., Lundholm, R., Soliman, M., 2016. Attracting attention in a limited attention world: exploring the causes and consequences of extreme positive earnings surprises. Management Science 62(10), 2871–2896. Konchitchki, Y., Xie, J., 2022. Undisclosed material inflation risk. Working paper, University of California Berkeley. Lawrence, A., 2013. Individual investors and financial disclosure. Journal of Accounting and Economics 56(1), 130–147. Lawrence, A., Ryans, J., Sun, E., Laptev, N., 2018. Earnings announcement promotions: a Yahoo finance field experiment. Journal of Accounting and Economics 66(2–3), 399–414. Li, F., 2008. Annual report readability, current earnings, and earnings persistence. Journal of Accounting and Economics 45(2–3), 221–247. Li, F.W., Wang, B., 2021. The gender effects of COVID-19 on equity analysts. Available at SSRN 3857376. Li, Y., Nekrasov, A., Teoh, S.H., 2020. Opportunity knocks but once: delayed disclosure of financial items in earnings announcements and neglect of earnings news. Review of Accounting Studies 25(1), 159–200. Liaukonytė, J., Žaldokas, A., 2022. Background noise? TV advertising affects real-time investor behavior. Management Science 68(4), 2465–2484. Lo, K., Ramos, F., Rogo, R., 2017. Earnings management and annual report readability. Journal of Accounting and Economics 63(1), 1–25. Louis, H., Sun, A., 2010. Investor inattention and the market reaction to merger announcements. Management Science 56(10), 1781–1793. Lu, Y., Ray, S., Teo, M., 2016. Limited attention, marital events and hedge funds. Journal of Financial Economics 122(3), 607–624. Madsen, J., Niessner, M., 2019. Is investor attention for sale? The role of advertising in financial markets. Journal of Accounting Research 57(3), 763–795. Miao, B., Teoh, S.H., Zhu, Z., 2016. Limited attention, statement of cash flow disclosure, and the valuation of accruals. Review of Accounting Studies 21, 473–515. Moss, A., 2022. How do brokerages’ digital engagement practices affect retail investor information processing and trading? Working Paper. Available at https://austinsmoss.github.io/austinmoss. me/Moss_JMP_How-Do-DEPs-Affect-Retail-Investor-Information-Processing-and-Trading. pdf.

M10794_HILARY&MCLEAN_v1.indd 34

18/07/23 5:24 PM

Limited attention and financial decision-making  ­35

Nekrasov, A., Teoh, S.H., Wu, S., 2021. Visuals and attention to earnings news on Twitter. Review of Accounting Studies. https://doi.org/10.1007/s11142-021-09630-8. Paredes, T.A., 2003. Blinded by the light: information overload and its consequences for securities regulation. Washington University Law Quarterly 81, 417–485. Peress, J., 2014. The media and the diffusion of information in financial markets: evidence from newspaper strikes. Journal of Finance 69(5), 2007–2043. Radin, A.J., 2007. Have we created financial statement disclosure overload? CPA Journal 77(11), 6–9. Rennekamp, K., 2012. Processing fluency and investors’ reactions to disclosure readability. Journal of Accounting Research 50(5), 1319–1354. Rogers, J.L., Skinner, D.J., Zechman, S.L., 2016. The role of the media in disseminating insidertrading news. Review of Accounting Studies 21(3), 711–739. Securities and Exchange Commission (SEC), 2007. Speech by SEC staff: feedback from individual investors on disclosure. Available at http://www.sec.gov/ news/speech/2007/spch011907ljs.htm. Song, H., Schwarz, N., 2008. If it’s hard to read, it’s hard to do: processing fluency affects effort prediction and motivation. Psychological Science 19(10), 986–988. Twedt, B., 2016. Spreading the word: price discovery and newswire dissemination of management earnings guidance. The Accounting Review 91(1), 317–346. Umar, T., 2022. Complexity aversion when seeking alpha. Journal of Accounting and Economics 73(2–3), 101477. White, M.J., 2013. The importance of independence. Securities and Exchange Commission. Available at https://www.sec.gov/News/Speech/Detail/Speech/1370539864016. Zhang, Y., 2008. Analyst responsiveness and the post-earnings-announcement drift. Journal of Accounting and Economics 46(1), 201–215.

M10794_HILARY&MCLEAN_v1.indd 35

18/07/23 5:24 PM

2.  Seasonality in stock returns and government bond returns Mark J. Kamstra and Lisa A. Kramer

1 INTRODUCTION In seeking to understand time variation in the rates of return earned by those who invest in financial securities, researchers have uncovered various empirical regularities. Some are seasonal in nature, tending to occur on a deterministic schedule. These include stock market regularities such as the Monday effect (see French, 1980; Gibbons and Hess, 1981; Rogalski, 1984), the tax-year effect (Rozeff and Kinney, 1976), and the daylight saving effect (Kamstra et al., 2000). Others arise non-deterministically, for example depending on random weather events (Saunders, 1993; Hirshleifer and Shumway, 2003) or the outcome of championship sports competitions (Edmans et al., 2007). The focus of this chapter is a regularity of the former category, arising on average on a deterministic schedule based on predictable daylight exposure through the seasons, and affecting financial markets evidently by altering investors’ moods and risk aversion. This regularity has come to be known as the seasonal affective disorder (SAD) effect. The underlying hypothesis is that, beginning around autumn, as the length of daylight shortens in non-equatorial locations, people tend to become more despondent, their moods even reaching the threshold for clinical depression among a subset of individuals. Medical research, including that by Rosenthal et al. (1984) and Lam (1998), documents the relationship between seasonality in daylight exposure and seasonality in people’s moods. In turn, experimental research by Kramer and Weber (2012) shows the relationship between seasonality in people’s moods and seasonality in their financial risk preferences. Overall, during the seasons when the amount of daylight at a given location is below the annual average, most individuals tend to be relatively more depressed and relatively more risk averse than they are during the rest of the year. The implications of these relationships for financial markets have been extensively documented. Kamstra et al. (2003) examine stock market index returns for nine countries at various latitudes in both hemispheres, and find statistically significant and economically large differences in returns across the seasons consistent with investors’ risk preferences varying with daylight exposure. Investors appear to demand a higher risk premium during seasons when daylight exposure is reduced, the result of which is that the equity risk premium in the US is about 6% higher per annum during the fall/winter seasons than in spring/summer. The amplitude of seasonal return variation is relatively greater for equity markets located at higher latitudes versus those closer to the equator, and the timing of the seasonality in returns is offset by six months in southern hemisphere countries relative to northern hemisphere countries, just like the seasons. More recent studies have documented an equity market SAD effect for larger groups of countries (see Dowling and Lucey, 2008; Kamstra et al., 2012). 36

M10794_HILARY&MCLEAN_v1.indd 36

18/07/23 5:24 PM

Seasonality in stock returns and government bond returns  ­37

For a regularity rooted in seasonally varying investor risk aversion, a reasonable question is whether it may also apply to securities other than equities. Kamstra et al. (2015) examine US Treasury securities and find statistically significant and economically large seasonality in government bond returns consistent with the SAD effect. Specifically, during the fall and winter seasons, when investors tend to be more risk averse, Treasury bond returns are on average lower than during spring and summer. The difference between the peak and trough in average monthly annualized government bond returns is about 80 basis points, which is large relative to the unconditional average return for this exceptionally safe investment class. Essentially, government bond investors demand a lower bond return during those seasons when they experience higher than average risk aversion. Turning to investment quantities as opposed to security rates of return, Kamstra et al. (2017) consider the flow of investor funds into and out of mutual funds in the US, Canada, and Australia. As the authors show, mutual fund flows are dominated by the decisions of retail investors, not large institutional investors, and so mutual fund flows plausibly reveal portfolio reallocation decisions of retail investors. Kamstra et al. (2017) find that flows into and out of risky versus safe mutual funds are consistent with retail investors experiencing seasonal variation in risk aversion due to seasonal variation in daylight. Aggregate fund flows are such that, on average, these investors prefer safe mutual funds in the fall and risky mutual funds in spring. After controlling for other known influences on mutual fund flows, in the month of September the typical magnitude of flows out of risky funds is $13 billion, while the typical magnitude of flows into safe funds is US$3 billion (flows into cash holdings like bank accounts make up much of the difference). The directions of these flows are then reversed in spring. Additional key papers on the SAD effect include the following: Garrett et al. (2005) document SAD-related seasonal variation in the price of risk in the context of a conditional version of the capital asset pricing model estimated using equity market index data for the US, Sweden, New Zealand, the UK, Japan, and Australia; and Kamstra et al. (2014) find that plausible magnitudes of seasonal changes in risk aversion are able to generate the observed magnitudes of seasonal changes in equity and government bond returns in the US in the context of a consumption-based asset pricing model. In this chapter, we take a deep dive into seasonality as it arises in equity markets and government bond markets. Our novel contributions include: consideration of the broadened selection of maturities of government bond returns for the US; the first-ever analysis of the SAD effect in equity returns across size-sorted deciles for the US, Canada, the UK, Germany, and Australia; the first-ever consideration of the SAD effect based on disaggregated firm-by-firm stock return data for the US, Canada, the UK, Germany, and Australia; and the development of a new proxy for the SAD effect based on Google searches (supplementing the existing measure based on clinical onset of and recovery from symptoms among SAD patients). We also present international evidence on the weakening of the Monday effect and tax-year effect over time.

2 DATA Our initial analysis focuses on daily US value-weighted size-sorted decile stock returns retrieved from Ken French’s data library;1 daily value-weighted size-sorted decile stock

M10794_HILARY&MCLEAN_v1.indd 37

18/07/23 5:24 PM

38  Handbook of financial decision making

return data for Canada, Germany, the UK, and Australia compiled from Datastream firm-level return data; and monthly 2-, 5-, 7-, 10-, 20-, and 30-year US Treasury note and bond return data sourced from the Center for Research in Security Prices (CRSP).2 We end all of our data samples before the year 2020 to avoid price volatility associated with the Covid-19 pandemic. The US equity data series starts on January 2, 1926; the international data series starts on January 1, 1990 because there are very few firms with complete data available prior to 1990; and the Treasury bond data start in January 1972 because, until 1971, notes and bonds were offered only in fixed-price sales (see Garbade, 2007).3 We perform additional tests using firm-level daily US and international stock return data, making use of fixed effects for firm and year, with standard errors clustered by firm and day. We measure the SAD effect initially using the clinical onset of and recovery from seasonal depression developed by Kamstra et al. (2015), and later using a new instrument we develop based on Google search data. Table 2.1 contains summary statistics for daily value-weighted size-sorted decile stock return data and monthly government bond return data, and Table 2.2 contains summary statistics for firm-level stock return data. In Table 2.1 we see that mean stock returns and volatility are generally higher for smaller firms (Decile 1 corresponds to the smallest firms), with daily mean decile stock returns in the range of 2–25 basis points and volatility around 1 or 2% for all the countries. In Table 2.2, the firm-level summary statistics reveal much higher share volume and daily mean returns for US firms than for other countries, as well as roughly ten times the number of firm/day observations. Table 2.1  Summary statistics Series

N

Mean

Std. Dev.

Min.

Max.

US 30-Year Treasury US 20-Year Treasury US 10-Year Treasury US 7-Year Treasury US 5-Year Treasury US 2-Year Treasury US 1-Year Treasury US 90-Day Treasury US 30-Day Treasury

576 576 576 576 576 576 576 576 576

0.671 0.685 0.600 0.605 0.557 0.486 0.459 0.412 0.376

3.54 3.06 2.20 1.90 1.54 0.86 0.56 0.32 0.29

−14.738 −10.593 −6.682 −7.039 −5.802 −3.695 −1.721 −0.013 −0.004

17.220 15.235 9.999 10.749 10.612 8.420 5.606 2.131 1.516

25010 25010 25010 25010 25010 25010 25010 25010 25010 25010

0.107 0.099 0.101 0.097 0.093 0.095 0.090 0.088 0.083 0.074

2.44 2.26 2.08 1.98 1.90 1.83 1.76 1.71 1.64 1.50

−34.300 −33.280 −31.120 −31.500 −30.450 −32.380 −30.030 −31.060 −32.360 −28.000

120.990 93.340 75.960 67.470 56.240 59.820 52.150 53.200 46.860 35.170

7565 7552

0.104 0.045

1.13 0.88

−9.248 −17.143

45.355 10.360

US Decile 1 US Decile 2 US Decile 3 US Decile 4 US Decile 5 US Decile 6 US Decile 7 US Decile 8 US Decile 9 US Decile 10 Canada Decile 1 Canada Decile 2

M10794_HILARY&MCLEAN_v1.indd 38

18/07/23 5:24 PM

Seasonality in stock returns and government bond returns  ­39

Table 2.1  (continued) Series

N

Mean

Std. Dev.

Min.

Max.

Canada Decile 3 Canada Decile 4 Canada Decile 5 Canada Decile 6 Canada Decile 7 Canada Decile 8 Canada Decile 9 Canada Decile 10

7555 7554 7552 7556 7555 7554 7553 7551

0.044 0.041 0.035 0.035 0.029 0.034 0.038 0.038

0.93 0.85 0.89 0.93 0.91 0.90 0.86 0.99

−27.610 −10.969 −13.045 −10.381 −8.940 −9.226 −7.007 −9.569

16.682 7.968 9.091 9.760 7.145 8.605 7.800 9.635

Germany Decile 1 Germany Decile 2 Germany Decile 3 Germany Decile 4 Germany Decile 5 Germany Decile 6 Germany Decile 7 Germany Decile 8 Germany Decile 9 Germany Decile 10

7583 7581 7581 7581 7581 7581 7581 7581 7581 7581

0.062 0.039 0.021 0.027 0.028 0.033 0.039 0.056 0.048 0.030

1.39 1.22 1.10 1.08 1.12 1.11 1.15 1.22 1.16 1.33

−12.528 −9.474 −8.893 −7.027 −10.297 −10.527 −10.558 −10.642 −8.179 −9.206

18.382 24.282 11.037 11.702 9.464 10.881 11.504 12.835 11.825 17.348

UK Decile 1 UK Decile 2 UK Decile 3 UK Decile 4 UK Decile 5 UK Decile 6 UK Decile 7 UK Decile 8 UK Decile 9 UK Decile 10

7581 7580 7580 7580 7580 7580 7580 7580 7580 7580

0.069 0.035 0.034 0.049 0.038 0.039 0.048 0.045 0.045 0.036

2.29 0.69 0.67 0.68 0.75 0.83 0.93 0.97 1.05 1.08

−5.283 −5.960 −6.744 −5.211 −7.704 −7.757 −8.674 −8.125 −7.155 −8.805

188.220 5.488 5.443 5.228 8.184 7.526 9.657 8.496 8.929 9.935

Australia Decile 1 Australia Decile 2 Australia Decile 3 Australia Decile 4 Australia Decile 5 Australia Decile 6 Australia Decile 7 Australia Decile 8 Australia Decile 9 Australia Decile 10

7593 7591 7591 7591 7591 7592 7591 7591 7591 7591

0.093 0.066 0.051 0.044 0.051 0.048 0.045 0.042 0.041 0.040

0.87 0.79 0.81 0.85 0.90 0.94 0.94 0.93 0.92 0.99

−10.992 −10.811 −14.591 −8.809 −10.287 −9.536 −9.152 −8.891 −8.130 −8.377

6.729 7.669 7.116 8.521 8.049 7.764 6.401 6.992 6.601 7.170

Notes: The sample period for monthly Treasury returns is January 1972–December 2019; for US daily size-sorted decile stock returns January 1926–December 2019; and for international daily size-sorted decile stock returns January 1990–December 2019. In all cases: Equity returns are value weighted; Decile 1 corresponds to the smallest firms and Decile 10 to the largest; returns are expressed as percentages.

M10794_HILARY&MCLEAN_v1.indd 39

18/07/23 5:24 PM

40  Handbook of financial decision making

Table 2.2 Summary statistics calculated on means of variables firm-by-firm, January 1990–December 2019 Country & Variable

N

Mean Std. Dev. Median Skew.

US Share Price (USD) Volatility Log Volume (shares/1000) Mkt. Cap. (USD billions) Return (%)

41434787 22.140 39417381 0.039 40553502 10.387 41434787 1.370 41416936 0.117

Canada Share Price (USD) Volatility Log Volume (shares/1000) Mkt. Cap. (USD billions) Return (%)

5582327 4855390 5579699 5579699 5582327

40.194 3.776 2.682 0.625 0.003

5.644 3.562 2.529 0.106 0.044

511.75 19.41 1.96 1.50 1.87 0.38 2.53 9.41 0.29 2.54

Germany Share Price (USD) Volatility Log Volume (shares/1000) Mkt. Cap. (USD billions) Return (%)

4400776 39.022 4002602 4.357 4398498 1.074 4398498 1.715 4400776 −0.024

14.584 3.789 0.767 0.112 0.041

97.43 10.48 2.36 2.05 1.81 1.16 6.82 7.28 0.42 0.01

UK Share Price (USD) Volatility Log Volume (shares/1000) Mkt. Cap. (USD billions) Return (%)

5015967 8.210 4490366 2.783 5013184 3.826 5013184 1.275 5015967 −0.008

2.718 2.512 3.472 0.186 0.031

Australia Share Price (USD) Volatility Log Volume (shares/1000) Mkt. Cap. (USD billions) Return (%)

2107727 1897210 2107000 2107000 2107727

1.822 2.636 4.452 0.231 0.034

4.763 2.966 4.427 0.877 0.003

16.349 0.035 10.451 0.193 0.071

Kurt.

Min.

19.37 4.33 45.36 5.00 0.03 1.68 13.08 0.00 2.11 −0.06 −0.39 1.77 7.08 18.52 491.11 0.00 0.49 32.76 2271.63 −9.22

Max. 454.96 0.54 17.77 279.56 40.55

408.22 1.05 12975.24 7.43 0.30 25.25 10.70 −0.19 −1.78 110.95 0.00 47.22 123.03 −3.80 7.27 182.56 1.09 9.29 0.23 1.84 −2.10 61.85 0.00 24.94 −2.97

2428.62 28.94 9.37 84.76 4.87

131.99 36.11 1342.69 1.00 1.43 4.11 46.55 0.43 2.07 0.46 −0.06 −1.90 5.99 13.19 222.90 0.00 0.34 19.01 771.24 −2.74

5232.22 26.40 11.26 126.48 13.68

13.06 6.51 1.46 1.71 2.18 −0.13 3.02 10.89 0.28 −3.00

50.15 1.01 7.10 0.19 −0.33 −2.09 150.98 0.00 57.92 −4.53

165.56 16.85 10.55 55.41 2.89

Notes:  Volatility is calculated using daily high/low prices: Volatility = 200 * (High − Low) / (High + Low). These summary statistics are calculated in a two-step fashion. First, for each firm, the mean share price, volatility, etc. are calculated. The statistics are then calculated such that, for example, the reported median is the average median over firms.

3 ANALYSIS Here we estimate the SAD effect in the US equity and government bond data and in the international equity data, controlling for established regularities.

M10794_HILARY&MCLEAN_v1.indd 40

18/07/23 5:24 PM

Seasonality in stock returns and government bond returns  ­41

3.1 Evidence Based on US Size-Sorted Decile Stock Return and Government Bond Return Data For the US equity data, the model we estimate is:

ri,t = αi + βi,Mon ⋅ Mont + βi,Tax ⋅ Taxt + βi,OR ⋅ ORt + ϵt,

(2.1)

where rt is the daily return for a given equity return series indexed by i; ORt is the SAD onset/recovery variable from Kamstra et al. (2015);4 Mont is an indicator variable for trading days that occur on a Monday; Taxt is an indicator variable set to equal one for trading days in the first month of the tax year (January for the US); ϵ is a disturbance term; and i ranges from 1 to 10 in the case of decile regressions. We estimate the model as  a panel/time-series regression with MacKinnon and White (1985) bootstrap ­heteroskedasticity-consistent standard errors. Before we turn to the regression results, we wish to enable economic interpretation of the coefficient estimate on the SAD onset/recovery variable by examining the variable itself in greater detail. Panel A of Figure 2.1 contains a plot of the SAD onset/recovery variable, represented as a thick solid line. (We consider other the series plotted in Panel A and Panels B–F later.) The SAD onset/recovery variable represents the change in the proportion of people actively suffering from seasonal depression at a given point in time. It is positive when (on balance) people are succumbing to symptoms and negative when (on balance) people are recovering. With SAD-sufferers tending to first experience seasonal depression after the summer solstice, the SAD onset/recovery variable assumes a small positive value beginning around July. The variable then increases in magnitude to a peak around the autumn equinox and declines to around zero around the winter solstice, at which point a small fraction of SAD-sufferers begin recovering from seasonal depression. This leads to a negative value for the SAD onset/recovery variable beginning in January, reaching an annual low around the spring equinox and then rising back to zero around the summer solstice. Panel A of Table 2.3 contains regression results for the daily US value-weighted sizesorted decile equity returns. The SAD onset/recovery coefficient estimates, βˆOR, are uniformly negative across deciles, which indicates that the SAD effect is associated with relatively lower equity returns in summer/fall and relatively higher equity returns in winter/spring. This is consistent with equity returns being influenced by seasonally varying risk aversion arising due to seasonally varying length of day. The SAD onset/ recovery coefficient estimates are generally larger for smaller (riskier) firms and are significant with t-test statistics ranging from 2.0 to 2.5, with the single exception of the onset/ recovery coefficient for the largest decile. An (untabulated) joint test of significance of the onset/recovery coefficient estimates across deciles has a p-value less than 0.1%, indicative of a significant SAD effect for the US equity market overall. The Monday and tax-year coefficient estimates are also strongly significant and are similarly larger in magnitude for smaller firm-size deciles. For the US government bond return data, the model we estimate is:

M10794_HILARY&MCLEAN_v1.indd 41

ri,t = αi + βi,OR ⋅ ORt + ϵt,

(2.2)

18/07/23 5:24 PM

42  Handbook of financial decision making

Panel A: Monthly Averages

Panel B: US

Panel C: Germany

Panel D: Canada

Panel E: UK

Panel F: Australia

Notes: Panel A: The thick solid line (“Clinical”) is the monthly SAD onset/recovery variable developed by Kamstra et al. (2015). The dotted lines represent the monthly average of the full time-series of country-specific Google SVI SAD onset/recovery proxies in Panels B–F. Panel B: This plots the Google SVI onset/recovery proxy for the US. Panels C–F: These plot the Google SVI onset/recovery proxy for Germany, Canada, the UK, and Australia respectively.

Figure 2.1  SAD onset/recovery measures

M10794_HILARY&MCLEAN_v1.indd 42

18/07/23 5:24 PM

43

M10794_HILARY&MCLEAN_v1.indd 43

18/07/23 5:24 PM

.0070

.276 .028 9.73 .000

.0105

βTax Std. Err. t-test p-value

R2

.161 .027 5.98 .000

−.26 .022 −12 .000

βMon Std. Err. t-test p-value

−.25 .023 −11 .000

−.10 .041 −2.3 .010

−.10 .039 −2.5 .007

βOR Std. Err. t-test p-value

2

1

Decile → Statistic ↓

.0066

.131 .026 5.08 .000

−.24 .022 −11 .000

−.09 .039 −2.4 .009

3

Panel A: US Size-Sorted Decile Equity Returns

.0063

.100 .025 4.06 .000

−.23 .021 −11 .000

−.08 .038 −2.1 .018

4

.0065

.085 .024 3.48 .000

−.23 .021 −11 .000

−.08 .038 −2.2 .014

5

.0058

.068 .023 2.90 .002

−.22 .020 −11 .000

−.08 .036 −2.2 .013

6

.0064

.052 .023 2.27 .011

−.23 .020 −11 .000

−.08 .037 −2.3 .010

7

.0055

.036 .023 1.61 .054

−.21 .020 −11 .000

−.07 .036 −2.0 .024

8

Table 2.3 Regression results based on US size-sorted decile equity returns and government bond returns

.0049

.037 .022 1.67 .047

−.19 .019 −9.9 .000

−.07 .035 −2.0 .025

9

.0030

.002 .022 .095 .462

−.15 .019 −7.9 .000

−.05 .034 −1.6 .058

10

44

M10794_HILARY&MCLEAN_v1.indd 44

18/07/23 5:24 PM

.0071

.0090

R2

.0114

1.11 .431 2.58 .005

10-year

.0126

1.01 .349 2.89 .002

7-year

.0155

.908 .290 3.13 .001

5-year

.0114

.435 .167 2.60 .005

2-year

Notes: Panel A results are based on estimating Equation (2.1) using US value-weighted size-sorted decile return data in a panel/time-series model for January 1926–December 2019. βOR is the coefficient estimate on the SAD onset/recovery variable developed by Kamstra et al. (2015); βMon is an indicator variable set to equal one if a trading day occurs on a Monday; and βTax is an indicator variable set to equal one if a trading day occurs in the first month of the US tax year, January. Panel B results are based on estimating Equation (2.2) using US government bond return data in a panel/time-series model for January 1972–December 2019. Decile 1 is the smallest firm-size decile and Decile 10 is the largest. In all cases, we calculate standard errors using MacKinnon and White’s (1985) bootstrap heteroskedasticity-consistent method. We omit the intercept from the tabulated results for brevity.

1.22 .581 2.10 .018

1.59 .668 2.38 .009

βOR Std. Err. t-test p-value

20-year

30-year

Maturity→ Statistic ↓

Panel B: US Government Bond Returns

Table 2.3  (continued)

Seasonality in stock returns and government bond returns  ­45

where rt is the monthly return for a given Treasury bond series indexed by i (for the 2-, 5-, 7-, 10-, 20-, and 30-year maturities), ORt is the SAD onset/recovery variable from Kamstra et al. (2015), and ϵ is a disturbance term. We estimate the model as a panel/timeseries regression with MacKinnon and White (1985) bootstrap heteroskedasticity-­ consistent standard errors. For simplicity we include only the SAD onset/recovery as an explanatory variable; we find qualitatively identical results if we include additional controls, consistent with Kamstra et al. (2015), who consider alternative regression model specifications for estimating the SAD effect in Treasury notes and bonds. Panel B of Table 2.3 contains regression results for the monthly US Treasury return regressions.5 The SAD onset/recovery coefficient estimates generally decline from the longest to the shortest Treasury maturities. Each is strongly statistically significant, as is an (untabulated) joint test of significance of the coefficients across maturities. The coefficient estimate for the SAD effect for Treasury returns is positive, in contrast to the negative coefficient estimate observed for equity returns. This suggests that the SAD effect has a different seasonal influence on Treasuries relative to equities: in the fall, when the SAD effect is associated with relatively lower equity returns, the marginal effect on Treasuries is positive; and in the new year, when the SAD effect is associated with relatively higher equity returns, the marginal effect on Treasuries is negative. These findings are consistent with seasonally varying investor preferences being such that the reduced daylight in the fall leads to greater risk aversion (and greater relative preference for safer securities) than in winter/spring. Figure 2.2 depicts the stability of key regression model coefficient estimates over ­rolling-window subsets of the full sample periods for the US equity and government bond data. Consider first Panels A–C, which are based on US equity return regressions. To produce the plots in Panels A–C we estimate Equation (2.1) sequentially using rolling windows of 60 years of US daily value-weighted size-sorted decile equity return data at a time, updated every 126 days, over the full sample period (January 1926 to December 2019). For ease of plotting, we constrain each of the model coefficients (SAD onset/ recovery, Monday, and tax-year) to be the same across the ten size-sorted deciles. Panels A–C contain the rolling-window estimates of βOR, βMon, and βTax, respectively. Notice from Panel A that the rolling onset/recovery coefficient estimates are consistently negative and significant, generally taking on larger negative values in more recent decades. The Monday effect has been consistently negative and significant over the full sample, appearing to become somewhat diminished over time. The tax-year effect has been consistently positive and significant over the full sample, also appearing to diminish over time. Turning to Panel D of Figure 2.2—the case of the Treasury returns data—to produce the plot, we estimate Equation (2.2) on a rolling-window basis using 25 years of monthly bond data at a time, updated every six months, over the full sample period (January 1972 to December 2019). Again, for ease of plotting, we constrain the SAD onset/recovery coefficient estimate to be the same across all the Treasury bond maturities. We see in Panel D that the coefficient estimate is uniformly positive for Treasuries across all subsamples. (Recall that the equity returns are daily and the Treasury returns are monthly, so the scale of the SAD onset/recovery coefficients in Panel A versus D is not directly comparable.) Overall, the sign of the SAD effect is opposite for equities versus Treasuries; and, while the magnitude of the effect varies somewhat over time, it is reliably negative (positive) for equities (Treasuries), and typically very strongly significant at conventional

M10794_HILARY&MCLEAN_v1.indd 45

18/07/23 5:24 PM

46  Handbook of financial decision making

Panel A: βOR, US Equities

Panel B: βMon, US Equities

Panel C: βTax, US Equities

Panel D: βOR, US Treasurys

Notes: Panels A, B, and C: coefficient estimates for the SAD onset/recovery variable, Monday effect, and tax-year effect, respectively, based on rolling periods of roughly 60 years of US value-weighted daily decile returns data at a time, with estimates updated every 126 periods. The three estimates are constrained to be equal across deciles. Panel C: coefficient estimates for the SAD onset/recovery variable based on rolling windows of 25 years of US monthly government bond return data, with estimates updated every 6 months. This estimate is constrained to be equal across the 30-, 20-, 10-, 7-, 5-, and 2-year maturities. Panels A–D: dashed lines represent a 90% confidence interval around the coefficient estimates.

Figure 2.2 Coefficient estimates for US equity and Treasury return rolling-window regressions significance levels. As we can see from Figure 2.2, however, the tax loss selling and Monday effects have been shrinking over time. 3.2  Evidence Based on International Size-Sorted Decile Equity Return Data We consider now size-sorted decile equity return data for Australia, Canada, Germany, and the UK. This selection of countries reflects a range of latitudes which may be helpful for identification, considering that the SAD effect arises due to seasonality in daylight exposure, and hence should vary in intensity depending on the latitude and hemisphere

M10794_HILARY&MCLEAN_v1.indd 46

18/07/23 5:24 PM

Seasonality in stock returns and government bond returns  ­47

of a given population. The mean center of population is approximately 37°N for the US, 48°N for Canada, 51°N for Germany, and 53°N for the UK. Because variations in daylight are more extreme in higher-latitude locations, we expect stronger SAD effects in the UK, Germany, and Canada than in the US. For Australia, the mean center of population is about 34°S. Because the seasons are offset by six months in the southern hemisphere, we expect the SAD effect to be likewise offset. The regression model we use for the international equity return data is Equation (2.1) with some explanatory variables modified as follows. The tax year begins in April in the UK and July in Australia, and hence the tax-year indicator variable is set to one for trading days in April in the UK and July in Australia. We shift SAD onset/recovery by six months for Australia to account for the fact that in the southern hemisphere the seasons are six months out of phase relative to the northern hemisphere. The decile data are value-weighted returns of individual equities in size-sorted portfolios, with Decile 1 containing the smallest firms. Summary statistics for the international decile returns appear in Table 2.2 and are described in Section 2 above. We estimate the international regression models as panel/time-series models one country at a time. Results appear in Tables 2.4 and 2.5. For each of the northern hemisphere countries, the onset/recovery coefficient is consistently negative for all deciles, just as reported above for the US equity decile returns data. The magnitude is generally largest and most statistically significant for the smallest firmsize decile, and the magnitude tends to decline as firm size increases. These findings suggest that, for the northern hemisphere countries, returns tend to be below average as SAD onset occurs in the fall and above average as people recover from SAD in winter, leading to above average returns for those investors holding equities through the winter. Turning to Australia, we see a strong, statistically significant SAD effect only for the smallest firm deciles; larger firm deciles show no statistical significance or even have a positive (generally insignificant) coefficient. For each country, an (untabulated) joint test of the SAD effect across deciles shows strong statistical significance, at the 0.2% level for Australia and better than 0.1% for the remaining countries, suggesting the SAD effect is significant overall for each country’s equity market. As reported above for the US, both Canada and Australia present mostly strong, statistically significant negative Monday effects, generally declining in magnitude as firm size increases. The Monday effect, however, is not evident in Germany, and in the UK we see a large, negative, statistically significant Monday effect only for mid-size firm deciles. In Canada and Germany, the tax-year effect is evident for smaller deciles but declines in magnitude and significance with firm size. For Australia and the UK, the tax-year effect is absent. The mixed findings across countries for the tax-year effect and the Monday effect are consistent with the findings of Kamstra et al. (2012), who report mixed evidence of these effects across dozens of international exchanges. 3.3  Evidence Based on Disaggregated Firm/Day-Level Equity Return Data To supplement the decile-based analysis reported above, we also employed modern panel/ time-series regressions using the full cross-section of firms rather than size-sorted portfolios. Using the full cross-section of firms in this context entails many millions of data points with cross-sections as large as several thousand firms over about 30 years of

M10794_HILARY&MCLEAN_v1.indd 47

18/07/23 5:24 PM

48

M10794_HILARY&MCLEAN_v1.indd 48

18/07/23 5:24 PM

.092 .036 2.55 .005 .0097

.148 .040 3.67 .000

.0084

βTax Std. Err. t-test p-value

R2

−.19 .030 −6.1 .000

−.19 .049 −3.9 .000

−.14 .030 −4.7 .000

−.25 .055 −4.6 .000

βOR Std. Err. t-test p-value

2

βMon Std. Err. t-test p-value

1

Decile → Statistic ↓

Panel A: Canada

.0101

.105 .035 3.02 .001

−.17 .028 −6.1 .000

−.21 .048 −4.4 .000

3

.0093

.063 .033 1.90 .029

−.18 .028 −6.3 .000

−.19 .049 −3.9 .000

4

.0076

.100 .036 2.80 .003

−.15 .030 −5.1 .000

−.18 .050 −3.7 .000

5

.0072

.050 .037 1.34 .091

−.17 .031 −5.5 .000

−.18 .053 −3.5 .000

6

.0062

.017 .036 .455 .324

−.17 .029 −5.7 .000

−.14 .052 −2.7 .004

7

.0041

.000 .036 .007 .497

−.12 .030 −4.0 .000

−.17 .052 −3.2 .001

8

.0046

−.00 .035 −.07 .470

−.13 .029 −4.5 .000

−.14 .049 −2.9 .002

9

.0004

.003 .041 .067 .473

−.02 .031 −.78 .218

−.08 .056 −1.5 .066

10

Table 2.4 Regression results for Canada and Germany based on size-sorted decile equity returns, January 1990–December 2019

49

M10794_HILARY&MCLEAN_v1.indd 49

18/07/23 5:24 PM

.260 .048 5.39 .000 .0051

.050 .041 1.23 .110

.278 .061 4.58 .000

.0060

βMon Std. Err. t-test p-value

βTax Std. Err. t-test p-value

R2

.0072

.230 .053 4.37 .000

.049 .035 1.40 .081

−.29 .063 −4.5 .000

3

.0043

.190 .048 3.97 .000

.056 .034 1.62 .053

−.17 .062 −2.7 .003

4

.0033

.157 .049 3.22 .001

.033 .036 .927 .177

−.19 .064 −3.0 .001

5

.0032

.143 .049 2.94 .002

.051 .036 1.40 .080

−.19 .063 −3.1 .001

6

.0031

.164 .048 3.38 .000

.044 .038 1.17 .121

−.18 .068 −2.6 .004

7

.0014

.093 .051 1.83 .034

−.01 .040 −.33 .369

−.16 .071 −2.2 .012

8

.0012

.035 .047 .740 .230

.006 .037 .154 .439

−.17 .068 −2.6 .005

9

.0007

−.01 .056 −.24 .407

.031 .043 .721 .236

−.15 .078 −1.9 .026

10

Notes: Results are based on estimating Equation (2.1) using Canadian and German value-weighted size-sorted decile return data for each country individually in a panel/ time-series model. βOR is the coefficient estimate on the SAD onset/recovery variable developed by Kamstra et al. (2015); βMon is an indicator variable set to equal one if a trading day occurs on a Monday; βTax is an indicator variable set to equal one if a trading day occurs in the first month of the tax year (January). We omit the intercept from the tabulated results for brevity. We calculate standard errors using MacKinnon and White’s (1985) bootstrap heteroskedasticityconsistent method. Decile 1 is the smallest firm-size decile and Decile 10 the largest.

.031 .037 .817 .207

−.20 .071 −2.7 .003

−.31 .076 −4.1 .000

βOR Std. Err. t-test p-value

2

1

Decile → Statistic ↓

Panel B: Germany

50

M10794_HILARY&MCLEAN_v1.indd 50

18/07/23 5:24 PM

.087 .026 3.36 .000 .0030

.063 .039 1.60 .055

.0003

βTax Std. Err. t-test p-value

R2

.033 .021 1.59 .056

−.11 .042 −2.7 .003

.053 .038 1.40 .080

−.13 .042 −3.1 .001

βOR Std. Err. t-test p-value

2

βMon Std. Err. t-test p-value

1

Decile → Statistic ↓

Panel A: UK

.0026

.063 .025 2.58 .005

.001 .021 .030 .488

−.13 .040 −3.2 .001

3

.0023

.059 .026 2.25 .012

−.03 .021 −1.2 .108

−.12 .040 −2.9 .002

4

.0041

.041 .028 1.47 .071

−.08 .024 −3.3 .000

−.16 .044 −3.7 .000

5

.0025

.030 .032 .953 .170

−.07 .026 −2.7 .003

−.13 .047 −2.8 .002

6

.0035

.014 .035 .395 .346

−.09 .029 −3.3 .001

−.19 .053 −3.5 .000

7

.0026

−.02 .038 −.54 .294

−.09 .030 −3.0 .002

−.16 .055 −3.0 .001

8

.0017

−.04 .039 −1.0 .158

−.08 .033 −2.4 .008

−.13 .060 −2.2 .015

9

10

.0005

−.08 .043 −1.8 .033

.005 .034 .147 .442

−.06 .064 −.99 .161

Table 2.5 Regression results for the UK and Australia based on size-sorted decile equity returns, January 1990–December 2019

51

M10794_HILARY&MCLEAN_v1.indd 51

18/07/23 5:24 PM

.009 .034 .280 .390 .0026

−.09 .026 −3.5 .000

.080 .041 1.93 .027

.0037

βMon Std. Err. t-test p-value

βTax Std. Err. t-test p-value

R2

.0026

.051 .034 1.50 .067

−.08 .026 −3.2 .001

−.10 .044 −2.2 .013

3

.0009

−.01 .035 −.36 .361

−.06 .025 −2.4 .008

−.02 .046 −.41 .341

4

.0015

−.04 .039 −.92 .178

−.08 .028 −3.0 .001

.023 .049 .480 .316

5

.0014

−.02 .041 −.37 .357

−.07 .029 −2.6 .005

.084 .049 1.73 .042

6

.0009

−.02 .040 −.40 .345

−.07 .029 −2.2 .012

.055 .050 1.10 .135

7

.0006

−.04 .041 −.96 .168

−.04 .029 −1.5 .071

.052 .050 1.05 .148

8

.0008

−.06 .040 −1.5 .065

−.04 .029 −1.5 .073

.063 .049 1.27 .102

9

.0004

−.02 .043 −.50 .307

.005 .030 .155 .439

.094 .053 1.77 .038

10

Notes: Results are based on estimating Equation (2.1) using UK and Australian value-weighted size-sorted decile return data for each country individually in a panel/ time-series model. βOR is the coefficient estimate on the SAD onset/recovery variable developed by Kamstra et al. (2015), shifted by six months for Australia to account for its location in the southern hemisphere; βMon is an indicator variable set to equal one if a trading day occurs on a Monday; βTax is an indicator variable set to equal one if a trading day occurs in the first month of the tax year (April in the UK, July in Australia). We omit the intercept from the tabulated results for brevity. We calculate standard errors using MacKinnon and White’s (1985) bootstrap heteroskedasticityconsistent method. Decile 1 is the smallest firm-size decile, and Decile 10 the largest.

−.10 .024 −4.1 .000

−.04 .043 −.89 .187

−.16 .048 −3.3 .000

βOR Std. Err. t-test p-value

2

1

Decile → Statistic ↓

Panel B: Australia

52  Handbook of financial decision making

daily data. We estimate the SAD effect with disaggregated firm-level daily data, including fixed effects for both firm and year, and with controls for firm price, market capitalization, return volatility, and trading volume, all lagged one period. We conduct robust inference, making use of standard errors clustered by date and firm. The regression model we estimate for each country is as follows:   

3

​​r​  i,t​​ = ​α​  i,Year​​ + ​βMon ​  ​​ ⋅ ​Mon​  t​​ + ​βTax ​  ​​ ⋅ ​Tax​  t​​ + ​  ∑   ​​β​  k,OR ​  ​​  ⋅ ​D​  i,k,t−1​​  ⋅ ​OR​  t​​​ ​ k=1

​ + β​ ME ​  ​​ ⋅ ​ME​  i,t−1​​ + ​βVol ​  ​​ ⋅ ​Vol​  i,t−1​​ + ​βVolat ​  ​​ ⋅ ​Volat​  i,t−1​​ + ​βP​  ​​ ⋅ ​P​  i,t−1​​ + ϵ​ ​ i,t​​.​

(2.3)

Here ri,t is the time-series of returns for firm i; Mont and Taxt are defined as stated above to capture the Monday and tax-year effects; and Di,k,t−1 are indicator variables to capture the market capitalization tercile of a firm i at time t−1. (Di,k,t−1 equals 1 if firm i is in the kth firm market capitalization tercile at time t−1 and 0 otherwise, with Tercile 1 ­corresponding to the smallest firms.) By interacting the market capitalization tercile dummy variable with the SAD onset recovery variable, ORt, we allow the SAD effect to vary across firms in different firm-size terciles. MEi,t−1 is firm i’s market capitalization at time t−1 and Voli,t−1 is firm i’s share trading volume at time t−1 (in thousands). Volati,t−1 is firm i’s return volatility at time t−1, calculated using daily high/low prices: Volatility = 200*(High−Low)/ (High+Low). Pi,t−1 is firm i’s firm price in USD at time t−1. The intercept captures firm and year fixed effects. Tables 2.6–2.8 contain results for the US, Canada, Germany, the UK, and Australia. Column (1) in each country’s set of results presents the simplest model, including only the SAD onset/recovery variables. Column (2) contains results based on the model most comparable to the decile regressions considered above (although here the Monday and tax-year effects are constrained to be equal across all firms, unlike the case with the decile model, which allows variation in these effects across deciles). Column (3) contains results for the most parameterized model, including many additional controls. All of these regressions include fixed effects for year and firm, and we calculate clustered standard errors for each model, clustered by date and firm. In each of the tables we see the SAD effect is statistically significant and strongest for the smallest size tercile (with the exception of the UK, for which estimates are similar across terciles). The SAD effect is smallest in magnitude and significance for the US and Australia, the two countries closest to the equator. The effect is larger and more significant for the three most northerly countries (Canada, Germany, and the UK), even for the most heavily parameterized specification. The SAD effect coefficient magnitudes roughly match those reported above for the decile data. For the US, the Monday effect is weaker than we see in the decile regressions, and the tax-year effect disappears completely. This is likely a result of the restriction that the coefficient value be the same for all firms, and the fact that we are estimating the model over a period for which the Monday and tax-year effects were waning (1990–2019). For Australia, the Monday effect is strongly significant, in contrast to the decile regressions. Canada’s tax-year and Monday effect coefficients are roughly equivalent to averages of those coefficients from the decile regressions. Germany continues to exhibit very strong tax-year effects and no Monday effect, and the UK exhibits stronger tax-year and Monday effects than with the decile regressions.

M10794_HILARY&MCLEAN_v1.indd 52

18/07/23 5:24 PM

Seasonality in stock returns and government bond returns  ­53

Table 2.6 Regression results for the US with firm-level data, January 1990–December 2019

β1,OR β2,OR β3,OR βMon

(1)

(2)

(3)

−0.080∗∗ (0.039) −0.074 (0.060) −0.090 (0.062)

−0.077∗∗ (0.039) −0.071 (0.060) −0.087 (0.062) −0.104∗∗∗ (0.029) 0.036 (0.037)

−0.076∗ (0.041) −0.079 (0.060) −0.093 (0.062) −0.103∗∗∗ (0.030) 0.023 (0.038) −0.0004∗∗ (0.0002) 0.021∗∗∗ (0.003) 2.537∗∗∗ (0.554) −0.002∗∗∗ (0.0002)

Y

Y

Y

41,416,936 0.002 0.002

41,416,936 0.002 0.002

38,873,910 0.002 0.002

βTax βME βVol βVolat βP Firm & Year Fixed Effects N R2 Adjusted R2

Notes: Results are based on estimating Equation (2.3) using US disaggregated firm-level daily data in panel/timeseries models. β1,OR, β2,OR, and β3,OR are estimates that capture the SAD effect for the smallest through largest terciles of firms, respectively, based on interacting tercile indicator variables with the SAD onset/recovery variable developed by Kamstra et al. (2015). βMon is the coefficient estimate on an indicator variable set to equal one if a trading day occurs on a Monday, and βTax is the coefficient estimate on an indicator variable set to equal one if a trading day occurs in the first month of the US tax year. βME is the coefficient estimate on firm market capitalization, and βVol is the coefficient estimate on firm share trading volume. βVolat is the coefficient estimate on firm return volatility calculated using daily high/low prices: Volatility = 200 (High − Low) / (High + Low). βP is the coefficient estimate on firm share price. The regressions include fixed effects for year and firm, and we calculated clustered standard errors for each model, clustered by date and firm. *, **, and *** indicate statistical significance at the 10, 5, and 1% levels, respectively.

Altogether, the results based on the individual firm-level data broadly confirm the findings from the decile regressions. There is a robust SAD effect in the equity markets, strongest in far northerly latitudes and in smaller firms; and there is inconsistent but intriguing evidence of tax-year and Monday effects across countries. The longer time-series of US data, which permits long windows for rolling-window analysis, suggests that some of the variability in the Monday and tax-year effects across countries might arise from the relatively shorter samples in countries other than the US. These two effects appear to be

M10794_HILARY&MCLEAN_v1.indd 53

18/07/23 5:24 PM

54  Handbook of financial decision making

Table 2.7 Regression results for Canada and Germany with firm-level data, January 1990–December 2019 Canada (1) β1,OR β2,OR β3,OR βMon

(2)

(3)

(1)

(2)

(3)

−0.221 (0.046) −0.182∗∗∗ (0.051) −0.125∗∗ (0.052)

−0.221 (0.046) −0.183∗∗∗ (0.051) −0.125∗∗ (0.052) −0.149∗∗∗ (0.027) 0.067∗∗ (0.032)

−0.214 (0.052) −0.173∗∗∗ (0.054) −0.123∗∗ (0.053) −0.156∗∗∗ (0.029) 0.075∗∗ (0.035) −0.002∗∗ (0.001) 0.071∗∗∗ (0.003) 0.009∗∗ (0.004) 0.001∗∗∗ (0.0002)

−0.201 (0.066) −0.182∗∗ (0.075) −0.150∗ (0.082)

−0.180 (0.066) −0.161∗∗ (0.075) −0.130 (0.082) 0.024 (0.039) 0.213∗∗∗ (0.052)

−0.133∗ (0.072) −0.136∗ (0.078) −0.102 (0.083) 0.025 (0.041) 0.216∗∗∗ (0.056) −0.003∗∗∗ (0.001) 0.061∗∗∗ (0.004) −0.0003 (0.005) 0.001∗∗∗ (0.0002)

Y

Y

Y

Y

Y

Y

5,582,327 0.002 0.001

5,582,327 0.002 0.002

4,855,390 0.003 0.003

4,400,776 0.003 0.002

4,400,776 0.003 0.002

4,002,602 0.004 0.003

∗∗∗

βTax βME

∗∗∗

βVol βVolat βP Firm & Year Fixed Effects N R2 Adjusted R2

Germany ∗∗∗

∗∗∗

∗∗∗

Notes: Results are based on estimating Equation (2.3) using disaggregated firm-level daily data in panel/time-series models. β1,OR, β2,OR, and β3,OR are estimates that capture the SAD effect for the smallest through largest terciles of firms, respectively, based on interacting tercile indicator variables with the SAD onset/recovery variable developed by Kamstra et al. (2015). βMon is the coefficient estimate on an indicator variable set to equal one if a trading day occurs on a Monday, and βTax is the coefficient estimate on an indicator variable set to equal one if a trading day occurs in the first month of the tax year. βME is the coefficient estimate on firm market capitalization, and βVol is the coefficient estimate on firm share trading volume. βVolat is the coefficient estimate on firm return volatility calculated using daily high/low prices: Volatility = 200 (High − Low) / (High + Low). βP is the coefficient estimate on firm share price. The regressions include fixed effects for year and firm, and we calculated clustered standard errors for each model, clustered by date and firm. *, **, and *** indicate statistical significance at the 10, 5, and 1% levels, respectively.

waning in the US over the last 40 years (see Figure 2.2), and perhaps also around the world. Furthermore, because these two effects are strongest for small firms (which are more prevalent in non-US equity markets), the variability in tax-year and Monday coefficients across countries may simply reflect differences in typical firm sizes across countries. As to why the tax-year and Monday effects are waning, we can only speculate that

M10794_HILARY&MCLEAN_v1.indd 54

18/07/23 5:24 PM

Seasonality in stock returns and government bond returns  ­55

Table 2.8 Regression results for the UK and Australia with firm-level data, January 1990–December 2019 UK (1) β1,OR β2,OR β3,OR βMon

Australia

(2)

(3)

−0.117 (0.037) −0.132∗∗∗ (0.041) −0.110∗ (0.057) −0.039∗ (0.022) 0.087∗∗∗ (0.030)

−0.102 (0.040) −0.129∗∗∗ (0.044) −0.106∗ (0.057) −0.046∗∗ (0.023) 0.088∗∗∗ (0.032) −0.002∗∗∗ (0.001) 0.062∗∗∗ (0.004) 0.017∗∗∗ (0.004) 0.004∗∗∗ (0.001)

Y

Y

Y

5,015,967 0.002 0.001

5,015,967 0.002 0.001

4,490,366 0.002 0.002

βTax βME

∗∗∗

βVol βVolat βP Firm & Year Fixed Effects N R2 Adjusted R2

(1)

−0.149 (0.035) −0.164∗∗∗ (0.040) −0.142∗∗ (0.055) ∗∗∗

∗∗

(2)

−0.079 (0.039) 0.033 (0.046) 0.056 (0.050)

(3)

−0.074 (0.039) 0.038 (0.046) 0.061 (0.050) −0.059∗∗ (0.024) 0.074∗∗∗ (0.028)

−0.079∗ (0.045) 0.037 (0.048) 0.064 (0.050) −0.061∗∗ (0.026) 0.075∗∗ (0.029) −0.001 (0.001) 0.045∗∗∗ (0.003) 0.008 (0.006) 0.001 (0.0005)

Y

Y

Y

2,107,727 0.003 0.002

2,107,727 0.003 0.002

1,897,210 0.003 0.003

∗∗

∗

Notes: Results are based on estimating Equation (2.3) using disaggregated firm-level daily data in panel/time-series models. β1,OR, β2,OR, and β3,OR are estimates that capture the SAD effect for the smallest through largest terciles of firms, respectively, based on interacting tercile indicator variables with the SAD onset/recovery variable developed by Kamstra et al. (2015), shifted by six months for Australia. βMon is the coefficient estimate on an indicator variable set to equal one if a trading day occurs on a Monday, and βTax is the coefficient estimate on an indicator variable set to equal one if a trading day occurs in the first month of the tax year. βME is the coefficient estimate on firm market capitalization, and βVol is the coefficient estimate on firm share trading volume. βVolat is the coefficient estimate on firm return volatility calculated using daily high/low prices: Volatility = 200 (High − Low) / (High + Low). βP is the coefficient estimate on firm share price. The regressions include fixed effects for year and firm, and we calculated clustered standard errors for each model, clustered by date and firm. *, **, and *** indicate statistical significance at the 10, 5, and 1% levels, respectively.

it may be an anomaly that is becoming cheaper to arbitrage over time as market liquidity and sophistication increase. Sudden awareness of these anomalies (say, with publication in journals) and subsequent efforts to arbitrage them undoubtedly figure into this ­phenomenon (see, for instance, McLean and Pontiff, 2016).

M10794_HILARY&MCLEAN_v1.indd 55

18/07/23 5:24 PM

56  Handbook of financial decision making

4 USING GOOGLE SEARCH VOLUME INDEX DATA TO MEASURE SAD The use of Google Search Volume Index (SVI) data, also known as Google Trends, has become prevalent in behavioral finance research, often to identify retail investor attention (e.g., Da et al., 2011). SVI data have also been used to identify a range of aspects of people’s current situations in a given location, including employment status, interest in purchasing real estate, cinema attendance, and health status.6 Given the utility of SVI data to identify otherwise difficult-to-assess time-varying characteristics of people by location, we anticipate they may be useful in proxying for seasonal depression, which varies by latitude and hemisphere. To form a new proxy for the prevalence of SAD in a given country over time, we collect data by country on Google searches related to “seasonal affective disorder,” selecting the subcategory “disorder” for greater specificity. For each country, we collect search volume data monthly back to the earliest availability (January 2004)—except for Australia, for which we go back to January 2007 due to sparsity of the Australian search data prior to that point. The SVI data potentially reflect the incidence, or prevalence, of SAD at a local country level. To create a proxy analogous to the SAD onset/recovery variable, which reflects changes in SAD incidence, we calculate the change in search volume and normalize the resulting series to match the scale of the SAD onset/recovery variable.7 The resulting SVI proxy for onset/recovery is highly correlated with our primary onset/recovery variable, with roughly 80–90% correlation for countries other than Australia and around 60% for Australia. Figure 2.3 displays the results of a Google Trends search for “seasonal affective disorder” (with the “disorder” subcategory selected, as shown in the top-left corner of the figure). The region is set to the US, the time range is between January 2004 and May 2022, and the search content is “Web Search.” We see a distinct annual seasonal pattern associated with the onset of and recovery from SAD similar to the seasonality we observe in data from the clinical studies Kamstra et al. (2015) used to construct the SAD incidence and SAD onset/recovery variables. The Google SVI output shown in Figure 2.3 also provides cross-sectional detail across US states in the map, showing much greater volume of searches in higher-latitude states, greatest in Alaska and lowest in southern states. Reassuringly, items appearing under “Related topics” are all health-focused (while some items do not pertain directly to SAD, the fact that they are all health-related confirms we are not picking up random searches), and items appearing under “Related queries” are all clearly linked to SAD. Our stock and government bond return samples begin before Google SVI search figures were available. For those pre-2004 dates for which SVI data are unavailable (pre-2007 for Australia), we replace missing SVI values with the average monthly or daily SVI values in the sample for 2004 and later (2007 and later for Australia). The average monthly SVI values we use for each country’s pre-2004 (or pre-2007) values appear in Panel A of Figure 2.1 (depicted by dotted lines), and each country’s SVI time-series for 2004 (or 2007) and later appears in Panels B–F. We re-estimate our decile stock return and Treasury return regression models, Equations (2.1) and (2.2), replacing the SAD onset/recovery variable derived from patients’ clinical symptoms with a country-specific Google SVI estimate of SAD onset/recovery. Table 2.9 contains results pertaining to the coefficient estimates for the SVI onset/recovery proxy.

M10794_HILARY&MCLEAN_v1.indd 56

18/07/23 5:24 PM

Seasonality in stock returns and government bond returns  ­57

Notes: The graph labeled “Interest over time” depicts search volume. The map labeled “Interest by sub-region” depicts search intensity by state, with darker shading corresponding to more intense search activity. The bottom panels list related topics and queries.

Figure 2.3 Google Trends search result for the US for “seasonal affective disorder” and “disorder” subcategory, January 2004–May 2022

M10794_HILARY&MCLEAN_v1.indd 57

18/07/23 5:24 PM

58  Handbook of financial decision making

Table 2.9  Regression results using country-specific Google SVI onset/recovery proxies Panel A: US Government Bond Returns Maturity → Statistic ↓

30yr

20yr

10yr

7yr

5yr

2yr

βOR:SVI t-test p-value

1.61 2.47 .007

1.27 2.20 .014

1.10 2.54 .005

.954 2.58 .005

.793 2.61 .005

.395 2.24 .012

Panel B: Equity Returns Decile → Country & Statistic ↓

1

2

3

4

5

6

7

8

9

10

US βOR:SVI t-test p-value

−.12 −.11 −.10 −.08 −.08 −.08 −.09 −.07 −.06 −.04 −2.9 −2.5 −2.4 −2.1 −2.1 −2.1 −2.3 −1.7 −1.7 −1.2 .002 .006 .008 .017 .017 .017 .012 .041 .042 .117

Canada βOR:SVI t-test p-value

−.42 −.33 −.36 −.33 −.31 −.31 −.25 −.26 −.22 −.12 −6.3 −5.6 −6.1 −5.6 −5.2 −4.8 −4.0 −4.1 −3.8 −1.8 .000 .000 .000 .000 .000 .000 .000 .000 .000 .034

Germany βOR:SVI t-test p-value

−.37 −.24 −.30 −.24 −.24 −.26 −.23 −.22 −.21 −.20 −4.2 −2.9 −4.2 −3.4 −3.3 −3.6 −3.1 −2.8 −2.8 −2.2 .000 .002 .000 .000 .001 .000 .001 .002 .002 .013

UK βOR:SVI t-test p-value

−.32 −.14 −.20 −.19 −.23 −.17 −.24 −.21 −.18 −.10 −3.1 −3.1 −4.2 −4.0 −4.5 −3.3 −4.0 −3.2 −2.6 −1.3 .001 .001 .000 .000 .000 .001 .000 .001 .004 .093

Australia βOR:SVI t-test p-value

−.24 −.11 −.11 −.07 −4.3 −2.0 −2.1 −1.2 .000 .021 .016 .122

−.05 −.84 .201

.054 .842 .200

.043 .678 .249

.007 .117 .454

.010 .157 .437

.059 .901 .184

Notes: Panel A: Results are based on estimating Equation (2.2) using US government bond return data in a panel/ time-series model for January 1972–December 2019. βOR:SVI is the coefficient estimate on the SAD onset/ recovery proxy based on Google SVI data. The intercept is omitted for brevity. Panel B: Results are based on estimating Equation (2.1) on a country-by-country basis using valueweighted size-sorted decile return data in a panel/time-series model (sample period January 1926–December 2019 for the US and January 1990–December 2019 for other countries). As in Panel A, βOR:SVI is the coefficient estimate on the SAD onset/recovery proxy based on Google SVI data. Regression models include a Monday indicator set to equal one if a trading day occurs on a Monday and tax-year indicator variable set to equal one if a trading day occurs in the first month of the tax year. Intercept, Monday, and tax-year coefficient estimates are omitted for brevity. Decile 1 is the smallest firm-size decile, and Decile 10 the largest. In all cases, we calculate standard errors using MacKinnon and White’s (1985) bootstrap heteroskedasticity-consistent method.

M10794_HILARY&MCLEAN_v1.indd 58

18/07/23 5:24 PM

Seasonality in stock returns and government bond returns  ­59

Recall that the SVI proxy for SAD onset/recovery is scaled to match the magnitude of Kamstra et al.’s (2015) SAD onset/recovery variable, facilitating the comparison of results based on the two different onset/recovery measures. For US Treasuries, appearing in Panel A of Table 2.9, there is very little change in results using the Google SVI onset/recovery proxy relative to Kamstra et al.’s onset/recovery measure; in Panel A we see slightly stronger results for the longest-term securities. The regression results using the Google SVI onset/recovery proxy for equity returns appear in Panel B. For the US size-sorted equities, relative to results considered above using the Kamstra et al. (2015) clinical onset/ recovery measure, we now see a slightly stronger SAD effect in the small-firm deciles and little change in the large-firm deciles. The Google SVI onset/recovery measure is strongly positively correlated with the clinical measure, with a correlation coefficient of 0.87, so perhaps the similarity of results is unsurprising. In the international data, we see uniformly stronger results using the country-specific Google SVI measure of SAD onset/ recovery. As before, the SAD effect is strongest for the smallest firms, declining in magnitude and significance as the size decile increases. Australia remains the weakest set of results (not unexpected because its population resides closest to the equator); but now all five of the smallest firm deciles present with negative coefficients, and the three smallest firm deciles show strong SAD effects—very strong for the smallest firm decile, with a t-test statistic value exceeding 4. The largest t-test statistics for US equities are close to 3 in magnitude, over 6 for Canada, and over 4 for Germany and the UK. We also re-estimate the regression model based on firm-level data, Equation (2.3), replacing the clinical onset/recovery variable with the Google SVI-derived measure. Results for the coefficient estimates on the Google proxy for onset/recovery are shown in Table 2.10. We see that the magnitude of the SAD effect based on use of the SVI data is similar in magnitude, and perhaps even a bit larger for each country, than seen in Tables 2.6–2.8. Overall, the SAD effect is evident in all of the countries, strongest and uniformly significant for the smallest third of the firms and generally declining with increasing firm size.

5  CONCLUDING REMARKS A major feature of people’s natural environments is daylight. The changing balance between daylight and darkness through the seasons has consequences for the mood and risk preferences of market participants as they make financial decisions at different points of the year, which in turn has implications for financial markets. In examining US Treasury returns and equity returns for the US, Canada, the UK, Germany, and Australia, we found large, statistically significant seasonal variation in returns consistent with market participants experiencing seasonal variation in daylight, mood, and risk aversion. We showed these effects in size-sorted stock return deciles and in individual firm-level data for all the countries in our sample. We also introduced a novel measure of seasonally varying investor risk aversion based on country-specific Google Trends data. While the seasonal affects we document vary by location, they are distinct from the influence of geography on financial decision making (summarized by Wang in Chapter 6 of this volume), which may arise for reasons such as ease of information acquisition, familiarity, and culture, as opposed to daylight exposure. The connections between

M10794_HILARY&MCLEAN_v1.indd 59

18/07/23 5:24 PM

60  Handbook of financial decision making

Table 2.10 Regression results using firm-level data and country-specific Google SVI onset/recovery proxies, January 1990–December 2019 Country → Statistic ↓

USA

Canada

Germany

UK

Australia

β1,OR:SVI

−0.100∗∗ (0.044) −0.083 (0.062) −0.079 (0.063)

−0.356∗∗∗ (0.062) −0.294∗∗∗ (0.064) −0.177∗∗∗ (0.064)

−0.213∗∗∗ (0.079) −0.191∗∗ (0.083) −0.084 (0.087)

−0.225∗∗∗ (0.047) −0.195∗∗∗ (0.045) −0.133∗∗ (0.059)

−0.095∗ (0.056) 0.031 (0.064) 0.060 (0.064)

β2,OR:SVI β3,OR:SVI

Notes: Results are based on estimating Equation (2.3) using disaggregated firm-level daily data in a panel/time-series model, one country at a time. β1,OR:SVI, β2,OR:SVI, and β3,OR:SVI are estimates that capture the SAD effect for the smallest through largest terciles of firms, respectively, based on interacting tercile indicator variables with the Google SVI proxy for SAD onset/recovery. The model includes controls for the Monday effect, tax effect, firm market capitalization, firm share trading volume, firm return volatility, and firm share price, with coefficient estimates omitted from the table for brevity. The regressions include fixed effects for year and firm, and we calculated clustered standard errors for each model, clustered by date and firm. *, *, and *** indicate statistical significance at the 10, 5, and 1% levels, respectively.

daylight and seasonal variation in mood and risk preferences are believed to arise in part via seasonal changes in neurotransmitters such as serotonin and dopamine (Sohn and Lam, 2005). For a broader discussion of neurofinance, including the role of neurotransmitters in financial decision making more generally, see Chapter 4 in this volume by Payzan-LeNestour. Overall, the seasonal connections between daylight, mood, risk preferences, and international financial market returns highlight the role of human nature in financial decisions at the micro level and aggregate market outcomes at the macro level.

ACKNOWLEDGEMENT We gratefully acknowledge the contributions of our co-authors Ian Garrett, Tan Wang, Mark Weber, Russ Wermers, and especially our late co-author Maurice D. Levi (Mo), with whom we collaborated on many projects over the course of more than two decades. We fondly recall Mo’s creativity, intellect, and bubbly enthusiasm.

NOTES 1. We thank Ken French for making this valuable resource freely available. 2. A lack of return data for international government bonds constrains us to investigate only US government bond data. 3. We apply the following filters on the international equity returns data. We exclude obvious data errors (e.g., price > daily high price, price   r  >  1 − q​(C1). Second, investors find it in their interest to evaluate all N ​ ​ projects.2 We start by determining the capital allocation and expected output given investors’ signals. Then, we compare the surplus under different arrangements for supplying those signals, corresponding to functions of the media, namely (i) distributing data, (ii) selecting projects, and (iii) creating novel information. 2.2  Capital Allocation In period 1, an investor maximizes expected output conditional on productivity signals​ ~​  ​​ ​s​  ​ ​ ​k​ ​, subject to her budget constraint ​∑ N  ​  k​ ​  ≤  K​and to non-negativity con∑ N   ​  E​ ​ ​ v  n=0 ( n| n) n n=0 n straints k ​​ n​ ​  ≥  0​. Under condition (C1) she allocates all her capital to projects deemed successful; if there are no such projects, she invests entirely in the safe project.3 The expected aggregate output, conditional on signals, equals ​qK​in the former case and ​rK​ in the latter. The probability that all projects are deemed failures equals ​​(1 / 2)​ N​​. Therefore, the period-0 expected output (or expected output henceforth) equals Y ​   =  ​(1 − ​(​ _1​2)​  ​ N​)​qK +  (​ _1​2)​  ​ N​ rK  =  Kq − ​K(​ _1​2)​  ​ N​(q − r)​​. The first term—the product of the capital stock with the signal precision—represents expected output if at least one project is deemed successful. Intuitively, a larger ​K​allows one to produce on a larger scale, and a larger ​q​to allocate capital more efficiently across the projects. Both complement each other in that the marginal product of capital rises in the quality of its allocation. The second term, which drops rapidly as the number of projects rises, represents the expected shortfall in case all projects are deemed failures. 2.3  Functions of the Media We turn now to the source of investors’ signals. We consider different arrangements for supplying those signals. In particular, we examine three media companies that produce and distribute signals, each representing a function of the media, and compare the social

M10794_HILARY&MCLEAN_v1.indd 194

18/07/23 5:24 PM

The role of media in financial decision-making  ­195

surplus generated in each case. The social surplus is defined as the sum of expected output minus the cost of producing and analyzing signals. The model is agnostic about how that surplus is shared between investors and the media company, but its implications do not depend on the sharing rule. 2.3.1  Benchmark: no media As a benchmark, we consider the case in which there is no media outlet, so investors gather their own signals. They each pay d​  + a​per project. The surplus equals ​ ​S0​  ​  =  I​(Y − Na − Nd)​​.

(9.1)

We consider next three different media outlets. 2.3.2  Propagation of information: media as distributor The most basic function of the media is to provide raw data. Accordingly, our first arrangement is one in which the media company retrieves data for all projects and makes all those data available. The cost of setting up the repository is ​D​. The surplus equals ​ S ​ 1​  ​  =  I​(Y − Na)​ − Nd − D​.

(9.2)

The data repository is useful because it avoids duplicating across investors and assets the data retrieval costs. The surplus difference relative to no media is ​​S1​  ​ − ​S0​  ​  =  ​(I − 1)​ Nd − D​ . This difference is positive provided the savings on data retrievals—​ ​(I − 1)​ Nd  ≈  INd​for a large population of investors—exceed the repository’s setup cost (​D​). With a fixed setup cost, the social gain grows in the number of investors and projects. 2.3.3  Selection of information: media as editor We now turn to an arrangement whereby a media outlet tells investors which projects to pay attention to. Specifically, it reveals the list of projects deemed successful, together with the associated data. This arrangement represents traditional outlets such as the Wall Street Journal (WSJ) or CNBC, which cover selected stories only. To make use of this information, the investor must analyze the data for the selected projects—she expects N ​  / 2​ such projects. The cost of setting up the outlet is ​E​. The surplus equals ​ ​S2​  ​  =  IY − ​(_​ 2I ​ + 1)​Na − Nd − E​.

(9.3)

In this expression, the analysis cost is paid for all N ​ ​projects by the media outlet and for only half by each investor; the data retrieval costs are borne solely by the media outlet. Thus, the media outlet allows investors to restrict their analysis to the selected projects, thereby cutting analysis costs by a half. The surplus difference relative to the data repository is ​​S2​  ​ − ​S1​  ​  =  (​ I / 2 − 1)​Na − ​(​E − D​)​​. It is positive if the savings on analysis costs, ​(I / 2 − 1)​Na​, exceed the difference in setup costs between the two media functions,​ E − D​.

M10794_HILARY&MCLEAN_v1.indd 195

18/07/23 5:24 PM

196  Handbook of financial decision making

2.3.4  Creation of new information: media as creator We have assumed so far that the media outlet has no technological advantage in producing information. Now we relax this assumption by considering an outlet that unearths original information, that is, information that investors cannot uncover alone. This includes the interpretation, contextualization and discovery of news about projects’ fundamentals (e.g., stock recommendations or investigative journalism) as well as information about investors (e.g., their sentiment vis-à-vis a stock). We model this function of the media by assuming that the signals it produces are both more precise than those available to investors and more costly. Specifically, the media outlet’s signal is correct with probability q​ ′ ​   >  q​, leading to an expected output ​Y′ ​   =  K​q′ ​  − ​K(​ _1​2)​  ​ N(​ ​q′ ​  − r)​  >  Y​. Its analysis cost rises to ​​a′ ​   >  a​, while the data retrieval and setup costs are unchanged at d​ ​and ​E​. Therefore, the surplus equals ​ ​S3​  ​  =  I​Y′ ​  − N​(_ ​Ia2 ​  + ​a′ ​) ​ − Nd − E​.

(9.4)

The surplus difference relative to news editing equals ​S3​  ​ − ​S2​  ​  =  I​(​Y′ ​  − Y)​ − N​(​a′ ​  − a)​  =  I​(1 − ​(​ _​12​)  ​ N​)​K​(​q′ ​  − q​)​)​ − N​(​a′ ​  − a)​. Here the gain lies not in cost savings, as with previous functions, but in the improved accuracy of the media, which allows each investor to allocate capital more efficiently. If the output gain, ​I(​ ​Y′ ​  − Y)​​, surpasses the extra analysis cost,​ N​(​a′ ​  − a)​, then the surplus with a news creator is larger than with a news editor. 2.3.5  Comparative statics For all three functions of the media, the surplus it generates scales with the size of the investor population, I​ ​. This is because each investor saves on data retrieval and analysis and allocates capital more efficiently. Hence media outlets with a broader audience have larger social value. In the case of news propagation and selection, the social gain increases (linearly) in the number of projects ​N​as well as in the costs of data retrieval and analysis, ​d​and ​a​. Intuitively, each additional project needs to go through a costly evaluation. This implies that the media is more valuable in economies with more investment opportunities, and for more complex firms—that is, firms with more data and data that are more difficult to analyze. In the case of news creation, the surplus generated by the media is increasing in the size of the capital stock ​K​because the superior information can be applied on a larger investment scale. Thus, the media adds more value to larger stocks.

3 PROPAGATION OF INFORMATION: MEDIA AS DISTRIBUTOR The most basic function of the media is to facilitate consumers’ access to (existing) information by reducing data retrieval costs (what we called ​d​in the model in Section 2). These media organizations do not create any information, and make little or no editorial decisions. Rather, they either actively disseminate information or provide a platform for users to access the stored information themselves, thus serving as information intermediaries. We consider three major types of media of this sort: (1) search platforms for information

M10794_HILARY&MCLEAN_v1.indd 196

18/07/23 5:24 PM

The role of media in financial decision-making  ­197

repositories; (2) internet-based discussion boards and social media websites; and (3) newswires. 3.1  Search Platforms Information is critical to good financial decision-making. As the amount of data about firms multiplies, search platforms that allow users to quickly filter information stored in a repository grow more valuable. Prior literature has considered two search platforms used by financial decision-makers: Bloomberg terminals and the US Securities and Exchange Commission (SEC) Electronic Data Gathering, Analysis, and Retrieval (EDGAR) system. 3.1.1  For-profit search platforms Bloomberg terminals provide real-time access to market data for Wall Street firms, and are the leading search portals for financial information used by professionals. ­Ben-Rephael et al. (2017) show that the most common users of Bloomberg are portfolio managers, analysts, and traders. Analyzing search behaviors, they find that institutional investors respond more quickly to news than do retail traders. Moreover, Bloomberg search volume is correlated with permanent price changes following earnings announcements and analyst recommendation changes. These results indicate that market efficiency is positively correlated with the search activity of finance professionals. Thomson Reuters is another large information intermediary. It publishes earnings information on its First Call service, as well as analyst forecasts on its Institutional Brokerage Estimate System (I/B/E/S). Schaub (2018) shows that abnormal returns and volume are lower on earnings announcement dates if they are entered into the First Call system with a delay, but larger on the entry date for delayed publications, suggesting that First Call increases market efficiency. Similarly, Akbas et al. (2018) find that forecasts with larger abnormal delays on I/B/E/S are related to slower price discovery, especially for small stocks with less analyst coverage. 3.1.2  The EDGAR search platform Another large repository of information relevant for financial decision-makers is the SEC’s online repository, which is accessed through its Electronic Data Gathering, Analysis, and Retrieval (EDGAR) system. This offers free access, and contains all electronic filings of disclosures mandated by the SEC. Thus, like the Bloomberg search platform, EDGAR serves as an information intermediary with no editorial role. Researchers exploit EDGAR’s server logs to understand whether search activities influence market outcomes. To separate human search activity from automated web scraping performed by robots, they typically design algorithms based on the volume and speed of download requests that originate from a particular IP address. Using a relatively loose filter for robot searches, Drake et al. (2015) show that the type of information requested is heavily skewed toward annual reports, representing 21% of all requests. Requests are more common when firms have negative abnormal returns, and for larger firms with more analysts and media coverage. Using a more restrictive algorithm, Loughran and McDonald (2017) document that the median Center for Research in Security Prices (CRSP) firm has only 1.68 per searches per day for its 10-K during the first

M10794_HILARY&MCLEAN_v1.indd 197

18/07/23 5:24 PM

198  Handbook of financial decision making

quarter after the filing. They also show that requests for annual reports are spread out over the year following their filing, compared to requests for initial public offering (IPO) filings, which are concentrated in time. Using a range of algorithms, Li et al. (2019) show that the number of non-robot EDGAR searches is negatively related to earnings surprise and post-announcement drift. Though this line of research shows that stock market outcomes are correlated with EDGAR searches, it cannot establish any causal role because omitted variables may affect both searches and outcomes. To address this problem, researchers exploit the staggered implementation of EDGAR. When the SEC introduced EDGAR, it randomly assigned firms into ten groups with staggered dates for mandated electronic filing over 1993–1996, thereby creating exogenous variation in a firm’s exposure to EDGAR. Researchers have shown that, thanks to EDGAR: (1) trades by individual investors are more informative about future returns, and sell-side analyst forecasts are more accurate (Gao and Huang, 2020); (2) the cost of capital decreases and equity financing increases (Goldstein et al., 2022); (3) investments are less sensitive to stock prices (Bird et al., 2021); and (4) the dispersion in analysts’ earnings forecasts, short interest, trading volume around earnings announcements, and stock price crash risk decrease (Chang et al., 2022). Overall, these studies show that greater access to information through EDGAR improves market ­efficiency. To identify EDGAR users, researchers match the IP addresses of users on EDGAR server logs to data on the ownership of IP addresses. Gibbons et al. (2021) find that analysts rely on EDGAR in 24% of their updates, and that EDGAR searches are related to lower forecast errors. Crane et al. (2022) show that hedge funds that make more requests on EDGAR earn higher returns, especially those that request information quickly or in larger volumes. In summary, research provides a consistent result: greater information acquisition, enabled by EDGAR and Bloomberg search platforms, leads to better financial decisions. Thus, through its most basic function of providing access to information, media contributes to improving decision-making. Looking forward, firms that release information are beginning to recognize the importance of automated text processing. Since, as discussed above, much of the search volume on EDGAR is performed by robots, firms alter the language they use in their filings to be more easily processed through artificial intelligence (AI) algorithms (Cao et al., 2022). For example, firms avoid words perceived as negative by computational algorithms but not by humans. 3.2  Social Media The advent of social media on the internet presents a new form of information intermediation. Compared to traditional media, the sources of information on social media are relatively unfiltered and decentralized. Thus, we think of social media as a platform that connects information creators with information seekers, with little editorial intervention. Sites hosting discussions about stocks include Yahoo! Finance, RagingBull.com, and TheMotleyFool.com launched in the early days of the Internet, as well as Seeking Alpha and Twitter later on.

M10794_HILARY&MCLEAN_v1.indd 198

18/07/23 5:24 PM

The role of media in financial decision-making  ­199

3.2.1  The beneficial effect of social media Early studies found that message board posts have little to no relationship with future returns, but they do predict future volume and volatility. For example, Antweiler and Frank (2004) report that the number and tone of posts on discussion boards has no substantial effect on future returns after considering transaction costs, but they do predict volatility and volume, even after controlling for articles in the WSJ. Similar results are presented in Tumarkin and Whitelaw (2001) and Dewally (2003) using different time periods and internet sites. In contrast, studies of more recent times indicate that social media posts may predict returns. Chen et al. (2014) find that the negative tone of posts on Seeking Alpha and of users’ replies negatively predict stock returns over the next three months (without subsequently reversing), controlling for analysts’ reports and the tone in the Dow Jones (DJ) News Service articles. Ding et al. (2022) report further that Seeking Alpha posts are correlated with less underreaction to earnings surprises, suggesting that they accelerate the incorporation of earnings news into stock prices. Nekrasov et al. (2022) show that adding visual components to tweets improves market efficiency. To alleviate endogeneity concerns, Farrell et al. (2022) exploit the delayed publication of Seeking Alpha articles, in a setting similar to those of Akbas et al. (2018) and Schaub (2018). Comparing trading behavior after an article has been submitted but before it is published to trading behavior immediately after the article is published, they find that Seeking Alpha articles lead to more informed trading, especially when their authors have stronger track records and academic backgrounds. In contrast to specialized outlets like Seeking Alpha, Twitter attracts a more general audience. On one hand, this allows a wider diversity of opinion. On the other, Twitter users may be less knowledgeable about finance and more prone to trend-chasing. Research indicates that Twitter’s broad base has predictive power for market outcomes. For example Bartov et al. (2018) find that the tone of Twitter posts about Russell 3000 firms predicts future earnings surprises and associated returns; and Gu and Kurov (2020) find that Twitter sentiment predicts a firm’s future stock returns over the next day, controlling for sentiment from traditional media sources. To identify a causal effect, Rakowski et al. (2019) exploit randomly occurring Twitter outages. Comparing periods when the platform is offline to when it is online, they show that Twitter causes increased trading activity and stock returns, regardless of the tone of the tweets. Thus, Twitter posts are not merely a sideshow, but do influence market outcomes. 3.2.2  The detrimental effect of social media Several papers indicate that herding is common on social media, thereby reducing the aforementioned beneficial role. Da and Huang (2020) finds evidence of herding on Estimize.com, an open platform that allows users to make earnings forecasts which are publicized on their website and on Bloomberg terminals. In a field experiment, the authors show that consensus estimates are more accurate when users are prevented from seeing other forecasts before they make their own. Heimer (2016) shows that exposure to social network messages on Myfxbook doubles the magnitude of traders’ disposition effect. Thus, trading mistakes are spread through messages conveyed by social networking sites. Chawla et al. (2022) instrument for the speed of retweeting on Twitter, and find that faster diffusion is correlated with lower spreads and positive price

M10794_HILARY&MCLEAN_v1.indd 199

18/07/23 5:24 PM

200  Handbook of financial decision making

pressure that are reversed the following day. Thus, they argue, Twitter tends to spread stale news. Cookson et al. (2023) study bias in information received by investors on the social network site Stocktwits. They find that users who are bullish on a particular stock are five times more likely to follow a user who is also bullish on the stock. This leads bullish users to see many more bullish messages. These results suggest that investors willingly sacrifice accuracy of financial information for the utility derived from conforming reports. Further discussion of the effect of word-of-mouth communication through social media platforms is presented by Hwang in Chapter 8 of this Handbook. Bernhardt, Ellison, Rennekamp, and White (Chapter 7) also discuss the role of social media platforms in the transmission of false information. 3.2.3  Incentives for companies to use social media In addition to posts by retail investors, the decentralized nature of social media makes it an attractive platform for firms to disseminate disclosures without depending on the editorial choices of traditional media. For example, Blankespoor et al. (2014) and Guindy (2021) show that social media reduces the cost of capital and adverse selection. On the other hand, firms may be strategic in their disclosures. In a sample of S&P 1500 firms, Jung et al. (2018) report that firms are less likely to disseminate bad news via Twitter than good news. 3.2.4  Social media as watchdog Finally, as discussed below in Section 5, traditional media can serve as a watchdog for corporate fraud. Goetzmann (2022) provides a speculative discussion of the role of social media for corporate governance using the meme-stock phenomenon of 2021. He notes that, on the one hand, social media provides a louder voice for small investors to influence corporate actions. On the other hand, social media is also rife with inaccurate rumors that could mislead investors. 3.3 Newswires Newswires are services that distribute to their subscribers press releases written by firms. Their main task is to be fast, and they typically have little editorial input. Some newswires publish nearly any press report for a fee paid by the report’s author. Others work on a subscriber-funded basis and provide some editorial selection of their coverage. Historically, newswires have been used by other media companies as sources of information. With the advent of electronic communication, newswires began distributing to end-users through Bloomberg terminals or directly through their own websites. Researchers generally report a positive association between the use of newswire services and market efficiency. Drake et al. (2014) find that, as the frequency of news flashes about earnings announcements published on newswires increases, the mispricing of cashflows decreases, though the mispricing of accruals is unaffected. Using the predicted coverage of management earnings guidance on the DJ Newswire, Twedt (2016) finds that newswire coverage is associated with larger initial price reactions and faster price discovery. To provide a causal inference, Soltes et al. (2010) use the amount of competing news published on the day of a press release to instrument for exogenous changes in newswire

M10794_HILARY&MCLEAN_v1.indd 200

18/07/23 5:24 PM

The role of media in financial decision-making  ­201

readership. They find that, when readership is higher, bid–ask spreads, share turnover, and idiosyncratic volatility are lower. Rogers et al. (2016) use the abrupt initiation of coverage of Form 4 insider trading filings via DJ newswires as a natural experiment to show that the dissemination of information by newswires influences stock prices and volume beyond the content of the information generated from a Form 4 filing. Boulland et al. (2017) exploit a policy change in Europe that mandated firms to communicate through Englishlanguage electronic newswire services. They document that earnings announcements generate a stronger initial reaction to earnings surprises, higher trading volume, and less subsequent stock price drift after a firm adopts a wire service.

4  SELECTION OF INFORMATION: MEDIA AS EDITOR The second key function of the media is the selection of news to publish. We organize our discussion on the selection role of media around two major questions. First, why does the media perform an editorial role at all? The answer to this question is rooted in investors’ limited ability to absorb all information. In response to this limitation, the media provides a valuable service by selecting the information that its audience will find most valuable. Second, given that the media can help alleviate attention limitations, how does it choose which information to publish and which to ignore? The answer to this question is based on the incentives of the media, which include both profit maximization and private benefits to the owners of media companies. 4.1  Why Media Companies Select News 4.1.1  Theory of limited attention There is now a well-developed body of theory analyzing the causes and consequences of limited attention. One stream, rational attention theory, views agents’ attention as a scarce resource that agents allocate optimally by weighing its opportunity cost against the trading profits they expect from higher attention and superior financial information (Sims, 2003; Van Nieuwerburgh and Veldkamp, 2009, 2010). Traditional models of the demand for information can also be interpreted in this light (e.g., Grossman and Stiglitz, 1980). Another theory stream assumes that inattentive people simply neglect some public signals (Hirshleifer and Teoh, 2003). In the model in Section 2, we capture investors’ limited attention as the analysis cost, a. Alternatively, behavioral theories assume that attention responds to stimuli and has a direct effect on agents’ utility, regardless of its effect on probability distributions and wealth (Loewenstein, 1987). For instance, agents prefer to ignore bad news, even though it improves their decision-making, and instead focus on attention to good news. For a more in-depth review of the literature on limited attention, see Nekrasov, Teoh, and Wu in Chapter 1 of this Handbook. 4.1.2  Evidence of the attention role of the media We start by reviewing studies that document an effect of media coverage on individual investor behavior, and then turn to those that report an effect on market outcomes.

M10794_HILARY&MCLEAN_v1.indd 201

18/07/23 5:24 PM

202  Handbook of financial decision making

4.1.2.1  Investor behavior and the media In an impactful study, Barber and Odean (2008) document that retail investors are net buyers of stocks featured in the DJ News Service, which they label “attention-grabbing stocks,” especially for purchases. This finding is consistent with the observation that retail investors rarely sell short and restrict their sells, unlike purchases, to the stocks they already own. Hence, purchases are chosen from a much larger set of stocks than are sells, and so are more sensitive to media coverage. Fang et al. (2014) show that professional investors too are sensitive to stocks’ media coverage. They document that mutual fund managers tend to buy stocks in the media more heavily than other stocks. Moreover, this tendency hurts their investment performance, consistent with the notion that it is driven by limited attention. Solomon et al. (2014) show that fund managers attract flows from investors by selecting stocks in the media limelight. Engelberg and Parsons (2011) are the first to rigorously show a causal relationship between media and attention. They document that retail trades respond more strongly to firms’ earnings releases when they are covered in the local media. They generate exogenous variations in coverage, holding fixed their information content, by considering, for example, extreme weather events that disrupt the delivery of daily newspapers. Peress (2014) also provides causal evidence by exploiting national newspaper strikes in several countries, exogenous to stock market movements. On strike days, trading volume, the dispersion of stock returns, and intraday volatility decrease. Moreover, the market appears to “miss a beat” and then catch up as returns on the strike’s eve have less predictive power for returns on the strike’s day and more power for the return on the day after the strike. 4.1.2.2  Stock market outcomes and media: the beneficial effect of media Moving from individual decisions to market outcomes, Stice (1991) studies differences in market responses to SEC filings versus delayed media coverage. He shows that stock prices respond to WSJ earnings announcements but not to SEC filings when the filings preceded the WSJ article. Thus, media coverage helps move prices to their full-­information value. Fang and Peress (2009) report that the presence of media coverage, regardless of the content, is associated with a risk discount, as proxied by negative future stock returns, especially among small stocks and stocks with high individual ownership, low analyst following, and high idiosyncratic volatility. Their interpretation is that the media alleviates informational frictions by reaching a broad population of investors, consistent with Merton’s (1987) “investor recognition hypothesis.” A similar finding exists for corporate bonds (Gao et al., 2020). The content of articles also matters. Tetlock et al. (2008) find that their pessimism predicts firms’ earnings and returns. To provide causal evidence of the pricing effects of the media, Klibanoff et al. (1998) use the Net Asset Value (NAV) of closed-end funds to compare fundamental values to prices. In a frictionless market, both NAVs and fund prices would respond equally to new information. Klibanoff et al. find that, while prices typically underreact to changes in NAVs, they do not when relevant news appears on the front page of the New York Times (NYT), indicating that media coverage improved investors’ attention to the news.

M10794_HILARY&MCLEAN_v1.indd 202

18/07/23 5:24 PM

The role of media in financial decision-making  ­203

Peress (2008) employs a related strategy in the context of earnings announcements. He forms pairs of announcements made by the same firm in the same year and generating similar earnings surprise (based on analysts’ consensus forecast) with one announcement covered in the WSJ. He finds that covered announcements display stronger price and trading volume reactions at the announcement and less subsequent drift. This finding suggests that media coverage speeds up the capitalization of earnings news into prices. Guest (2021) confirms the causal nature of these effects by exploiting restructuring events at the WSJ that changed its coverage of earnings news. Fedyk (forthcoming) provides evidence that the prominence of news coverage affects prices. Using exogenous variation in the placement of news stories on Bloomberg terminals, she shows that stocks with more prominent media coverage have higher trading volumes. Thus, the selection decision of media editors has a large, causal impact on financial decision-makers. Another line of research compares stock price behavior following extreme returns for firms with media coverage versus those without. Most studies find evidence that prices tend to reverse less in the presence of media coverage, suggesting that media coverage increases market efficiency (Pritamani and Singal, 2001; Larson and Madura, 2003; Tetlock, 2010; Chan, 2003). Vega (2006) shows that, if the media attracts informed investors, there is no overreaction; but if the news media attracts uninformed investors, stock prices overreact to the media coverage. 4.1.2.3  The detrimental effect of media Shiller (2000) writes that “the history of speculative bubbles begins roughly with the advent of newspapers” (p. 85). Indeed, not all evidence implies that media coverage improves market efficiency. In perhaps the earliest study of the role of media in finance, Niederhoffer (1971) shows that the S&P Composite Index overreacts to stories with large headlines reported in the NYT from 1950 to 1963. Interestingly, he finds that the type of news story—such as a natural disaster, presidential election, or medical breakthrough— does not influence the pattern of index returns. Frank and Sanati (2018) find a similar result in a study of S&P 500 firms covered in the Financial Times from 1982 to 2013. Dougal et al. (2012) report a similar pattern involving the pessimism of WSJ columnists. Because columnists are pre-scheduled, this finding suggests that journalists influence investor behavior. Hillert, Jacobs and Müller (2014) document that firms with higher newspaper coverage exhibit stronger momentum, which subsequently reverses; moreover, this effect is more pronounced if the article’s tone matches the firm’s return over the portfolio formation period. Further evidence of overreaction to television media coverage is presented in Meschke (2004) and Engelberg et al. (2012). The power of the media to influence investor behavior is also apparent when media coverage is misleading, for example, when the media reports stale information. Tetlock (2011) measures the novelty of news published in the DJ Newswire using textual analysis. He finds that stock prices react less to stale news initially but reverse over the following ten days. This suggests that investors overreact to stale news. A similar phenomenon is found for inaccurate news. In Ahern and Sosyura’s (2015) study of merger rumors published in the media, investors differentiate inaccurate rumors from accurate ones: for target firms, the publication date return is 6.9% in accurate rumors, compared to 3.0% for inaccurate rumors. However, these returns reverse

M10794_HILARY&MCLEAN_v1.indd 203

18/07/23 5:24 PM

204  Handbook of financial decision making

by −1.4% over the following ten days, indicating an overreaction. Thus, investors appear to overestimate the accuracy of the media. Schmidt (2020) models the publication of rumors and predicts when they are more likely credible. 4.2  How Media Companies Select News: Incentives Media companies are driven to maximize their profits by increasing revenues and reducing costs. At the same time, managers of media companies may also aim to maximize personal non-pecuniary benefits. Both factors help explain how the media chooses which stories to publish. 4.2.1  Media bias appeals to readers A media company that aims to maximize profits selects stories that it believes will appeal to its readers. In Mullainathan and Shleifer (2005), readers value accuracy but also derive disutility from reading news inconsistent with their prior beliefs. Gentzkow and Shapiro (2006) argue that, when readers are uncertain about the quality of a media outlet, they rationally view news conforming to their beliefs as a sign of quality. Thus, when people hold diverging beliefs, media companies have an incentive to segment the market and slant their coverage towards their priors. In a recent theoretical paper, Goldman et al. (2022) apply these incentives to the financial media. In equilibrium, firms manipulate their reports, but stock prices are not biased on average. Media bias is often manifested in political orientation (Gentzkow and Shapiro, 2010), and recent evidence shows that newspapers are more likely to slant their reporting of firms based on the political position of executives. Fos et al. (2022) find that, from 2008 to 2020, executive teams at S&P 1500 firms became more likely to be affiliated with the same political party. Relatedly, Goldman et al. (2021) show that newspapers are more likely to grant politically aligned firms more favorable coverage in longer articles. For example, WSJ articles are more positive for firms that donate more to the Republican Party. The disparity between the tone of media coverage of the WSJ and the NYT for politically extreme firms leads to an increase in disagreement and, hence, trading volume. Similarly, Rees and Twedt (2022) show that media outlets negatively slant their coverage of earnings announcements if the political affiliation of the firm differs from the media outlet’s affiliation. 4.2.2  Sensational stories appeal to readers Beyond slant, media companies also have an incentive to publish stories that are sensational, even if inaccurate, to appeal to readers’ interests. Shiller (2000) argues that the media is naturally attracted to financial reporting because it provides constant news flow with high stakes and casino-like outcomes. Jensen (1979) argues that the mass media is a producer of entertainment, not information. Core et al. (2008) find evidence that the media sensationalizes CEO compensation, consistent with Jensen’s view. They report that CEOs with high total annual pay are more likely to receive coverage by the media, regardless of whether the compensation exceeds expected pay based on economic fundamentals. Similarly, Ahern and Sosyura (2015) show that media outlets, including well-established sources such as the WSJ and the NYT, are willing to report questionable rumors about mergers if the featured firms have a large appeal.

M10794_HILARY&MCLEAN_v1.indd 204

18/07/23 5:24 PM

The role of media in financial decision-making  ­205

4.2.3  Advertising revenue Many media companies have a local audience, and accordingly report local news. Gurun and Butler (2012) document that local media uses fewer negative words to cover local companies compared to non-local companies. They find evidence that advertising expenses of local companies help explain the favorable reporting. Thus, media companies maximizing profit may also selectively cover local firms to attract advertising revenues. 4.2.4  Firm-originated media coverage A unique aspect of financial media is that it is quite difficult for journalists to uncover new facts. Firms are not, in general, obliged to answer journalists’ questions the way government officials are. This makes it costly for the media to acquire information on its own. Thus, much of the information that the financial media reports originates from firms themselves. This gives firms the power to influence media coverage. Cook et al. (2006) argue that investment bankers have an incentive to promote IPOs through media stories to attract retail investors. They find that the vast majority of news articles about IPOs are non-negative and descriptive. In addition, they show that investment bankers’ compensation is positively correlated with the number of press articles. Bushee and Miller (2012) find that small firms hire investor relations (IR) firms to garner institutional investors and more media coverage. Firms that hire IR firms see their disclosures covered more widely compared to a matched set of firms that do not hire IR firms. Solomon (2012) extends this work to show that IR firms generate more media coverage of more positive press releases. Ahern and Sosyura (2014) study how acquiring firms attempt to influence their stock price by issuing news releases before the announcement of a merger. They show that acquirers issue 10% more press releases during merger negotiations compared to baseline averages. The press releases are more likely to use positive language and words that indicate confidence or exaggeration. While these efforts lead to greater coverage in news media, the tone of the media does not follow the exaggerated tone of the firmoriginated news releases. Gurun (2020) documents that the presence of a media professional on a firm’s board of directors is correlated with receiving more positive media coverage. Alternative drivers of news coverage are activist hedge funds. Dyck et al. (2008) provide evidence that one such, the Hermitage Fund, used the media to increase public pressure on portfolio firms with corporate governance violations in Russia. In particular, AngloAmerican newspapers were more likely to cover a firm’s governance violation if the Hermitage Fund was one of its shareholders. 4.2.5  Managers seek non-pecuniary benefits Rather than maximize profits, media companies may promote their owners’ particular agenda, especially if the companies’ ownership is concentrated. Demsetz and Lehn (1985) argue that, of all industries, the mass media is most likely to have concentrated ownership so that owners can satisfy their preferences through the product they sell. Empirically, the authors find that media firms have significantly more ownership concentration than other firms, driven by family and individual owners. Wagner and Collins (2014) find that, after Rupert Murdoch purchased the WSJ from the Bancroft family in 2007, its editorial pages became more conservative. Kedia and Kim

M10794_HILARY&MCLEAN_v1.indd 205

18/07/23 5:24 PM

206  Handbook of financial decision making

(2021) show further that the WSJ increased its coverage and positive tone of firms connected to Murdoch through overlapping board memberships, relative to the coverage in the NYT.

5  NEW INFORMATION: MEDIA AS CREATOR Though the core role of the media is to distribute information that others have created, in some cases the media expands to news creation—what we modelled in Section 2 as a higher signal precision, q​ ′​. Thinking of content creation and distribution as two separate functions in a supply chain, a media organization performing both activities is vertically integrated. We hypothesize that the media vertically integrates when there are synergy gains from doing so. For instance, creating content in-house may provide synergies because the media firm may have better information on what its customers want. 5.1  Fundamental Information Created by the Media The media can help investors better understand firms’ fundamentals. Using textual analysis, Guest (2021) reports that WSJ articles whose tone is aligned with that of firms’ earnings releases but which differ in their language accelerate the incorporation of news into prices. Language differences include articles being longer, more readable, and specific— including more references to the industry and economy, quoting more sources, and repeating less “stale” news from previous WSJ articles. These findings suggest that articles with substantive editorial content provide information about firms’ earnings by contextualizing or simplifying firms’ disclosures. One type of reporting specific to media companies is investigative journalism. For example, Miller (2006) shows that the media plays a pivotal role in uncovering accounting fraud: in 29% of firms sanctioned by the SEC for accounting violations, the media published articles regarding fraud before a public acknowledgement by the firm or the SEC. Miller finds that the market reaction is significantly stronger for original reporting compared to outsourced news, suggesting that the media serves as an important watchdog for corporate fraud. Dyck et al. (2010) confirm this finding in a study of all reported fraud cases in large US firms from 1996 to 2004. They estimate that the media identified 13% of cases, with a bias towards larger cases. You et al. (2018) find that the media’s watchdog role relies on its independence. They show that Chinese state-controlled media is less critical than market-oriented Chinese media. Farrell and Whidbee (2002) report that investigative reporting influences firm behavior. In a sample of 79 firms over 1982–1997, firms with forced CEO turnovers were the subjects of 76% more WSJ stories about declining earnings than were matched firms without forced turnovers, even after controlling for performance measures. Joe et al. (2009) study firms’ responses to being named among the 25 worst corporate boards in Business Week. They find that, compared to similar firms, named firms subsequently experience a significant increase in the proportion of independent directors and in CEO turnover, and decreases in staggered boards. Relatedly, Kuhnen and Niessen (2012) provide evidence that media coverage influences CEO compensation. They measure public opinion of stock options using the tone of media articles from 1990 to 2010. When the tone is more

M10794_HILARY&MCLEAN_v1.indd 206

18/07/23 5:24 PM

The role of media in financial decision-making  ­207

negative, firms reduce option grants, though the size of overall compensation does not change. In contrast to the watchdog role of the media, Malmendier and Tate (2009) show that positive media coverage of CEOs can have detrimental effects. They document that CEOs who win awards granted by publications such as Business Week and Forbes subsequently underperform and spend more time on outside activities, while compensation and earnings of management increase. 5.2  Meta-Information Created by the Media The media is in a unique position to provide insight into its readers. Understanding the interests, emotions, preferences, or constraints of a large sample of the population is valuable for detecting deviations of prices from fundamentals (e.g., due to herding). It also helps assess whether deviations may widen further before closing, thereby exposing traders to interim losses (e.g., through a short squeeze), and whether a trade is getting crowded. One way that traditional media provides this “meta-information” (i.e., information not about fundamentals but about investors’ beliefs about fundamentals) is through opinion polls. Such polling provides information to readers but may also influence readers’ opinions. For instance, Morton et al. (2015) find that exposure to exit polls decreases voter turnout by a significant margin. With the advent of online media, new tools are available for the media to provide metainformation. For example, it is common for media companies to report the articles that received the most views or downloads. In addition, a feedback loop may be created where media companies tailor news to the preferences of their readers. Social media is also used to infer meta-information about investors’ beliefs and sentiment. Sul et al. (2017) and Liew and Budavari (2017) show that sentiment aggregated from Twitter about S&P 500 firms helps predict future stock returns. Bartov et al. (2018) show that aggregate opinion from Twitter posts predict a firm’s quarterly earnings and announcement returns. Social media can also be a self-reinforcing mechanism, such that investors incorrectly extrapolate what they perceive to be a widely held belief from a narrow set of users, as in the meme-stock episodes of 2021 (Pedersen, 2022). A distinct feature of social media is that it allows investors not only to learn about others’ views but also to explicitly coordinate their trades. Allen et al. (2021) show that the concerted actions of traders through social media platforms leads to price surges and short squeezes in meme stocks in 2021.

6 CONCLUSION This chapter reviewed the literature on the role of the media for financial decision-making. We define the media as the process of intermediation between information creators and information consumers. This definition spans a wide range of media organizations, including search engines for repositories of corporate filings, social media, and traditional print media. Within this definition, we argue that media has three main roles: (1) to distribute unedited information to users; (2) editorial selection of information in response to users’ limited attention; and (3) the creation of new content.

M10794_HILARY&MCLEAN_v1.indd 207

18/07/23 5:24 PM

208  Handbook of financial decision making

In broad brush strokes, our review of the literature suggests that media helps people make better decisions. Greater media coverage improves market efficiency, with faster price discovery and less adverse selection. Media coverage also serves as a monitor of corporate governance. Of course, there are cases in which the media’s own incentives can lead to suboptimal outcomes, such as market overreactions caused by inaccurate information or biased media coverage. Our historical review of the prior literature provides perspectives on a number of avenues for future research on the media and financial decision-making. First, the preponderance of existing research focuses on the role of the media in publicly traded securities. However, the media is also likely to play a large role in many household financial decisions, such as mortgages, school loans, and retirement savings. Second, as data expands exponentially and the cost of data storage falls, the role of the media as an editor is likely to become even more critical for financial decision-making. Third, as consumer data allows firms to more precisely target their customer base, the incentives of the media are likely to exaggerate polarization among the population. In this context, understanding the costs and benefits for consumers, the potential for spreading false information, and the ­potential effects on financial decision-making are more important than ever.

NOTES * We thank Naveen Gondhi, David Hirshleifer, David Solomon, Paul Tetlock, Brady Twedt, and Laura Veldkamp for helpful comments. 1. For instance, the probability of success conditional on a success signal equals ​0.5q / (​ ​0.5q +  0.5​(1 − q)​)  =  q​​ using Bayes’ law. Expected productivity conditional on a success signal, ~​  ​ | ​s​  ​  =  1)​, equals ​q × 1 + ​(​1 − q​)​ × 0  =  q​​. ​E​(​ v  n n 2. Intuitively, this requires the capital stock K ​ ​or the signal precision ​q​to be large, or the number of projects N ​ ​, the safe productivity r​ ​, or the signal cost a​  + d​to be low. The specific condition is K ​ ​(q − r)​ln​(2)​  ≥  (​ a + d)​ 2​ ​ N​​. 3. Under linear technologies and utility, investors are indifferent to how they allocate their capital across projects deemed successful. A small degree of risk-aversion or decreasing return to capital would break this indifference and lead them to distribute capital evenly across projects deemed successful.

REFERENCES Ahern, K., Sosyura D., 2014. Who writes the news? Corporate press releases during merger negotiations. Journal of Finance 69, 241–291. Ahern, K., Sosyura D., 2015. Rumor has it: Sensationalism in financial media. Review of Financial Studies 28, 2050–2093. Akbas, F., Markov, S., Subasi, M., Weisbrod, E., 2018. Determinants and consequences of information processing delay: Evidence from the Thomson Reuters Institutional Brokers’ Estimate System. Journal of Financial Economics 127, 366–388. Allen, F., Haas, M., Nowak, E., Pirovano, M. and Tengulov, A., 2021. Squeezing shorts through social media platforms. Swiss Finance Institute Research Paper. Antweiler, W., Frank, M. Z., 2004. Is all that talk just noise? The information content of internet stock message boards. Journal of Finance 59, 1259–1294. Barber, B., Odean, T., 2008. All that glitters: The effect of attention and news on the buying ­behavior of individual and institutional investors. Review of Financial Studies 21, 785–818.

M10794_HILARY&MCLEAN_v1.indd 208

18/07/23 5:24 PM

The role of media in financial decision-making  ­209

Bartov, E., Faurel, L., Mohanram, P., 2018. Can Twitter help predict firm-level earnings and stock returns? Accounting Review 93, 25–57. Ben-Rephael, A., Da, Z., Israelsen, R., 2017. It depends on where you search: Institutional investor attention and underreaction to news. Review of Financial Studies 30, 3009–3047. Bird, A., Karolyi, A., Ruchti, T., Truong, P., 2021. More is less: Publicizing information and market feedback. Review of Finance, 745–775. Blankespoor, E., Miller, G., White, H., 2014. The role of dissemination in market liquidity: Evidence from firms’ use of Twitter. Accounting Review 89, 79–112. Boulland, R., Degeorge, F., Ginglinger, E., 2017. News dissemination and investor attention. Review of Finance, 21(2), 761–791. Bukovina, J., 2016. Social media big data and capital markets: An overview. Journal of Behavioral and Experimental Finance 11, 18–26. Bushee, B., Miller, G., 2012. Investor relations, firm visibility, and investor following. Accounting Review 87, 867–897. Cao, S., Jiang, W., Yang, B., Zhang, A., 2022. How to talk when a machine is listening? Corporate disclosure in the age of AI. NBER Working Paper No. 27950. Chan, W., 2003. Stock price reaction to news and no-news: Drift and reversal after headlines. Journal of Financial Economics 70, 223–260. Chang, Y., Hsiao, P, Ljungqvist, A., Tseng, K., 2022. Testing disagreement models. Journal of Finance 77, 2239–2285. Chawla, N., Da, Z., Xu, J., Ye, M., 2022. Information diffusion on social media: Does it affect trading, returns, and liquidity? Working paper. Chen, H., De, P., Hu, Y., Hwang, B., 2014. Wisdom of crowds: The value of stock opinions transmitted through social media. Review of Financial Studies 27, 1367–1403. Cook, D., Kieschnick R., Van Ness, R., 2006. On the marketing of IPOs. Journal of Financial Economics 82, 35–61. Cookson, J., Engelberg, J., Mullins, W., 2023. Echo chambers. Review of Financial Studies 36, 450–500. Core, J., Guay, W., Larcker, D., 2008. The power of the pen and executive compensation. Journal of Financial Economics 88, 1–25. Crane, A., Crotty, K., Umar, T., 2022. Hedge funds and public information acquisition. Management Science, https://doi.org/10.1287/mnsc.2022.4466. Da, Z., Huang, X., 2020. Harnessing the wisdom of crowds. Management Science 66, 1847–1867. Demsetz, H., Lehn, K., 1985. The structure of corporate ownership: Causes and consequences. Journal of Political Economy 93, 1155–1177. Dewally, M., 2003. Internet investment advice: Investing with a rock of salt. Financial Analysts Journal 59, 65–77. Ding, R., Shi, Y., Zhou, H., 2022. Social media coverage and post-earnings announcement drift: Evidence from Seeking Alpha. European Journal of Finance, https://doi.org/10.1080/13518 47X.2021.2022508. Dougal, C., Engelberg, J., Garcia, D., Parsons, C., 2012. Journalists and the stock market. Review of Financial Studies, 25, 639–679. Drake, M., Guest, N., Twedt, B., 2014. The media and mispricing: The role of the business press in the pricing of accounting information. Accounting Review 89, 1673–1701. Drake, M., Roulstone, D., Thornock, J., 2015. The determinants and consequences of information acquisition via EDGAR. Contemporary Accounting Research 32, 1128–1161. Dyck, A., Morse, A., Zingales, L., 2010. Who blows the whistle on corporate fraud? Journal of Finance 65, 2213–2253. Dyck, A., Volchkova, N., Zingales, L., 2008. The corporate governance role of the media: Evidence from Russia. Journal of Finance 63, 1093–1135. Engelberg, J., Parsons, C., 2011. The causal impact of media in financial markets. Journal of Finance 66, 67–97. Engelberg, J., Sasseville, C., Williams, J., 2012. Market madness? The case of Mad Money. Management Science 58, 351–364.

M10794_HILARY&MCLEAN_v1.indd 209

18/07/23 5:24 PM

210  Handbook of financial decision making

Fang, L., Peress, J., 2009. Media coverage and the cross-section of stock returns. Journal of Finance 64, 2023–2052. Fang, L., Peress, J., Zheng, L., 2014. Does media coverage of stocks affect mutual fund trading and performance? Review of Financial Studies 27, 3441–3466. Farrell, M., Green, C., Jame, R., Markov, S., 2022. The democratization of investment research and the informativeness of retail investor trading. Journal of Financial Economics 145, 616–641. Farrell, K., Whidbee, D., 2002. Monitoring by the financial press and forced CEO turnover. Journal of Banking & Finance 26, 2249–2276. Fedyk, A., forthcoming. Front page news: The effect of news positioning on financial markets. Journal of Finance. Fos, V., Kempf, E., Tsoutsoura, M., 2022. The political polarization of corporate America. NBER Working Paper No. 30183. Frank, M., Sanati, A., 2018. How does the stock market absorb shocks? Journal of Financial Economics 129, 136–153. Gao, H., Wang, J., Wang, Y., Wu, C., Dong, X., 2020. Media coverage and the cost of debt. Journal of Financial and Quantitative Analysis 55, 429–471. Gao, M., Huang, J., 2020. Informing the market: The effect of modern information technologies on information production. Review of Financial Studies 33, 1367–1411. Gentzkow, M., Shapiro, J., 2006. Media bias and reputation. Journal of Political Economy 114, 280–316. Gentzkow, M., Shapiro, J., 2010. What drives media slant? Evidence from US daily newspapers. Econometrica 78, 35–71. Gibbons, B., Iliev, P., Kalodimos, J., 2021. Analyst information acquisition via EDGAR. Management Science 67, 769–793. Goldstein, I., Yang, S., Zuo, L., 2022. The real effects of modern information technologies: Evidence from the EDGAR implementation. NBER Working Paper No. 27529. Goetzmann, W., 2022. Shareholder democracy, meet memocracy. Financial Analysts Journal 78, 5–8. Goldman, E., Gupta, N., Israelsen, R., 2021. Political polarization in financial news. Working paper. Goldman, E., Martel, J., Schneermeier, J., 2022. A theory of financial media. Journal of Financial Economics 145, 239–258. Grossman, S., Stiglitz, J., 1980. On the impossibility of informationally efficient markets. American Economic Review 70, 393–408. Gu, C., Kurov, A., 2020. Informational role of social media: Evidence from Twitter sentiment. Journal of Banking and Finance 121, 105969. Guest, N., 2021. The information role of the media in earnings news. Journal of Accounting Research, 59(3), 1021–1076. Guindy, M., 2021. Corporate Twitter use and cost of equity capital. Journal of Corporate Finance 68, 101926. Gurun, U., 2020. Benefits of publicity. Quarterly Journal of Finance 10, 2050016. Gurun, U., Butler, A., 2012. Don’t believe the hype: Local media slant, local advertising, and firm value. Journal of Finance 67, 561–597. Heimer, R., 2016. Peer pressure: Social interaction and the disposition effect. Review of Financial Studies 29, 3177–3209. Hillert, A., Jacobs, H. and Müller, S., 2014. Media makes momentum. Review of Financial Studies, 27(12), 3467–3501. Hirshleifer, D., Teoh, S., 2003. Limited attention, information disclosure, and financial reporting. Journal of Accounting and Economics 36, 337–386. Jensen, M., 1979. Toward a theory of the press. In: Bruner, K. (Ed.), Economics and Social Institutions. Martinus Nijhoff, Dordrecht. Joe, J., Louis, H., Robinson, D., 2009. Managers’ and investors’ responses to media exposure of board ineffectiveness. Journal of Financial and Quantitative Analysis 44, 579–605. Jung, M., Naughton, J., Tahoun, A., Wang, C., 2018. Do firms strategically disseminate? Evidence from corporate use of social media. Accounting Review 93, 225–252. Kedia, S., Kim, G., 2021. Impact of media ownership on news coverage. Working paper.

M10794_HILARY&MCLEAN_v1.indd 210

18/07/23 5:24 PM

The role of media in financial decision-making  ­211

Klibanoff, P., Lamont, O., Wizman, T., 1998. Investor reaction to salient news in closed-end country funds. Journal of Finance 53, 673–699. Kuhnen, C., Niessen, A., 2012. Public opinion and executive compensation. Management Science 58, 1249–1272. Larson, S., Madura, J., 2003. What drives stock price behavior following extreme one-day returns? Journal of Financial Research 26, 113–127. Li, R., Wang, X., Yan, Z., Zhao, Y., 2019. Sophisticated investor attention and market reaction to earnings announcements: Evidence from the SEC’s EDGAR log files. Journal of Behavioral Finance 20, 490–503. Liew, J., Budavari, T., 2017. The “sixth” factor: A social media factor derived directly from tweet sentiments. Journal of Portfolio Management 43, 102–111. Loewenstein, G., 1987. Anticipation and the valuation of delayed consumption. Economic Journal 97, 666–684. Loughran, T., McDonald, B., 2016. Textual analysis in accounting and finance: A survey. Journal of Accounting Research 54, 1187–1230. Loughran, T., McDonald, B., 2017. The use of EDGAR filings by investors. Journal of Behavioral Finance 18, 231–238. Malmendier, U., Tate, G., 2009. Superstar CEOs. Quarterly Journal of Economics 124, 1593–1638. Marty, T., Vanstone, B., Hahn, T., 2020. News media analytics in finance: a survey. Accounting & Finance 60, 1385–1434. Merton, R., 1987. A simple model of capital market equilibrium with incomplete information. Journal of Finance 42, 483–510. Meschke, F., 2004. CEO interviews on CNBC. Working paper. Miller, G., 2006. The press as a watchdog for accounting fraud. Journal of Accounting Research 44, 1001–1033. Morton, R., Muller, D., Page, L., Torgler, B., 2015. Exit polls, turnout, and bandwagon voting: Evidence from a natural experiment. European Economic Review 77, 65–81. Mullainathan, S., Shleifer, A., 2005. The market for news. American Economic Review 95, 1031–1053. Nekrasov, A., Teoh, S., Wu, S., 2022. Visuals and attention to earnings news on Twitter. Review of Accounting Studies 27, 1233–1275. Niederhoffer, V., 1971. The analysis of world events and stock prices. Journal of Business 44, 193–219. Pedersen, L., 2022. Game on: Social networks and markets. Journal of Financial Economics 146, 1097–1119. Peress, J., 2008. Media coverage and investors’ attention to earnings announcements. Working paper. Peress, J., 2014. The media and the diffusion of information in financial markets: Evidence from newspaper strikes. Journal of Finance 69, 2007–2043. Pritamani, M., Singal, V., 2001. Return predictability following large price changes and information releases. Journal of Banking & Finance 25, 631–656. Rakowski, D., Shirley, S., Stark, J., 2019. Twitter activity, investor attention, and the diffusion of information. Financial Management 50, 3–46. Rees, L., Twedt, B., 2022. Political bias in the media’s coverage of firms’ earnings announcements. Accounting Review 97, 389–411. Rogers, J., Skinner, D., Zechman S., 2016. The role of the media in disseminating insider-trading news. Review of Accounting Studies 21, 711–739. Schaub, N., 2018. The role of data providers as information intermediaries. Journal of Financial and Quantitative Analysis 53, 1805–1838. Schmidt, D., 2020. Stock market rumors and credibility. Review of Financial Studies 33, 3804–3853. Shiller, R., 2000. Irrational Exuberance. Princeton University Press, Princeton, NJ. Sims, C., 2003. Implications of rational inattention. Journal of Monetary Economics 50, 665–690. Solomon, D., 2012. Selective publicity and stock prices. Journal of Finance 67, 599–638. Solomon, D. Soltes, E., and Sosyura, D., 2014. Winners in the spotlight: Media coverage of fund holdings as a driver of flows. Journal of Financial Economics, 113(1), 53–72.

M10794_HILARY&MCLEAN_v1.indd 211

18/07/23 5:24 PM

212  Handbook of financial decision making

Soltes, E., Leuz, C., Smith, A., Kisin, R., Levitt, S., Solomon, D., 2010. Disseminating firm disclosures. Working paper. Stice, E., 1991. The market reaction to 10-K and 10-Q filings and to subsequent The Wall Street Journal earnings announcements. Accounting Review 66, 42–55. Sul, H., Dennis, A., Yuan, L., 2017. Trading on Twitter: Using social media sentiment to predict stock returns. Decision Sciences 48, 454–488. Tetlock, P., 2010. Does public financial news resolve asymmetric information? Review of Financial Studies 23, 3250–3557. Tetlock, P., 2011. All the news that’s fit to reprint: Do investors react to stale information? Review of Financial Studies 24, 1481–1512. Tetlock, P., 2014. Information transmission in finance. Annual Review of Financial Economics 6, 365–384. Tetlock, P., 2015. The role of media in finance. In: Handbook of Media Economics, Vol. 1, 701–721. Tetlock, P., Saar-Tsechansky, M., Macskassy, S., 2008. More than words: Quantifying language to measure firms’ fundamentals. Journal of Finance 63(3), 1437–1467. Tumarkin, R., Whitelaw, R., 2001. News or noise? Internet message board activity and stock prices. Financial Analysts Journal 57, 41–51. Twedt, B., 2016. Spreading the word: Price discovery and newswire dissemination of management earnings guidance. Accounting Review 91, 317–346. Van Nieuwerburgh, S., Veldkamp, L., 2009. Information immobility and the home bias puzzle. Journal of Finance 64, 1187–1215. Van Nieuwerburgh, S., Veldkamp, L., 2010. Information acquisition and under-diversification. Review of Economic Studies 77, 779–805. Vega, C., 2006. Stock price reaction to public and private information. Journal of Financial Economics 82, 103–133. Wagner, M., Collins, T., 2014. Does ownership matter? The case of Rupert Murdoch’s purchase of the Wall Street Journal. Journalism Practice 8, 758–771. You, J., Zhang, B., Zhang, L., 2017. Who captures the power of the pen? Review of Financial Studies 31, 43–96.

M10794_HILARY&MCLEAN_v1.indd 212

18/07/23 5:24 PM

PART III INSTITUTIONS, FRAMEWORKS, AND TOOLS

M10794_HILARY&MCLEAN_v1.indd 213

18/07/23 5:24 PM

Part III.1 Institutions

M10794_HILARY&MCLEAN_v1.indd 214

18/07/23 5:24 PM

10. Disclosure regulation: past, present, and future S.P. Kothari, Liandong Zhang, and Luo Zuo

1 INTRODUCTION This chapter provides an overview of the theories of financial market regulation, with a focus on corporate disclosures and financial reporting. In a market economy, financial markets play a central role in allocating economic resources, and financial reporting and disclosure are critical to the efficient operation of the financial market. Therefore, the regulation of financial reporting and disclosure is the cornerstone of financial market regulation that potentially shapes financial market efficiency, which, in turn, affects real economic efficiency and welfare. In our description of the key issues of disclosure and reporting regulation, we discuss selected academic research to illustrate the key points and recent developments. However, this is not a comprehensive review of the academic literature on disclosure regulation.1 1.1  What Is Corporate Disclosure Regulation? Broadly defined, disclosure regulation creates and implements disclosure and reporting rules imposed by an authority – that is, a government, legislature, or regulatory agency such as the Securities and Exchange Commission (SEC) in the United States. This definition is consistent with recent work in the accounting literature. For example, Leuz and Wysocki (2016, p. 527) define disclosure regulation as including “a central authority formally creating and interpreting disclosure and reporting rules, monitoring compliance with these rules, and enforcing and imposing penalties for deviation from the rules.” Typically, governments delegate the responsibility of creating and enforcing disclosure rules to certain agencies. For example, in the United States, the SEC is the primary government agency that sets and enforces rules stipulating what information to disclose, when to disclose it, and how to disclose it. In some countries, such as Singapore, the government delegates the major role of disclosure regulation to stock exchanges. 1.2  A Brief History of U.S. Disclosure Regulation The 1929 stock market crash, which resulted in steep investor losses, prompted the U.S. Congress to focus on the role of corporate disclosures in the efficient pricing of securities. Prior to the establishment of the SEC in 1934, the New York Stock Exchange (NYSE) had already started in 1930 to consult the American Institute of Accountants (AIA), forerunner of the American Institute of Certified Public Accountants (AICPA), about its requirements on listed firms’ financial statements. In 1934, a blue-ribbon committee of the AIA put forward a set of five broad accounting principles that were “so generally accepted they should be followed by all listed companies” (Carey 1969, p. 177). The AIA approved these principles and added another principle in 1934. 215

M10794_HILARY&MCLEAN_v1.indd 215

18/07/23 5:24 PM

216  Handbook of financial decision making

The U.S. Congress passed the Securities Act and the Securities Exchange Act in 1933 and 1934, respectively, and the SEC was created by the latter act. Since then, the SEC has delegated the responsibility of formulating Generally Accepted Accounting Principles to the Committee on Accounting Procedures (CAP) from 1939 to 1959, to the Accounting Principles Board (APB) from 1959 to 1973, and to the Financial Accounting Standards Board (FASB) from 1973 to the present. The Sarbanes–Oxley Act of 2002 led to the creation of the Public Company Accounting Oversight Board (PCAOB) to oversee audit firms and their audits. The SEC administers various statutes, including the Securities Act of 1933, the Securities Exchange Act of 1934, the Trust Indenture Act of 1939, the Investment Company Act of 1940, the Investment Advisers Act of 1940, the Securities Acts Amendments of 1964 and 1975, the Williams Act of 1968, the Sarbanes–Oxley Act of 2002, the Dodd–Frank Wall Street Reform and Consumer Protection Act of 2010, and the Jumpstart Our Business Startups Act of 2012. The SEC’s governing statutes establish its authority to make rules that have the force of law within certain limits. These governing statutes require that the SEC justify its proposed rules as necessary to protect investors and promote efficiency, competition, and capital formation. For a more comprehensive discussion of securities regulation in the United States, see Watts and Zuo (2016), Mahoney (2021), and Zeff (2021). 1.3 Organization The remainder of this chapter is organized as follows. In Section 2, we discuss the theories of disclosure regulation, including the justifications for and the political economy of disclosure regulation. Section 3 presents several examples of disclosure regulation and empirical evidence on economic consequences. Section 4 discusses the current debate on environment, society, and governance (ESG) disclosure regulation. Finally, in Section 5, we conclude by offering some tentative thoughts on emerging issues of disclosure regulation.

2  THEORIES OF DISCLOSURE REGULATION In this section, we discuss (1) potential justifications for disclosure regulation in the financial market, and (2) the political economy of disclosure regulation. 2.1  Justifications for Disclosure Regulation In this subsection, we first describe the information asymmetry problems in the financial market that form the basis of potential demands for disclosure regulation. Then, we discuss whether market forces, private contracts and enforcement, and alternative non-regulatory monitoring mechanisms can help solve the information problems. Finally, we discuss disclosure costs and externalities as potential justifications for disclosure regulation.

M10794_HILARY&MCLEAN_v1.indd 216

18/07/23 5:24 PM

Disclosure regulation: past, present, and future  ­217

2.1.1  Information problems In information economics, the demand for information stems from two major intertwined problems of information asymmetry: adverse selection and moral hazard. Under adverse selection, corporate managers might have better information about firm value than outside investors have. This constitutes an information advantage for the managers. Armed with this information advantage, self-interested managers have an incentive to inflate the performance and prospects of their firms, and thereby maximize their own wealth (via stockholdings, insider trading, etc.). As a result, investors might not be able to tell the difference between good and bad firms. Rational investors therefore would pay the “average” price for all firms. Good firms would then be undervalued and would exit, which meant that only bad firms would remain. In an extreme case, this “lemons” problem could lead to the collapse of the capital market (Akerlof, 1970; Myers and Majluf, 1984). Such a collapse can be averted if investors receive information that enables them to discriminate between good and bad firms. The moral hazard problem in the corporate context occurs when corporate managers’ actions are unobservable to investors after investors have invested in a firm and the managers have incentives to expropriate the invested funds (Berle and Means, 1932; Jensen and Meckling, 1976). For example, managers might use the proceeds from the sale of equity in a firm to pay themselves excessively, indulge in excessive perquisites and employment benefits (for example, using corporate jets), or overinvest to build a corporate empire. Alternatively, if investors acquire a debt stake in the firm, shareholder-managers can expropriate debt holders by issuing more senior debt or taking excessive risks. These actions are beneficial to managers and harmful to the interests of outside investors. Investors protect themselves against expropriation through contracts and through access to reliable information that helps them monitor the behavior of corporate managers and protect their interests in the firm. 2.1.2 Information production: market forces, private contracts and enforcement, and non-traditional players However, even if we recognize the importance of information for a well-functioning capital market, it is not straightforward that mandatory disclosures are necessary or desirable. As noted by an SEC economist, “a common mistake in evaluating the net benefits of a regulation to society is to assume that a state of laissez-faire would arise in the absence of regulation” (Alexander and Lee, 2004, p. 418). Similarly, in Mahoney (2021, p. 3), the “unregulated” market is subject to “generally applicable rules of property, contract, tort, agency, and criminal law enforced by general-purpose police, prosecutors, and courts.” Therefore, even in the absence of government regulation, market forces and private litigation can potentially address the information problems. In a competitive managerial labor market, a manager’s reputation will be damaged and her compensation might decrease if the market discovers that the manager has released incomplete, biased, or false information to the capital market. In addition, the capital market may reward firms that make truthful disclosures with higher share prices and lower cost of financing. These market forces can serve as disciplinary or reward mechanisms that motivate managers to voluntarily disclose information. Moreover, the corporate control market and the product market can also deter managers from taking v­ alue-destroying actions (e.g., Gompers et al., 2003; Giroud and Mueller, 2011).

M10794_HILARY&MCLEAN_v1.indd 217

18/07/23 5:24 PM

218  Handbook of financial decision making

Firms and capital providers frequently enter into private contracts to address the problem of information asymmetry (Coase, 1960; Roberts, 2015; Schoenfeld, 2020). For example, securities issuers offer private contracts to commit to full disclosure and guarantee its accuracy. Private contracts can help achieve an efficient equilibrium as long as the courts can enforce these contracts. Similarly, to address the moral hazard problem, investors and managers can sign shareholder or debt contracts based on observable information and employ information intermediaries such as external auditors to enhance or assure the accuracy of the information used in those contracts. In addition to market forces and private contracts, recent research suggests that some non-traditional players in the capital market, such as employees, can help alleviate the information asymmetry problems. These players’ incentives for monitoring and information production stem from their implicit claims in the firm. They can also be driven by certain monetary or non-monetary rewards. For example, employees have incentives to monitor because their working conditions, job security, and prospects hinge on firms’ actions. They are also motivated to blow the whistle when there is a large financial reward for doing so.2 Moreover, employees’ cost of obtaining information might be low because such information is often a by-product of their normal work. Dyck et al. (2010) show that employees play an important role in detecting and revealing corporate fraud. Huang et al. (2020) find that employees’ predictions of their companies’ business outlook from Glassdoor.com are informative in predicting future operating performance. The media and the short sellers are among other prominent non-traditional players in corporate monitoring and information discovery. The media are motivated to uncover and disseminate newsworthy, often negative information on public firms to increase the circulation and readership of their products or services (Miller, 2006). Using local newspaper closures as an exogenous shock, Heese et al. (2022) show that the local press is an effective monitor of corporate misconduct. Short sellers also have financial incentives to discover and disseminate negative information about firms (Boehmer et al., 2020). Fang et al. (2016) show that (the prospect of) short selling helps curb earnings management, detect fraud, and improve the informativeness of stock prices. Therefore, for disclosure regulation to be desirable, it should be the case that market forces, private contracts and courts, and various non-traditional players must be unable to fully resolve the information asymmetry problems. Incidents of corporate misreporting and fraud appear to be consistent with the existence of residual information problems in the capital market. For a broader population of U.S. firms, Kothari et al. (2009) find that firms on average tend to withhold or delay the disclosure of bad news relative to good news (despite the existing regulatory system). Market frictions, sometimes introduced by firms themselves, can prevent market forces from eliminating information asymmetry problems. Limited arbitrage in the financial market and the bounded rationality of market participants are examples of frictions that hinder the market’s ability to discipline corporate managers and motivate fair and full disclosure. Corporate managers can also adopt various anti-takeover provisions to shield themselves from the discipline of the corporate-control market. Moreover, Narayanan (1985) and Stein (1989) observe that managers concerned about their labor-market reputation or current share prices might take actions, such as real earnings management, to boost measures of short-term performance at the expense of long-run shareholder value.

M10794_HILARY&MCLEAN_v1.indd 218

18/07/23 5:24 PM

Disclosure regulation: past, present, and future  ­219

In addition, optimal private contracts are difficult to achieve because it is difficult and costly to specify the rights, obligations, and remedies of the contracting parties in every possible state of nature. Worse still, managers can distort the information used in contracts, and the incentive problems of auditors themselves may hinder their ability to ensure reliability (e.g., Ege and Stuber, 2022). Even if a complete contract can be written, enforcement by the courts can be costly, biased, corrupted, or politically motivated (George, 1998; Shleifer, 2012). The monitoring and information roles of non-traditional monitors are also unlikely to fully address the residual information problems left by market forces and private contracts and enforcement. First, because these players are mostly outsiders, they cannot have access to all information. Second, even if they have information, corporate managers can find ways to increase the cost of information collection and dissemination. For example, recent decades have seen an increase in so-called strategic lawsuits against public participation (SLAPPs) – lawsuits brought by corporations against the news media, short sellers, and individuals with the primary purpose of stifling the public dissemination and discussion of negative information about firms (e.g., Lee et al., 2023). 2.1.3  Costs of disclosures and externalities Even though market forces, private contracts, and non-traditional players can ensure optimal information production from an individual firm’s perspective, there can be an underproduction of information from a society’s perspective as disclosure is costly and has externalities. Therefore, the cost-benefit optimal level of disclosure for a firm may not be the cost-benefit optimal level for society. In Verrecchia’s (1983) partial disclosure theory, the cost of disclosure is an important friction to full disclosure. In addition to the direct monetary and time costs (managerial opportunity costs) of collecting, processing, certifying, and disseminating information, disclosures also carry indirect costs, such as proprietary cost. Proprietary cost of disclosure refers to the concern that a firm’s disclosures to the capital market can damage the firm’s competitive position in the product market. It is arguably the most important indirect cost of corporate disclosure, and the proprietary cost hypothesis predicts that firms with higher proprietary costs tend to disclose less. However, earlier research on the association between product market competition and disclosure provides largely mixed evidence on the proprietary cost hypothesis (Beyer et al., 2010). A growing body of recent literature uses quasi-natural experiments and refined proxies of proprietary costs and disclosure to study the causal effects of proprietary costs on the level of corporate disclosure (Lang and Sul, 2014). For example, Li et al. (2018) exploit the staggered adoption of the inevitable disclosure doctrine (IDD) as an exogenous shock that increases the proprietary cost of disclosure. They find that firms respond to IDD adoption by reducing the disclosure of major customer identities. Using the Uniform Trade Secrets Act setting, Glaeser (2018) finds that firms that rely more heavily on trade secrecy disclose less proprietary information but more non-proprietary information. However, the total effect of trade secrecy is to decrease corporate transparency. Bernard et al. (2018) find that private firms manage firm size downward to avoid size-based disclosure requirements, providing evidence on the economic significance of the proprietary cost of financial statement disclosures. Overall, recent literature provides clear evidence of the significance of proprietary cost and its negative effect on information production.

M10794_HILARY&MCLEAN_v1.indd 219

18/07/23 5:24 PM

220  Handbook of financial decision making

While it is theoretically sound that corporate disclosures have externalities and spillover effects, the evidence on disclosure externalities has been relatively sparse. Nevertheless, it is accumulating (e.g., Badertscher et al., 2013; Bernard et al., 2020; Breuer, 2021; Durnev and Mangen, 2020; Glaeser and Omartian, 2022; Kim and Valentine, 2021, 2023). For example, Bernard et al. (2020) show that rivals’ disclosures help facilitate investment and product decisions, including acquisition and product differentiation strategies. Kim and Valentine (2021) find an increase in innovations for firms whose rivals reveal more information in patent disclosures and a decrease in innovation for firms whose own disclosures are divulged to competitors, consistent with both positive externalities and proprietary costs of disclosure. Kim and Valentine (2023) find that public firm innovation-relevant disclosures have a positive effect on future patent sales by other parties that consume these disclosures, consistent with financial disclosures generating positive information externalities. Overall, recent studies have provided some initial evidence on the existence of externalities of corporate disclosure. Since disclosing firms bear all the costs of disclosure but cannot collect revenue from parties that benefit from disclosure externalities (public goods), they tend to disclose less information than the socially optimal level. This information underproduction problem caused by externalities represents another important justification for disclosure regulation. 2.1.4 Summary Overall, although market forces, private contracts and enforcements, and various other non-regulator stakeholders can help mitigate the information problems in the capital market, they are unlikely to be able to fully resolve the problems. In addition, the public good nature of corporate disclosure combined with its non-trivial costs suggests that the optimal level of disclosure for firms is unlikely to be optimal from a society’s perspective. Under these circumstances, disclosure regulation might be justified. 2.2  The Political Economy of Disclosure Regulation Assuming that disclosure regulation is necessary to resolve residual information problems in the capital market, the next question is whether the regulatory system can be trusted to design and enforce disclosure rules that maximize social welfare. In this subsection, we briefly review key theories of economic regulation and discuss some salient evidence on their relevance for corporate disclosure and financial reporting regulation. 2.2.1  Public interest theory of regulation The public interest theory emerges naturally from the justifications for regulation discussed in Section 2.1. It assumes that regulators can resolve or strive to resolve the residual information problems by designing and enforcing regulation that maximizes social welfare (Pigou, 1920). According to this theory, governments and their regulatory agencies are benign, free from private interests, and capable of estimating the social costs and benefits of regulation. While the public interest theory might not be a precise description of what governments do, it does prescribe an ideal of how regulation should be carried out. More importantly, it serves as an overarching motivation for researchers to evaluate the aggregate costs and benefits of disclosure regulation. Prior literature in accounting and finance tends to examine the effect of disclosure regulation on firm-level

M10794_HILARY&MCLEAN_v1.indd 220

18/07/23 5:24 PM

Disclosure regulation: past, present, and future  ­221

costs and benefits, and there is limited evidence on the net aggregate effect of disclosure regulation. One notable exception is Breuer (2021), who examines the effect of reporting and auditing mandates on aggregate resource allocation. The author finds that financial reporting mandates facilitate ownership dispersion in capital markets and spur competition in product markets. However, the effects of reporting mandates on aggregate productivity and growth are ambiguous. 2.2.2  Capture theory of regulation The capture theory of regulation challenges the public interest theory’s assumption of a benevolent and competent government (Stigler, 1971; Peltzman, 1976). Under the capture theory, self-serving regulators are captured by various interest groups who compete for and against regulation (Becker, 1983). The resultant regulation represents regulators’ own utility-maximizing choice, depending on the relative pressure applied by different interest groups. Therefore, the regulation is unlikely to be socially optimal under the capture theory. In the context of disclosure regulation, the accounting and finance literature has examined whether firms can influence the SEC’s regulatory decisions (e.g., Ramanna, 2008; Correia, 2014; Heese et al., 2017; Mehta and Zhao, 2020; Thompson, 2022). For example, consistent with regulatory capture, Correia (2014) finds that firms use political contributions and lobbying to establish long-term connections to powerful politicians, and these politicians then influence SEC enforcement actions to the benefit of their connected firms. In contrast, Heese et al. (2017) find that political connections are positively associated with the likelihood of firms receiving SEC comment letters, inconsistent with SEC capture. Thompson (2022) examines the relation between political connections and Confidential Treatment (CT) Orders, which are regulatory exemptions issued by the SEC that permit firms to redact information. The author explores a regime shift triggered by the Congressional investigation and hearing into the SEC’s CT process in late 2009 and early 2010 following American International Group’s CT request to redact its Troubled Asset Relief Program contract with the Federal Reserve Bank of New York. Thompson finds that CT requests from politically connected firms are less likely to be rejected before the regime shift but more likely to be rejected following the regime shift. Overall, the evidence for SEC capture is somewhat mixed, depending on the type and time of the SEC decisions examined. 2.2.3  Ideology theory of regulation The ideology theory of regulation posits that regulators are neither as benevolent as suggested by public interest theory nor as self-serving as assumed in capture theory (Grossman and Helpman, 1994; Austen-Smith, 1995). Under the ideology theory, regulators have ideologies but are open to lobbying from constituents with specific knowledge (Kothari et al., 2010). Bischof et al. (2020) provide some of the first evidence that political ideology plays a role in the politics of accounting standard setting. They show that, in addition to special interest pressure, ideology explains politicians’ stance in the debate about fair value accounting and the expensing of employee stock options. In addition, political ideology is likely to play a stronger role in disclosure and accounting rules that have strong real or social consequences, such as bank accounting rules that determine regulatory interventions (e.g., Yue et al., 2022).

M10794_HILARY&MCLEAN_v1.indd 221

18/07/23 5:24 PM

222  Handbook of financial decision making

2.2.4 Summary Overall, evidence from the accounting and finance literature appears to provide partial support for all three theories of regulation in the context of setting and enforcing disclosure and financial reporting rules. However, the empirical literature on the political economy of disclosure regulation is still relatively nascent in the disclosure literature. This area of investigation may merit further efforts, particularly about the aggregate effects of disclosure regulation.

3  EXAMPLES OF DISCLOSURE REGULATION This section uses three examples to illustrate the economic consequences of disclosure regulation. These three disclosure mandates relate to the production, dissemination, and presentation of corporate disclosures and are widely studied in academic research. We use them to highlight how disclosure mandates affect not only firms’ information environment but also corporate investment decisions. As mentioned earlier, the existing research focuses primarily on the firm-level effects of disclosure mandates, but there is some limited evidence on the aggregate effects of disclosure regulation. 3.1  Regulation of Information Production How often should firms report their financial statements? While the United States has required quarterly reporting since 1970, regulators around the world still vigorously debate whether to adopt or abolish quarterly reporting. For example, the European Union started to require companies to produce narrative interim management statements on a quarterly basis in 2004 but stopped this requirement in 2013.3 Singapore had required quarterly reporting for some of its firms since 2003 but scrapped it in 2020.4 In 2018, U.S. President Donald J. Trump asked the SEC to evaluate the quarterly reporting system and consider moving back to the semi-annual reporting system.5 Given the practical relevance of this topic, a large academic literature has emerged to decipher the determinants and economic consequences of financial reporting frequency. Early research examines whether a firm’s voluntary provision of interim financial reports prior to SEC regulation reflects the cost-benefit tradeoffs facing managers under an agency framework (Leftwich et al., 1981). This work is largely descriptive and yields some puzzling results. For example, semi-annual reporters on the American Stock Exchange in 1948 exhibit a higher leverage ratio than both annual and quarterly reporters. This result is inconsistent with the agency view that firms with a higher leverage ratio need to commit to more intensive monitoring by reporting more frequently. Researchers attribute these puzzles to methodological problems in association studies. To overcome these problems, researchers often utilize the reporting frequency change in the United States as the empirical setting. The SEC started to require listed firms to provide annual reports in 1934, semi-annual reports in 1955, and quarterly reports in 1970. Researchers can use a sample over this period (e.g., 1951–1973) for empirical analysis. Such a sample has at least two desirable features. First, there is significant cross-sectional and time-series variation in reporting frequency. Second, since some firms had voluntarily adopted more frequent reporting prior to the SEC mandate,

M10794_HILARY&MCLEAN_v1.indd 222

18/07/23 5:24 PM

Disclosure regulation: past, present, and future  ­223

researchers can use these voluntary adopters as controls in a difference-in-differences analysis. Capital markets research has examined how the frequency of financial reporting affects the information content of annual reports, earnings timeliness, and the cost of equity. Evidence shows that the stock return variability at the annual report announcement date in the “annual-plus-quarterly-reports” environment is lower than that in the “annualreport-only” environment (McNichols and Manegold, 1983). This finding is consistent with the theoretical prediction that interim reporting leads to a reduction in the information content of annual reports. However, semi-annual reporters and quarterly reporters do not exhibit significant differences in earnings timeliness, that is, the speed with which accounting information is reflected in stock prices (Butler et al., 2007). Further evidence shows that firms that voluntarily adopted quarterly reporting experienced an increase in earnings timeliness but firms that increased reporting frequency due to the SEC mandate did not (Butler et al., 2007). Follow-up work shows that more frequent reporting leads to a reduction in information asymmetry and the cost of equity for both mandatory and voluntary adopters (Fu et al., 2012). Overall, there is evidence that more frequent financial reporting improves the information environment. Another line of work aims to understand the real effects of financial reporting frequency (Roychowdhury et al., 2019). There are two opposing forces of increasing financial reporting frequency. On the one hand, an increase in reporting frequency can enhance transparency and monitoring. On the other hand, frequent reporting can reduce managers’ decision horizon and induce myopia. Research has examined various managerial decisions, including capital investment (Kraft et al., 2018; Kajüter et al., 2019; Nallareddy et al., 2021), real activities manipulations (Ernstberger et al., 2017), cash holdings (Downar et al., 2018), banks’ loan portfolio quality (Balakrishnan and Ertan, 2018), and corporate innovation (Fu et al., 2020). The evidence is mixed. For example, U.S. firms decreased their capital investment levels after a reporting-frequency increase (Kraft et al., 2018), but no such evidence is found for firms in the United Kingdom or Singapore (Kajüter et al., 2019; Nallareddy et al., 2021). While prior research often focuses on the firm-level effects of reporting frequency mandates, more recent work has started to pay attention to the externalities of reporting frequency mandates on firms whose reporting frequency is not affected by the mandate. As we have discussed earlier, this step is important as evidence on aggregate effects and externalities from regulation is crucial to justifying regulations (Leuz and Wysocki, 2016). Fu et al. (2020) is one example of a study in which the researchers explicitly consider and examine the externalities and aggregate effects of reporting frequency mandates. The authors find no evidence that the externality effect on industry peers is statistically significant, and conclude that the aggregate effect of frequent reporting on total innovation (including both treatment effects and spillover effects on peer firms) appears to be negative. Overall, academic research suggests that disclosure regulation has both benefits and costs. While the benefits of enhanced disclosure and transparency for directly affected firms are relatively straightforward, it is less clear how to assess the positive and negative externalities and aggregate effects. In addition, it is not the case that “more is better,” since more disclosures not only entail certain compliance costs but also may pressure managers to meet investors’ short-term expectations. Adopting more realistic assumptions about

M10794_HILARY&MCLEAN_v1.indd 223

18/07/23 5:24 PM

224  Handbook of financial decision making

the behavior of investors and managers can help us better understand and predict the effects of enhanced disclosures (Hanlon et al., 2022). 3.2  Regulation of Information Dissemination Information dissemination technologies have greatly increased the timeliness of firm disclosures and reduced the costs of accessing them. With technological advances, the SEC has implemented a series of regulatory changes to improve the accessibility of firm disclosures to the public. For example, in 1993 the SEC began to mandate electronic submission of corporate filings through the Electronic Data Gathering, Analysis, and Retrieval (EDGAR) system, and in 2013 the SEC allowed companies to use social media outlets (e.g., Facebook and Twitter) to announce key information. These fundamental changes in information dissemination brought by modern technologies affect not only capital markets but also firms themselves. In this subsection, we use the EDGAR implementation as an example to illustrate the economic consequences of enhanced information dissemination (Gao and Huang, 2020; Chang et al., 2023; Goldstein et al., 2023). Before the EDGAR system was implemented in 1993, the SEC required firms to submit paper copies of filings, which were then stored in its public reference rooms located in Washington DC, New York, and Chicago. In each location, one or two copies of the same filing were available for the public to access. A New York Times article described these public reference rooms as “a zoo” in which “files are often misplaced or even stolen” (see Noble, 1982). Instead of physically visiting these reference rooms, investors could ­subscribe to commercial data vendors, which often charged a non-trivial fee. These paper filings also entailed a significant production lag for data aggregators such as Standard & Poor’s (D’Souza et al., 2010). Because of this restricted and delayed access to SEC filings, there was significant information asymmetry among investors even though these filings were “public.” On February 23, 1993, the SEC began to require registered firms to submit their filings electronically. In the EDGAR system’s phase-in schedule, the SEC divided all registered firms into ten groups based on firm size, and required each group to file electronically after a specified implementation date (SEC Release No. 33-6944 and No. 33-6977). All firms in the first group (CF-01) were required to file electronically in April 1993, and firms in the last group (CF-10) could wait until May 1996 to do so. Gao and Huang (2020) exploit this staggered timing of the EDGAR implementation and provide some causal evidence on the capital market effects of information ­dissemination. Specifically, they find that the EDGAR implementation leads to an increase in information production by individual investors and analysts. After the EDGAR ­implementation, individual investors’ stock trades become more informative about future returns, and sell-side analysts’ forecasts also become more accurate. In addition, stock pricing efficiency improves after a firm starts filing electronically via the EDGAR system. Together, these findings indicate that greater and broader information dissemination facilitated by the EDGAR system improves forecasting price efficiency, that is, the extent to which stock prices reflect all publicly available information. Using the EDGAR shock, Goldstein et al. (2023) show that the EDGAR implementation leads to a lower cost of equity capital and an increase in the level of equity financing and corporate investment. They further show that these effects are concentrated in value

M10794_HILARY&MCLEAN_v1.indd 224

18/07/23 5:24 PM

Disclosure regulation: past, present, and future  ­225

firms, consistent with the EDGAR shock primarily affecting corporate disclosures about assets in place. In addition, the authors provide evidence suggesting that greater information dissemination facilitated by the EDGAR system leads to a decrease in revelatory price efficiency (i.e., the extent to which prices reveal new information to managers) and managerial learning from prices. Overall, these findings indicate that information dissemination technologies entail a tradeoff between improved equity financing and reduced managerial learning from prices. Overall, the findings highlight both the benefits and costs of broader information dissemination. While it is intuitive that greater information dissemination facilitated by technologies can benefit less-sophisticated investors and level the playing field in the market, it does not necessarily make firms themselves better off if doing so reduces investors’ incentives to acquire information that can be useful for firm managers. These findings have implications for various FinTech innovations. FinTech innovations often make it easy for the investing public to get a huge amount of data at low cost, thereby improving forecasting price efficiency. However, FinTech innovations can also dampen sophisticated investors’ incentives to acquire information and reduce revelatory price efficiency. It would be ­interesting to evaluate the tradeoffs brought by various modern information technologies. 3.3  Regulation of Information Presentation The SEC has long required firms to clearly present information and use plain English in all financial disclosures (SEC, 1998). In the past, the SEC requirements were concerned primarily with making disclosures clear and accessible to human readers. More recently, however, the SEC has started to require firms to present information in a way that can be easily accessed and processed by machines. For example, in April 2009, the SEC required registered firms to file financial statements in the eXtensible Business Reporting Language (XBRL) format. There is also a trend of adopting XBRL worldwide; as of October 2019, more than 50 countries had adopted it.6 Several U.S. states (e.g., Florida, California) have either adopted or are planning to adopt XBRL for the reporting of governmental financial information.7 This recent trend engenders a lot of research on whether and how the presentation of firm disclosures (e.g., XBRL format) affects capital markets and firms. The XBRL mandate requires firms to tag numerical values in the financial statements using either standard tags defined in the XBRL taxonomy or customized tags. This process is not trivial to firms, and firms often make mistakes when they begin to tag numerical values. Amit Varshney, a top research analyst at Credit Suisse, once stated: I thought the XBRL documents were created to promote the mass consumption of financial reporting data, but that’s not the case because of the inconsistent tagging. If I have to collect the data for a handful of companies, I still find it easier to hand collect it from the HTML document. (quoted in Harris and Morsfield, 2012, p. 36)

Similar to the EDGAR implementation, the XBRL implementation has also been staggered. Firms were divided into three tiers based on their public float, that is, the market value of publicly tradable shares. The “large accelerated filers” with a public float of over $5 billion were tier-1 firms and were required to start XBRL filings in fiscal year 2009. The remaining “large accelerated filers” with a public float of between $700 million and

M10794_HILARY&MCLEAN_v1.indd 225

18/07/23 5:24 PM

226  Handbook of financial decision making

$5 billion were tier-2 firms and were required to start XBRL filings in fiscal year 2010. The remaining smaller filers were tier-3 firms and were required to start XBRL filings in fiscal year 2011. During the first year of XBRL adoption, firms only needed to tag the numerical values in the financial statements and tag each footnote as a text block. After the initial year, firms were required to adopt detailed tagging, that is, tagging each numerical value (whether in financial statements or footnotes) separately. A growing body of research examines the effects of XBRL on capital markets and firms in the fields of both information systems and accounting (Blankespoor et al., 2020). Blankespoor et al. (2014) provide evidence that firms adopting XBRL exhibit higher bid–ask spreads, lower liquidity, and lower trading volume in the year right after the XBRL mandate. These findings suggest that more-sophisticated investors can leverage their resources to derive greater benefits from XBRL than less-sophisticated investors. Thus, there is greater information asymmetry among investors after the XBRL implementation. Blankespoor (2019) utilizes the setting of XBRL detailed tagging requirements and examines how investors’ information processing costs affect firms’ disclosure choices. The author finds that firms increase their quantitative footnote disclosures after the XBRL detailed tagging mandate. This result suggests that firms do consider the information processing costs of investors while making their disclosure choices. A more recent study by Li et al. (2021) examines whether and how a firm’s XBRL adoption affects its financial statement readability. The authors use a difference-in-differences approach and show that the XBRL mandate leads to a decrease in the readability of the HTML-formatted annual reports, especially for those adopters with more quantitative disclosures, those with smaller firm size, and those with a higher level of financial complexity. The authors further show that the annual reports of XBRL adopters contain more grammatical violations but not more words. Overall, these findings suggest that XBRL adoption diverts managers’ attention in the initial years and can negatively affect managerial decision-making. Overall, evidence suggests that financial reporting technologies such as XBRL significantly affect not only how investors process firm information but also how firms construct and disseminate quantitative and qualitative disclosures. The existing studies largely focus on the relatively short-term impacts of XBRL adoption, given that it is a recent phenomenon. Future research could examine the long-term effects of XBRL adoption and the dynamic adjustment processes of investors and firms. It would be also interesting to understand how the Inline XBRL format (i.e., embedding XBRL data directly into an HTML document) required by the SEC in 2018 affects capital markets and firms.8

4  MANDATORY ESG DISCLOSURES On March 21, 2022, the SEC proposed rules to enhance and standardize climate-related disclosures in firms’ registration statements and periodic reports. The proposed rules would require firms to provide “information about climate-related risks that are reasonably likely to have a material impact on their business, results of operations, or financial condition, and certain climate-related financial statement metrics in a note to their audited financial statements.”9 In particular, the proposed rules would require firms to provide disclosures of direct greenhouse gas emissions (Scope 1) and indirect emissions

M10794_HILARY&MCLEAN_v1.indd 226

18/07/23 5:24 PM

Disclosure regulation: past, present, and future  ­227

from purchased energy (Scope 2). Firms would also be required to disclose greenhouse gas emissions from the upstream and downstream value chains (Scope 3) when these emissions are material or when firms have set an emission target that includes these emissions. Similar ESG-related disclosure mandates have already been implemented in the European Union, such as the Non-Financial Reporting Directive in 2014, the Sustainable Finance Disclosure Regulation in 2019, and the Taxonomy Regulation in 2020. In 2022, Singapore also introduced a phased approach to mandatory climate reporting. In the following, we first summarize the rationale for such disclosures in a market economy, and then discuss the potential economic consequences of such disclosures. Adam Smith (1776) noted long ago that individuals’ pursuit of self-interest can lead to the maximization of social welfare through the invisible hand (i.e., the free-market mechanism). This is true in a perfectly competitive market with perfect information and no externalities. In this setup, it is assumed that people act as if they fully understand the costs and benefits of their choices. However, we know that people make decisions based on their perceptions, which can (and in fact often do) deviate from reality (Graham, 2022; Hanlon et al., 2022). In many situations, people do not fully understand or internalize how their decisions will harm their own future or future generations. Here are a few quotes from early economists, as noted by Thaler (2016): The pleasure which we are to enjoy ten years hence, interests us so little in comparison with that which we may enjoy today. Smith (1759, p. 273) Our telescopic faculty is defective and … we therefore see future pleasures, as it were, on a diminished scale. Pigou (1920, p. 21) This is illustrated by the story of the farmer who would never mend his leaky roof. When it rained, he could not stop the leak, and when it did not rain, there was no leak to be stopped! Fisher (1930, p. 82)

The traditional capital budgeting model in corporate finance and other valuation courses often focuses on forecasting cash flows over the next five to ten years, and it assumes that, once firms reach the steady state, these cash flow patterns will last forever (e.g., Lundholm and Sloan, 2019). Some criticize this as an intrinsic limitation of our current forecasting approach – if something is going to happen in the distant future, we ignore it in our financial modelling. On the other hand, the approach might be understandable (and arguably “reasonable” or “acceptable”) because the impact of such “distant” risks (e.g., climate risk) on a firm’s financial performance is very difficult, if not impossible, to quantify. And there is a heated debate on whether and what information related to such distant risks can be credibly disclosed in a firm’s SEC filings (Bolton et al., 2021; Karpoff et al., 2022; LoPucki, 2022). On carbon emissions, while it is relatively straightforward to require firms to disclose their operational emissions (i.e., Scope 1 emissions), and perhaps it is straightforward for firms to report their emissions from their energy consumption (Scope 2 emissions), it is more difficult for firms to accurately compute and report emissions attributed to their suppliers and customers (Scope 3 emissions). These supply-chain emissions are often much more significant than the operational emissions. For example, Apple stated that it emitted 47,430 tons of greenhouse gases in the fiscal year ending September 26, 2020, and that the emissions by its suppliers and customers were 475 times larger, at 22 million tons.10 Thus, ignoring a firm’s Scope 3 emissions can significantly underestimate its climate impact. In addition, the production of some green products may create significant

M10794_HILARY&MCLEAN_v1.indd 227

18/07/23 5:24 PM

228  Handbook of financial decision making

Scope 3 emissions. For example, the solar industry relies heavily on coal-burning power plants to produce polysilicon, an essential component in most solar panels.11 While the use of solar panels can materially reduce carbon emissions in electricity generation relative to the use of fossil fuels, the production of solar panels can entail a significant amount of carbon emissions. Proponents of climate-related disclosures claim that requiring firms to consider and report these different scopes of carbon emissions has several benefits, despite the difficulty of gathering accurate data. After all, the goal of this disclosure requirement is not simply to get data on carbon emissions, but also to alter firm and individual behavior (Christensen et al., 2021). First, people (including corporate managers) have limited attention, and asking them to report those carbon data in their firms’ SEC filings can force managers to allocate their attention to this important aspect of production processes. Like sending warnings to those who smoke, explicitly requiring firms to disclose carbon data can enhance the self-awareness of the potential climate impact of their actions. Second, the goal of mandatory disclosure is not only to give investors the information that they want (for green or sustainable investing), but also to induce managers to choose more desirable actions, although unanimity on what is desirable is highly unlikely. If the market assigns a high cost to carbon emissions, the manager of the firm has an economic incentive to adopt alternatives to achieve the firm’s commercial objectives – that is, the manager must make an optimal cost-benefit decision. Third, emission information might enable major corporate customers to discipline their suppliers through pricing and purchase decisions (e.g., Chen et al., 2023). For example, suppliers of a major corporate customer that faces stringent environmental regulation and enforcement may alter their behavior toward environmental protection even if these suppliers are subject to less stringent legal requirements or enforcement. This potential outcome is offered as a justification for Scope 3 disclosures. If all firms, public or private, faced the same level of legal requirement and enforcement for environmental protection, then Scope 3 disclosures might be redundant. However, given that private firms and firms in less developed countries are often under loose legal requirements regarding environmental protection, Scope 3 disclosures are hypothesized to be helpful in achieving global environmental protection through a market mechanism. Because it can be difficult to accurately track the emissions by suppliers or customers, a science-based disclosure framework might be more objective, and therefore more credible. For example, instead of requiring firms to track the carbon emissions of their suppliers, a potentially simpler approach that can generate verifiable and auditable data is to require firms to report the generic production processes of their suppliers (e.g., coalburning or solar). This type of disclosure is more detailed than industry averages but nevertheless requires no specific technological know-how. This kind of generic, inputbased disclosure framework does not lead to concerns about disclosing the proprietary information of suppliers, and it produces data that can be verified by a third party. However, understanding what data are potentially important to estimate the amount of carbon emissions requires regulators and preparers to make choices among divergent scientific data. It is also important to consider the expected direct and indirect costs of the regulation. Direct costs would include compliance costs for firms to meet the disclosure requirements. Indirect costs would include potential litigation risk or leakage of proprietary information. This cost consideration leads the SEC to exempt smaller reporting companies from the Scope 3 emissions disclosure requirement in the proposed rules.

M10794_HILARY&MCLEAN_v1.indd 228

18/07/23 5:24 PM

Disclosure regulation: past, present, and future  ­229

Overall, we believe that accounting researchers can leverage their institutional knowledge and contribute significantly to non-financial disclosure regulation, just as they contribute to financial disclosure regulation, by examining the benefits, costs, externalities, and aggregate effects of the regulation. In this process, it is also important to bring scientists, economists, and business practitioners into the discussion.

5  CONCLUDING REMARKS In our review of the theories and empirical evidence of disclosure regulation, we conclude that some extent of disclosure regulation is likely justified. First, market forces, private contracts and enforcement, and various non-traditional, non-regulator monitors are unlikely to resolve all the information problems in the capital market. Second, disclosure carries substantial costs for the disclosing firms, which cannot enjoy the full benefit of disclosure. This leads to the underproduction of information from a society’s point of view. The literature has provided mounting evidence on the firm-level benefits of disclosure regulation. However, we still have limited evidence on the net aggregate effect of disclosure regulation, which is crucial to evaluating whether disclosure mandates are indeed necessary or desirable. Moreover, the recent development of technology (such as regulation technology or RegTech) likely changes the direct costs of information production and regulatory compliance, thereby re-shaping the welfare analysis of disclosure regulation. It would be interesting to examine whether and how the development of technology makes disclosure regulation more or less effective in improving the financial market and real economic efficiency. Since the call of Kothari et al. (2010), we have seen a growing number of studies on the political economy of disclosure and reporting regulation. From the empirical evidence, our overall impression is that all three theories of regulation (public interest, capture, and ideology) are relevant in explaining the current status of disclosure regulation. Going forward, it would be interesting to examine how the three theories interact in explaining the creation, enforcement, and consequences of disclosure regulation. For example, researchers could examine how the relative power of interest groups and the ideological congruence between interest groups and regulators shape the creation and enforcement of disclosure regulation. Moreover, given that political factors likely influence the creation and enforcement of disclosure regulation, it is conceivable that these same factors also affect the aggregate financial and real economic effect of disclosure regulation. Currently, researchers tend to rely on explicit political contributions and lobbying to identify firm connections to politicians. It would be interesting to expand the analysis by identifying less explicit links between politicians and their constituencies for a more complete picture of political influence on disclosure regulation. For example, researchers could examine social networks and charitable giving (e.g., Bertrand et al., 2020). Regarding emerging issues of disclosure regulation, we have offered some thoughts in Section 4 on ESG disclosure mandates, which are arguably among the most important topics in today’s disclosure regulation. We also note that the pandemic of the past three years has accelerated the adoption of digital technologies in various aspects of business. This creates many new challenges for disclosure and reporting regulation. For example, a growing number of companies are incorporating cryptocurrencies (e.g., Bitcoin) and

M10794_HILARY&MCLEAN_v1.indd 229

18/07/23 5:24 PM

230  Handbook of financial decision making

non-fungible tokens (NFTs) into their business. Companies currently account for these digital assets as indefinite-lived intangible assets based on some non-binding guidelines. However, there are no specific accounting or disclosure rules for these digital assets from accounting standard setters. Therefore, it is important for regulators and researchers to understand the nature of these assets and their market dynamics, and to evaluate whether current accounting or disclosure rules apply to various digital assets – and, if not, what new rules are needed. Another prominent development in the capital market is that companies and investors have increased their reliance on social media to acquire and disseminate information. However, due to the largely unregulated nature of social media, the reliability of information is a big challenge. Thus, it is important to understand how information on social media increases or decreases market efficiency, and how regulators can step in to improve the situation. Finally, we note that most disclosure regulation literature is based on U.S. settings. As North (1994, p. 366) notes, “economies that adopt the formal rules of another economy will have very different performance characteristics than the first economy because of different informal norms and enforcement.” Therefore, we need more research to understand how various informal norms and enforcement affect the implications of otherwise similar disclosure regulation (e.g., Hail et al., 2018).

NOTES   1. We direct interested readers to Healy and Palepu (2001), Kothari et al. (2010), and Leuz and Wysocki (2016) for extensive reviews of the literature.  2. https://www.qui-tam-attorney.com/10-largest-qui-tam-whistleblower-rewards.html.  3. https://ec.europa.eu/commission/presscorner/detail/fr/MEMO_13_544.  4. https://www.reuters.com/article/sgx-regulations/singapore-exchange-scraps-compulsory-quar​ terly-reporting-for-companies-idUSL4N29E1B1.  5. https://www.wsj.com/articles/trump-directs-sec-to-study-six-month-reporting-for-public-com​ panies-1534507058.  6. https://www.forbes.com/sites/forbesfinancecouncil/2019/10/21/how-well-do-you-speak-this-fin​ ancial-language.  7. https://www.govtech.com/biz/Can-Standardized-Financial-Data-Help-Government-Save-Mo​ ney.html.  8. https://www.sec.gov/structureddata/osd-inline-xbrl.html.  9. https://www.sec.gov/news/press-release/2022-46. 10. https://www.wsj.com/articles/climate-disclosure-poses-thorny-questions-for-sec-as-rules-weig​ hed-11645180200. 11. https://www.wsj.com/articles/behind-the-rise-of-u-s-solar-power-a-mountain-of-chinese-coal-​ 11627734770.

REFERENCES Akerlof, G.A., 1970. The market for “lemons”: Quality uncertainty and the market mechanism. Quarterly Journal of Economics 84, 488–500. Alexander, C.R., Lee, Y.H.A., 2004. The economics of regulatory reform: Termination of airline computer reservation system rules. Yale Journal on Regulation 21, 369–439. Austen-Smith, D., 1995. Campaign contributions and access. American Political Science Review 89, 566–581.

M10794_HILARY&MCLEAN_v1.indd 230

18/07/23 5:24 PM

Disclosure regulation: past, present, and future  ­231

Badertscher, B., Shroff, N., White, H.D., 2013. Externalities of public firm presence: Evidence from private firms’ investment decisions. Journal of Financial Economics 109, 682–706. Balakrishnan, K., Ertan, A., 2018. Banks’ financial reporting frequency and asset quality. The Accounting Review 93, 1–24. Becker, G.S., 1983. A theory of competition among pressure groups for political influence. Quarterly Journal of Economics 98, 371–400. Berle, A., Means, G., 1932. The Modern Corporation and Private Property. New York: Macmillan. Bernard, D., Blackburne, T., Thornock, J., 2020. Information flows among rivals and corporate investment. Journal of Financial Economics 136, 760–779. Bernard, D., Burgstahler, D., Kaya, D., 2018. Size management by European private firms to minimize proprietary costs of disclosure. Journal of Accounting and Economics 66, 94–122. Bertrand, M., Bombardini, M., Fisman, R., Trebbi, F., 2020. Tax-exempt lobbying: Corporate philanthropy as a tool for political influence. American Economic Review 110, 2065–2102. Beyer, A., Cohen, D.A., Lys, T.Z., Walther, B.R., 2010. The financial reporting environment: Review of the recent literature. Journal of Accounting and Economics 50, 296–343. Bischof, J., Daske, H., Sextroh, C.J., 2020. Why do politicians intervene in accounting regulation? The role of ideology and special interests. Journal of Accounting Research 58, 589–642. Blankespoor, E., 2019. The impact of information processing costs on firm disclosure choice: Evidence from the XBRL mandate. Journal of Accounting Research 57, 919–967. Blankespoor, E., deHaan, E., Marinovic, I., 2020. Disclosure processing costs, investors’ information choice, and equity market outcomes: A review. Journal of Accounting and Economics 70, 101344. Blankespoor, E., Miller, B.P., White, H.D., 2014. Initial evidence on the market impact of the XBRL mandate. Review of Accounting Studies 19, 1468–1503. Boehmer, E., Jones, C.M., Wu, J., Zhang, X., 2020. What do short sellers know? Review of Finance 24, 1203–1235. Bolton, P., Reichelstein, S., Kacperczyk, M.T., Leuz, C., Ormazabal, G., Schoenmaker, D., 2021. Mandatory corporate carbon disclosures and the path to net zero. Management and Business Review, Fall 2021. Breuer, M., 2021. How does financial-reporting regulation affect industry-wide resource allocation? Journal of Accounting Research 59, 59–110. Butler, M., Kraft, A., Weiss, I.S., 2007. The effect of reporting frequency on the timeliness of earnings: The cases of voluntary and mandatory interim reports. Journal of Accounting and Economics 43, 181–217. Carey, J.L., 1969. The Rise of the Accounting Profession: From Technician to Professional, 1896–1936. New York: American Institute of Certified Public Accountants. Chang, Y.C., Ljungqvist, A., Tseng, K., 2023. Do corporate disclosures constrain strategic analyst behavior? Review of Financial Studies, forthcoming. Chen, J., Su, X., Tian, X., Xu, B., Zuo, L., 2023. The disciplinary role of major corporate customers. Working paper. Christensen, H.B., Hail, L., Leuz, C., 2021. Mandatory CSR and sustainability reporting: Economic analysis and literature review. Review of Accounting Studies 26, 1176–1248. Coase, R., 1960. The problem of social cost. Journal of Law and Economics 3, 1–44. Correia, M.M., 2014. Political connections and SEC enforcement. Journal of Accounting and Economics 57, 241–262. D’Souza, J.M., Ramesh, K., Shen, M., 2010. The interdependence between institutional ownership and information dissemination by data aggregators. The Accounting Review 85, 159–193. Downar, B., Ernstberger, J., Link, B., 2018. The monitoring effect of more frequent disclosure. Contemporary Accounting Research 35, 2058–2081. Durnev, A., Mangen, C., 2020. The spillover effects of MD&A disclosures for real investment: The role of industry competition. Journal of Accounting and Economics 70, 101299. Dyck, A., Morse, A., Zingales, L., 2010. Who blows the whistle on corporate fraud? Journal of Finance 65, 2213–2253. Ege, M.S, Stuber, S.B., 2022. Are auditors rewarded for low audit quality? The case of auditor ­lenience in the insurance industry. Journal of Accounting and Economics 73, 101424.

M10794_HILARY&MCLEAN_v1.indd 231

18/07/23 5:24 PM

232  Handbook of financial decision making

Ernstberger, J., Link, B., Stich, M., Vogler, O., 2017. The real effects of mandatory quarterly reporting. The Accounting Review 92, 33–60. Fang, V.W., Huang, A.H., Karpoff, J.M., 2016. Short selling and earnings management: A controlled experiment. Journal of Finance 71, 1251–1294. Fisher, I., 1930. The Theory of Interest: As Determined by Impatience to Spend Income and Opportunity to Invest It. New York: Macmillan. Fu, R., Kraft, A., Tian, X., Zhang, H., Zuo, L., 2020. Financial reporting frequency and corporate innovation. Journal of Law and Economics 63, 501–530. Fu, R., Kraft, A., Zhang, H., 2012. Financial reporting frequency, information asymmetry, and the cost of equity. Journal of Accounting and Economics 54, 132–149. Gao, M., Huang, J., 2020. Informing the market: The effect of modern information technologies on information production. Review of Financial Studies 33, 1367–1411. George, T.E., 1998. Developing a positive theory of decisionmaking on US Courts of Appeals. Ohio State Law Journal 58, 1635–1696. Giroud, X., Mueller, H.M., 2011. Corporate governance, product market competition, and equity prices. Journal of Finance 66, 563–600. Glaeser, S., 2018. The effects of proprietary information on corporate disclosure and transparency: Evidence from trade secrets. Journal of Accounting and Economics 66, 163–193. Glaeser, S., Omartian, J.D., 2022. Public firm presence, financial reporting, and the decline of U.S. manufacturing. Journal of Accounting Research 60, 1083–1128. Goldstein, I., Yang, S., Zuo, L., 2023. The real effects of modern information technologies: Evidence from the EDGAR implementation. Working paper. Gompers, P., Ishii, J., Metrick, A., 2003. Corporate governance and equity prices. Quarterly Journal of Economics 118, 107–156. Graham, J., 2022. Presidential address: Corporate finance and reality. Journal of Finance 77, 1975–2049. Grossman, G.M., Helpman, E., 1994. Protection for sale. American Economic Review 84, 833–850. Hail, L., Tahoun, A., Wang, C., 2018. Corporate scandals and regulation. Journal of Accounting Research 56, 617–671. Hanlon, M., Yeung, K., Zuo, L., 2022. Behavioral economics of accounting: A review of archival research on individual decision makers. Contemporary Accounting Research 39, 1150–1214. Harris, T.S., Morsfield, S.G., 2012. An evaluation of the current state and future of XBRL and interactive data for investors and analysts. Working paper. Healy, P.M., Palepu, K.G., 2001. Information asymmetry, corporate disclosure, and the capital markets: A review of the empirical disclosure literature. Journal of Accounting and Economics 31, 405–440. Heese, J., Khan, M., Ramanna, K., 2017. Is the SEC captured? Evidence from comment-letter reviews. Journal of Accounting and Economics 64, 98–122. Heese, J., Pérez-Cavazos, G., Peter, C.D., 2022. When the local newspaper leaves town: The effects of local newspaper closures on corporate misconduct. Journal of Financial Economics 145, 445–463. Huang, K., Li, M., Markov, S., 2020. What do employees know? Evidence from a social media platform. The Accounting Review 95, 199–226. Jensen, M.C., Meckling, W.H., 1976. Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics 3, 305–360. Kajüter, P., Klassmann, F., Nienhaus, M., 2019. The effect of mandatory quarterly reporting on firm value. The Accounting Review 94, 251–277. Karpoff, J.M., Litan, R., Schrand, C., Weil, R.L., 2022. What ESG-related disclosures should the SEC mandate? Financial Analysts Journal 78, 9–18. Kim, J., Valentine, K., 2021. The innovation consequences of mandatory patent disclosures. Journal of Accounting and Economics 71, 101381. Kim, J., Valentine, K., 2023. Public firm disclosures and the market for innovation. Journal of Accounting and Economics, forthcoming. Kothari, S.P., Ramanna, K., Skinner, D.J., 2010. Implications for GAAP from an analysis of ­positive research in accounting. Journal of Accounting and Economics 50, 246–286.

M10794_HILARY&MCLEAN_v1.indd 232

18/07/23 5:24 PM

Disclosure regulation: past, present, and future  ­233

Kothari, S.P., Shu, S., Wysocki, P.D., 2009. Do managers withhold bad news? Journal of Accounting Research 47, 241–276. Kraft, A.G., Vashishtha, R., Venkatachalam, M., 2017. Frequent financial reporting and managerial myopia. The Accounting Review 93, 249–275. Lang, M., Sul, E., 2014. Linking industry concentration to proprietary costs and disclosure: Challenges and opportunities. Journal of Accounting and Economics 58, 265–274. Lee, J., Ng, S., Yoo, I.S., Zhang, L., 2023. Freedom of expression protection and corporate concealment of bad news: Evidence from state anti-SLAPP laws. Working paper. Leftwich, R.W., Watts, R.L., Zimmerman, J.L., 1981. Voluntary corporate disclosure: The case of interim reporting. Journal of Accounting Research 19, 50–77. Leuz, C., Wysocki, P.D., 2016. The economics of disclosure and financial reporting regulation: Evidence and suggestions for future research. Journal of Accounting Research 54, 525–622. Li, X., Zhu, H., Zuo, L., 2021. Reporting technologies and textual readability: Evidence from the XBRL mandate. Information Systems Research 32, 1025–1042. Li, Y., Lin, Y., Zhang, L., 2018. Trade secrets law and corporate disclosure: Causal evidence on the proprietary cost hypothesis. Journal of Accounting Research 56, 265–308. LoPucki, L.M., 2022. Corporate greenhouse gas disclosures. UC Davis Law Review, 56, 405–466. Lundholm, R., Sloan, R., 2019. Equity Valuation and Analysis (5th Edition). Washington, DC: Kindle Direct Publishing. Mahoney, P.G., 2021. The economics of securities regulation: A survey. Foundations and Trends® in Finance 13, 1–94. McNichols, M., Manegold, J.G., 1983. The effect of the information environment on the relationship between financial disclosure and security price variability. Journal of Accounting and Economics 5, 49–74. Mehta, M.N., Zhao, W., 2020. Politician careers and SEC enforcement against financial misconduct. Journal of Accounting and Economics 69, 101302. Miller, G.S., 2006. The press as a watchdog for accounting fraud. Journal of Accounting Research 44, 1001–1033. Myers, S.C., Majluf, N.S., 1984. Corporate financing and investment decisions when firms have information that investors do not have. Journal of Financial Economics 13, 187–221. Nallareddy, S., Pozen, R., Rajgopal, S., 2021. Consequences of more frequent reporting: The U.K. experience. Journal of Law, Finance, and Accounting 6, 51–88. Narayanan, M.P., 1985. Managerial incentives for short-term results. Journal of Finance 40, 1469–1484. Noble, K.B., 1982. S.E.C. data: Difficult hunt. New York Times, May 19. North, D.C., 1994. Economic performance through time. American Economic Review 84, 359–368. Peltzman, S., 1976. Toward a more general theory of regulation. Journal of Law and Economics 19, 211–240. Pigou, A.C., 1920. The Economics of Welfare. London: Macmillan. Ramanna, K., 2008. The implications of unverifiable fair-value accounting: Evidence from the political economy of goodwill accounting. Journal of Accounting and Economics 45, 253–281. Roberts, M.R., 2015. The role of dynamic renegotiation and asymmetric information in financial contracting. Journal of Financial Economics 116, 61–81. Roychowdhury, S., Shroff, N., Verdi, R.S., 2019. The effects of financial reporting and disclosure on corporate investment: A review. Journal of Accounting and Economics 68, 101246. Schoenfeld, J., 2020. Contracts between firms and shareholders. Journal of Accounting Research 58, 383–427. Securities and Exchange Commission (SEC), 1998. A Plain English Handbook: How to Create Clear SEC Disclosure. SEC Office of Investor Education and Assistance. https://www.sec.gov/ pdf/handbook.pdf (accessed October 20, 2022). Shleifer, A., 2012. The Failure of Judges and the Rise of Regulators. Cambridge, MA: MIT Press. Smith, A., 1759. The Theory of Moral Sentiments. Reprint edited by D.D. Raphael and A.L. Macfie. Indianapolis: Liberty Classics, 1981. Smith, A., 1776. An Inquiry into the Nature and Causes of the Wealth of Nations. Reprint edited by R.H. Campbell and A.S. Skinner. Indianapolis: Liberty Classics, 1981.

M10794_HILARY&MCLEAN_v1.indd 233

18/07/23 5:24 PM

234  Handbook of financial decision making

Stein, J.C., 1989. Efficient capital markets, inefficient firms: A model of myopic corporate behavior. Quarterly Journal of Economics 104, 655–669. Stigler, G.J., 1971. The theory of economic regulation. Bell Journal of Economics and Management Science 2, 3–21. Thaler, R.H., 2016. Behavioral economics: Past, present, and future. American Economic Review 106, 1577–1600. Thompson, A.M., 2022. Political connections and the SEC confidential treatment process. Journal of Accounting and Economics 74, 101511. Verrecchia, R.E., 1983. Discretionary disclosure. Journal of Accounting and Economics 5, ­179–194. Watts, R.L., Zuo, L., 2016. Understanding practice and institutions: A historical perspective. Accounting Horizons 30, 409–423. Yue, H., Zhang, L., Zhong, Q., 2022. The politics of bank opacity. Journal of Accounting and Economics 73, 101452. Zeff, S.A., 2021. Evolution of U.S. regulation and the standard-setting process for financial ­reporting: 1930s to the present. Foundations and Trends® in Accounting 15, 263–372.

M10794_HILARY&MCLEAN_v1.indd 234

18/07/23 5:24 PM

11. The audit in a modern economy

W. Robert Knechel and Eddie Thomas

Auditing is one of the oldest professions known to most societies. At the core, auditing is about the verification of information that is provided by one party which is to be used by another in making certain decisions. The adage of “garbage in, garbage out” is a shorthand way of saying that if bad or inaccurate information is provided, poor decisions are likely to follow. Consequently, the conceptual foundation of auditing is grounded on the relatively simple idea that the accuracy of information from one party to another can be improved through third-party verification – namely, an audit.

1  THE ROLE OF TRUST IN THE DEMAND FOR AUDIT Capital markets depend upon significant investor trust to function (Knack and Keefer, 1997). Investors – shareholders and lenders – provide capital to companies, but most investors are outsiders to company operations. Thus, they may have little influence on managerial decision-making and must rely almost completely on public information provided by managers, typically in the form of quarterly and annual financial statements. These reports are designed to inform investors and other stakeholders as to whether a company is putting its resources to appropriate and profitable use. When (potential or actual) investors have doubts about the quality of an investment, they will either avoid the investment altogether1 or protect themselves by (a) reducing the price they are willing to pay for its stock (shareholders) or (b) increasing the interest rate demanded (lenders).2 Companies that cannot earn the trust of investors therefore will find it harder to raise capital to fund growth opportunities. From a societal point of view, underinvestment leads to decreased production of goods and services. In general economics, this loss of production is often referred to as “agency costs” – the potential value lost because of information asymmetry between various stakeholders who have an interest in the organization. Consequently, companies, investors, and society have incentives to promote investor trust. Several institutional practices exist that empower investors to have some control over the managers who manage their investments. Of most importance is the ability of shareholders to elect the board of directors, which then hires top management, provides general guidance to management, and monitors the success of the company. A subset of the board, the audit committee, is generally charged with oversight of the financial reporting process. Many companies also have an internal audit function to determine that processes are well-designed to achieve company objectives. Other stakeholders have various ways to influence a company’s operations – for example, a lender can write loan covenants that limit managers’ ability to take on excessive risk. Perhaps the greatest protections for investors, however, and the ones most relevant to their individual decision-making, are the disclosures that companies make about their 235

M10794_HILARY&MCLEAN_v1.indd 235

18/07/23 5:24 PM

236  Handbook of financial decision making

operations and financial position. Companies have many such venues for disclosure, such as the business press and conference calls with analysts. The most important disclosures, however, are company financial reports. Financial reports provide historical information about recent results of operations and overall financial position. Investors can use this information to corroborate their prior understanding of the company and as a basis for projections of future performance. However, to be of use to investors, financial information must be reliable, timely, and complete. Financial reporting is so important for promoting investor trust that a number of institutional structures exist to enhance the quality of financial reporting. Regulatory bodies, such as the U.S. Securities & Exchange Commission (SEC), exist to deter fraud and enforce laws and policies that make markets fairer and more transparent. Standard-setters establish accounting standards as to what information is required and how it should be reported so that investors are confident they can understand the financial information being presented. Other parties, such as the media and financial analysts, also watch the activities of an organization closely. The most direct manner of checking on a company’s behavior and information is through the external audit, which is designed to provide independent verification of the information contained in annual financial reports. Audits exist to provide “reasonable assurance” that financial reports have been properly prepared in accordance with the relevant accounting framework that has been established by accounting standard-setters. By increasing investor confidence in a company’s financial reporting, auditors increase investor confidence in the company itself (Brydon, 2019, Section 5.1.40). Research indicates that auditing improves the value of financial information for investors, as evidenced by reduced risk to external shareholders (Newman et al., 2005), reduced cost of capital for companies (Mansi et al., 2004; Pittman and Fortin, 2004; Botosan and Plumlee, 2005), and investors responding more efficiently to reported earnings (Aobdia et al., 2015). Audits are successful both at reducing accounting errors (Kinney et al., 2004) and reducing fraud risk (Carcello and Nagy, 2004; Lobo and Zhao, 2013). Given the role that audits have in promoting trust in financial disclosures, and thereby the companies that prepare those statements, it would be expected that investors demand, and companies provide, higher-quality auditing in situations where investors’ information risk is higher. Research supports this expectation across several contexts. First, companies in need of external financing have been shown to be more willing to bear the costs of an audit or of higher-reputation auditors than companies with less need (Knechel et al., 2008). Second, lenders utilize debt covenants to limit their risk, and many debt covenants rely on financial information. Research indicates that organizations with relatively higher debt levels either choose to be audited (even if not legally required) or employ an auditor with a higher reputation for quality (Chow, 1982; Knechel et al., 2008). Alternatively, the investors in a company with concentrated ownership (i.e., a small number of investors own a large percentage of the company’s stock) may have enough influence and incentive to directly obtain information about managerial actions, resulting in less demand for highquality audits (Mitra et al., 2007; Hope et al., 2012). Given that the demand for auditing is higher where the risks of being misinformed are higher, one might expect that auditing would be in higher demand in countries (and cultures) where people inherently have less trust in each another. Research indicates, however, that countries with higher levels of societal trust and civic cooperation also have

M10794_HILARY&MCLEAN_v1.indd 236

18/07/23 5:24 PM

The audit in a modern economy  ­237

a higher demand for quality auditing (Knechel et al., 2019). Why might this be? Two reasons stand out. First, since the quality of audits is very hard to objectively determine (Knechel et al., 2013), in societies with low levels of trust, investors may be no more likely to trust the work of auditors than they trust the managers of the firm being audited. Organizations in a low trust society may rely more on narrower social networks (e.g., familial and tribal relations) to obtain financing (Fukuyama, 2002). Second, high trust societies have more to lose from managerial misrepresentations. High trust societies are associated with less earnings management (Nanda and Wysocki, 2011) and tax avoidance (Felix et al., 2017), which would explain why investors respond more strongly in these countries to financial disclosures (Pevzner et al., 2015). Accordingly, these societies would have a greater need to identify (and “punish”) managers who are unworthy of trust because distrust may be contagious and trust is hard to develop but easy to lose (Balliet and Van Lange, 2013). Therefore, the auditing profession provides greatest value within a particular social and economic context. In societies where there are generally higher levels of trust, investors are more willing to provide capital to organizations where they have limited private access to information and limited influence. These investors face heightened vulnerability that can be mitigated by having audit professionals assure the accuracy of reported financial information. When organizations have greater access to capital, the costs of capital are lower, allowing for a more efficient and productive use of society’s resources. Societies with higher general levels of trust can therefore obtain greater wealth as well.

2  THE VERIFICATION PROCESS 2.1  The Collaborative Audit Network Auditors provide assurance to those who rely on financial statements by first obtaining assurance for themselves through the accumulation of audit evidence, and then sharing their conclusions with others using a report of the auditors’ opinion on the material fairness of the audit client’s financial statements. Like the investors who rely on an organization’s financial statements, auditors are external to the organization being audited. The independence of the auditor from the audit client is critical for investors to be able to trust the auditor’s work. Unlike external investors, however, auditors have access to the audited organization. The terms of the audit engagement require management: (1) to give the auditor access to all information that is relevant for the preparation and presentation of the financial statements; and (2) unrestricted access to personnel within the organization. There are different interpretations of how auditors operate (or, at least, should operate) in the context of a specific client environment. Some believe that auditors are properly akin to the police, monitoring and enforcing audit client compliance with accounting standards. From this point of view, the auditor is close enough to obtain evidence verifying the client’s financial position and results of operations, but somehow also distant enough from the client to avoid threats to independence or objectivity. This view imagines the ideal auditor–audit client relationship on a spectrum from detached to adversarial. The term used in auditing to refer to independent thought about a client is

M10794_HILARY&MCLEAN_v1.indd 237

18/07/23 5:24 PM

238  Handbook of financial decision making

“professional skepticism.” Consistent with an adversarial viewpoint, an auditor exercises professional skepticism through a presumption of the audit client’s “guilt” – that is, a presumption that the client’s financial statements are materially misstated (Nelson, 2009). Auditors themselves prefer not to have the role of “policing” an audit client (Gibbins et al., 2001). How an auditor approaches a specific client depends partially on the attitude of the client themselves. An audit requires a degree of cooperation between various parties to the process (i.e., auditor, audit committee, management, internal auditors, etc.). An audit is likely to be most efficient and effective when all parties in the process share a common goal of improving the reliability of financial reporting. Unfortunately, that may not always be the case for all clients. If management is antagonistic toward the auditor, the auditor may take on more of a policing role within the audit process. Managers might do this for a variety of reasons. They might consider the audit an intrusion on their time. They might consider the financial reporting to belong to them, and thus do not consider the work of the audit to be especially valuable. Or they might be trying to conceal fraudulent reporting. Whatever the reason, auditors who are kept at a distance are likely to engage the audit in a more adversarial manner while working to improve the relationship with the client to make the relationship less adversarial (Gibbins et al., 2001).3 An alternative to the adversarial approach to professional skepticism is to take a “neutral” position – not expecting a client to have material errors, but also withholding judgment until evidence supports a specific conclusion. This view emphasizes the benefits of close interaction between auditors and the different parties within the audit client. From this perspective, auditors are primarily service providers who work closely with the audit client to produce credible financial disclosures. In this arrangement, management has primary responsibility for the financial statements, while the auditor has primary responsibility for providing assurance. The quality of the financial statements is improved by the work of the auditors, and the quality of the audit is improved by appropriate cooperation of management and other parties in the audit client.4 Audit quality depends upon: (a) the competencies (and other resources) that the auditor and the different parties within the audit client bring to the audit process; and (b) how well those competencies (and other resources) are integrated into the audit process (Knechel et al., 2020). The need for cooperation is most evident between the auditor and management. In each stage of the audit, cooperation with management is vital. In the planning stage, auditors rely on inquiry with management to develop an understanding of the client as part of assessing risks of material misstatements. Managers who are more forthcoming with auditors will help auditors to better understand where to focus their attention. In the testing stage, auditors rely on management to provide information and to help them interpret audit evidence. While auditors cannot rely exclusively on discussions with management as audit evidence,5 the auditor utilizes management’s knowledge of the organization to help make sense of the results of other audit procedures. Management’s cooperation is particularly important when auditing managerial estimates (Fornelli, 2013), since auditors typically test these by testing management models and assumptions that go into the calculations on which an estimate is based (Griffith et al., 2015). In the final stage of the audit, auditors may need to confront management about discovered misstatements. Research describes this confrontation as a negotiation between the auditor and management, and the better the relationship between management and the auditor, the more

M10794_HILARY&MCLEAN_v1.indd 238

18/07/23 5:24 PM

The audit in a modern economy  ­239

likely the company is to accept the conclusions of the auditor (Gibbins et al., 2001). Auditing standards such as from the Public Company Accounting Oversight Board (PCAOB) highlight the importance of management cooperation with the auditor: (1) managers have a better understanding of their firm’s operations than does the auditor;6 (2) management’s representations are considered part of the audit evidence;7 and (3) lack of cooperation from management is one reason why a well-conducted audit might still fail to uncover a material misstatement.8 Audits also function better when the auditors work well with other members of the financial reporting ecosystem, such as the audit committee and internal auditors. Audit committees and internal auditors are also best understood as co-creators (rather than simply monitors) in the financial reporting process. Effective audit committees will actively assist the external auditor, and thereby contribute to financial reporting by taking an active role in setting agendas, helping determine an organization’s accounting policies, reviewing managerial accounting estimates, assessing the risk of fraud, and providing a governance role for the internal audit function (Beasley et al., 2009). When audit committees are ineffective, auditors must attempt to compensate by increasing their own work (Cohen et al., 2002). Similarly, an effective audit will integrate internal auditor competencies in identifying control deficiencies.9 Greater coordination between external and internal auditors has been shown to lead to greater detection and disclosure of material weaknesses in controls (Lin et al., 2011). The importance of close working relationships between the auditor and others in the financial reporting ecosystem has only increased over time. Improvements in technology have led to more sophisticated accounting information systems and business processes, requiring more sophisticated forms of corporate governance and internal controls. The contemporary audit places an emphasis on the auditor’s understanding and testing of internal controls, especially for the audit of large publicly traded companies where an audit of internal controls is required. As auditors focus on understanding and evaluating audit client processes, rather than just year-end balances in the financial statements, they find themselves more embedded within the audit client (Knechel et al., 2020). Does the close interaction of auditors with different parties in the audit client create a threat to auditor independence? Potentially. The American Institute of Certified Public Accountants (AICPA) Code of Professional Conduct identifies a familiarity threat to auditor independence, which is “the threat that, due to a long or close relationship with a client, a member will become sympathetic to the client’s interest or too accepting of the client’s work or product” (AICPA, 2014, Section 1.000.12). At the same time, the close interaction of auditors with different parties associated with the client potentially creates safeguards for auditor independence as well (Driskill et al., 2022). One safeguard is that the auditor’s specialized knowledge (expertise) gives the auditor power in the auditor/ client relationship, a power which increases over the auditor’s tenure with the client. Insofar as audits involve auditors applying general accounting and auditing concepts to a particular audit context, auditors who have extended experience with a client are more aware of the client’s particular risks of material misstatement, and have more precise expectations of what would be abnormal in the client’s financial reporting. Another safeguard is that auditors and their clients create co-investments in the working relationship over time, such that it becomes difficult for a different auditor to take over and provide the same level of service. Auditors with shorter tenure have less client-specific knowledge

M10794_HILARY&MCLEAN_v1.indd 239

18/07/23 5:24 PM

240  Handbook of financial decision making

(Stice, 1991), less ability to get clients to accept audit adjustments (Iyer and Rama, 2004), and less sharing of information (Bell et al., 2015; DeFond and Zhang, 2014). Yet, auditors must be independent of their clients – both in appearance and in mindset; but, due to the important role the audit client has in the production of the audit, auditors must also have a certain level of trust in their clients. An auditor who has an active distrust in a company’s management would be foolish to accept the engagement because normal auditing procedures are unlikely to be sufficient to overcome the client’s risk of fraudulent financial reporting, especially if management is actively working to hide things from the auditor. Professional skepticism requires the auditor to be open to the possibility that the auditor’s trust in the client is misplaced; but the work of the audit each year is to obtain sufficient appropriate evidence to justify, and renew, that trust. 2.2  Audit Quality and Audit Efficiency Quality in the production of economic services is different than quality in the production of tangible economic goods. With the production of tangible goods (i.e., manufacturing), quality is typically understood in terms of the number of deviations from product specifications. Manufacturing quality depends upon the design and operation of processes that minimize defective products based on a system of rigorous standardization. In a manufacturing context, quality and efficiency go hand in hand. The more identical units the system can produce, the more efficient it is considered. Audit firms, regulators, and academics sometimes apply similar logic to consideration of audit quality and the audit process, believing that audit quality can be understood in terms of how well an audit complies with auditing standards and inspection processes. At one time, auditing standards were considered to be a set of minimum expectations for what constituted an acceptable audit, with the assumption that auditors following best practices would exceed those expectations. Now, auditors are more likely to define best practices by the auditing standards themselves (Catasús et al., 2013, pp. 32–60). Likewise, auditors are increasingly likely to choose procedures and documentation practices to receive a passing grade from regulatory inspectors such as the PCAOB, possibly at some sacrifice to their own critical thinking (Westermann et al., 2019). However, in spite of this focus on standards, audit quality does not follow primarily from the standardization of processes for a number of reasons. First, there are differences in the demand for auditing, meaning that different stakeholders will have different perceptions about what constitutes sufficient assurance from the auditor. Second, each client has inherently different characteristics, including idiosyncrasies in business plans, operations, accounting knowledge, transactions, controls systems, and willingness to cooperate with the auditor. Third, auditing involves significant professional judgment, suggesting that differences will arise across individual auditors in how they plan and conduct an audit and evaluate audit evidence. Accordingly, for audit services there is a natural tension between audit quality and audit efficiency. There are, however, efficiencies peculiar to economic services that are available to auditing firms. While auditing firms have limited opportunity to take advantage of economies of scale in the way that manufacturing firms can, they can take advantage of economies of scope.10 Economies of scope occur when offering one set of services makes it more efficient to offer other services as well, even if they are different from the audit. This can

M10794_HILARY&MCLEAN_v1.indd 240

18/07/23 5:24 PM

The audit in a modern economy  ­241

occur because, once the auditor has learned about the internal operations of a client, they are well positioned to provide potentially valuable insights into many different areas of the organization because of their general overall level of service expertise. More specifically, the auditor’s knowledge of business, accounting, auditing, and the audit client puts the audit firm in a position to provide service offerings that other firms would have trouble duplicating without investing in a very significant and steep learning curve. The end result is an economic rationale for bundling of services with one provider. The trend towards sustainability or environmental, social, and governance (ESG) reporting is one example where an auditor’s expertise can be extended to other types of information. This comparative advantage may also create potential conflicts in the audit process, however. In some situations, an auditor’s provision of other services could potentially undermine the objectives of the audit itself. Regulators and academics have expressed concerns that non-audit services impair auditor independence by developing economic bonds between the auditor and the client. A common belief among regulators and many commentators is that an auditor is less likely to challenge the audit client’s financial reporting out of fear of losing other business. The preponderance of research evidence suggests that this does not happen in most auditor–client relationships. Nevertheless, there is also evidence that suggests that investors believe that non-audit services actually impair auditor independence (Francis and Ke, 2006; Khurana and Raman, 2006). Such a belief can reduce the perceived assurance that investors obtain from the audit, reducing the external value of the audit. Consequently, laws and professional practices that put some limits on auditors’ ability to expand their service offerings with their clients can create economic value for outsiders despite the potential internal economic value that might follow from bundling some services.

3  AUDITING AND OTHER ASSURANCE SERVICES The limitations on non-audit services discussed in the last section may be less applicable in areas that result in expanded assurance and which rely on an auditor’s expertise in verifying the accuracy of information available to an organization and its stakeholders. As discussed in the last section, economies of scope can lead audit firms to develop nonaudit services for their audit clients. Utilizing their knowledge of verification procedures, audit firms also can enjoy economies of scope by developing assurance engagements for other subject matters than historical financial statements. International Standard on Assurance Engagements (ISAE) 3000 lists a number of possible subject matters for assurance other than historical financial statements: prospective financial performance, non-financial performance (e.g., sustainability reporting), physical characteristics (e.g., size of a facility), systems and processes (e.g., internal controls, IT systems), and behavior (e.g., compliance with regulation or contract). 3.1  The General Structure of Assurance Services Auditing is often interpreted as a special case of “assurance” as applied to financial statements. More generally, an assurance engagement is defined by the International Auditing and Assurance Standards Board as:

M10794_HILARY&MCLEAN_v1.indd 241

18/07/23 5:24 PM

242  Handbook of financial decision making

an engagement in which a practitioner aims to obtain sufficient appropriate evidence in order to express a conclusion designed to enhance the degree of confidence of the intended users other than the responsible party about the subject matter information (that is, the outcome of the measurement or evaluation of an underlying subject matter against criteria). (IAASB, 2004, p. 7)

In the typical assurance engagement there is a “responsible party” that is making an assertion that some subject matter has been measured or evaluated properly in relation to relevant criteria. The practitioner performs procedures to obtain either high (“reasonable”) or moderate (“limited”) assurance that the responsible party’s assertion is materially correct, and reports on the practitioner’s conclusion that is supported by those procedures.11 In some cases the engaging party that hires the practitioner may be different from the responsible (“reporting)” party. Finally, the resulting report will be issued to one or more “users.” As an example, imagine a manufacturing company is seeking to insure its inventory at its fair market value. The manufacturing company asserts to the insurance company that its inventory is worth a certain amount at fair market value. The valuation of the inventory is the subject matter and fair market pricing is the criteria by which the subject matter is measured. The insurance company seeks limited assurance on this assertion, and hires a practitioner to perform appropriate procedures to verify the values and issue a report. The manufacturing company is responsible for the inventory (and is thus the “responsible party”), while the insurance company is the “engaging party.” Because the engagement is for the purpose of informing a specific party (i.e., the insurance company), the practitioner issues a report where the insurance company is considered the primary user. Audits are assurance engagements, and thus follow the above formula. In an audit, the management of an organization is asserting that its historical financial statements (i.e., the subject matter) have been measured and presented in accordance with a relevant accounting framework (i.e., the criteria). Since auditors are typically hired by the organizations they audit, the responsible party and the engaging party are the same. The auditor (practitioner) performs appropriate procedures to obtain a high level of assurance on management’s assertions and to report a conclusion (“opinion”) that is supported by the evidence. Since historical financial statements are usually made available to a wide range of parties, there are many possible users. We will focus on two specific subject matters for assurance: non-GAAP financial information and sustainability reporting. Looking at other forms of assurance engagements is helpful in two ways. First, just as learning a second language helps you understand your own language better, so a consideration of other forms of assurance services will clarify audit assurance in general. Second, as information becomes more readily available through explosions in technology, investors and other stakeholders will likely demand greater assurance on information other than historical financial statements (Knechel, 2021). Given their experience with verification procedures, auditors are naturally suited to provide that assurance. However, as we will see, assurance over other subject matters can be thornier than with historical financial statements. 3.2  Non-GAAP Financial Measures While companies must follow an acceptable accounting framework – typically U.S. or international Generally Accepted Accounting Principles (GAAP) – when preparing

M10794_HILARY&MCLEAN_v1.indd 242

18/07/23 5:24 PM

The audit in a modern economy  ­243

their financial statements, it is common for companies to present some measures on a non-GAAP basis in other contexts, such as in quarterly earnings announcements and in the management discussion and analysis in a company’s annual report. Non-GAAP measures can be thought of as GAAP measures with adjustments (CFA Institute, 2016). A common adjustment in a non-GAAP measure is to eliminate items that are considered non-recurring, and thus are less indicative of future earnings (Black et al., 2021). Companies might also make other adjustments to present financial information on a basis that either adds information beyond the GAAP measurements or is consistent with measures used in executive compensation. For example, in its 2021 annual report, Walmart reports a Return on Investment measure (analogous to Return on Assets) that adds back accumulated depreciation and amortization so that its long-lived assets are calculated at their original historical cost. There is evidence that non-GAAP measures generally provide additional information to investors (Bradshaw et al., 2018; Black et al., 2021); but there is also concern that some companies use these non-GAAP measures to mislead investors by making the companies’ performance look better than it is. Given the significant increase in non-GAAP reporting over the last 30 years (Zhang, 2019), it is not surprising then that there has been regulation to protect investors. Current guidance in the U.S. requires companies when presenting non-GAAP measures to also provide a reconciliation with reported GAAP measures so that investors have a basis to judge the relevance of the non-GAAP measures. The SEC has also provided guidance to keep companies from giving their non-GAAP measures greater prominence than their GAAP measures (Zhang, 2019). Concern about the credibility of non-GAAP measures creates a call from both regulators and investors for auditors to provide some assurance on the credibility of these measures (Zhang, 2019; CFA Institute, 2016). Under current guidance in the U.S., auditors are only required to consider non-GAAP measures when they are included in documents containing audited financial statements. In these circumstances, auditors are required to determine if there is: (a) a material inconsistency between these measures and information presented in the financial statements; (b) a misstatement of fact; or (c) the measures are presented in a way that is misleading because information needed to evaluate the measures has been omitted or presented obscurely.12 It is conceivable that the call for assurance over non-GAAP measures may increase, whether by regulation or market demand. In terms of the general structure of assurance services, however, there is an obvious problem. If GAAP is the criteria by which practitioners provide assurance on historical financial information, what would be the criteria by which practitioners could provide assurance on non-GAAP measures, especially since different companies calculate non-GAAP measures differently? Are practitioners expected to simply provide assurance that the measures have been calculated correctly (given the way that the company defines the non-GAAP measure) and properly disclosed, or would practitioners also provide assurance that the way that the company defines the nonGAAP measure is itself reasonable? Would assurance on non-GAAP measures lead to more credible disclosures? On the one hand, there is evidence that companies with higher-quality auditors also have higherquality non-GAAP measures (Feng et al., 2023). On the other hand, having practitioners provide assurance only over calculations and disclosures (as opposed to the reasonableness of the non-GAAP measures themselves) could actually create the unintended

M10794_HILARY&MCLEAN_v1.indd 243

18/07/23 5:24 PM

244  Handbook of financial decision making

consequence of having investors rely on non-GAAP measures more than is warranted. Experimental research indicates that investors rely more on non-GAAP measures with low information content when those measures have auditors’ assurance than when not assured, presumably because they interpret the auditors’ assurance as extending to the relevance of the non-GAAP measures (Anderson et al., 2022). This is a novel example of the “expectations gap” that arises when users of financial information have a misunderstanding of what auditors are providing assurance over. 3.3  Environmental, Social, and Governance (ESG) Reporting When considering investment options, investors often consider non-financial information in addition to financial information. An example of non-financial information that is increasingly important is companies’ reporting on their social values and how their operations affect environmental, social, and governance concerns. Environmental concerns include such matters as energy efficiency, waste, greenhouse gas emissions, water usage, and pollution. Social concerns include employee safety, fair wages, employee diversity, exploitation of minority communities or lesser-developed countries, and data privacy protection. Governance concerns include board composition, how well environmental and social concerns are promulgated and monitored within an organization, managerial compensation, political lobbying, and illegal acts. Companies have a lot to gain from demonstrating their commitment to social responsibility. At the beginning of 2020, investors using ESG criteria to screen investments had $16.6 trillion invested in U.S. companies, representing nearly a third of all U.S. assets in professionally managed funds. This represented a 43% increase over the $11.6 trillion invested in 2018 (Choe, 2022). Companies disclose ESG performance in different locations, such as company websites, presentations in conference calls, annual reports (where ESG information is included with financial information), and stand-alone reports. In 2020, 90% of the top 250 companies in the world and 80% of the top companies in each nation engaged in sustainability reporting (KPMG, 2020). Given the importance, it is not surprising that companies obtain assurance for their ESG reporting. Survey evidence indicates that investors have little confidence in unaudited ESG information provided by companies given that companies tend to portray themselves in a flattering light (Cohen et al., 2011). As a result, by 2020, 62% of the top 250 companies that engage in sustainability reporting had some form of assurance over their reporting (KPMG, 2020). Unlike financial reporting, assurance often comes from non-accounting firms (Cohen and Simnett, 2015). Nonetheless, while the ESG subject matter may require expertise not found in the typical accounting education, the verification approach in the audit of financial statements – which includes risk-assessment procedures, tests of assertions, and reporting of findings – works well for the assurance of ESG reporting too (Huggins et al., 2011; Pflugrath et al., 2011).13 Assurance on ESG reporting, however, is even thornier than assurance on non-GAAP reporting. Referring again to the general structure of assurance engagements, there are a number of critical challenges. First, as with non-GAAP reporting, one issue is what criteria will be used to judge the adequacy of ESG disclosures. Currently, there are a number of reporting frameworks. The most commonly used standards come from the Global Reporting Initiative, but there are competing standards from the Sustainability

M10794_HILARY&MCLEAN_v1.indd 244

18/07/23 5:24 PM

The audit in a modern economy  ­245

Accounting Standards Board and the International Standards Organization (KPMG, 2020). Second, and perhaps more importantly, it is unclear what the subject matter of ESG reporting should be. Whereas in historical financial reporting the subject matter is straightforward (namely, the required financial statements and associated notes), in ESG reporting the subject matter depends on the company. In some cases, companies issue comprehensive ESG reports, but there are also reports on single topics (Cohen and Simnett, 2015). In fact, unlike financial reporting, where more disclosure is generally considered better than less disclosure, in ESG reporting there is a concern that firms will overemphasize ESG matters that are less important or performing well to obscure underperformance in areas that are more important (Lyon and Maxwell, 2011; O’Dwyer, 2011). Third, in any assurance engagement, practitioners only provide assurance that assertions about a subject matter are materially in accordance with relevant criteria – that is, any discrepancies between what has been asserted and reality would not affect anyone’s decision-making. Determining materiality in the realm of ESG reporting is challenging (Canning et al., 2019). With audits of financial information, auditors typically operationalize materiality by determining numerical thresholds such that errors in the financial statements are deemed immaterial if they fall below that threshold. However, with the non-financial information in ESG reporting, this approach may not be effective. First, some non-financial information is not quantitative to begin with, such as information about a company’s governance structures. Second, unlike financial statements – where all items are measured by a single unit of measure (i.e., money) – non-financial information that is quantitative will have different metrics. For example, air pollution might be measured in terms of toxic materials per cubic volume of air, whereas employee safety might be measured in terms of accidents per employee per year. Third, ESG reporting likely has a wider audience than financial reporting, making it harder to settle upon single measures of materiality. Altogether, assurance in ESG reporting is more likely to determine materiality in qualitative terms that differ with each subject matter (Edgley et al., 2015; Moroney and Trotman, 2016). Finally, there is the question of the level of assurance provided for ESG reporting. While ESG assurance standards generally use the same binary options used in assurance of historical financial information (i.e., “limited” vs. “reasonable” assurance), it appears that, in practice, ESG assurance falls along a continuum. Accordingly, it is common for ESG assurance reports to provide some level of detail on the procedures performed to give an indication of the level of assurance provided (Deegan et al., 2006). 3.4  Evolution of Assurance Processes In accounting research, audit quality has traditionally been understood as a function of auditor competence and auditor independence: the auditor must have the competence to have a reasonable likelihood of discovering material misstatements and the independence to pressure the client to correct those misstatements, or report that the client’s financial statements are materially misstated (DeAngelo, 1981). From the perspective described earlier that audit is a kind of economic service, we can generalize to a broader definition of audit quality: the quality of an audit service is a function of; (a) the competencies and other resources that all parties involved in the audit process have to offer; and (b)

M10794_HILARY&MCLEAN_v1.indd 245

18/07/23 5:24 PM

246  Handbook of financial decision making

how well those competencies and other resources are integrated into the audit service offering. While the auditor must have ultimate responsibility for the conduct of the audit, audit quality depends upon the contributions of the auditor, management, the audit committee, and potentially other parties as well (Knechel et al., 2020). As we broaden our considerations to other assurance services, we need to think about how these other assurance services challenge both the competencies available and the nature of their integration into assurance service offerings. As we will see, these considerations may also lead us to rethink the traditional audit of historical financial statements as well. 3.4.1  Competencies (and other resources) in assurance services With audits of historical financial information, there are two general auditor competencies that are relevant: knowledge of appropriate financial reporting (determined largely through a knowledge of accounting standards); and knowledge of appropriate auditing procedures (determined by experience and a knowledge of auditing standards). Both of these competencies are part of the core in an accounting education, and accounting practitioners cannot obtain a license without passing tests related to each competency. These competencies are usually supplemented with the experience of conducting audits in specific industries. However, with other assurance services, it is less likely that an accounting education will provide the relevant expertise in the subject matter being assured and the criteria by which the subject matter is evaluated. ESG reporting, for example, calls upon a wide range of expertise, including areas like engineering and environmental science. Not many accounting graduates will possess that expertise, so accounting firms performing assurance on ESG reporting will need to hire non-accounting personnel either as employees of the firm or as independent contractors for specific assurance engagements (Knechel, 2021). For non-accounting firms providing assurance on ESG reporting, the opposite may be true: they may need people trained in auditing practices, who are likely to come from an accounting background, to supplement the subject matter expertise of the firm. The divergence between subject matter competencies and competence in verification procedures raises the question of whether auditing (and assurance generally) should be housed exclusively within an accounting education. Perhaps auditing should be treated as a stand-alone discipline (Brydon, 2019). At the present time, that may not make much sense given that audits of historical financial statements are by far the dominant form of assurance services. If regulators and markets increasingly seek assurance on other kinds of information, however, a change might be appropriate (Knechel, 2022). Another issue regarding competencies has to do with the competencies of the organizations whose representations are being assured. In the standard model for assurance services, there is a responsible party making assertions about some subject matter according to some criteria, and the practitioner provides assurance that those assertions are materially correct. The standard model presupposes that the responsible party has the competency to make credible assertions. Even with historical financial statements, however, this presupposition may be dubious. While SEC regulations prohibit auditing firms from preparing a client’s financial statements (e.g., Regulation S-X),14 AICPA standards (which would apply to the audits of non-publicly traded organizations) do allow for auditing firms to prepare the financial statements if there are safeguards present to protect the auditor’s independence (AICPA, 2014). The AICPA standards reflect a realism that many

M10794_HILARY&MCLEAN_v1.indd 246

18/07/23 5:24 PM

The audit in a modern economy  ­247

organizations lack sufficient accounting knowledge to credibly put together their own financial statements, and that it is not cost-beneficial to hire another accounting firm to do that work. The issue of an organization’s competency to make credible assertions likely becomes more pronounced for subject matters other than financial reporting. For example, while larger organizations may have the resources to perform their own ESG reporting (or hire a third party to do so), smaller organizations may not. For organizations with limited resources, the standard assurance model may be inadequate. Accordingly, the AICPA has issued standards on a new kind of assurance engagement (“Direct Examination”) where the responsible party is not required to measure or evaluate the underlying subject matter or provide a written assertion to the practitioner.15 Instead, the responsible party relies upon the expertise of the assurance provider to measure or evaluate the underlying subject matter directly and issue a report. The AICPA explicitly created the standard due to ­complexity of new forms of non-financial information (Tysiac, 2020). 3.4.2  Integrating competencies (and other resources) into assurance processes As stated before, assurance on ESG reporting is provided by both accounting and nonaccounting firms. Researchers who have compared the assurance provided by each have noticed important differences. Not surprisingly, accounting firms rely on auditing practices to model their assurance on ESG reporting. Non-accounting firms, on the other hand, tend to blur the line between assurance and consulting, with greater emphasis on making recommendations for improving ESG performance and reporting (O’Dwyer and Owen, 2005; Casey and Grenier, 2015). Also, while both kinds of firms tend to provide limited assurance (rather than reasonable assurance), accounting firms are more careful in their reporting to signal the lower level of assurance (O’Dwyer and Owen, 2005). Given the long-standing concerns in the audit profession about how non-audit services, especially consulting, potentially impair auditor independence, it is to be expected that there would also be concerns about the independence of non-accounting firms (O’Dwyer and Owen, 2005). Deegan et al. (2006) find that, in contrast to accounting assurers, non-­ accounting assurers in their assurance reports often praise their clients for their commitment to ESG values. This practice certainly looks strange to people from an audit background. Nonetheless, that non-accounting firms find it intuitive to focus more on recommendations (than do accounting firms) raises the question of whether the concerns about independence are excessive, especially if these concerns inhibit the value that assurance engagements can provide. It is worth considering that many companies do not release the assurance reports on their ESG reporting, suggesting that companies often obtain assurance more for internal purposes (Cohen and Simnett, 2015). Even with the audits of financial statements, the evidence that non-audit services hurt audit quality has never been well established empirically (Ashbaugh et al., 2003; Habib, 2012). If we think of independence within the larger question of how well integrated the competencies of various parties are into the assurance process, then restricting assurers from providing recommendations may appear to compromise the assurance process as much as traditional threats to independence do. 3.4.3  Will the assurance model hold up? To restate, the traditional model of assurance is that there is a responsible party that makes assertions about whether a subject matter is properly measured or evaluated in

M10794_HILARY&MCLEAN_v1.indd 247

18/07/23 5:24 PM

248  Handbook of financial decision making

relation to relevant criteria. A practitioner provides assurance to third parties on the assertions by performing procedures designed to probe the credibility of the assertions and reporting a conclusion based upon those procedures. From the prior discussion, there are a number of reasons why the development of assurance practices is challenging this model: ●

●

●

●

●

Subject matter: With historical financial statements, the subject matter is settled by practice and regulation. With other kinds of information, the subject matter may not be settled – which may call into question whether the assurance provider not only needs to provide assurance on the subject matter identified by management, but also on whether management has identified the subject matter appropriately. This is especially relevant for assurance on ESG reporting. Relevant criteria: With historical financial information, the relevant criteria will depend on the appropriate accounting framework (e.g., U.S. or international GAAP). While auditors are required to determine whether financial statements have been prepared in accordance with an appropriate accounting framework, this determination is typically not challenging. In other kinds of assurance engagements, however, the criteria may be less settled. At an extreme, such as non-GAAP financial measures, there may be no criteria independent of what the responsible party chooses. Assurance providers may find increasingly that, not only should they provide assurance based upon the criteria identified by management, but also that they will need to provide assurance on the proper selection of criteria. Logically, assurance on criteria requires a kind of judgment not itself based upon criteria. Responsible party: With historical financial statements, management is the party responsible for the financial statements. This is commonly understood to mean that management is responsible for the assertion that it has measured its historical financial information in relation to a relevant accounting framework. With newer kinds of non-financial information, management may not have the competence to perform the measurement or evaluation of the subject matter, and it may not be cost-effective for management to hire one party to perform the measurements/ evaluations and another party to provide the assurance over the measurements/ evaluations. Accordingly, assurance providers might be performing assurance engagements like direct examination engagements, where they measure or evaluate the subject matter directly without relying on managerial assertions. Management is still responsible for the underlying subject matter, but not for assertions about the underlying subject matter. Assurance: Traditionally, assurance is understood to be provided on two levels: reasonable and limited. As subject matters for assurance increase, it is likely that users of information may have difficulty understanding what the level of assurance represents. Assurance on ESG reporting suggests that assurance might develop more along a continuum, with greater transparency from the assurance provider on the procedures employed to communicate the level of assurance provided. Third parties: The assurance model presupposes that assurance is being provided to external stakeholders who lack direct access to an organization’s information. As subject matters become more complex and diverse, however, management (and boards of directors) may find it increasingly important to obtain assurance for

M10794_HILARY&MCLEAN_v1.indd 248

18/07/23 5:24 PM

The audit in a modern economy  ­249

internal purposes. As mentioned before, many firms obtain assurance on their ESG reporting but do not release the assurance reports. It may be the case that the lines between external assurance and internal assurance, as well as between assurance and consulting, become increasingly blurred.

4 CONCLUSIONS The audit of traditional financial statements has been the raison d’être of the profession since its inception. However, as assurance practices evolve in response to the everexpanding information demands of the market, accounting firms will be challenged to expand their competencies and organizational structures to take advantage of new demands, and there are important implications for the practice of audit. One possibility is that new assurance practices will be seen primarily as threats to the integrity of audits, as is often the case with other non-audit services. A more progressive interpretation is to consider the expansion of the information environment as an opportunity for the profession. The expansion of assurance opportunities may challenge the traditional model of audit, but a changing model of assurance might bring changes to audits too. As accounting firms, their clients, professional bodies, regulators, and other stakeholders wrestle with these issues, the experience might provide a basis for meaningful innovation in audit practices (Humphrey et al., 2021). Consider the following possibilities: ●

The steps in a traditional audit involve management preparing its financial statements, the auditor auditing those financial statements, and management and auditor jointly agreeing what information should be presented in the financial statement. From this perspective, management “produces” the information and auditors “verify” the information. While the principle is reasonable, it may not work well as an absolute standard. As discussed before, auditors are increasingly embedded with their clients, not only in terms of the ongoing nature of audit procedures but also as a potential source of advice to a client. Insofar as auditors and their audit clients have long-term relationships, issues that are settled in one year will likely have a long-term effect on the financial statements. The jointness of reporting decisions continues from year to year, and the audit profession and regulators might benefit from a new perspective on what it means for an auditor to be “independent”, recognizing that preparation and audit of the financial statements requires co-­ production (Knechel et al., 2020). ● Auditors may benefit from considering a wider range of participants and stakeholders in the reporting and assurance process (or ecosystem). For example, decisions could be influenced by the type of information being assured and through interactions with a broader range of stakeholders (Edgley et al., 2015). Also, the espoused level of assurance may be conditional on understanding a broader range of needs from stakeholders, potentially closing the so-called “expectations gap.” Active involvement of stakeholders in the assurance process brings their needs into consideration in conducting and reporting on the audit process (Knechel et al., 2020; Driskill et al., 2022).

M10794_HILARY&MCLEAN_v1.indd 249

18/07/23 5:24 PM

250  Handbook of financial decision making ●

It is likely that the users of financial information will increasingly wish to understand financial information within a larger context, including various forms of non-financial disclosures. Under current practice, auditors limit their opinion to the financial statements, with only modest expectations of the assurance provided for disclosures that accompany the financial statements. This convention is probably due to a narrow focus on accounting standards as the basis for reporting. However, knowledge of verification processes transcends financial accounting and can be used to provide broader assurance over non-financial information as well.

If there is innovation in audit practices, the innovation is likely to come from above (standard-setters and regulators) and below (audit firms and other audit stakeholders). In both cases, there may be a need for some relinquishing of control. In the case of standardsetters and regulators, it might be helpful to understand themselves as part of the financial (and non-financial) reporting ecosystem rather than as standing over it. Humphrey et al. (2021) draw an interesting analogy to the world of electronics, where standardization is not about regulation of behaviors and processes but about enhancing the “compatibility” and “interoperability” of products. In the case of audit firms, innovating the audit service likely involves some combination of developing a better understanding of the needs of audit stakeholders and better integrating the knowledge (and other resources) of audit stakeholders. This may involve resource liquefaction – the availability of knowledge (and other resources) to all participants in the audit network – which would promote resource density, that is, the ability to focus relevant resources to solve problems and create opportunities in the audit process (Lusch and Nambisan, 2015). Such a possibility may require participants in the audit process to relinquish some of their control to the benefit of all in the form of enhanced assurance in a world of increasingly complex financial and nonfinancial reporting.

NOTES   1. “Adverse selection” describes situations where sellers know more about the quality of goods or services offered than buyers. At an extreme, buyers will respond to this information asymmetry by avoiding the market altogether.  2. “Moral hazard” describes situations where workers (agents) have incentives to misbehave because their actions are not observable to those they work for (principals). When agent behavior is completely unobservable, principals are likely to assume the worst and reduce the compensation they offer.   3. In the extreme, an auditor might resign if faced with a client who is extremely reticent and non-cooperative. In most jurisdictions, an auditor can terminate the relationship with a client “at will.” Resigning is the ultimate threat an auditor can use to foster cooperation with a client; but a relationship that reaches that point is probably not going to end well.   4. As is discussed in more detail below, the issues of “cooperation” and “independence” are inherently intertwined. While an auditor needs a certain level of cooperation from the client, they must also maintain an acceptable level of professional skepticism (i.e., objectivity and independence) when dealing with the client.   5. Discussions with management are referred to as “client inquiries” in the auditing standards.   6. Public Company Accounting Oversight Board (PCAOB). Auditing Standard (AS) 1001.03, Responsibilities and functions of the independent auditor.   7. PCAOB, AS 2805.02, Management representations.

M10794_HILARY&MCLEAN_v1.indd 250

18/07/23 5:24 PM

The audit in a modern economy  ­251

  8. PCAOB, AS 1015.12, Due professional care in the performance of work.   9. PCAOB, AS 2605, Consideration of the internal audit function. 10. Since every engagement has different characteristics, scaling up for more engagements means adding more unique activities rather than more standard outputs. 11. Reasonable or high assurance refers to the outcome of a standard audit of information, so it is considered highly reliable. Limited assurance refers to the application of a reduced set of audit procedures to information such that the information is not considered as reliable as in an audit. 12. American Institute of Certified Public Accountants (AICPA). AU-C Section 720, The auditor’s responsibilities relating to other information included in annual reports. 13. In early 2022, the SEC released a proposal to make some climate-related disclosures mandatory for listed companies. The proposal included provisions for assurance over the information (SEC, 2022). 14. SEC, Regulation S-X: Accounting rules form and content of financial statements. Rule 2-01(c) (4)(i)(B). 15. AICPA, AT-C Section 206, Direct examination engagements.

REFERENCES AICPA, 2014. AICPA Code of Professional Conduct. New York: American Institute of Certified Public Accountants. Anderson, S.B., Hobson, J.L., Sommerfeldt, R.D., 2022. Auditing non-GAAP measures: Signaling more than intended. Contemporary Accounting Research 39 (1), 577–606. Aobdia, D., Lin, C.-J., Petacchi, R., 2015. Capital market consequences of audit partner quality. The Accounting Review 90 (6), 2143–2176. Ashbaugh, H., LaFond, R., Mayhew, B.W., 2003. Do nonaudit services compromise auditor independence? Further evidence. The Accounting Review 78 (3), 611–639. Balliet, D., Van Lange, P.A.M., 2013. Trust, punishment, and cooperation across 18 societies: A meta-analysis. Perspectives on Psychological Science 8 (4), 363–379. Beasley, M.S., Carcello, J.V., Hermanson, D.R., Neal, T.L., 2009. The audit committee oversight process. Contemporary Accounting Research 26 (1), 65–122. Bell, T.B., Causholli, M., Knechel, W.R., 2015. Audit firm tenure, non-audit services, and internal assessments of audit quality. Journal of Accounting Research 53 (3), 461–509. Black, D.E., Christensen, T.E., Ciesielski, J.T., Whipple, B.C., 2021. Non-GAAP earnings: A consistency and comparability crisis? Contemporary Accounting Research 38 (3), 1712–1747. Botosan, C.A., Plumlee, M.A., 2005. Assessing alternative proxies for the expected risk premium. The Accounting Review 80 (1), 21–53. Bradshaw, M.T., Christensen, T.E., Gee, K.H., Whipple, B.C., 2018. Analysts’ GAAP earnings forecasts and their implications for accounting research. Journal of Accounting and Economics 66 (1), 46–66. Brydon, D., 2019. Assess, Assure and Inform: Improving Audit Quality and Effectiveness. Report of the Independent Review into the Quality and Effectiveness of Audit. London: UK Government. Canning, M., O’Dwyer, B., Georgakopoulos, G., 2019. Processes of auditability in sustainability assurance: The case of materiality construction. Accounting and Business Research 49 (1), 1–27. Carcello, J.V., Nagy, A.L., 2004. Audit firm tenure and fraudulent financial reporting. Auditing: A Journal of Practice & Theory 23 (2), 55–69. Casey, R.J., Grenier, J.H., 2015. Understanding and contributing to the enigma of corporate social responsibility (CSR) assurance in the United States. Auditing: A Journal of Practice & Theory 34 (1), 97–130. Catasús, B., Hellman, N., and Humphrey, C., 2013. Revisiones Roll I Bolagsstyminge (The Role of Auditing in Corporate Governance). Stockholm, Sweden: SNS Förlag.

M10794_HILARY&MCLEAN_v1.indd 251

18/07/23 5:24 PM

252  Handbook of financial decision making

CFA Institute, 2016. Bridging the gap: Ensuring effective non-GAAP and performance reporting. https://www.cfainstitute.org/-/media/documents/article/position-paper/bridging-the-gap-ensuringnon-gaap-and-performance-reporting.ashx. Choe, S., 2022. Explainer: ESG investing and the debate surrounding it. Associated Press. May 19. https://www.brproud.com/business/ap-business/explainer-esg-investing-and-the-debate-surround​ ing-it/. Chow, C.W., 1982. The demand for external auditing: Size, debt and ownership influences. The Accounting Review 57 (2), 272–291. Cohen, J.R., Holder-Webb, L., Nath, L., Wood, D., 2011. Retail investors’ perceptions of the decision-usefulness of economic performance, governance, and corporate social responsibility disclosures. Behavioral Research in Accounting 23 (1), 109–129. Cohen, J.R., Krishnamoorthy, G., Wright, A.M., 2002. Corporate governance and the audit process. Contemporary Accounting Research 19 (4), 573–594. Cohen, J.R., Simnett, R., 2015. CSR and assurance services: A research agenda. Auditing: A Journal of Practice & Theory 34 (1), 59–74. DeAngelo, L.E., 1981. Auditor size and audit quality. Journal of Accounting and Economics 3 (3), 183–199. Deegan, C., Cooper, B.J., Shelly, M., 2006. An investigation of TBL report assurance statements: UK and European evidence. Managerial Auditing Journal 21 (4), 329–371. DeFond, M.L, Zhang, J., 2014. A review of archival auditing research. Journal of Accounting and Economics 58 (2), 275–326. Driskill, M., Knechel, W.R., Thomas, E., 2022. Financial auditing as an economic service. Current Issues in Auditing 16 (2), 39–50. Edgley, C., Jones, M.J., Atkins, J., 2015. The adoption of the materiality concept in social and environmental reporting assurance: A field study approach. British Accounting Review 47 (1), 1–18. Felix, R., Gaynor, G., Pevzner, M., Williams, J.L., 2017. Societal trust and the economic behavior of nonprofit organizations. Advances in Accounting 39, 21–31. Feng, Z., Francis, J.R., Shan, Y., Taylor, S.L., 2023. Do high-quality auditors improve non-GAAP reporting? The Accounting Review 98 (1), P215–P250. Fornelli, C., 2013. CAQ provides perspectives on understanding audit quality. Sarbanes-Oxley Compliance Journal. http://www.s-ox.com/dsp_getFeaturesDetails.cfm?CID=3260. Francis, J.R., Ke, B., 2006. Disclosure of fees paid to auditors and the market valuation of earnings surprises. Review of Accounting Studies 11, 495–523. Fukuyama, F., 2002. Social capital and development. SAIS Review (1989–2003) 22 (1), 23–37. Gibbins, M., Salterio, S., Webb, A., 2001. Evidence about auditor-client management negotiation concerning client’s financial reporting. Journal of Accounting Research 39 (3), 535–563. Griffith, E.E., Hammersley, J.S., Kadous, K., 2015. Audits of complex estimates as verification of management numbers: How institutional pressures shape practice. Contemporary Accounting Research 32 (3), 833–863. Habib, A., 2012. Non-audit service fees and financial reporting quality: A meta-analysis. Abacus 48 (2), 121–157. Hope, O.-K., Langli, J.C., Thomas, W.B., 2012. Agency conflicts and auditing in private firms. Accounting, Organizations, and Society 37 (7), 500–517. Huggins, A., Green, W.J., Simnett, R., 2011. The competitive market for assurance engagements on greenhouse gas statements: Is there a role for assurers from the accounting profession? Current Issues in Auditing 5 (2), A1–A12. Humphrey, C., Sonnerfeldt, A., Komori, N., Curtis, E., 2021. Audit and the pursuit of dynamic repair. European Accounting Review 30 (3), 445–471. International Auditing and Assurance Standards Board (IAASB), 2004. International Standard on Assurance Engagements 3000: Assurance Engagements other than Audits or Reviews of Historical Financial Information. New York. Iyer, V.M., Rama, D., 2004. Clients’ expectations on audit judgments: A note. Behavioral Research in Accounting 16, 63–74. Khurana, I.K., Raman, K.K., 2006. Litigation risk and the financial reporting credibility of Big 4

M10794_HILARY&MCLEAN_v1.indd 252

18/07/23 5:24 PM

The audit in a modern economy  ­253

versus non-Big 4 audits: Evidence from Anglo-American countries. The Accounting Review 79 (2), 473–495. Kinney, Jr., W.R, Palmrose, Z.-V., Scholz, S., 2004. Auditor independence, non-audit services, and restatements: Was the U.S. government right? Journal of Accounting Research 42 (3), 561–588. Knack, S., Keefer, P., 1997. Does social capital have an economy payoff ? A cross-county investigation. Quarterly Journal of Economics 112 (4), 1251–1288. Knechel, W.R., 2021. The future of assurance in capital markets: Reclaiming the economic imperative of the auditing profession. Accounting Horizons 35 (1), 133–151. Knechel, W.R., 2022. Who stands for audit? A commentary on the ‘Brydon report.’ International Journal of Auditing 26 (1), 27–31. Knechel, W.R., Krishnan, G.V., Pevzner, M., Shefchik, L.B., Velury, U., 2013. Audit quality: Insights from the academic literature. Auditing: A Journal of Practice & Theory 32 (Supplement 1), 385–421. Knechel, W.R., Mintchik, N., Pevzner, M., Velury, U., 2019. The effects of generalized trust and civic cooperation on the Big N presence and audit fees across the globe. Auditing: A Journal of Practice & Theory 38 (1), 193–219. Knechel, W.R, Niemi, L., Sundgren, S., 2008. Determinants of auditor choice: Evidence from a small client market. International Journal of Auditing 12 (1), 65–88. Knechel, W.R., Thomas, E., Driskill, M., 2020. Understanding financial auditing from a service perspective. Accounting, Organizations and Society 81, 1–23. KPMG, 2020. The KPMG Survey of Sustainability Reporting 2020. https://assets.kpmg/content/ dam/kpmg/xx/pdf/2020/11/the-time-has-come.pdf. Lin, S., Pizzini, M., Vargus, M., Bardhan, I.R., 2011. The role of the internal audit function in the disclosure of material weaknesses. The Accounting Review 86 (1), 287–323. Lobo, G.J., Zhao, Y., 2013. Relation between audit effort and financial report misstatements: Evidence from quarterly and annual restatements. The Accounting Review 88 (4), 1385–1412. Lusch, R.F., Nambisan, S., 2015. Service innovation. MIS Quarterly 39 (1), 155–176. Lyon, T.P., Maxwell, J.W., 2011. Greenwash: Corporate environmental disclosure under threat of audit. Journal of Economics & Management Strategy 20 (1), 3–41. Mansi, S.A., Maxwell, W.F., Miller, D.P., 2004. Does auditor quality and tenure matter to investors? Evidence from the bond market. Journal of Accounting Research 42 (4), 755–793. Mitra, S., Hossain, M., Deis, D.R., 2007. The empirical relationship between ownership characteristics and audit fees. Review of Quantitative Finance and Accounting 28 (3), 257–285. Moroney, R., Trotman, K.T., 2016. Differences in auditors’ materiality assessments when auditing financial statements and sustainability reports. Contemporary Accounting Research 33 (2), 551–575. Nanda, D., Wysocki, P., 2011. The relation between trust and accounting quality. Working paper, University of Miami. Nelson, M.W., 2009. A model and literature review of professional skepticism in auditing. Auditing: A Journal of Practice & Theory 28 (2), 1–34. Newman, D.P., Patterson, E.R., Smith, J.R., 2005. The role of auditing in investor protection. The Accounting Review 80 (1), 289–313. O’Dwyer, B., 2011. The case of sustainability assurance: Constructing a new assurance service. Contemporary Accounting Research 28 (4), 1230–1266. O’Dwyer, B., Owen, D.L., 2005. Assurance statement practice in environmental, social and sustainability reporting: A critical evaluation. British Accounting Review 37 (2), 205–229. Pevzner, M., Xie, F., Xin, X., 2015. When firms talk, do investors listen? The role of trust in stock market reactions to corporate earnings announcements. Journal of Financial Economics 117 (1), 190–223. Pflugrath, G., Roebuck, P., Simnett, R., 2011. Impact of assurance and assurer’s professional affiliation on financial analysts’ assessment of credibility of corporate social responsibility information. Auditing: A Journal of Practice & Theory 30 (3), 2011. Pittman, J.A., Fortin, S., 2004. Auditor choice and the cost of debt capital for newly public firms. Journal of Accounting and Economics 37 (1), 113–136.

M10794_HILARY&MCLEAN_v1.indd 253

18/07/23 5:24 PM

254  Handbook of financial decision making

Securities and Exchange Commission (SEC), 2022. Proposed rule: The enhancement and standardization of climate-related disclosures for investors. https://www.sec.gov/rules/pro​posed/2022​/ 33-11042.pdf. Stice, J., 1991. Using financial and market information to identify pre-engagement factors associated with lawsuits against auditors. The Accounting Review 66 (3), 516–533. Tysiac, K., 2020. “Direct examination” engagement created by SSAE No. 21. Journal of Accountancy, September 30. https://www.journalofaccountancy.com/news/2020/sep/aicpa-ssaeno-21-direct-examination-engagement.html. Westermann, K., Cohen, J., Trompeter, G., 2019. PCAOB inspections: Public accounting firms on “trial”. Contemporary Accounting Research 36 (2), 694–731. Zhang, J., 2019. Learning from the current research on non-GAAP financial measures. CPA Journal 89 (7), 32–37.

M10794_HILARY&MCLEAN_v1.indd 254

18/07/23 5:24 PM

Part III.2 Frameworks

M10794_HILARY&MCLEAN_v1.indd 255

18/07/23 5:24 PM

12.  Accounting and prices Steven J. Monahan*

1 INTRODUCTION I am convinced that if it were the result of deliberate human design, and if the people guided by the price changes understood that their decisions have significance far beyond their immediate aim, [the price system] would have been acclaimed as one of the greatest triumphs of the human mind. (Friedrich August von Hayek—The use of knowledge in society)

Although Hayek (1945) focuses on prices of goods and services, his arguments are equally valid for the prices of securities. As discussed in Arrow (1964), security prices play a central role in determining the amount that investors pay to smooth their consumption across time (e.g., saving for retirement) and states of nature (e.g., buying an index fund). Security prices also determine the amount of funding allocated to entrepreneurs, and thus they affect incentives to be entrepreneurial. Finally, as discussed in Bond et al. (2012), security prices serve as information signals that people can use when: (1) making real decisions (e.g., learning from the stock price reaction to a merger announcement); (2) evaluating contractual performance (e.g., stock-based compensation); and (3) conducting regulatory oversight (e.g., estimating the probability that a financial institution will fail). Given its importance, it is not surprising that a number of institutions have emerged and/or evolved to facilitate the security price system. In this chapter, I focus on the financial reporting system. I do this for three reasons. First, a primary objective of financial reporting is to facilitate security price formation.1 Second, analytical and empirical evidence implies that financial reporting plays a central role in determining enterprise and equity value. Finally, despite its importance, fundamental questions about the role that the financial reporting system plays in facilitating security price formation remain unanswered. I begin by discussing valuation. I propose three criteria for judging a valuation model, which are: (1) generalizability; (2) links between payoffs and operating performance; and (3) forecast horizon and continuing value estimate. Then I describe the longstanding dividend discount model (DDM). The key point of this discussion is that, although the DDM is generalizable, it fails to meet the second two criteria. The reason for this is that, from a first-order perspective, dividend policy is irrelevant. Hence, forecasting dividends is an arbitrary process, and thus the DDM is impracticable. This paradoxical result is crucial because it opens the door for accounting-based valuation models. Colloquially speaking, if forecasting dividends doesn’t make sense, investors will have to forecast something else—and accounting numbers (e.g., earnings) are a natural candidate. With the above in mind, I discuss accounting-based valuation models. I focus on the residual income valuation (RIV) model, which is well known and intuitive. I emphasize two results: (1) the RIV model is generalizable; and (2) it expresses value as a function of expected future accounting numbers. The importance of the first result is obvious. 256

M10794_HILARY&MCLEAN_v1.indd 256

18/07/23 5:24 PM

Accounting and prices  ­257

The second result is important for two reasons. First, it implies that, unlike the DDM, the RIV model does not rely on forecasts that are based on arbitrary assumptions about companies’ dividend policies. Second, it leaves open the possibility that the RIV model is practicable. Specifically, if accounting numbers are linked to operating outcomes that drive value creation and destruction, RIV-based value estimates are a function of: (1) forecast payoffs that can be meaningfully predicted ex ante and meaningfully evaluated ex post; and (2) a continuing value estimate that meaningfully embeds a constant-growth assumption. After discussing the a priori advantages of accounting-based valuation models, I turn to the empirical evidence. First, I discuss the use and usefulness of accounting-based valuation models. I focus on three issues: (1) the relative accuracy of accounting-based value estimates vis-à-vis estimates based on the DDM and the discounted cash flow (DCF) model, which is a special case of the RIV model; (2) the usefulness of historical accounting numbers for forecasting future accounting numbers; and (3) the role that accounting numbers play in assessing risk. Second, I discuss the price impact of higher ­accounting quality, which, broadly speaking, manifests itself either through its effect on expected future earnings or its effect on the cost of capital. Finally, I discuss price informativeness, which, in my opinion, is a central issue that we know too little about. Throughout my discussion of the empirical evidence, I emphasize that, although a lot has been learned, our understanding of fundamental issues remains incomplete.

2 VALUATION Values [in a rational and perfect economic environment] are determined solely by “real” considerations in this case the earning power of the firm’s assets and its investment policy and not by how the fruits of the earning power are “packaged” for distribution. (Merton H. Miller and Franco Modigliani—Dividend policy, growth, and the valuation of shares)

In this section, I propose three criteria for evaluating a valuation model, and then discuss how the DDM, RIV, and DCF models perform with respect to these criteria. 2.1  Criteria for Judging a Valuation Model The three criteria are: 1. Generalizability. 2. Links between payoffs and operating performance. 3. Forecast horizon and continuing value estimate. Two comments about these criteria are warranted. First, they are not absolutes. Rather, they should be applied on a relative basis with the objective of identifying the best model not the perfect model—because there probably isn’t one. For example, it is unlikely that any model will have payoffs that are perfectly linked to the company’s operating performance. However, the payoffs for some models will have a more direct link to operating performance than the payoffs of other models.

M10794_HILARY&MCLEAN_v1.indd 257

18/07/23 5:24 PM

258  Handbook of financial decision making

Second, the objective underlying the criteria is to obtain a model that is realistic and practicable. A model that is based on realistic assumptions is generalizable, and thus, when implemented correctly, it will yield an accurate estimate of value. However, the model must also be practicable. For instance, if the payoffs that determine value cannot be meaningfully forecast ex ante and meaningfully evaluated ex post, the model is not practicable. Similarly, a model is not practicable if to implement it the user must develop an explicit forecast of every future payoff. 2.1.1 Generalizability A model is generalizable if it can be applied to all cases. Generalizability is ultimately a function of the assumptions underpinning the model. If the assumptions reflect reality, the model is generalizable. Within the context of valuation, two standard assumptions are: (1) no arbitrage and (2) financial policy irrelevance. No arbitrage implies that investors cannot create a zero-net-investment (negativenet-investment) portfolio on date ​t​that will subsequently earn a guaranteed positive (nonnegative) return. Colloquially speaking, there are no free lunches. Although there is a growing body of literature that challenges the no arbitrage assumption, it remains the centerpiece of most valuation models. The reason for this is two-fold. First, there is no well-accepted alternative. Second, although arbitrage opportunities may exist at any point in time, they are fleeting. Colloquially speaking, free lunches get eaten. Financial policy irrelevance implies that, from a first-order perspective, choices about dividend policy and capital structure are irrelevant. These are bedrock assumptions that date back to seminal results in Miller and Modigliani (1961) and Modigliani and Miller (1958). The basic intuition for dividend policy irrelevance (DPI) is straightforward: Although a company that creates value will pay larger dividends ceteris paribus, paying a larger dividend does not create value. Rather, value is created from operating activities. To elaborate, if a company pays a dividend, it can still fund its operations by issuing shares; and, if its current shareholders prefer to maintain their ownership percentage, they can purchase additional shares with the dividends they receive. Alternatively, if the company does not pay a dividend, it can invest the funds in zero net present value (NPV) projects; and, if its shareholders want cash immediately, they can create “homemade” dividends by selling shares. Hence, dividend policy choices have no effect on value—that is, they are zero NPV decisions. Capital structure irrelevance (CSI) implies that whether the company uses debt or equity to fund its operations does not matter. Combining DPI and CSI leads to the conclusion that value is created by the operations, and that the source of funds used to finance operating activities is irrelevant. Like the no arbitrage assumption, the assumption that financial policy is irrelevant is a first approximation. That said, the underlying intuition is compelling: A dollar obtained from a debt issuance will buy just as much as a dollar obtained from an equity issuance or a dollar of cash that is held in the company’s bank account. Colloquially speaking, a company’s operations do not care how they are funded. 2.1.2  Link between payoffs and operating performance The amount investors pay for a security is a function of the payoffs the security is expected to generate. Hence, forecasting future payoffs is the essence of valuation. Because operating

M10794_HILARY&MCLEAN_v1.indd 258

18/07/23 5:24 PM

Accounting and prices  ­259

activities drive value, it is essential that investors forecast payoffs that are the result of operating performance. Moreover, payoffs that are directly linked to operating performance can be meaningfully forecast ex ante and meaningfully evaluated ex post. For example, by evaluating market conditions, the competitive landscape, supplier relations, managerial talent, and so on investors can combine forecasts of sales growth, margins, cost structure, productivity, and working capital utilization into a forecast of earnings. Subsequently, this forecast can be compared to realized earnings, and the reasons for the forecast error can be identified, learned from, and used as a basis for revising the forecasts of earnings that remain unrealized. 2.1.3  Forecast horizon and continuing value estimate It is impossible to form an infinite sequence of explicit forecasts. Rather, an assumption must be made that, after a specific future date ​T​, which I refer to as the horizon date, the payoffs will evolve in a well-defined manner. A commonplace assumption is to assume that, subsequent to date T ​ ,​ the payoffs generated by company i​ ​, ​​Xi,t ​  ​, will grow at a constant rate ​Gi,T ​ X ​​.  If true, and if company ​i’​s gross discount rate, R ​ i​ ​, is neither time-varying nor ​𝔼​ 0​[​Xi,T ​  ​]​ × ​Gi,T ​ X ​​  ​𝔼​ 0​[​Xi,T+1 ​  ​ ​ stochastic, the continuing value estimate is equal to ​​_   ​  =  _ ​​R​ ​ − ​G​ X ​] ​​   ​ (​𝔼 ​ ​ 0[​⋅]​​is the ​Ri​ ​ − ​Gi,T ​ X ​ ​    i i,T expectation operator conditional on information available on date zero, which is the date on which the value estimate is being developed). Two comments about this criterion are warranted. First, it does not simply require that X ​ 0​[​Xi,T+1 ​  ​]​ ˆ ​  ​​ such that the continuing value estimate can be stated as ​​𝔼_ there exist a number ​ G ​ .​​ X    i,T ˆ ​Ri​ ​ − ​​ G ​​ i,T ​​ 

This restriction is too lax because it would be met in almost all cases.2 For example, if the X continuing value estimate is positive and ​𝔼​ 0​[​Xi,T ​  ​]​    ​Ri​ ​​ that will suffice. Rather, the criterion requires that there exists a date T ​ ​such that ∀ ​  t  >  0, ​​​ 𝔼​ 0​[​Xi,T+t ​  ​]​  =  ​𝔼​ 0​[​Xi,T+t−1 ​  ​]​ × ​Gi,T ​ X ​​ and ​Gi,T ​ X ​ ​