357 122 10MB
English Pages [491] Year 2021
THE ROUTLEDGE HANDBOOK OF FINTECH
The Routledge Handbook of FinTech offers comprehensive coverage of the opportunities, challenges and future trends of financial technology. This handbook is a unique and in-depth reference work. It is organised in six thematic parts. The first part outlines the development, funding, and the future trends. The second focuses on blockchain technology applications and various aspects of cryptocurrencies. The next covers FinTech in banking. A significant element of FinTech, mobile payments and online lending, is included in the fourth part. The fifth continues with several chapters covering other financial services, while the last discusses ethics and regulatory issues. These six parts represent the most significant and overarching themes of FinTech innovations. This handbook will appeal to students, established researchers seeking a single repository on the subject, as well as policy makers and market professionals seeking convenient access to a one-stop guide. K. Thomas Liaw is a Professor of Finance in the Economics and Finance Department at St. John’s University, New York, USA.
THE ROUTLEDGE HANDBOOK OF FINTECH Edited by K. Thomas Liaw
First published 2021 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN and by Routledge 52 Vanderbilt Avenue, New York, NY 10017 Routledge is an imprint of the Taylor & Francis Group, an informa business © 2021 selection and editorial matter, K. Thomas Liaw; individual chapters, the contributors The right of K. Thomas Liaw to be identified as the author of the editorial material, and of the authors for their individual chapters, has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data Names: Liaw, K. Thomas, editor. Title: The Routledge handbook of fintech / edited by K. Thomas Liaw. Description: Abingdon, Oxon ; New York, NY : Routledge, 2021. | Series: Routledge international handbooks | Includes bibliographical references and index. Identifiers: LCCN 2020048981 (print) | LCCN 2020048982 (ebook) | ISBN 9780367263591 (hardback) | ISBN 9780429292903 (ebook) Subjects: LCSH: Finance—Technological innovations. | Financial services industry—Technological innovations. Classification: LCC HG173 .R684 2021 (print) | LCC HG173 (ebook) | DDC 332—dc23 LC record available at https://lccn.loc.gov/2020048981 LC ebook record available at https://lccn.loc.gov/2020048982 ISBN: 978-0-367-26359-1 (hbk) ISBN: 978-0-367-76008-3 (pbk) ISBN: 978-0-429-29290-3 (ebk) Typeset in Bembo by codeMantra
CONTENTS
List of figures List of tables List of contributors
ix xi xiii
PART I
FinTech innovations and trends
1
1 FinTech innovations and financial markets: an introduction K. Thomas Liaw
3
2 FinTech venture capital Douglas J. Cumming and Armin Schwienbacher
11
3 The future of finance: why regulation matters Mark Fenwick and Erik P.M. Vermeulen
38
4 Digital currencies: what role in our financial system? Grégory Claeys and Maria Demertzis
51
PART II
Blockchain and cryptocurrencies
67
5 Decentralized autonomous risk transfer on the blockchain Alexander Braun, Niklas Haeusle and Stephan Karpischek 6 Distributed ledger technologies and blockchain for FinTech: principles and applications Raghava Rao Mukkamala and Ravi Vatrapu v
69
79
Contents
Contents
vii
.
(
Taylor & Francis Taylor & Francis Group http://taylorandfrancis.com
FIGURES
ix
Figures
1 6.6 16.7 17.1 17.2 17.3 18.1 18.2 20.1 20.2 21.1 21.2 21.3 22.1 22.2 23.1 23.2 23.3 23.4 23.5 25.1 25.2 25.3 25.4 25.5 25.6 25.7
Average number of bankcards by age and year of credit entry Average bankcard balance by age and year of credit entry Basic ROSCA structure Lending Circle participant score improvements MAF client-reported income strategies Minimal spanning tree representation of the borrowing companies network Network of Italian SMEs based on trade flows Minimum Spanning Tree drawn from the RMT filtered correlation matrix Dynamic net pairwise Spillover Indexes associated to the 8 analysed market exchanges Annual issuance in par amount Muni FinTech innovation New issue pricing analysis from DIVER Pricing and Scales platform Differences and overlaps between PropTech and FinTech Selected companies in the shared economy When to use Machine Learning Banking Big Data source and applications ML impact to market for prediction modelling Simple decision tree Perceptron (simplest ANN) A news database on cryptocurrency-related policies Bitcoin intraday price reaction to two news events News impact on intraday bitcoin price Legal status news and bitcoin returns AML/infrastructure and interoperability news and bitcoin returns Premia and trading volume Compliance process using embedded supervision
x
315 316 324 326 330 339 341 361 363 372 380 381 386 389 398 400 401 402 403 457 458 459 460 460 464 465
TABLES
xi
Tables
1 4.3 Types of restricted licensing phases 14.4 Gatekeeping measures: monetary thresholds 14.5 “Gatekeeping” measures: local ownership A14.1 Australia’s restricted ADI licensing regime A14.2 Hong Kong’s virtual bank licensing regime A14.3 Malaysia’s virtual bank licensing regime A14.4 Singapore’s digital bank licensing regime A14.5 APRA approved restricted and full ADIs A14.6 HKMA virtual bank approvals A14.7 Potential applicants interested in Malaysia’s licensing regime A14.8 Applicants for the MAS digital full or wholesale banking licenses 15.1 Blockchain technology in major banks 15.2 FinTech layout of domestic listed banks 15.3 FinTech innovation in banking channels 15.4 Definition of variables 15.5 Descriptive analysis of variables 15.6 Estimated results 15.7 Estimated results 15.8 Estimated results 15.9 Estimated results 15.10 Estimated results 15.11 Estimated results 16.1 The growth of general purpose credit and debit payments, 2003–2015 (transactions in billions) 18.1 Description of variables included in the dataset 18.2 Performance measures for the estimated credit scoring models: comparison between baseline and network-based specification 18.3 Performance measures for the estimated credit scoring models: comparison between baseline and network-based specification 20.1 ETFs by asset classes 20.2 ETF classes summary statistics 20.3 Portfolio summary results for filtered and unfiltered covariance matrix (absolute values) 25.1 The price impact of regulatory news: regression results 25.2 Response of prices and network volumes across cryptocurrencies
xii
240 241 243 249 250 250 251 253 254 255 256 271 274 276 288 289 293 294 295 295 296 297 309 338 340 342 360 361 362 462 463
CONTRIBUTORS
Arianna Agosto, PhD in Statistics, is Research fellow in Statistics at the Department of Economics and Management of the University of Pavia, Italy. She is Data Scientist at the Fintech Lab of University of Pavia, which carries out research and consulting projects for the European Union, fintech companies and financial institutions. Her main research topics concern statistical and econometric models for credit risk, financial risk, systemic risk and contagion. She also has professional experience in bank risk management. Tom Akana is an Advisor and Research Fellow, Consumer Finance Institute, Federal Reserve Bank of Philadelphia. Raphael Auer is Principal Economist in the Innovation and the Digital Economy unit at the Bank for International Settlements, Switzerland. Raphael Auer previously spent three years in the Monetary and Economic Department’s Monetary Policy unit. Prior to that, he worked for 10 years at the Swiss National Bank, including as Deputy Head and Economic Advisor of the International Trade and Capital Flows division. In 2009–10, he was Globalization and Governance Fellow at Princeton School of Public and International Affairs and visiting fellow at the Federal Reserve Bank of New York. He holds a PhD in economics from MIT and serves as president of the Central Bank Research Association. His current policy work focuses on issues related to cryptocurrencies, stablecoins and CBDC. On these topics, he contributes to BIS policy publications and management speeches, as well as to various international forums, including the G20-CPMI Cross-Border Payments Taskforce. He has published extensively in the field of international economics, monetary policy and digital currencies. Alexander Braun is Director at the Institute of Insurance Economics of the University of St. Gallen (HSG) in Switzerland, where he also holds the Associate Professorship in Insurance and Capital Markets. In addition, he is an Affiliated Scholar at the Wharton Risk Center (University of Pennsylvania). Mr. Braun obtained a PhD in Finance at the University of St. Gallen and held visiting scholar positions at the Wharton School (Penn) and Fox School of Business (Temple University) in the US. His focus areas comprise Natural Catastrophe Risk, Insurance-Linked Securities (ILS), Digital Insurance and Sustainable Insurance. Mr. Braun started his professional career in the Capital Markets Division of Lehman Brothers in xiii
Contributors
London. Overall, he looks back on more than 12 years of combined scholarly and practical experience in risk management, insurance and the financial markets. Paola Cerchiello is Associate Professor of Statistics at the Department of Economics and Management of the University of Pavia. Her research activity is mainly devoted to the study of statistical models for unstructured and complex data in economics and finance. From an applied viewpoint, she focuses on text data analysis, systemic risk, reputational risk, initial coin offerings, cyber risk, sentiment analysis and deep learning. Zhuming Chen – School of Business, Sun Yat-sen University, China. Stijn Claessens is Head of Financial Stability Policy and Deputy Head of the Monetary and Economic Department at the Bank for International Settlements (BIS), Switzerland. Stijn Claessens represents the BIS externally in senior groups, including the Financial Stability Board, the Basel Committee on Banking Supervision and the G20. Within the BIS, he leads policy-based analyses of financial sector issues and oversees the work of the Committee on the Global Financial System and other committee secretariats. Between 1987 and 2006, he worked at the World Bank in various positions. From 2007 to 2014, he was Assistant Director in the Research Department of the International Monetary Fund. From 2015 to early 2017, he was Senior Adviser in the Division of International Finance of the Federal Reserve Board. He holds a PhD in business economics from the Wharton School of the University of Pennsylvania and a master’s degree from Erasmus University, Rotterdam. He taught at the New York University business school and the University of Amsterdam. Grégory Claeys is Senior Fellow at Bruegel (Brussels, Belgium) and Associate Professor at the Conservatoire National d’Arts et Métiers, Paris, France. Ryan Clements, BA (Honors, First Class), LLB (Distinction), LLM (Magna Cum Laude), SJD (Duke) is an assistant professor, and Chair in Business Law and Regulation at the University of Calgary Faculty of Law, Canada. Douglas J. Cumming, J.D., PhD, CFA, is the DeSantis Distinguished Professor of Finance and Entrepreneurship at the College of Business, Florida Atlantic University, and Visiting Professor of Finance at Birmingham Business School. Douglas has published over 195 articles in leading journals, such as the Academy of Management Journal, Journal of Financial Economics, Review of Financial Studies, and Journal of International Business Studies, and has been cited over 18,500 times. He is currently the Managing Editor-in-Chief of the Review of Corporate Finance (2021–) and the British Journal of Management (2020–). He is the former Managing Editor-in-Chief of the Journal of Corporate Finance (2018–2020). Krishnan Dandapani, Professor of Finance, Florida International University, USA. Maria Demertzis is Deputy Director at Bruegel (Brussels, Belgium). Mark Fenwick is Professor of International Business Law at the Faculty of Law, Kyushu University, Fukuoka, Japan. His research interests are in the fields of white-collar and corporate crime, and business regulation in a digital age. Recent publications include Regulating Fintech in Asia (Springer, 2020) and Smart Contracts: Technological, Business and Legal Perspectives xiv
Contributors
(Hart, 2021). He has a master’s and a PhD degree from the University of Cambridge, and has been a Visiting Professor in China, Hong Kong, Singapore, Vietnam and the Netherlands. Marc Feth holds a B.Sc. in Economics from the University of Groningen, the Netherlands and a M.Sc. in Finance from the Warwick Business School, United Kingdom. After working in Corporate Finance and Private Equity in Amsterdam and London, he worked as a consultant in the German real estate industry and as a board member for Europe’s biggest property manager (by number of units managed). He invests directly in PropTech startups as an Angel Investor, and has various points of contact with start-ups, small and medium size enterprises, international corporations, investors and service providers in the real estate industry. Niklas Haeusle is a PhD student at University St. Gallen and project manager at the Institute of Insurance Economics. His research interests lie at the intersection of digitalization, insurance economics and organization economics. He received a MSc in Business Economics from the University of Ulm, Germany. Robby Houben – University of Antwerp, Belgium, Faculty of Law, Research Group Business & Law; lawyer at the Antwerp Bar. Nicole G. Iannarone is an assistant professor at Drexel University’s Thomas R. Kline School of Law, USA. She holds a JD from Yale Law School, USA, and a B.S. in History and Political Science, summa cum laude, from Brenau University Women’s College, USA. Her scholarship focuses on the regulation of financial intermediaries, consumer investor protection, the consumer investor’s experience in resolving securities disputes, and professional ethics. Mohammad Hashemi Joo – North Carolina Agricultural and Technical State University, USA Tatja Karkkainen – Adam Smith Business School, University of Glasgow, UK. Stephan Karpischek – Decentralized Insurance Foundation, Zug, Switzerland. Anil Savio Kavuri is a research associate at Loughborough University, and research fellow at Australian National University. He has a PhD in economics from Australia National University, an MPhil in Finance from University of Cambridge, an M.I.A from Columbia University and a First Class BSc in Economics from University College London. He graduated top of the class in the most technical subjects, including game theory, experimental economics, money & banking at the University College of London and economics of energy, marine energy transportation, petroleum markets & trading at Columbia University. His professional background includes being a project financier at Scotia Capital on Wall Street, New York and a consultant at the World Bank. His research focuses on FinTech and InsurTech with a focus on business models and disruption, data innovation, AI, big data, technology partnerships and DLT. Furthermore, he is currently developing a 2-sided electronic marketplace that will enable AI & data sharing between start-ups and insurance firms to spur innovation in the insurance industry. Jing Li – Foshan University, China. Xin Li – School of Business, Sun Yat-sen University, China. xv
Contributors
K. Thomas Liaw is a Professor of Finance in the Economics and Finance Department at St. John’s University. He was chairman of the department. His areas of teaching and research are capital markets, investments, fintech, and green finance. Professor Liaw was president of the Chinese American Academic and Professional Society. His professional experiences include corporate assignments and directorships. Professor Liaw holds his PhD from Northwestern University. Joseph E. McPhail is Founder and CEO of WEquil, an umbrella company that invests in startups and builds software to address financial and education technology problems. He often teams with his daughters Sumay and Aila who recently founded WEquil School, a mobile app designed to help tailor education to each individual’s unique strengths and interests. Joseph won a Young Alumni Award from Iowa State University where he graduated and continues to help graduates match their passions with career opportunities. He currently lives in Washington DC, USA with his wife Lihong, the co-author of his chapter. Lihong L. McPhail is an executive at the Commodity Futures Trading Commission, Washington DC, USA. She is passionate about turning data into value to solve real world problems. She has published numerous articles on financial markets, risk management and emerging technologies. She was a Chartered Financial Analyst (CFA) and Financial Risk Manager (FRM) charter holder, and received a PhD in economics from Iowa State University. Alistair Milne is Professor of Financial Economics at the School of Business and Economics, Loughborough University, UK. Previously he was senior lecturer and then reader in banking and finance at Cass Business School, City University of London; he has also worked at the Bank of England, the University of Surrey, London Business School, HM Treasury and for the Government of Malawi. He has been a regular visitor to the Monetary Policy and Research Department of the Bank of Finland. He researches financial infrastructures and financial technology; bank regulation and capital management; and the role of banks and credit in the macroeconomy. He has over 100 journal publications on these and other topics and is the author of a comprehensive account of the global credit crisis “The Fall of the House of Credit”. He holds a PhD in economics from the London School of Economics. His current research is focused on regulation, RegTech, economics of central bank issue of digital currencies, open banking, ‘peer to peer’ lending platforms and the application of AI, distributed ledgers and related data technologies in financial service. Raghava Rao Mukkamala is the director of the Centre for Business Data Analytics and an associate professor at the Department of Digitalization, Copenhagen Business School (CBS), Denmark. Raghava holds a PhD in Theoretical Computer Science, and his current research focuses on Big Data Analytics, data science, blockchain technologies and cyber security. His current research program seeks to develop new algorithms for big data analytics by combining formal/mathematical modelling approaches with machine learning techniques. Raghava has many years of programming and IT development experience from the Danish IT industry before moving to research. Jackson Mueller is Director of Policy and Government Relations at Securrency. He advises the Chief Strategy Officer and the Chief Legal Officer on public policy and regulatory positions. Jackson has previously served as an associate director at the Milken Institute Center for Financial Markets and head of the Institute’s Financial Technology program, where he xvi
Contributors
focused on the potential of technology in promoting capital formation, financial inclusion, and cost-efficient compliance. He has also served as an assistant vice president at the Securities Industry and Financial Markets Association (SIFMA), focusing on capital markets policy. He received his bachelor’s degree in political science from the University of Richmond and a master’s degree in public policy from American University. Jackson lives and works in Washington, DC. Yuka Nishikawa – College of Charleston, USA. Paolo Pagnottoni – FinTech Laboratory, Department of Economics and Management, University of Pavia, Italy. Gloria Polinesi – Department of Economic and Social Sciences, Università Politecnica delle Marche, Italy. Anna Omarini – Bocconi University and SDA Bocconi School of Management, Italy. Ana García Rodríguez is the general counsel of Grupo Kutxabank, the most solvent bank in Spain. She has been for many years an experienced Partner with a demonstrated history of working as legal advisor and consultant in the banking and financial service industry. Skilled in Fintech, Insurtech, Financial and Banking Regulation, Public and sophisticated M&A, Corporate Governance, CIS and Private Equity and Hedge Funds, Structure Finance and Derivatives. She is a professor at Instituto de Empresa and a recognized speaker in banking and finance conferences. Armin Schwienbacher is Professor of Entrepreneurial Finance at SKEMA Business School since 2010. He obtained his PhD in 2003 at the University of Namur, Belgium, on exit strategies of venture capital funds. Armin is specialized in the economic and regulatory areas of crowdfunding and venture capital finance. His work has been published in a large number of international academic journals. He is currently Director of the Research Center FAIRR in finance and accounting at SKEMA and Editor at the FT-ranked Journal Entrepreneurship Theory and Practice. Alexander Snyers, University of Antwerp, Belgium, Faculty of Law, Research Group Business & Law. Timothy J. Stevens, CFA is the President, Chief Operating Officer and co-founder of Lumesis, Inc.,Stamford, CT, USA, a financial technology firm that serves the Municipal Fixed Income Market. Prior to co-founding Lumesis, Tim spent over 14 years with Ambac Financial Group, Inc. where, most recently, he was a Senior Managing Director and led all Capital Markets activity including management of the company’s investment portfolios, the Guaranteed Investment Contract business, the Global Interest Rate Derivatives business and the Global Secondary Markets Group. Prior to Ambac, Tim worked in the Audit Services group of Deloitte. Tim holds a B.S. in Business and Economics from Lehigh University. He currently holds the Chartered Financial Analyst designation and was a licensed Certified Public Accountant during the years that he practiced accounting. He was also registered with FINRA as a General Securities Representative (Series 7) and a General Securities Principal (Series 24). xvii
Contributors
Anca Mirela Toma graduated from the University of Pavia in Economics, where she felt in love with behavioural economics topics. She just finished the three-year PhD in Applied Economics and Management with a research focus on fintech development and blockchain applications. Her research activity is mainly devoted to statistical models for unstructured data. From an applied viewpoint, she focuses on text data analysis, credit risk, fraud detection, initial coin offerings, sentiment analysis and behavioural economics. Ravi Vatrapu – Copenhagen Business School, Denmark; Department of Technology, Kristiania University College, Norway; Ted Rogers School of Management, Ryerson University, Canada. Erik P. M. Vermeulen is a Professor of Business and Financial Law at Tilburg University in the Netherlands, Senior Legal Counsel at Signify (formerly known as Philips Lighting) and an Innovation Advisor at a law firm. Erik can best be described as an innovator. His thought-provoking and innovative views have attracted international attention. He regularly serves as an expert advisor to international organizations. He is a board/advisory member of several companies/organizations and has appeared at numerous conferences as a featured or keynote speaker. Erik has a blog at erikpmvermeulen.medium.com. Shihan Wang - School of Business, Sun Yat-sen University, China. Jinghong Zeng – School of Business, Sun Yat-sen University, China. Weihan Zhang – School of Business, Sun Yat-sen University, China. Xiangyu Zhang – School of Business, Sun Yat-Sen University, China.
xviii
PART I
FinTech innovations and trends
.
(
Taylor & Francis Taylor & Francis Group http://taylorandfrancis.com
1 FINTECH INNOVATIONS AND FINANCIAL MARKETS An introduction K. Thomas Liaw FinTech is the use of financial technology to automate processes in financial services. Financial firms have been using technology to transform their businesses. FinTech will continue to be a significant force in the global financial marketplace that presents financial institutions and professionals with opportunities and challenges. As such, there has been a dramatic increase in research and publication on FinTech. The offerings of FinTech courses and FinTech programs at business schools are rising as well. The Routledge Handbook of FinTech is a comprehensive guide for students, researchers, and market professionals on the opportunities and challenges of FinTech. The Routledge Handbook of FinTech is organised into six thematic parts and each is composed of topical and high-quality chapters. The first part covers an introduction, FinTech venture capital funding, the future of finance with respect to regulation, and the role of digital currencies in the financial system. The second part focuses on the blockchain technology applications and various aspects of cryptocurrencies. The section includes two chapters on blockchain and distributed ledger technology. In addition, two chapters provide statistical analyses on initial coin offerings. The remaining three chapters analyse price discovery in Bitcoin futures and cash markets, trading and regulation of cryptoassets, and cryptoassets in financial crime. The next part covers FinTech in banking. The chapters provide research findings in the open industry of banking, FinTech opportunities and challenges in banking, virtual banks in the Asia-Pacific region, and FinTech in China’s banking. A significant element in FinTech, payment and lending, is included in the fourth part. Coverage includes the impact of technology on consumer payment preferences, technology in lending circles, and lending risk management. The fifth part continues with several chapters on FinTech in other financial services. Empirical results and theoretical discussions are on robot advisory, municipal markets, and the real estate market. The last part discusses regulatory issues. This part begins with a chapter on the implications of machine learning on banking regulation. The following chapter discusses FinTech regulations in the USA and Canada. Part 6 concludes with a research on cryptocurrency market reactions to regulatory news.
3
K. Thomas Liaw
1. FinTech innovation and trends Chapters in the first part discuss FinTech from the broad finance perspective. The first chapter, “FinTech innovations and financial markets: An introduction,” outlines the objectives and coverage of this handbook. Chapter 2, “FinTech venture capital,” examines FinTech venture capital investments and the role of institutional factors on the international allocation of FinTech venture capital. The chapter documents a notable change in the pattern of FinTech venture capital investments around the world relative to other types of investments after the global financial crisis. The chapter shows that FinTech venture capital investments are relatively more common in countries with weaker regulatory enforcement and without a major financial center after the financial crisis. The chapter also shows the FinTech boom is more pronounced for smaller private limited partnership venture capitalists that likely have less experience with prior venture capital booms and busts. These FinTech venture capital deals are substantially more likely to be liquidated, especially when located in countries without a major financial center. Chapter 3, “The future of finance: Why regulation matters,” explores the various challenges of regulating technology-driven change in financial services. As technology “eats the world,” it creates a plethora of new opportunities, but it also creates tremendous challenges that require some form of government intervention. In a world where agility is essential, governments are sluggish and disconnected. This creates potential risks, most obviously for the consumers of financial services. In developing answers to these regulatory dilemmas, the authors argue that traditional approaches to regulating financial services are obsolete and that new thinking is required. State actors are at an enormous informational disadvantage and lack the capacities and resources to keep up with the fast-moving actors that dominate the sector. New and more innovative approaches need to be found. The chapter describes several such approaches, including regulatory “co-creation,” policy experimentation, and technology-driven regulation (so-called RegTech). Chapter 4, “Digital currencies: What role in our financial system?,” reviews the emergence of different forms of digital currencies and whether they fulfil a good role as currencies. To this end, the chapter also examines how and whether they might challenge traditional currencies as well as the role guardians of money, price and financial stability, namely central banks, play. Furthermore, the chapter discusses the need for central banks to also adapt by becoming more digital and to what effect.
2. Blockchain and cryptocurrencies Chapter 5, “Decentralized autonomous risk transfer on the blockchain,” explains the functionality and operating principles of this new method of risk transfer. The chapter provides an overview of the corresponding institutional arrangement and its challenges. The chapter identifies two main challenges, namely ensuring the quality of the product and gaining a critical network size. A feasible solution for these problems is the usage of system-specific tokens, which are used as a form of digital collateralization. In a short case study, the authors scrutinize how these theoretical concepts can be translated into practice and show the potential of the idea. While flight delay risk is already handled through decentralized autonomous risk transfer, the biggest benefits can be realized by decentralizing crop and hurricane insurance, including a seamless transfer of the risk to capital markets. Chapter 6, “Distributed ledger technologies for FinTech: principles and applications,” presents the underlying technical principles of distributed ledger technology (DLT) and 4
Fintech innovations and financial markets
blockchain technology and outlines their practical applications in FinTech. In recent years, DLT and blockchain technologies in general and cryptocurrencies in particular have attracted substantial attention from both researchers and practitioners due to their unique foundational aspects such as the lack of centralized control and high level of anonymity. Because of the disruptive nature, DLT and blockchain have led to the evolution of decentralized applications in multiple domains such as finance, health care, supply chains, etc. This chapter first outlines basic principles and foundations underpinning the DLT and blockchain technologies. Then, the chapter discusses several applications in the FinTech domain such as cryptocurrencies, smart contracts, risk management, corporate finance, governance, crowdfunding, and derivative markets. In addition, the chapter summarizes the different kinds of initiatives in terms of regulation, compliance and governance for cryptocurrencies and other DLT applications in the financial domain. Finally, the chapter concludes with some reflections on the impact of DLT and blockchain technologies on the FinTech domain in the near future. The next several chapters are related to cryptocurrencies. Chapter 7, “Initial coin offerings: a statistical analysis of the main characteristics,” describes a statistical approach to detect which characteristics of an initial coin offering (ICO) are significantly related to fraudulent behaviors. The authors leverage several different variables like entrepreneurial skills, number of people chatting on Telegram on the given ICO, and relative sentiment, type of business, and token pre-sale price. Through logistic regression, multinomial regression and sentiment analysis, the results shed a light on the riskiest ICOs. Chapter 8, “Initial coin offerings: a new trend in the market,” studies a fundraising method, the initial coin offerings (ICOs). Despite its detractors, critics and associated risks, the success is uncontestable. Tokens and cryptocurrencies offerings are made using the blockchain technology and based on the use of Bitcoin or Ethers as a payment method. Currently these ICOs are not expressly subject to any regulation in any country; however, it seems that this situation will soon change. American, Asian and European regulators have already portrayed that this phenomenon will be regulated in the future in line with the current regulations applicable to securities markets taking into account the specificities of these offerings. Chapter 9, “Price discovery in the Bitcoin futures and cash markets,” analyses the Bitcoin futures mid-quote data from the Chicago Board Options Exchange (CBOE) and Bitcoin market index to examine price discovery in Bitcoin markets. Furthermore, the chapter seeks to assess the Bitcoin market microstructure. The results drawn on the intraday prices show that the futures are leading the price discovery at different frequencies even with comparably low futures trading volumes. This supports the extant literature of futures-spot market price discovery and the role of informed traders in the futures market. Chapter 10, “Trading and regulation of cryptocurrencies, stablecoins and other cryptoassets,” examines the regulation of cryptocurrencies, stablecoins and other cryptoassets. The chapter starts with an overview of these new assets. After this, the authors discuss the legal and regulatory classification and the application of regulatory requirements for Know Your Customer (KYC) and Anti-Money Laundering (AML). The authors then assess the appropriate regulatory regime, arguing that there is no need for bespoke regulation, over time most cryptoassets and the exchanges on which they trade will move within existing regulatory frameworks. Those remaining cryptoassets that fall outside the regulatory framework will be a small niche which can be safely left outside the regulatory perimeter. Chapter 11, “Cryptoassets and financial crime: a European Union perspective,” looks at the market from a different perspective. This chapter scrutinizes the use of cryptoassets in the context of financial crime and regulatory responses in the European Union (EU). The 5
K. Thomas Liaw
focus is on the EU legal context. The chapter starts by assessing what exactly cryptoassets are. Secondly, it identifies the key actors in cryptoasset schemes. Thirdly, the challenges cryptoassets bring from the perspective of combating financial crime are scrutinized. Subsequently the EU regulatory framework relating to money laundering and terrorist financing via cryptoassets is critically assessed and suggestions for future law-making are made. The chapter concludes with a brief note on cybersecurity and some thoughts on the right level of law-making as regards cryptoassets.
3. FinTech in banking Chapters 12–15 provide interesting research findings of FinTech applications in banking. Chapter 12, “FinTechs: unbundling to rebundling in the open industry of banking,” argues that digital transformation is the driving force of the new business revolution, where the increased usage of digital devices is transforming the way customers do their banking, changes market expectations, and in so doing value creation and value delivery in the banking and financial industry are both under a deep transformation. The first part of the chapter explores how far the evolution of FinTech in the banking arena is going and how regulation and technology innovation are changing the industry towards an open banking paradigm. In the second part, business platforms and ecosystems are analysed. It is pointed out that strategists no longer take their value chains as a given but strive to cope with a market where evolving and being resilient are part of mandatory strategic planning. Chapter 13, “Is FinTech a threat or a promise to banks?,” discusses whether or not FinTech is a threat to well-established banking systems, or are these disruptions strengthening and creating new opportunities and prospects for the banking industry. The research reviews the developments in FinTech with respect to banking. The chapter also investigates the benefits and challenges that FinTech has brought to the banking industry. Chapter 14, “Virtual rush: the race for virtual banks in the Asia-Pacific region,” shows that there has been a significant focus and willingness on the part of both policymakers and regulators to review longstanding parameters and requirements that govern banking. The proliferation of the internet, mobile phone penetration, and the capture and responsible dissemination of vast quantities of data have uprooted traditional financial norms, opened up new, alternative options to transact, and created a potentially more responsive financial services ecosystem capable of delivering a seamless, unique experience for the end user. The way we “bank” continues to change and evolve, so much so that regulators and policymakers around the world are adapting legacy licensing regimes in support of virtual banking. This chapter examines several emerging virtual bank licensing regimes in the Asia-Pacific region, with a focus on licensing developments in Australia, Hong Kong, Malaysia, and Singapore. Particular attention is paid to why these regimes are being implemented now, the similarities and differences between the regimes, and what to make of this ongoing race for virtual bank supremacy in the region. Chapter 15, “The restructure of China’s banking industry by artificial intelligence and FinTech,” investigates FinTech and banking in China. With the rapid development of digital technologies such as big data, cloud computing, artificial intelligence, and the Internet of Things etc., the banking industry is facing the pressure of strategic transformation. The rise of FinTech plays a role in restructuring the financial industry, and also provides an effective solution for the transformation of the banking industry. Through qualitative analysis, this paper evaluates the use of FinTech in the Chinese banking industry to explore the degree of transformation and upgrading of the industry. It also discusses how the banking industry 6
Fintech innovations and financial markets
should adapt to the trend of the new generation of FinTech. The chapter also provides a case study on two banks which have outstanding performance in this FinTech reform: China Agricultural Bank and WeBank of China. In the empirical analysis, the data of Chinese listed banks in the past ten years show that the application of FinTech can improve the performance of commercial banks.
4. FinTech in payment and lending Chapters in this part examine applications of FinTech in payment, lending, and risk management. Chapter 16, “Consumer payment preferences and the impact of technology and regulation: insights from the Visa Payment Panel Study,” provides discussions on consumer payment preferences from a workshop by The Consumer Finance Institute in August 2018. Michael Marx, senior director at Visa, Inc., discusses recent data from the Visa Payment Panel, highlighting the evolution of consumer payment preferences since the Great Recession and the passage of the Credit Card Accountability Responsibility and Disclosure (CARD) Act of 2009. A number of intriguing trends were discussed. Debit card adoption and growth have shown signs of slowing, even as regulatory changes have increased its prevalence recently among younger consumers. Credit card usage continues to grow and has shifted largely to rewards-based products. Payment preferences for younger consumers appear to be influenced by the availability of financial products (driven by social and regulatory influences) as well as the advent of mobile wallets and person-to-person (P2P) technologies. Chapter 17, “Meet people where they are: Building formal credit using informal financial traditions,” looks into the unique private lending circles and how technology contributes to such lending. The Consumer Finance Institute hosted a workshop in February 2019 featuring José Quiñonez, chief executive officer, and Elena Fairley, programs director, of Mission Asset Fund (MAF) to discuss MAF’s approach to building technology and products to help its clients improve access to mainstream financial markets. MAF’s signature program, Lending Circles, adapts a traditional community-based financial tool known as a rotating savings and credit association (ROSCA) to help establish or expand credit reports for participants who may not be able to do so through traditional means. Lending Circles have served more than 10,000 clients since 2007 and have expanded well beyond MAF’s core constituency in the Mission District of San Francisco. Quiñonez and Fairley discussed MAF’s approach to working with the communities it serves and shared the key successes and challenges that MAF has encountered. This chapter provides an overview of the information shared in the workshop and additional interviews with Ramya Gopal, director of MAF Lab, regarding MAF’s technology and development philosophy. Chapter 18, “Peer-to-peer lending risk management,” shows that the growth of FinTech is changing the way in which credit is granted and is redefining the role of financial intermediation. In the context of credit services, FinTech peer-to-peer (P2P) lenders have introduced many opportunities, ensuring higher speed, better customer experience and reduced costs with respect to traditional lending. However, P2P lending platforms lead to higher credit risk, fully borne by the lenders, and systemic risk, due to the high and direct interconnectedness among borrowers and lenders generated by the platform. This calls for new and more accurate credit risk models to protect consumers and preserve financial stability. A factor that penalizes the accuracy of P2P credit scoring models is that the platforms often do not have access to borrowers’ data usually employed by banks, such as account transaction data, financial data and credit bureau data. For these reasons, the accuracy of credit risk estimates provided by P2P lenders may be poor. However, P2P platforms involve their users and, in 7
K. Thomas Liaw
particular, the borrowers, in a continuous networking activity. Data from such activity can be effectively used not only for commercial purposes, as it is customarily done, but also to improve credit risk accuracy.
5. FinTech in other financial services Chapter 19, “Rethinking automated investment adviser disclosure,” explores the parameters of the pre-existing regulatory regime as applied to a new landscape of investing with a focus on the key regulatory tool of disclosure, examining the challenges of delivering appropriate disclosures when advice is provided in a new format. It further suggests a reconceptualization of the aim of disclosure to capitalize on robo-advisers’ ability to truly know their clients and shift the burden of investor education from the customer to a fiduciary level adviser with the ability to tailor advice and education to each investor at an individualized level of understanding. The chapter concludes with a recommendation: robo-advisers and investor collaboration via an active and iterative disclosure regime that encourages investor comprehension, understanding, and learning. Chapter 20, “Market risk for robot advisory,” discusses whether robot advisory platforms that involve the provision of automated consultant and investment services with virtually no human contact may underestimate risks, causing a mismatch between investors’ expected and actual risk. Cryptocurrencies are a new asset class considered by robo-advisors in the near future. In this nascent market it is fundamental to understand the price dynamics in order to investigate in which exchange platforms the price formation process takes place and how the latter are interconnected. The chapter proposes an asset allocation strategy that takes individual user’s preferences into account improving robot advisory portfolio allocation. In particular, random matrix theory filter and network metrics are combined in the minimum variance portfolio model in order to construct portfolios overperforming in terms of risk and realized risk with respect to the Markowitz model. Also, price discovery and interconnectedness of cryptocurrency market exchanges are studied in order to help investors in choosing the most suitable trading platforms to place profitable trades depending on their own strategy. Chapter 21, “Financial technology in the US municipal fixed income market,” is unique. The municipal market has been transformed by the application of financial technology over the past 40 years. Important improvements have occurred in several areas including debt structuring, book building and order allocation, competitive bidding, electronic trading platforms, portfolio management, and transactional market data. The availability of data and the emergence of new technologies, coupled with increased regulatory scrutiny, entrepreneurial ambition, and competitive pressure, have helped fuel this innovation. Financial technology has greatly enhanced information transparency, encouraged a level playing field in the market, improved the ability of issuers to access the capital markets and lowered their cost of funding, and has assisted investors by increasing market liquidity and improving the ability to manage the risk/return profile of their portfolios. Current municipal market financial technology trends include enhancements in issuer and investor communication, pricing analytics, suggestive searches to optimize portfolio building, algorithmic trading and the application of artificial intelligence and machine learning to these and other endeavors. Chapter 22, “Proptech: the real estate industry in transition,” examines FinTech in property markets. In the past, real estate investments were considered safe and long-term. Technological innovations influenced the sector only slowly and decision makers were unwilling to change their customary business practice. However, in recent years technological 8
Fintech innovations and financial markets
developments set off a wave of innovations. The chapter provides an overview of the current transitions in the industry. Hereby, the three Proptech subcategories Real estate FinTech, Smart real estate, and Shared economy real estate are explained in more detail and underlined by real life examples.
6. FinTech regulatory issues The final three chapters examine FinTech and regulatory issues. Chapter 23, “Machine Learning implications for banking regulations,” discusses that Machine Learning (ML) automates prediction, making it cheaper and more accurate. The amount and variety of financial data will continue to increase, and with it the value of ML. A key implication for regulators is that the banking industry is likely to rely increasingly on ML methods for decisions that, by design, cannot be fully understood by their developers. As a result, regulators at all levels will increasingly confront ML models they can’t fully comprehend. Examination is impacted through the need for supervisors to opine on model risk. ML models contain complex features. Examiners may need to understand the implications of ML for transparency and associated operational risks. Use of historical data to train models may also have fair lending implications. Some banks and FinTech firms are already using ML for a broad range of banking services such as fraud detection, risk management and pricing. Policy may be impacted through at least two channels: operational risk and market behavior. ML has a direct impact on model risk, a component of operational risk. Banks are subject to model risk management regulatory guidance which has not been updated since April 2011. Some aspects of this guidance may be challenging to apply to ML tools due to their “black-box” nature. ML could also be changing the very nature of market behavior for some liquid assets. The chapter provides an overview of ML and explores these and other implications for banking regulation. Chapter 24, “Regulating FinTech in Canada and the United States: Comparison, challenges and opportunities,” shows that FinTech has many potential benefits and it could transform banking, lending, payments, investing and other financial services through the internet, smartphones, artificial intelligence, blockchain and cryptocurrencies, and many other current and future digital technologies. Such benefits include lower costs, an enhanced scope of products and services, and the possibility of reaching and offering previously underserved customers greater credit and financial services. Policymakers in Canada and the US should encourage these positive developments, foster innovation and competition, and reduce barriers to entry, while ensuring adequate safeguards are established for the stability of the financial system and necessary consumer protections are in place. The market environment and regulatory approaches in Canada and USA are similar but not uniform. Each jurisdiction faces different challenges and opportunities. The speed and complexity that this new wave of FinTech has expanded throughout North America and the world has created regulatory challenges for authorities in the US and Canada. FinTech has many potential risks. If this revolution is not managed well, the results could be serious, including the risk of destabilizing the financial system. In both jurisdictions, policymakers must be mindful of FinTech’s unique risk propositions and its benefits; both when it is adopted internally by existing financial institutions under regulatory oversight, and when FinTech originates from new, consumer-facing market entrants. They must also ensure that regulatory efforts are coordinated with international best-practices and be mindful of any potential unintended effects of regulatory action given an increasingly complex and interconnected financial market. Chapter 25, “Cryptocurrency market reactions to regulatory news,” shows that cryptocurrency valuations, transaction volumes and user bases react substantially to news about 9
K. Thomas Liaw
regulatory actions. The impact depends on the specific regulatory category to which the news relates: events related to general bans on cryptocurrencies or to their treatment under securities law have the greatest adverse effect, followed by news on combating money laundering and the financing of terrorism, and on restricting the interoperability of cryptocurrencies with regulated markets. News pointing to the establishment of specific legal frameworks tailored to cryptocurrencies and initial coin offerings coincides with strong market gains. These results suggest that cryptocurrency markets rely on regulated financial institutions to operate and that these markets are segmented across jurisdictions.
10
2 FINTECH VENTURE CAPITAL Douglas J. Cumming and Armin Schwienbacher1
“…without the financial crisis and the popular anger it spawned against the whole banking system, there would be no fintech” — “Fintech’s Wakeup Call”, Bloomberg, February 22, 20162 “After the 2008 crisis, banks faced additional capital adequacy requirements and also came under fire for non-compliance on existing rules. So while fintech startups are still subject to many of the same rules as their traditional counterparts, they don’t have the added burdens that come with litigation, fines and other penalties that several large institutions have had to deal with in recent years.” — “An Inside Look At Fintech Marketplace Lenders”, Forbes, February 27, 20163
Introduction Recent research has shown that corporate governance issues are fundamentally different in young entrepreneurial firms than in large, well established ones, and that these issues may vary across countries (Armitage et al., 2017; Bjørnskov and Foss, 2013; Djankov et al., 2002; Luo and Junkunc, 2008; Wright et al., 2007). While the latter has been studied extensively, the former has recently attracted much interest in connection with entrepreneurial firms active on a global scale (Zahra, 2014). Firms may obtain a comparative advantage when located in countries with more favorable regulation, since it can reduce their costs and enable developing innovations that are more difficult to implement in countries with more stringent regulation (Bozkaya and Kerr, 2014; Braun et al., 2013; Dharmapala and Khanna, 2016; Hornuf and Schwienbacher, 2017; Levine et al., 2015; Wang and Wang, 2012;). Empirical evidence suggests that differential enforcement of law may drive the structure of corporate governance and patterns of start-up activity. In this chapter, we examine a specific industry – ‘FinTech’, or financial technology – to see whether or not differential enforcement of banking rules around the world affects the financing patterns of FinTech start-ups, including the shareholder structure and type of investors participating in the financing. We further examine the impact on these firms to go public. Recent years have seen an increasing amount of hype about FinTech, and venture capital (VC) FinTech in particular, around the world. The hype in some camps is sufficient to remind some practitioners of the dot com bubble from 1998–2000. For example, at a March 11
D.J. Cumming, A. Schwienbacher
2016 seminar at the law offices of McCarthy Tétrault in Toronto, one senior practitioner was overheard saying that some of the junior roundtable participants should “settle down!” because “they are not old enough to remember the dot com bubble”. The new FinTech wave is also driven by ventures and no longer just by investments by incumbents in internal projects. According to Accenture (2015), the worldwide investment volume in FinTech ventures amounted to USD 12.21 billion in 2014 and, while precise estimates with more recent data are difficult, our understanding is that it has grown substantially since that time. This new wave has been attributed by some researchers and practitioners to the financial crisis for at least two reasons (Arner et al., 2015; Kelly, 2014). First, many skilled employees of banks and other financial institutions left their job (or even were fired) and sought new opportunities by undertaking entrepreneurial initiatives, leading to an increased supply of investment opportunities by venture capital funds (and thus increased demand for venture capital by FinTech projects). Second, since incumbents have been subject to stronger regulation and scrutiny by regulators since the start of the financial crisis, FinTech ventures that develop products and services that are outside the scope of financial regulators (such as crowdfunding platforms and alternative payment systems) have become more attractive relative to incumbents. A striking example are crowdfunding platforms that are often structured so that many of the services that are subject to strong regulatory oversight are either outsourced to other service suppliers (e.g., Paypal for payments) or not offered at all (e.g., they may not provide “investment advice” as a means of avoiding compliance with Markets in Financial Instruments Directive (MiFiD) regulation). Thus, many platforms are able to operate under light regulation and in some countries without any significant license (Hornuf and Schwienbacher, 2017). Another illustrative example relates to alternative payments systems. This may have spurred more investments into FinTech ventures, hoping it will reduce the costs of financial intermediation for the economy as a whole (Philippon, 2016). We investigate whether the financial crisis has changed investment behavior made by venture capital funds in FinTech ventures, and how. More specifically, in this chapter we address the issue of whether or not there has been a change in the pattern of FinTech VC investments since the financial crisis. In the spirit of the Bloomberg quote above, we expect that there is a spike in FinTech VC. In the view of different enforcement levels of banking rules around the world for large established organizations versus start-ups, as documented by Forbes and quoted above, we expect the rise in FinTech is more pronounced in countries which do not have a major financial center where FinTech start-ups are more able to flourish with less risk of regulatory oversight. Similarly, we expect smaller VC funds with less experienced management to have a greater spike in FinTech investments as such fund managers are less likely to see the hype in the context of the recent history of booms and busts such as experienced in the dot com bubble or the recent financial crisis. In view of the apparent boom, and tendency to overinvest in the latest fads, we expect that FinTech investments on average are less likely to result in successful initial public offering (IPO) and acquisition exits and are more likely to be written off. To test these propositions, we extract all VC investments from VentureXpert from 1/1/1990 to 12/31/2015 and record as FinTech ventures all VC investments active in the financial services industry, leading to a final sample of 2,678 investment rounds in 747 distinct FinTech ventures. Similarly we obtain a sample of 277,994 VC investments in non-FinTech ventures during the same time period that we use as control group. The data examined are consistent with our expectations. Controlling for other things being equal, we estimate that round investment amounts for FinTech went up in the full sample. We observe investment rounds at higher levels for the subsample of deals done by 12
Fintech venture capital
independent VCs, but no material change for corporate or financial institution affiliated VCs. FinTech VC round investment amounts went up by a small amount among large VC funds, and by a much larger amount among small VC funds. FinTech VC round investment amounts significantly increased in countries without a major financial center, but were unchanged in countries with a major financial center. Also, controlling for other things being equal, we observe round syndicate size went up for FinTech VC for full sample, and the subset of early stage rounds, and the subsample excluding early stage rounds. Further, FinTech VC round syndicate size increased more for small VC funds than large VC funds. Round syndicate size for FinTech increased substantially in countries without a major financial center, but were unchanged in countries with a major financial center. Finally, the data indicate that controlling for other things being equal, FinTech VC deals are less likely to be write-offs if they originated after the financial crisis, except for investments made in ventures located in countries without a major financial center. In these cases, we observe a substantial increase in the probability of having write-offs for FinTechs after the financial crisis. This chapter is related to a growing literature on booms and busts in VC, and regional analyses of VC. First, with respect to booms and busts in general, booms and busts in VC investment have been documented by Gompers and Lerner (1999), Cumming et al. (2005) and Buzzacchi, Scellato and Ughetto (2015), among others. Second, region specific patterns of VC activity have been documented by Bertoni et al. (2015) and Bozkaya and Kerr (2014) for Europe. International analyses of VC investment patterns linked to differences in institutional settings have been examined, among others, by Bertoni and Groh (2014), Bonini, Alkan and Salvi (2012), Dai and Nahata (2016), Guler and Guillén (2010), Johan, Schweizer and Zhan (2014), Li and Zahra (2012), Schertler and Tykvova (2012), Schwienbacher (2008), Tykvova and Schertler (2014), and Wang and Wang (2012). Prior work has not examined investment VC investment cycles of FinTech. This extension is not trivial, as we are able to examine whether enforcement of financial regulation can spur investment activity into start-ups, as well as whether or not and how the VC cycle can seemingly and surprisingly repeat itself despite repeated booms and bust cycles. An exception is the work by Haddad and Hornuf (2016), who examine economic and technological determinants of FinTech start-ups, without however examining the investment cycle nor the impact of the recent financial crisis. This chapter is organized as follows. The next section reviews related literature and develops testable hypotheses. Thereafter we present the data, summary statistics, and multivariate analyses. The last section offers concluding remarks and policy implications.
Hypotheses There are costs and benefits associated with developing a start-up FinTech company that is away from a financial center. On the benefit side, there is a dearth of enforcement of banking rules in countries without a major financial center, which in turn encourages innovation and exploitation of risky opportunities for FinTech start-ups.4 The reason why enforcement is more pronounced in regions with major financial centers is that there are economies of scale in prudential supervision (Cassard, 1994). FinTech start-ups thereby benefit in the spirit of regulatory arbitrage with differential enforcement of banking rules. However, there are also costs of being away from a financial center, and one of these costs is being less well connected with the main industry players. For instance, there is a gain to entrepreneurs from being close to Silicon Valley where significant human and financial resources are located. 13
D.J. Cumming, A. Schwienbacher
Similarly, FinTech ventures may have an advantage by being located in New York or London (United Kingdom) as opposed to Louisiana. A number of prior studies link weaker regulation with innovation and entrepreneurial risk-taking. First, Saxenian (2000) argues that a key reason why Silicon Valley developed into a vibrant and highly innovative tech area as we know it today is because of the weaker labor regulation in California as opposed to other US states. It reduced failure costs of startups, in case they needed to lay off employees at short notice. Similarly, it created a more liquid labor market so that the costs for employees associated with increased risk of being fired was mitigated by the higher chances of finding a new job quickly in the highly liquid labor market in Silicon Valley. The more employer-favorable labor market regulation in turn encourages innovation and entrepreneurial risk taking. Comparing across countries, Bozkaya and Kerr (2014) find consistent evidence in the Europe, and Wang and Wang (2012) find consistent evidence across other countries including Asia. Second, Dharmapala and Khanna (2016) document a positive stock market reaction to the announcement of already listed companies in the US that opt for the weaker disclosure and compliance obligations under the JOBS Act of 2012. That is, companies that recently listed on the stock market and that fall in the category of “emerging growth companies”, as defined in the JOBS Act, may comply with the weaker regulation of the JOBS Act. The fact that the choice to opt for the weaker regulation leads to a positive stock market reaction means the extra expected costs of weaker investor protection are lower than the cost savings associated with reduced needs for the company to disclose and comply with the more stringent securities regulation. Third, Levine, Lin and Shen (2015) find that cross-border acquisitions generate lower abnormal returns in countries with stronger labor protection regulations due to higher costs related to such regulation. Stronger labor protection also leads to fewer cross-border acquisitions in these countries. Assuming these cross-border acquisitions to be value-creating, they find that tougher labor regulation reduces efficiency. Fourth, minimum capital requirements can strongly impact choices of newly incorporated firms. In an empirical study of five different European countries that made significant changes in the minimum capital requirements for newly incorporate firms, Braun, Eidenmüller, Engert and Hornuf (2013) find that the reduction or abolishment of minimum capital requirements increased entrepreneurship activities and thus ultimately risk taking by entrepreneurs. Fifth, Hornuf and Schwienbacher (2017) argue that tailored regulation is needed to encourage equity crowdfunding, which would not be able to develop well under strict securities regulation that applies to large, established firms. This leads to establishing weaker investor protection under equity crowdfunding (so that transaction costs are reduced) than under the issuance of new shares by large firms (who can afford the higher transactions costs associated with stronger investor protection and information disclosure requirements). Sixth, using macro-level OECD data, Blind (2012) studies the impact of different forms of regulation (economic, social and institutional regulations) on innovation. While the literature often attributes ambivalent effects of regulation on innovation, Blind shows that regulations that generate compliance costs deter innovation, while regulations that create extra incentive effects encourage innovation. Several examples are provided by Blind in the theoretical analysis, and the empirical evidence is consistent. In sum, there is substantial prior theory and evidence that is consistent with the view that strong regulation and enforcement offers better investor protection, but at higher transaction costs. Compliance is costly, which deters risk taking and may particularly affect risky 14
Fintech venture capital
start-ups. We therefore expect that growth in FinTech VC activity will be more pronounced in countries without strong enforcement of banking rules. Hypothesis 1: Growth in FinTech VC investment has been more pronounced in countries without a major financial center since the global financial crisis. Next, VCs in countries without a major financial center are less likely to be experienced (Nahata et al., 2014; Schertler and Tykvová, 2012; Schwienbacher, 2008), since there are fewer exit opportunities there. This lack of exit opportunities reduces investments. VC experience is a key determinant of VC success, as the networks, due diligence, and value added services provided by VCs (such as providing strategic, financial, human resource, and marketing advice, and facilitating connections to customers, suppliers, accountants, lawyers and investment banks) is substantially higher among experienced VCs (Cumming and Johan, 2013; Hege et al., 2009; Nahata, 2008; Nahata et al., 2014). Inexperienced VCs are more likely to overinvest in boom periods and invest without the knowledge of prior investment cycles (Gompers and Lerner, 1999) and hence be subjected to problems in subsequent busts (Nahata, 2008; see also Chaplinsky and Gupta-Mukherjee, 2009, Nahata et al., 2014; Nielsen, 2008, 2010, among others). The boom in FinTech, therefore, is more likely to be amongst inexperienced VCs, and less likely to result in successful exit outcomes. Hypothesis 2: Growth in FinTech VC investment has been more pronounced among less experienced VCs since the global financial crisis. Hypothesis 3: FinTech VC investments are less likely to result in successful exit outcomes, particularly since the global financial crisis.
Data We collected all VC investments from the VentureXpert database from 1/1/1990 to 12/31/2015. To ensure that we only focus on true venture capital deals, we excluded investment rounds that are categorized as “Buyout/Acquisition”, “Real Estate” and “Other”. We classify VC investments with “Company VE Primary Industry Sub-Group 1” to be “Financial Services” as FinTech investments. We obtained a final sample of 2,678 investment rounds in 747 distinct FinTech ventures. A manual check on a selected number of individual cases included in the category confirms our classification, even though the sample of FinTechs is large. Similarly we obtained a sample of 277,994 round investments in nonFinTech ventures that comprises all remaining industry sectors. We further collected different m acro-economic data related to GDP from the World Bank database. To assess whether a venture is located in a country with a major financial center, we rely on the 2015 Global Financial Centres Index 18 (GFCI 18) that is computed by the Z/Yen Group Limited on an annual basis. While this index captures the overall market capitalization of stock markets and banks, it takes a more comprehensive perspective. It also includes competitiveness of the center, among others. While some considered factors are quantitative measures taken from various sources such as the World Bank, the OECD, and the Economist Intelligence Unit, others are assessed through thousands of questionnaire responses received. More specifically, the GFCI provides ratings for financial centers calculated by a “factor assessment model” that uses these two distinct sets of measures. We rely on this index as it better captures the type of proxy than restricting to, say, the size of the public equity market, since FinTech activities covers the full spectrum of financial activities (not just stock 15
D.J. Cumming, A. Schwienbacher
market activities). We consider the following list of countries to have a major financial center: United Kingdom (London), United States (New York, San Francisco, Washington DC, Chicago, Boston), China (Hong Kong), Singapore (Singapore), Japan (Tokyo), South Korea (Seoul), Switzerland (Zurich, Geneva), Canada (Toronto), Germany (Frankfurt), and Australia (Sydney). The dummy variable Financial Center takes a value of 1 for ventures located in any of these countries, and 0 for all the others. While the index is provided annually, we only use the values of 2015, since the definition of the index varies almost every year. This ensures comparability over the time period in terms of definition of a financial center.5 All the variables are defined in Appendix Table 2.1. The summary statistics are provided in Table 2.1. Average round amounts in FinTech ($17,937,200), however, have been larger than non-FinTech ($13,279,000). Syndicate size is on average larger for non-FinTech (4.6 syndicated partners on average) than FinTech (3.2). FinTech start-ups are older on average (7.5 years) than non-FinTech (5.1 years). For the entire sample years, a greater proportion of FinTech deals are in the expansion stage (49.6%) versus non-FinTech (40.9%), while relatively more non-FinTech deals are in the seed, start-up, and later stages (7.6%, 27.3%, and 24.2%, respectively) than their FinTech counterparts (5.9%, 23.0%, and 21.6%, respectively). A higher proportion of FinTech deals exit as IPOs (30.0%) than non-FinTech deals (22.1%). A lower proportion of FinTech deals exit as trade sales (51.2%) relative non nonFinTech deals (61.4%). There is no statistically significant difference in the proportion of deals that exit as write-offs between FinTechs (15.7%) and non-FinTechs (14.5%). The country distribution shows FinTech is mostly in the US (62.8%), followed by the UK (6.2%), China (6.3%), Canada (1.8%) and France (1.8%). The country distribution shows non-FinTech is mostly in the US (75.9%), Canada (3.9%), the UK (3.5%), France (3.0%), and China (2.9%). Financial (bank-affiliated) VCs are more likely to finance FinTech (10.2%) than nonFinTech (6.6%), while private limited partnerships are more likely to finance non-FinTech (65.0%) than FinTech (63.6%), corporate VCs are more likely to finance non-FinTech (7.7%) than FinTech (5.8%). Older VC firms are more likely to finance FinTech (average age is 21.8 years) than non-FinTech (average age is 20.2 years). Similarly, FinTech ventures are more often financed by larger VC funds (average fund size of USD 395.3 million) than non-FinTech ventures (USD 292.0 million). Table 2.2 provides summary statistics of all FinTech deals by country of origin for the two main dependent variables, (ln(Round Amount) and Syndicate Size (No. Investors)). It offers a comprehensive picture of the distribution of our international sample of FinTech deals. Statistics are reported for all FinTech deals, and for the subsamples of FinTech deals after (Post-Crisis = 1) and before (Post-Crisis = 0) the financial crisis. Deals tend to be concentrated in a small set of countries, mainly large ones. However, no specific country outlier strikes out from this table. Still, in the multivariate analyses, we will include country dummies to control for any time-invariant country specificities. Figure 2.1 shows that the round investment size (i.e., the amount invested in a single round) increased after the financial crisis for FinTech in non-financial centers relative to non-FinTech and relative to financial centers for several years in a row. All values are relative to 2007. Similarly, Figure 2.2 shows that the number of rounds for FinTech increased in financial centers relative to non-FinTech and relative to financial centers. The number of FinTech VC investment rounds increased to around 40 from 2007 to 2012 (starting from a benchmark of 0 at 2007) among non-financial centers versus non-financial centers, and had a similar pattern among non-FinTech VC for non-financial centers versus financial centers. 16
Mean
0.3412 0.0096 13,323.8 8.5001 4.5392 5.1569 0.0759 0.2726 0.4096 0.2419 0.2216 0.6129 0.8345 0.1453 0.7527 0.0385 0.0348 0.0298 0.0295
Startup characteristics: Post-crisis (dummy) FinTech start-up (dummy) Round amount (x1000 USD) ln(round amount) Syndicate size (no. investors) Startup age (in years) Seed-stage dev. (dummy) Early-stage dev. (dummy) Expansion-stage dev. (dummy) Later-stage dev. (dummy) IPO exit (dummy) Trade sale exit (dummy) Successful exit (dummy) Bankruptcy exit (dummy) Start-up country: USA (dummy) Start-up country: Canada (dummy) Start-up country: United Kingdom (dummy) Start-up country: China (dummy) Start-up country: France (dummy)
Full sample
Variable
Table 2.1 Summary statistics of the sample
0.0000 0.0000 6,000.0 8.6995 4.0000 4.0000 0.0000 0.0000 0.0000 0.0000 0.0000 1.0000 1.0000 0.0000 1.0000 0.0000 0.0000 0.0000 0.0000
Median
0.4741 0.0977 39,161.4 1.6260 3.2081 8.7989 0.2648 0.4453 0.4918 0.4283 0.4153 0.4871 0.3717 0.3524 0.4314 0.1925 0.1832 0.1699 0.1691
Std. dev.
0.3170 1.0000 17,937.2 8.5672 3.1538 7.5358 0.0586 0.2296 0.4955 0.2162 0.3000 0.5118 0.8118 0.1567 0.6281 0.0183 0.0616 0.0631 0.0176
Mean
0.3414 0.0000 13,279.0 8.4995 4.5527 5.1338 0.0760 0.2730 0.4087 0.2422 0.2209 0.6137 0.8347 0.1452 0.7539 0.0387 0.0345 0.0294 0.0296
Mean
FinTech sample only Non-FinTech sample only
0.008 -0.000 0.032 0.000 0.000 0.001 0.000 0.000 0.002 0.000 0.000 0.029 0.249 0.000 0.000 0.000 0.000 0.000
p-value
(Continued)
Diff. mean test
Fintech venture capital
17
18
0.0665 0.6497 0.0772 20.2090 293.01 9.6895 38.4823 2.8160 277,994
VC fund characteristics: Financial VC fund (dummy) Private VC fund (dummy) Corporate VC fund (dummy) VC firm age (in years) Fund size (million USD)
Market conditions: GDP (in USD) GDP per capita (in USD) GDP growth No. obs. 10.2848 38.1139 2.6660
0.0000 1.0000 0.0000 15.0000 106.60
Median
5.3009 12.3711 2.2646
0.2491 0.4771 0.2668 18.0797 945.00
Std. dev.
7.9656 31.4577 3.7273 2,678
0.1019 0.6355 0.0579 21.8096 395.26
Mean
9.7062 38.5506 2.8071 275,316
0.0661 0.6499 0.0773 20.1935 292.00
Mean
FinTech sample only Non-FinTech sample only
0.000 0.000 0.000
0.000 0.122 0.000 0.000 0.000
p-value
Diff. mean test
Note: This table shows summary statistics for the full sample and for different subsamples (fintech versus non-fintech, post-crisis versus pre-crisis). The number of observations reported in the last line corresponds to the number for most of the variables. For some variables, fewer observations are available. For the variables, IPO Exit, Trade Sale Exit, Successful Exit, and Bankruptcy Exit, statistics are reported for the subsample of exited ventures only.
Mean
Variable
Full sample
D.J. Cumming, A. Schwienbacher
–0.0088***
0.0129 0.0205*** 0.0811* Yes Yes Yes
–0.0111***
0.6345*** 0.4794*** 0.8071*** 0.0048***
0.0164 0.0204*** 0.0799*
No Yes Yes
277994
VC fund characteristics: Financial VC fund (dummy) Private VC fund (dummy) Corporate VC fund (dummy) VC firm age (in years)
Market conditions: GDP (in USD) GDP per capita (in USD) GDP growth
Industry dummies Stage of dev. dummies Start-up country dummies
No. obs.
19 277994
No Yes Yes
0.0164 0.0205*** 0.0799*
0.6344*** 0.4794*** 0.8067*** 0.0048***
–0.1398 –0.0622 0.3133*** –0.0111***
[3]
277994
Yes Yes Yes
0.0129 0.0205*** 0.0811*
0.6359*** 0.4674*** 0.7818*** 0.0046***
–0.1030 0.4419*** 0.2733*** –0.0088***
[4]
96874
No Partial Yes
0.0095 0.0141** 0.0556
0.7702*** 0.6105*** 0.8302*** 0.0068***
–0.1214 0.0991 0.2946** –0.0065
[5]
96874
Yes Partial Yes
0.0073 0.0148** 0.0618*
0.7661*** 0.5926*** 0.8031*** 0.0066***
–0.0819 0.3644* 0.2597* –0.0069
[6]
Early/seed-stage rounds only
181120
No Partial Yes
0.0195 0.0244*** 0.0933*
0.5624*** 0.3943*** 0.7671*** 0.0038***
–0.1409 –0.1185 0.2759** –0.0114***
[7]
181120
Yes Partial Yes
0.0162 0.0238*** 0.0922*
0.5663*** 0.3845*** 0.7395*** 0.0036***
–0.1019 0.4545** 0.2272** –0.0087***
[8]
Early/seed-stage rounds excluded
Note: This table shows OLS regressions on the determinants of investment amounts. The dependent variable is ln(Round amount), where Round amount is the value in USD (x1000). Standard errors are clustered by year of investment. Significance levels: *, **, *** for 10%, 5%, 1%, respectively.
277994
0.6359*** 0.4674*** 0.7821*** 0.0046***
–0.0999 0.5278***
[2]
–0.1364 0.0353
[1]
Start-up characteristics: Post-crisis (dummy) FinTech start-up (dummy) Post-crisis * FinTech Startup age (in years)
Variable
Full sample
Table 2.2 Determinants of round investment amounts
Fintech venture capital
D.J. Cumming, A. Schwienbacher 60 40 20
-20
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
0
-40 -60 -80 -100 -120 Fintech Sample in Non-Financial Centers - Fintech Sample in Financial Centers Non-Fintech Sample in Non-Financial Centers - Non-Fintech Sample in Financial Centers
Figure 2.1 Total size of all VC rounds, scaled 2007=0 100 80
60 40 20
-20
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
0
-40
-60 -80 Fintech Sample in Non-Financial Centers - Fintech Sample in Financial Centers Non-Fintech Sample in Non-Financial Centers - Non-Fintech Sample in Financial Centers
Figure 2.2 Number of VC rounds, scaled 2007=0
Figure 2.3 shows that the number of investment rounds increased significantly after the financial crisis for FinTech for small funds relative to non-FinTech, and relative to large funds, consistent with Hypothesis 2. Similarly, Figure 2.4 shows that the annual volume of VC investments increased dramatically for FinTech for small funds relative to large funds and relative to non-FinTech, again consistent with Hypothesis 2. Among small funds, FinTech was far more important relative to non-FinTech, as compared to large funds.
20
Fintech venture capital 150 100 50
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
0 -50 -100 -150 -200 Fintech Sample among Small Funds - Non-Fintech Sample in Small Funds Fintech Sample among Large Funds - Non-Fintech Sample in Large Funds
Figure 2.3
Number of investment rounds, scaled 2007=0
600 500
400 300 200 100
-100
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
0
-200 -300 -400 Fintech Sample among Small Funds - Non-Fintech Sample in Small Funds Fintech Sample among Large Funds - Non-Fintech Sample in Large Funds
Figure 2.4 Annual volume of round investments, scaled 2007=0
Multivariate analyses Our multivariate analyses proceed as follows. First, we present analyses of the impact of the financial crisis on VC round investment amounts in Table 2.2 to establish the broad patterns in the FinTech after the financial crisis around the world as a base case without testing specifically for differences in countries as predicted by Hypotheses 1 and 2. Thereafter, in Tables
21
D.J. Cumming, A. Schwienbacher
2.3–2.5 we examine subsets of the data to directly test Hypotheses 1 and 2. Finally, in Table 2.6 we provide an analysis of exits to test Hypothesis 3. Table 2.2 presents OLS regressions of the determinants of round amounts (in natural logarithm) with an analysis for FinTech relative to non-FinTech after the financial crisis. Our main variable of interest is therefore the interaction term Post-Crisis*FinTech. Several other explanatory variables are included to control for alternative factors. The inclusion of fixed effects helps controlling for different common effects across industry, stage of development and country of start-up that may not be captured by our other variables. Among other things, it is meant to capture possible differences in economies of scale and scope across countries, and differences in time-invariant business laws, level of corruption and misreporting practices across countries (Djankov et al., 2002; Johan and Zhang, 2020). The data examined are consistent with our expectations that FinTech investments increased after the financial crisis. Controlling for other things being equal, we estimate that round investment amounts for FinTech went up, relative to the average amounts for all investments in the sample, by 3.2% to 3.7% for the full sample (Models 3–4), relative to the average deal sizes in the sample.6 FinTech went up by 3.0% to 3.4% for the subset of early stage VC FinTech deals (Models 5–6), and by 2.7% to 3.2% for the early/seed stage rounds excluded (Models 7–8). Table 2.3 presents an analysis of round investment amounts by type of VC fund. Round investment amounts for FinTech post-crisis went up by 5.4% in countries without a major financial center (Model 2), but were unchanged in countries with a major financial center (Model 1), consistent with Hypothesis 1. Here, we are unable to test the prediction in a single regression due to the inclusion of country fixed effects (which would eliminate the variable Financial Center, since it is invariant over time within countries). Table 2.3 also shows that round investment amounts went up for FinTech post-crisis by 4.6% for independent VCs (Model 4), but there was no change for corporate VCs (Model 5) or financial institution affiliated VCs (Model 3).7 Round investment amounts for FinTech post-crisis went up by 2.0% among large VC funds (Model 7), and 9.1% among small VC funds (Model 6), consistent with Hypothesis 2. To offer further support to Hypothesis 2, we use a second measure of experience, which is fund size. We define small VC funds as those with a fund size less than 100 million USD. We obtain similar conclusions with an alternative measure of experience (we consider VC firms with age less than 15 years as “young” and the others as “old”), namely age of the VC firms; indeed, the impact on FinTech in the post-crisis period is only significant for the younger VC firms (Models 8–9). Table 2.4 presents robustness analyses of FinTech rounds investment amounts excluding the US (Models 1 and 2), and within the US only at the state level (Models 3 and 4). Showing robustness without the US is important, since about 75% of the sample are investments made in that country (Table 2.1). As for the second one, we there provide an analysis at the US State level (see accompanying note in Table 2.4 for more details), where we expect no significant difference due to the same regulation applied across all US States. Thus, a lack of finding for the US State level analysis offers support to our argument. The evidence in Models 1 and 2 is consistent with Table 2.3 and shows a pronounced increase in FinTech after the financial crisis among countries without a financial center and not in countries with a financial center, consistent with Hypothesis 1. By contrast, within the US, there is no material difference in state-level investments depending on whether or not the state had a financial center (as defined in Table 3.4) in Models 3 and 4. This evidence in Models 3 and 4, and in the context of Models 1 and 2 and the evidence in Tables 2.2 and 2.3, supports our analysis at the country-level where financial regulation and enforcement takes place. There are scant 22
[1]
23
0.4558*** 0.6341*** 0.0054***
0.4702*** 0.7955***
0.0044***
0.0094 0.0254*** 0.0902*
Market conditions: GDP (in USD) GDP per capita (in USD) GDP growth
30036
No. obs.
18484
Yes Yes Yes
0.1086*** 0.0012 0.0934**
0.0053***
–– ––
––
–0.0219 0.7236** –0.0016 –0.0112***
[3]
0.0005
–– ––
––
–0.0107 –0.3014 –0.4767 –0.0144***
[5]
180621
Yes Yes Yes
21448
Yes Yes Yes
62973
Yes Yes Yes
0.0338 0.0236*** 0.0849*
0.0074***
0.3656*** 0.7443***
0.4811***
–0.3192 0.1979 0.7737*** –0.0152***
[6]
Small VC Corporate VCs funds
0.0301* –0.0175 0.0234*** 0.0170** 0.0746* 0.0914**
0.0053***
–– ––
––
–0.2295 0.3845** 0.3931*** –0.0090**
[4]
Financial VCs Private VCs
215021
Yes Yes Yes
–0.0078 0.0162** 0.0708**
0.0019***
0.5320*** 0.7853***
0.7674***
–0.0185 0.4799** 0.1741** –0.0075**
[7]
Large VC funds
0.6814***
0.0246 0.3218 0.1041 –0.0091***
[9]
0.0030***
139482
Yes Yes Yes
138512
Yes Yes Yes
0.0310 –0.0151 0.0220*** 0.0215** 0.0830* 0.0746*
–0.0063*
0.4630*** 0.5075*** 0.8890*** 0.6856***
0.6417***
–0.2072 0.4730*** 0.4167** –0.0083**
[8]
Younger VC Older VC firms firms
Note: This table shows OLS regressions on the determinants of investment amounts. The dependent variable is ln(Round amount), where Round amount is the value is USD (x1000). Small (Large) VC funds (Models 6 and 7) are classified based on the cut-off level of USD 100 million. Younger (Older) VC firms (Models 8 and 9) are classified based on the cut-off level of 15 years between time of investment and the VC firm’s incorporation date. Standard errors are clustered by year of investment. Significance levels: *, **, *** for 10%, 5%, 1%, respectively.
247958
Yes Yes Yes
Industry dummies Yes Stage of dev. dummies Yes Start–up country dummies Yes
0.1716 –0.0102 0.0371*
0.4506***
0.6937***
0.2739 0.6952** 0.4607** –0.0019*
[2]
VC fund characteristics: Financial VC fund (dummy) Private VC fund (dummy) Corporate VC fund (dummy) VC firm age (in years)
Start–up characteristics: Post–crisis (dummy) –0.1548 FinTech start–up (dummy) 0.3938** Post–crisis * FinTech –0.0232 Startup age (in years) –0.0122***
Variable
Major financial No major center financial center
Table 2.3 D eterminants of round investment amounts
Fintech venture capital
24
0.451*** 0.456*** 0.634*** 0.00536*** 0.172 –0.0102 0.0371* Yes Yes Yes
0.720*** 0.669*** 0.981*** 0.00167
0.0495** 0.0104* 0.0741*** No Yes Yes
38707
No. obs.
131853
–0.883 0.353 0.114* Yes Yes No
0.563*** 0.334*** 0.617*** 0.00396***
0.176 0.193 0.263 –0.0235***
[3]
77398
–1.738** 0.651** 0.112* Yes Yes No
0.775*** 0.482*** 0.824*** 0.00682***
0.281 0.579** –0.0947 –0.0140***
[4]
33968
–0.188 0.0196* 0.0565* Yes Yes Yes
0.459*** 0.516*** 0.745*** 0.00162
0.373 0.184 –0.181 –0.00269
[5]
244026
0.0820*** 0.00606 0.0736 Yes Yes Yes
0.650*** 0.455*** 0.774*** 0.00506***
–0.357 0.461*** 0.401*** –0.0148***
[6]
US ventures in non-financial centers only High internet Low internet (state level) usage usage
35830
0.0480 0.00759 0.00471 Yes Yes Yes
0.508*** 0.482*** 0.801*** 0.00407**
0.267 0.293 –0.318 –0.00164
[7]
242164
–0.00777 0.0304*** 0.0894* Yes Yes Yes
0.649*** 0.461*** 0.767*** 0.00466***
–0.172 0.512** 0.360*** –0.0153***
[8]
High mobile phone Low mobile phone subscription rate subscription rate
Note: This table shows OLS regressions on the determinants of investment amounts. The dependent variable is ln(Round amount), where Round amount is the value is USD (x1000). In Models 3 and 4, US states with a financial center are New York (New York), California (San Francisco), District of Columbia (Washington DC), Illinois (Chicago), and Massachusetts (Boston), following the definition of the country-level list of US states (see definition of Financial Center in Table 2.1). Models 5 and 6 show results for the subsamples of observations for which the country’s Internet Usage is above and below 75%. Models 7 and 8 show results for the subsamples of country’s mobile phone subscription is above and below 100. The standard errors are clustered by year of investment. Significance levels: *, **, *** for 10%, 5%, 1%, respectively.
30036
0.274 0.695** 0.461** –0.00193*
[2]
0.404** 1.036*** –0.890*** –0.00396*
[1]
Non-US ventures in non- US ventures in financial centers financial centers only only (state level)
Startup characteristics: Post–crisis (dummy) FinTech start–up (dummy) Post–crisis * FinTech Startup age (in years) VC fund characteristics: Financial VC fund (dummy) Private VC fund (dummy) Corporate VC fund (dummy) VC firm age (in years) Market conditions: GDP (in USD) GDP per capita (in USD) GDP growth Industry dummies Stage of dev. dummies Start–up country dummies
Variable
Non-US ventures in financial centers only
Table 2.4 Robustness analyses on the determinants of round investment amounts
D.J. Cumming, A. Schwienbacher
Fintech venture capital
differences in enforcement of banking regulations within a country, but marked differences in enforcement of banking regulations across countries. Hence, the marked differences in FinTech VC after the crisis are observed across countries where there is, and is not, a major financial center, consistent with Hypothesis 1. One possibly alternative explanation for the results obtained is that the impact on FinTech post-crisis being more pronounced in certain countries than others, may be due to differences in the level of democratization of digitalization. Countries where digitalization is more widely spread among the population may be more attractive for FinTech start-ups, since the adoption rate of their products and services may in turn be higher. To check the plausibility of this alternative explanation, we run different subsample analyses for countryyear observations where digitalization differs significantly. We use two measures of democratization of digitalization. One is the degree of internet usage (Models 5–6); the other one is the subscription rate of mobile phones (Models 7–8). High degrees of digital adoptions are associated with a greater internet usage and higher mobile phone subscription rate in the population. To ensure that we capture the digitalization trend of the most recent years that match the FinTech emergence, we take a cut-off of 90% of the distribution of these variables. Note also that these variables are time-varying, which means that observations of a given country are generally found in both subsamples (the more recent observations are more likely to be in the high digitalization subsample). However, we obtain the opposite results, in that the effect of the interaction term is only significant and positive when digitalization is low, not high. This offers no support for this alternative explanation. Table 2.5 presents an analysis of syndicate size. Controlling for other things being equal, round syndicate size went up for FinTech after the financial crisis, relative to the average syndicate size in the full sample, by 23.7% for full sample (Model 1), by 18.3% for early stage rounds (Model 2), 26.5% excluding early stage rounds (Model 3), 31.8% for small VC funds (Model 4), 22.7% for large VC funds (Model 5), and 37.1% for older VC firms. Also, the data indicate that syndicate size went up by 25.4% in countries without a large financial center (Model 9), and unchanged in countries with a large financial center (Model 8), consistent with Hypothesis 1 on investment behavior post-crisis. Table 2.6 presents an analysis of exit outcomes. There, we control for market conditions at time of exit, and whether a VC fund with an early-stage investment mandate participated in the financing of the VC-backed start-up. The data for the full sample indicate that controlling for other things being equal, FinTech VC deals are 4.6% less likely to be write-offs if they originated after the financial crisis (Model 3), but 8.3% more likely to be either an IPO or a trade sale (Model 2), consistent with Hypothesis 3. We find similar impacts on exit outcomes for FinTech after the crisis amongst the countries with a major financial center (Models 4–6). Focusing on the exits of ventures located in a country without a major financial center (Models 7–9), there is a material increase in liquidations post-crisis for FinTech, as the effect is now positive and economically large. The probability is now higher by 3.3% (Model 9), and a slight reduction in the probability of successful exits (IPOs or trade sales). Overall, the data indicate that changes in FinTech investments are more pronounced in countries without a major financial center after the crisis (Table 2.3), and that these ventures are more likely to fail after the crisis relative to non-FinTech ventures (Table 2.6). We note that we could have added other variables to the exit regressions, such as syndication. In Table 2.5 we noted that FinTech deals are more likely to be syndicated in regions without a financial center. Syndication of VC investment can facilitate larger investment amounts, and is often associated with more value-added advice by investors (Casamatta, 2003; Casamatta and Haritchabalet, 2007; Gompers and Lerner, 1999; Sevilir, 2010). Including syndication is 25
26
–0.310 0.346 1.077*** –0.061***
0.871*** 0.471*** 1.165*** 0.007***
–0.049** 0.006 0.046 Yes Yes Yes
277994
Start-up characteristics: Post-crisis (dummy) FinTech start-up (dummy) Post-crisis * FinTech Startup age (in years)
VC fund characteristics: Financial VC fund (dummy) Private VC fund (dummy) Corporate VC fund (dummy) VC firm age (in years)
Market conditions: GDP (in USD) GDP per capita (in USD) GDP growth Industry dummies Stage of dev. dummies Startup country dummies
No. obs.
96874
0.018 –0.010* 0.042 Yes Yes Yes
0.661*** 0.392*** 0.845*** 0.008***
0.281 0.323 0.833** –0.022***
[2]
181120
–0.089*** 0.014 0.049 Yes Yes Yes
0.971*** 0.526*** 1.326*** 0.007***
–0.640*** 0.258 1.202** –0.063***
[3]
[5]
Large VC funds
62973
–0.052** 0.012 0.046 Yes Yes Yes
0.537*** 0.263** 0.913*** 0.013***
215021
–0.041 0.004 0.042 Yes Yes Yes
0.954*** 0.500*** 1.203*** 0.006***
–0.550*** –0.299 0.275 0.376 1.444*** 1.030** –0.059*** –0.062***
[4]
Early/seed-stage Early/seed-stage Small VC rounds only rounds excluded funds
138512
–0.061** 0.004 0.027 Yes Yes Yes
1.087*** 0.675*** 1.291*** 0.008***
–0.280 0.197 0.502 –0.058***
[6]
Younger VC firms [8]
139482
–0.040* 0.009 0.061 Yes Yes Yes
0.407*** 0.023 0.750*** –0.016***
247958
–0.060** 0.013 0.047 Yes Yes Yes
1.011*** 0.524*** 1.251*** 0.007***
30036
–0.125 –0.019** 0.034 Yes Yes Yes
0.165 0.088 0.473*** 0.007***
0.200 0.718 1.154*** –0.028***
[9]
Large financial No large center financial center
–0.305 –0.359* 0.332 0.298 1.686*** 0.432 –0.063*** –0.066***
[7]
Older VC firms
Note: This table shows Poisson regressions on the determinants of syndicate size. The dependent variable is the number of investors involved in the financing of a given round. Reported coefficient values are marginal effects. Standard errors are clustered by year of investment. Significance levels: *, **, *** for 10%, 5%, 1%, respectively.
[1]
Variable
Full sample
Table 2.5 Determinants of round syndicate size
D.J. Cumming, A. Schwienbacher
27
–0.0554*** 0.0062 Yes Yes 14593 14593
Early-stage VC fund Industry dummies No. obs.
[5]
0.0030
0.0017* –0.0011
0.0810***
–0.0550*** 0.0013 Yes Yes 12564 12564
–0.0093*** 0.0121**
–0.0030*
0.0329
0.1524*** 0.0635*** –0.1168 –0.0007
[4]
0.0133*** Yes 12564
–0.0017
0.0004 0.0001
–0.0502***
–0.0558*** –0.0031
[6]
0.0748***
0.0476*** 0.0583
[8]
–0.0461 Yes 2029
0.0321***
0.0306** Yes 2029
0.0082***
0.0787*** –0.0034 –0.0077*** –0.0022***
0.0381
0.0313 –0.1165
[7]
0.0011 Yes 2029
–0.0007*
–0.0027* 0.0003***
0.0328**
–0.0056** 0.0190***
[9]
Note: This table shows Probit regressions on the determinants of exit outcomes as measure of success. The dependent variable “IPO Exit” is a dummy variable equal to 1 for exited ventures that made an IPO, “Successful Exit” that made either an IPO or a trade sale (TS), and “Bankruptcy Exit” that went bankrupt. The variable “Early-Stage VC Fund” is a dummy variable taking the value of 1 if at least one participating VC fund has an early-stage investment mandate, and 0 otherwise. Reported coefficient values are marginal effects. Standard errors are clustered by year of investment. Significance levels: *, **, *** for 10%, 5%, 1%, respectively.
0.0103*** Yes 14593
–0.0017
0.0043*
0.0182***
–0.0462***
0.0008* 0.0005
0.0828***
0.0074
–0.0519*** –0.0055
[3]
–0.0052*** 0.0027*** –0.0077*** –0.0017***
0.0625*** 0.0197
[2]
0.1421*** –0.1192*
[1]
Successful exit Bankruptcy exit
Ventures not located in country with a major financial center
Successful exit Bankruptcy exit IPO exit
Ventures located in country with a major financial center
Successful exit Bankruptcy exit IPO exit
Market conditions: GDP at exit (in USD) GDP per capita at exit (in USD) GDP growth at exit
Post-crisis (dummy) FinTech startup (dummy) Post-crisis * FinTech
Startup characteristics:
Variable
IPO exit
Full sample
Table 2.6 Determinants of exit outcomes
Fintech venture capital
D.J. Cumming, A. Schwienbacher
arguably endogenous to exit performance, and as such we do not include syndication as explanatory variable in Table 2.6. But doing so only strengthens our results that FinTech deals performed worse after the financial crisis in regions without a financial center, particularly relative to the expected performance in view of investment syndication and relative to performance of comparable investments in regions with a financial center. As robustness check, we further run Heckman probit regressions that control for the possible non-randomness of the sample of exited ventures. Indeed, the set of ventures that were started before and after the financial crisis may not be the same if exits through some of the routes take longer. We find that even after controlling for this possible self-selection bias, FinTech ventures started after the financial crisis have significantly higher risks of failing if located away from a major financial center. Thus, these extra tests confirm our earlier conclusion.
Alternative explanations and future research We use the Global Financial Centres Index 18 (GFCI 18) as computed by the Z/Yen Group as a main focus of international difference in banking regulatory enforcement. The advantages with this index are that (1) it is specific to the finance industry (unlike the World Bank Doing Business Indices, for example, that are not industry specific), (2) we can examine a stable set of countries over time to enable us to track what is going on in one consistent set of countries versus another, and (3) it pertains to financial regulation enforcement, which appears to be the most pertinent based on industry insights from Forbes and others (see Footnote 2 and accompanying text). There are hundreds of indices to measure differences across countries including but not limited to the World Bank indices. We believe our index is directly appropriate for our research question. Perhaps future research with better and more fine-tuned data will come to other insights with other country groupings. There could be alternative explanations for the patterns that we observe in the data. First, as pointed out before, there may be differences in the democratization of digitalization in developed versus developing countries. Our analysis however provides no support for this alternative explanation, although clearly it is a necessary condition for making FinTech investments attractive from the market perspective. A greater adoption of digital products and services will ensure greater market prospects for FinTechs. Second, the term FinTech covers the entire scope of services and products traditionally provided by the financial services industry (Haddad and Hornuf, 2016). It comprises several areas such as finance and investment, operations and risk management, payments and infrastructure, data security and monetization, and customer interface. As such, some of the results may be sensitive to the nature of FinTech start-ups at a level that is more granular than what we observe with the industry classifications in the Securities Data Company (SDC) Platinum VentureXpert dataset. Third, there may be economics of scale and scope in FinTech to different degrees in developed versus developing countries. These potential explanations, and possibly others, are not immediately obvious with the data that we have access to. Nevertheless, we have included fixed effects for industry, stage of development, and start-up country domicile in our regression specifications to capture many of these unobserved potential effects. Also, we use an approach akin to difference-indifferences specification to compare before and after the critical event, which further controls for these competing explanations insofar as they remain stable over time. We welcome and encourage future research on the topic with more detailed data, if and when they become available, to shed more light on the issues that we have examined for the first time in this chapter.
28
Fintech venture capital
Conclusion FinTech start-ups have been able to raise significant amounts of capital in the most recent years. We provide evidence that this change is more pronounced amongst smaller, private independent limited partnership VCs and in countries without a major financial center. This pattern is consistent with a differential enforcement of rules pertaining to financial institutions, where enforcement is more likely for major financial institutions and not startups, and hence FinTech start-ups are more likely to receive larger financing in countries without a major financial center. At the same time, these ventures are more likely to fail. Taken together, these findings may suggest inefficient VC investments in some regions of the world. Syndicate sizes for FinTech VCs have likewise become larger after the financial crisis, which would normally imply that FinTech deals would do better. But the evidence to date shows that FinTech deals are substantially less likely to be acquired after the financial crisis, and more likely to result in liquidation. The data thus point to exuberance in FinTech VC investment, and a short memory of other recent boom and bust cycles in VC. The data suggest a number of practitioner and policy implications. First, VCs should be concerned about excessive pushes into hot industries, fueled in part by media hype. With too much money chasing too few quality deals, there may be a commensurate reduction in the average quality of such deals, consistent with our evidence on FinTech exits herein. Second, policymakers should be aware that differential enforcement of financial regulations can spur economic activity in different directions that are away from the watchful eye of regulators. Public policies spurring VC investment patterns have been documented by Armour and Cumming (2006), Kanniainen and Keuschnigg (2003, 2004) and Keuschnigg and Nielsen (2003, 2004). However, prior work has not examined the intersection of enforcement policies and government programs. The evidence in this chapter suggests further work is warranted. FinTech VC is a burgeoning area because financial regulators are prone to focus on large financial institutions. Regulators could be given incentives to pay attention to the broader marketplace to not only protect consumers but also to ensure that innovations and new technologies are efficiently developed and not a response to differential regulatory oversight.
Notes
29
D.J. Cumming, A. Schwienbacher
References Accenture (2015). The future of fintech and banking: Digitally disrupted or reimagined? Accenture Consulting. Available at: https://www.accenture.com/us-en/insight-future-fintech-banking.aspx Armitage, S., Sarkar, S., Talaulicar, T. (2017). Corporate governance challenges in emerging economies. Corporate Governance: An International Review 25, 148–154. Armour, J., Cumming, D. (2006). The legislative road to Silicon Valley. Oxford Economic Papers 58, 596–635. Bertoni, F., Colombo, M.G., Quas, A. (2015). The patterns of venture capital investment in Europe, Small Business Economics 45 (3), 543–560. Bertoni, F., Groh, A.F. (2014). Cross-border investments and venture capital exits in Europe. Corporate Governance: An International Review 22 (2), 84–99. Bjørnskov, C., Foss, N. (2013). How strategic entrepreneurship and the institutional context drive economic growth. Strategic Entrepreneurship Journal 7 (1), 50–69. Bonini, S., Alkan, S., Salvi, A. (2012). The effects of venture capitalists on the governance of firms. Corporate Governance: An International Review 20 (1), 21–45. Bozkaya, A., Kerr, W. (2014). Labor regulations and European venture capital. Journal of Economics & Management Strategy 23 (4), 776–810. Braun, R., Eidenmüller, H., Engert, A., Hornuf, L. (2013). Does charter competition foster entrepreneurship? A difference-in-difference approach to European company law reforms. Journal of Common Market Studies 51, 399–415. Buzzacchi, L., Scellato, G., Ughetto, E. (2015). Investment stage drifts and venture capital managerial incentives. Journal of Corporate Finance 33, 118–128. Casamatta, C. (2003). Financing and advising: optimal financial contracts with venture capitalists. Journal of Finance 58, 2059–2085. Casamatta, C., Haritchabalet, C. (2007). Experience, screening and syndication in venture capital investments. Journal of Financial Intermediation 16, 368–398. Cassard, M. (1994). The role of offshore centers in international financial intermediation. IMF Working Papers Series, WP/94/107-EA. Washington: International Monetary Fund. Chaplinsky, S., Gupta-Mukherjee, S. (2009). The rise of M&A exits in venture capital: implications for capital recovery, Darden Working Paper. Cumming, D.J., Johan, S. (2013). Venture Capital and Private Equity Contracting: An International Perspective. Second Edition. Burlington, MA: Academic Press.
30
Fintech venture capital Cumming, D.J., Groh, A.P., Johan, S.A. (2018). Same rules, different enforcement: market abuse in Europe. Journal of International Financial Markets, Institutions and Money 54, 130–151. Cumming, D.J., Fleming, G., Schwienbacher, A. (2005). Liquidity risk and venture finance. Financial Management 34, 77–105. Dai, N., Nahata, R. (2016). Cultural differences and cross-border venture capital syndication. Journal of International Business Studies 47, 140–169. Dharmapala, D., Khanna, V. (2016). The costs and benefits of mandatory securities regulation: Evidence from market reactions to the JOBS Act of 2012. Journal of Law, Finance and Accounting 1, 139–186. Djankov, S., La Porta, R., Lopez-De-Silanes, F., Shleifer, A. (2002). The regulation of entry. Quarterly Journal of Economics 117, 1–37. Gompers, P.A., Lerner, J. (1999). The Venture Capital Cycle. Cambridge, MA: MIT Press. Guler, I., Guillén, M.F. (2010). Institutions and the internationalization of US venture capital firms. Journal of International Business Studies 41, 185–205. Haddad, C., Hornuf, L. (2016). The emergence of global Fintech market: economic and technological determinants. Working paper. Available on SSRN: http://ssrn.com/abstract=2830124. Hege, U., Palomino, F., Schwienbacher, A. (2009). Venture capital performance: the disparity between Europe and the United States. Finance, 30 (1), 7–50. Hornuf, L., Schwienbacher, A. (2017). Should securities regulation promote crowdinvesting? Small Business Economics 49, 579–593. Johan, S.A., Schweizer, D., Zhan, F. (2014). The changing latitude: labor-sponsored venture capital corporations in Canada. Corporate Governance: An International Review 22 (2), 145–161. Johan, S.A., Zhang, M. (2020). Information asymmetries in private equity: Reporting frequency, endowments, and governance. Journal of Business Ethics, forthcoming. Kanniainen, V., Keuschnigg, C. (2003). The optimal portfolio of start-up firms in venture capital finance. Journal of Corporate Finance, 9, 521–534. Kanniainen, V., Keuschnigg, C. (2004). Start-up investment with scarce venture capital support. Journal of Banking and Finance 28, 1935–1959. Kelly, G. (2014). The digital revolution in banking, G30 Occasional Paper 89. Available at: http:// www.centerforfinancialstability.org/research/OP89.pdf Keuschnigg, C., Nielsen, S.B. (2003). Tax policy, venture capital and entrepreneurship. Journal of Public Economics 87, 175–203. Keuschnigg, C., Nielsen, S.B. (2004). Start-ups, venture capitalists and the capital gains tax. Journal of Public Economics 88, 1011–1042. Li, Y., Zahra, S.A. (2012). Formal institutions, culture, and venture capital activity: A cross-country analysis. Journal of Business Venturing 27, 95–111. Levine, R., Lin, C., Shen, B. (2015). Cross-border acquisitions and labor regulations. NBER Working Paper No. 21245. Luo, Y., Junkunc, M. (2008). How private enterprises respond to government bureaucracy in emerging economies: the effects of entrepreneurial type and governance. Strategic Entrepreneurship Journal 2 (2), 133–153. Nahata, R. (2008). Venture capital reputation and investment performance. Journal of Financial Economics 90, 127−151. Nahata, R., Hazarika, S., Tandon, K. (2014). Success in global venture capital investing: Do institutional and cultural differences matter? Journal of Financial and Quantitative Analysis 49, 1039–1070. Nielsen, K. (2008). Institutional investors and private equity. Review of Finance 12, 185–219. Nielsen, K. (2010). The return to direct investment in private firms: new evidence on the private equity premium puzzle. European Financial Management 17, 436–463. Philippon, T. (2016). The FinTech opportunity. NBER Working Paper No. 22476. Saxenian, A. (2000). Regional Advantage: Culture and Competition in Silicon Valley and Route 128. Cambridge, MA: Harvard University Press. Sevilir, M. (2010). Human capital investment, new firm creation and venture capital. Journal of Financial Intermediation 19, 483–508. Schertler, A., Tykvová, T. (2012). What lures cross-border venture capital inflows? Journal of International Money and Finance 31(6), 1777–1799. Schwienbacher, A. (2008). Venture capital investment practices in Europe and in the United States. Financial Markets and Portfolio Management 22(3), 195–217.
31
D.J. Cumming, A. Schwienbacher Schroeder, W.R. (2001). Money laundering: a global threat and the international community’s response. FBI Law Enforcement Bulletin, May, 1–7. Tykvova, T., Schertler, A. (2014). Does syndication with Local Venture Capitalists moderate the effects of geographical and institutional Distance? Journal of International Management 20(4), 406–420. Wang, L., Wang, S. (2012). Economic freedom and cross-border venture capital performance. Journal of Empirical Finance 19, 26–50. Wright M., Westhead, P., Ucbasaran, D. (2007). The internationalization of SMEs and international entrepreneurship: a critique and policy implications. Regional Studies 41(7): 1013–1029. Zahra, S.A. (2014). Public and corporate governance and young global entrepreneurial firms. Corporate Governance: An International Review 22(2): 77–83.
32
33
Start-up country: Canada (dummy)
Start-up country: USA (dummy)
Bankruptcy exit (dummy)
Successful exit (dummy)
Trade sale exit (dummy)
IPO exit (dummy)
Later-stage dev. (dummy)
Expansion-stage dev. (dummy)
Early-stage dev. (dummy)
Round amount (x1000 USD) ln(Round amount) Syndicate size (no. investors) Startup age (in years) Seed-stage dev. (dummy)
Startup characteristics: Post-crisis (dummy) FinTech start-up (dummy)
Variable
Table A2.1 Definition of variables
(Continued)
Dummy variable equal to one if the round investment takes place in 2008 or later. Dummy variable equal to one if the VentureXpert variable “Company VE Primary Industry Sub-Group 1” is “Financial Services”, and zero otherwise (Source: SDC Platinum VentureXpert). Total amount invested by VCs in a given round, in thousands of USD (Source: SDC Platinum VentureXpert). Natural logarithm of the variable “Round amount”. Number of VC funds participating in the financing round (Source: SDC Platinum VentureXpert). Age in years of the venture at time of the financing round (Source: SDC Platinum VentureXpert). Dummy variable equal to one if the venture’s financing round is in the seed stage, and zero otherwise (Source: SDC Platinum VentureXpert; companystagelevel1 == “Startup/Seed”). Dummy variable equal to one if the venture’s financing round is in the early stage, and zero otherwise (Source: SDC Platinum VentureXpert; companystagelevel1 == “Early Stage”). Dummy variable equal to one if the venture’s financing round is in the expansion stage, and zero otherwise (Source: SDC Platinum VentureXpert; companystagelevel1 == “Expansion”). Dummy variable equal to one if the venture’s financing round is in the later stage, and zero otherwise (Source: SDC Platinum VentureXpert; companystagelevel1 == “Later Stage”). Dummy variable equal to one if the exit from the venture was through an IPO, and zero otherwise (Source: SDC Platinum VentureXpert; companysituation == “Went Public” or “In Registration”). Dummy variable equal to one if the exit from the venture was through a trade sale, and zero otherwise (Source: SDC Platinum VentureXpert; companysituation == “Acquisition” or “Merger” or “Pending Acquisition”). Dummy variable equal to one if the exit from the venture was through either an IPO or a trade sale, and zero otherwise (Source: SDC Platinum VentureXpert). Dummy variable equal to one if the exit from the venture was through a liquidation, and zero otherwise (Source: SDC Platinum VentureXpert; companysituation == “Bankruptcy - Chapter 11” or “Bankruptcy - Chapter 7” or “Defunct”). Dummy variable equal to one if the venture is located in the United States, and zero otherwise (Source: SDC Platinum VentureXpert). Dummy variable equal to one if the venture is located in Canada, and zero otherwise (Source: SDC Platinum VentureXpert).
Definition
Appendix
Fintech venture capital
Definition
34
Internet usage Mobile phone subscriptions
GDP growth at exit
GDP per capita at exit (in USD)
GDP at exit (in USD)
GDP growth
GDP per capita (in USD)
GDP (in USD)
Financial center (dummy)
Market conditions:
VC firm age (in years) Fund size
Corporate VC fund (dummy)
Private VC Fund (dummy)
Financial VC fund (dummy)
VC fund characteristics:
Dummy variable equal to one if the venture is located in the United Kingdom (London), the United States (New York, San Francisco, Washington DC, Chicago, Boston), China (Hong Kong), Singapore (Singapore), Japan (Tokyo), South Korea (Seoul), Switzerland (Zurich, Geneva), Canada (Toronto), Germany (Frankfurt) or Australia (Sydney), and zero otherwise (Source: 2015 Global Financial Centres Index 18 (GFCI 18) computed by the Z/Yen Group Limited). Gross domestic product in billions of USD (in current prices) in the year of the financing round (Source: World Bank; subject code == NGDPD). Gross domestic product per capita in USD (in current prices) in the year of the financing round (Source: World Bank; subject code == NGDPDPC). Annual percentages of constant price GDP are year-on-year changes in the year of the financing round (Source: World Bank; subject code == NGDP_RPCH). Gross domestic product in billions of USD (in current prices) in the year of exit (Source: World Bank; subject code == NGDPD). Gross domestic product per capita in USD (in current prices) in the year of exit (Source: World Bank; subject code == NGDPDPC). Annual percentages of constant price GDP are year-on-year changes in the year of exit (Source: World Bank; subject code == NGDP_RPCH). Individuals using the Internet, in % of population (Source: World Bank) Mobile cellular subscriptions, per 100 people (Source: World Bank)
Dummy variable equal to one if the VC fund is affiliated to a financial institution, and zero otherwise (Source: SDC Platinum VentureXpert; variable fundtypeshort == “FINCORP” or “IBANK”). Dummy variable equal to one if the VC fund is a private limited partnership fund, and zero otherwise (Source: SDC Platinum VentureXpert; variable fundtypeshort == “PRIV”). Dummy variable equal to one if the VC fund is affiliated to a non-financial corporation, and zero otherwise (Source: SDC Platinum VentureXpert; variable fundtypeshort == “CORPVEN”). Age in years of the VC firm at time of the financing round (Source: SDC Platinum VentureXpert). Size of the VC fund in million of USD (Source: SDC Platinum VentureXpert).
Start-up country: United Dummy variable equal to one if the venture is located in the United Kingdom, and zero otherwise (Source: SDC Platinum VentureXpert). Kingdom (dummy) Start-up country: China (dummy) Dummy variable equal to one if the venture is located in China, and zero otherwise (Source: SDC Platinum VentureXpert). Start-up country: France Dummy variable equal to one if the venture is located in France, and zero otherwise (Source: SDC Platinum (dummy) VentureXpert).
Variable
D.J. Cumming, A. Schwienbacher
Algeria Australia Bahrain Belgium Bolivia Brazil Bulgaria Burkina Faso Cambodia Canada Chile China Colombia Costa Rica Cyprus Denmark Egypt Finland France Germany Ghana Greece Guatemala Hungary India
Country (in alphabetical order)
1 35 1 3 1 11 1 1 1 49 1 169 1 1 1 2 2 3 47 4 2 1 1 2 267
No. obs. FinTech deals 8.306 8.808 10.916 5.373 8.006 8.424 8.975 5.182 8.103 5.848 8.038 9.752 9.210 9.903 8.987 10.653 7.090 6.862 8.271 6.878 8.059 4.787 7.653 6.516 9.276
All -10.373 10.916 --8.267 --8.103 4.117 -9.841 9.210 -8.987 10.653 7.090 -8.587 -8.517 --2.534 9.481
Post-crisis
Mean of ln(Round amount)
Table A2.2 Summary statistics of dependent variables, by country
8.306 8.344 -5.373 8.006 9.999 8.975 5.182 -6.349 8.038 9.701 -9.903 ---6.862 7.994 6.878 7.601 4.787 7.653 10.498 8.489
Pre-crisis 1.000 2.257 1.000 1.667 1.000 3.364 1.000 1.000 1.000 3.041 1.000 2.645 1.000 1.000 1.000 2.000 2.000 1.667 3.638 1.000 1.000 1.000 1.000 1.000 3.487
All -3.750 1.000 --3.600 --1.000 3.545 -3.484 1.000 -1.000 2.000 2.000 -2.909 -1.000 --1.000 3.736
Post-crisis
Mean of syndicate size (no. investors)
1.000 1.815 -1.667 1.000 1.000 1.000 1.000 -2.895 1.000 2.159 -1.000 ---1.667 4.280 1.000 1.000 1.000 1.000 1.000 2.527 (Continued)
Pre-crisis
Fintech venture capital
35
Indonesia Ireland Israel Italy Japan Kazakhstan Kenya South Korea Lebanon Luxembourg Malaysia Mexico Mongolia Mozambique The Netherlands New Zealand Nigeria Norway Pakistan Paraguay Poland Portugal Romania Russia Saudi Arabia Sierra Leone
Country (in alphabetical order)
4 26 5 6 19 1 6 33 3 1 2 13 2 4 20 4 3 4 1 1 5 3 4 10 1 3
No. obs. FinTech deals 9.688 8.575 8.130 9.129 5.994 9.210 8.587 7.797 9.122 8.366 9.945 8.784 9.306 6.716 8.024 8.542 8.473 9.111 7.550 8.132 9.311 7.663 5.724 8.507 10.916 9.477
9.903 8.775 8.746 -5.067 9.210 9.192 8.884 8.700 8.366 -8.784 -8.352 9.511 8.233 9.210 -7.550 8.132 10.916 7.663 -8.726 10.916 9.477
9.616 8.549 7.719 9.129 6.667 -7.378 7.727 9.966 -9.945 -9.306 6.171 7.859 9.469 8.105 9.111 --8.241 -5.724 8.361 ---
2.500 3.615 2.200 1.667 1.105 1.000 2.000 3.545 1.667 1.000 2.000 4.385 2.000 1.000 1.300 1.500 1.000 1.500 1.000 1.000 1.400 1.000 1.000 1.200 1.000 3.000
1.000 3.000 1.000 -1.000 1.000 2.000 2.000 2.000 1.000 -4.385 -1.000 1.000 1.667 1.000 -1.000 1.000 2.000 1.000 -1.500 1.000 3.000
Post-crisis
All
Pre-crisis
All
Post-crisis
Mean of syndicate size (no. investors)
Mean of ln(Round amount)
3.000 3.696 3.000 1.667 1.182 -2.000 3.645 1.000 -2.000 -2.000 1.000 1.333 1.000 1.000 1.500 --1.000 -1.000 1.000 ---
Pre-crisis
D.J. Cumming, A. Schwienbacher
36
37
9 1 3 4 9 4 1 3 165 1682 1 4
10.536 8.517 6.494 7.778 8.008 5.960 9.798 5.478 8.862 8.465 8.144 9.219
7.496 -8.248 -7.350 5.960 -7.223 8.311 8.600 8.144 9.116
10.916 8.517 5.617 7.778 10.309 -9.798 4.605 9.169 8.426 -9.254
4.556 1.000 1.000 1.000 3.444 4.000 1.000 1.667 3.788 3.229 1.000 1.000
1.000 -1.000 -3.857 4.000 -1.000 4.797 2.842 1.000 1.000
5.000 1.000 1.000 1.000 2.000 -1.000 2.000 3.226 3.343 -1.000
Note: This table presents sample means of the two main dependent variables (ln(Round amount) and Syndicate size (no. investors)) of all FinTech deals, by country of origin. The sample used here is restricted to the countries with at least one FinTech deal. Statistics are reported for all FinTech deals, and for the subsamples of FinTech deals after (Post-crisis = 1) and before (Post-crisis = 0) the financial crisis. “No. obs.” gives the number of FinTech deals in each country in the sample.
Singapore South Africa Spain Sweden Switzerland Thailand Togo Ukraine United Kingdom United States Uzbekistan Vietnam
Fintech venture capital
3 THE FUTURE OF FINANCE Why regulation matters Mark Fenwick and Erik P.M. Vermeulen
The FinTech revolution Over the last decade, FinTech – broadly defined as the use of new technology and innovation to compete in the marketplace of financial institutions and intermediaries – has disrupted the financial services sector in several ways (Chishti and Barberis, 2016; King, 2018; McMillan, 2014; Sironi, 2016). First, new technologies have allowed incumbent financial service providers to offer a range of new services that remove intermediaries in order to make transactions more effective and less prone to error (Haycock and Richmond, 2015). In this way, financial services are decentralized and made flatter. Most obviously, there is the growth of so-called mobile banking that allows customers to perform a wide range of transactions online. Networked access to financial services facilitates quicker access to all manner of transactions from checking financial status, making payments, and withdrawing and transferring funds. “Behind the scenes” activities of financial institutions are similarly transformed. This involves, for example, the use of Big Data to deliver a more efficient service, but it also allows firms to use technology to manage legal risks more effectively. The fallout from the 2008–9 Financial Crisis resulted in vast swaths of new banking regulation (Sironi, 2018). One effect of this additional regulatory burden has been the increased use of technology to help banks comply with new regulatory requirements. Sometimes referred to as “Regtech,” this involves using technology to comply with regulatory requirements (Arner et al., 2017). There are several areas of compliance and reporting where technology can have a significant benefit, such as anti-money laundering requirements (e.g., know your customer requirements), risk data aggregation, and real-time transaction monitoring. Second, FinTech has also facilitated the emergence of tech start-ups that offer an alternative source of financial services (Pratskevich, 2019). In particular, “app-based” companies are emerging everywhere. They challenge and disrupt incumbents, such as traditional banks, by supporting a range of financial services, for example, marketplace lending platforms, equity crowdfunding platforms, insurance services, algorithm-driven “robo-advisors” offering smarter, more personalized financial advice, as well as blockchain-based crypto-currency and payment systems. 38
The future of finance
For Millennial consumers, in particular, these alternative service providers (“challenger banks”) are attractive (West, 2018). Banks have traditionally failed to respond to the perception that banks are untrustworthy, profit-driven machines, associated with a selfish and unsustainable version of capitalism. If traditional financial institutions don’t meet these needs, then Millennial consumers simply move to younger, new providers that can. Finally, FinTech leverages technology to improve access to financial services for individuals that have traditionally been excluded, in particular in emerging economies (Blakstad, 2018; Buckley and Webster, 2016; Realini and Mehta, 2015). Driving this change is the global proliferation of smartphones. Smartphone penetration is expanding quickly around the world, with 6.1 billion users expected by 2020. Many start-ups are now leveraging that global reach and providing access to various financial services (most obviously, credit) in markets in Africa, South America, and South-East Asia. The range of services that are offered is expanding, as locally based start-ups proliferate.
The regulatory trilemma From even the briefest of surveys, it is evident that FinTech is disrupting every aspect of financial services. The FinTech revolution has proven enormously disruptive for two groups of actors, in particular, namely incumbent financial service providers and regulators and other policymakers. Incumbents face new and aggressive competition from young, agile start-ups that leverage digital technologies to deliver a smoother, more customer-focused experience. However, incumbents also face competition from larger, well-established technology companies that see opportunities in the financial sector. As such, the traditional silos between financial service companies, technology companies, and media and telecommunications companies have broken down as lines between financial services firms and other types of business become blurred (Price Waterhouse Cooper, 2014). This disruptive competition exposes inadequacies in traditional business models and practices and has compelled incumbents to innovate. The arrival of these two groups of non-traditional actors into the financial services sector is a big part of the disruption for policymakers. Regulators and other policymakers face various new challenges in designing and implementing a regulatory response to the technologydriven changes in financial services. As technology “eats the world,” it creates a plethora of new business opportunities, but it also creates tremendous challenges that require some form of state regulatory intervention. In a world where agility is essential, and “technology is faster than the law,” governments are often sluggish and disconnected (Fenwick, Kaal, and Vermeulen, 2017). This creates potential new risks, most obviously for the consumers of financial services, but also for the integrity of the financial system as a whole. The structural importance of financial services, particularly banks, in the operation of a developed capitalist economy has traditionally justified high levels of state intervention and regulation to ensure that banks and related actors do not undertake excessive risk. Historically, banks have been regarded as a unique form of business. On the one hand, they should be run privately for profit. On the other hand, however, they also perform a public utility-type function in that they provide credit, and this credit is vital to the health of the economy. A sophisticated regulatory system emerged to manage this risk and balance these different goals. In the context of FinTech, however, the risks are often uncertain or unknown, fall outside existing regulatory schemes, or both. 39
Mark Fenwick, Erik P.M. Vermeulen
Writing in the 1980s, the legal theorist Gunther Teubner famously identified a “regulatory trilemma” facing all regulators in late capitalism (Teubner, 1986). Teubner argued that any regulatory intervention faces three types of risk: the risk of the regulatory action not working, (i.e., the regulation misses the target or is otherwise ineffective); the risk of breaking the thing it seeks to regulate, (i.e., the regulation removes any incentive to engage in the activity that is being regulated); and the risk of undermining the law (i.e., the regulation undermines the doctrinal integrity of the law and legal system, more generally). The “trilemma” that Teubner described can be re-formulated as a question: How can we ensure that any regulatory intervention is effective, responsive, and legally coherent? This seems to be a particularly pressing problem in the context of FinTech. It is essential that any regulatory action is effective – that consumer interests and the integrity of the financial system as a whole are adequately protected. Moreover, it is vital that regulation isn’t overly burdensome and “kills” innovation, leading to an exodus of start-ups and talent to other jurisdictions that offer a “friendlier” regulatory environment. In a global economy where the transaction costs of relocating a business are reduced, regulatory competition is an important consideration. Finally, regulation needs to be consistent with other features of the legal system and the existing capacities, “know-how,” and experience of regulators. In developing answers to this regulatory trilemma in a FinTech context, it is important to acknowledge that state actors operate at a significant informational disadvantage (particularly when compared with the disruptive FinTech firms and large technology companies) and lack the capacities, resources and experience to keep up with the fast-moving actors that dominate the sector. Under such conditions of information and resource disadvantages, new and more innovative approaches need to be found. The chapter describes several such approaches and the different considerations that inform such approaches. Regulation matters, but it is only by embracing new approaches that a regulatory environment can be developed that fosters the responsible and safe deployment of financial innovation. Bad regulation seems likely to kill innovation and place a country at a significant economic disadvantage.
The sustainable financial service ecosystem of the future One way of approaching the regulation of FinTech is to think about the desired end state: What would we like the sector to look like in the future? What organizational structures seem most likely to deliver sustained and responsible innovation, and, based on this desired end state, what regulatory approach seems most likely to facilitate and encourage such businesses? What seems apparent is that incumbent providers cannot ignore the disruption. In a financial services context, incumbents have found themselves confronted with an unprecedented combination of new pressures as a result of this shift. Crucially, all of this disruption involves technology at some level. These new challenges include developing more customerfriendly services to attract more customers and deepen relationships with existing customers to retain them; rethinking distribution models and internal organization; responding to disruptive competition from “challenger” banks and new entrants to the market (start-ups but also corporations from other sectors, most obviously the technology sector); rebuilding trust with all stakeholders, especially customers; and managing new regulatory, capital, and security risks (Price Waterhouse Cooper, 2014). The crucial point here is that all of these challenges require engagement with digital technologies. Technology facilitates the delivery of better services and forms the crucial infrastructure for managing costs and risks. As such, 40
The future of finance
digital technologies have now become the primary engine of change in every aspect of financial services. Organizing-for-innovation is no longer optional. So, how then can incumbents respond to this new challenge? In other work, we have developed the argument that the most innovative companies in the world have responded to the unprecedented pressures and challenges of doing business in a digital age by reinventing themselves as more open, inclusive and “flatter” ecosystems (Fenwick and Vermeulen, 2015; Fenwick and Vermeulen, 2019a; Fenwick and Vermeulen, 2019b). The suggestion is that the closed, hierarchical, modern company – which has dominated the global economy for the last two centuries– is ill-equipped to respond to the challenges required by the FinTech revolution. We are living through the beginning of the “end of the corporation,” at least companies organized as closed, hierarchical systems that operate as proceduralized bureaucracies. The most well-known theory for explaining the failure of large corporations to innovate is Clayton Christensen’s The Innovator’s Dilemma (1997). Christensen’s argument was that, over time, all organizations inevitably develop habits and procedures for making decisions and allocating resources. In larger organizations – like modern corporations – such systems tend to be highly formalized. The result? Corporations get locked into decision-making and resource allocation models that focus on existing products and services. When they see something new – even if they have a strong sense that it will disrupt their industry – they are too focused on existing products or services to adapt. Corporations thus have an inherent tendency towards tunnel vision in which, by looking to satisfy their existing customers, they fail to notice how the world is quickly changing around them. This was Christensen’s key insights and the basis of the innovator’s dilemma. A corporation that had been innovative once often struggles to innovate the next time. Corporations may be very good at what they do, but that focused excellence is what kills them. It is not that bureaucratic procedures are inherently wrong – they may be an effective mechanism for serving existing customers and managing complexity in a large, possibly transnational organization – but such practices push companies to keep doing what they have done before. But that maintenance of the status quo leaves them open to disruption from more innovative rivals. To respond to this dilemma, new ways of organizing business have developed, and understanding the unique features of these alternative business forms and thinking about how to design a regulatory environment to facilitate these new ways of operating a business have become crucial tasks for all businesses, as well as policymakers. As such, we should no longer think in terms of traditional corporate structures. Company boundaries have become more open. Traditional corporate organizations with their fixed roles, static procedures, closed departments, and hierarchical relationships between different groups of stakeholders are all changing as companies adapt to a new operating environment. To make sense of this change in how firms organize themselves in a digital age, the concept of a business “ecosystem” can provide an alternative. In brief, such ecosystems combine the following features (Fenwick and Vermeulen, 2015): • •
•
Leveraging the unique characteristics of software technologies (e.g., low marginal costs) to deliver a powerful, frictionless experience to end-users. Adopting a flatter, fluid, and more inclusive style of organization built around networks of unbundled, high-performance, creative teams in which job roles and functions are evolving dynamically in response to the evolving business needs of the firm. Embracing a more open, transparent approach to communication and information management that relies on new computer-mediated communications, such as social media. 41
Mark Fenwick, Erik P.M. Vermeulen
•
•
Implementing a new style of digital leadership that focuses on creating an environment that facilitates creativity rather than an exclusive focus on supervising compliance or managing legal risk. Utilizing open collaboration with all stakeholders, especially multiple external partners, to feed the requirement of continually innovating – what we would term partnering-for-innovation.
In an age of hyper-competitive, technology-driven markets, every company needs to consider reinventing itself as an ecosystem of this kind. Such ecosystems are better placed to deliver that kind of innovation necessary to succeed in a technology-driven economy.
Partnering-for-innovation An important option for incumbent financial service providers – particularly for more conservative incumbents that struggle with “intra-preneurship” (i.e., internal innovation) – is to engage in open collaboration with external partners (particularly tech start-ups) to develop and reinvigorate their services. The best companies realize that their future will be determined by developments in technology and that “learning” from technology start-ups is often the most effective way to achieve this objective, especially when incumbents lack the capacities for technology-driven innovation. One way to achieve this is for incumbents to acquire or invest in start-ups, i.e., corporate venturing (Fenwick and Vermeulen 2016). The crucial point is that incumbent providers must be open to the possibility of obtaining knowledge and technology from an acquired start-up firm. The aim of such acquisitions is not assimilation, in that a start-up is simply absorbed into a larger corporate identity. Instead, the objective of a more open style of partnering is a dynamic relationship in which opportunities for mutual learning are emphasized. It is in this sense that we can talk (as was done in the Financial Times) of incumbent providers “borrowing the Start-up Genie’s Magic” (Newton, 2015). This contrasts with an earlier style of corporate acquisition in which assimilation was emphasized, and any learning was conceptualized as one-way (i.e., from corporate to the acquired entity). An open, inclusive, and fluid ecosystem-style organization can help larger, established firms in meeting the complex (and unprecedented) business challenges of delivering a different kind of financial service. This new style of partnering can be beneficial in the case of financial institutions as they look to respond to the disruptive challenge of FinTech. Banks, for instance, already engage in such partnering-for-innovation, and examples can be seen everywhere (Macheel, 2019). For example, many banks have now established partnerships with FinTech companies. For example, J.P. Morgan Chase, has partnered with OnDeck to offer fast approval and funding of small business loans. Another FinTech company, Prime Revenue, provides supply chain finance through a cloud-enabled platform to banks, including Barclays. Spain’s second-largest bank BBVA has actively engaged in FinTech acquisition. They became a major shareholder in British start-up Atom Bank and acquired Holbi, a Finnish based FinTech innovator, specializing in small business payments. Visa’s acquisition of Plaid can be understood similarly. Many banks now see “FinTech partnering” of this kind as a core competency that they need to develop to stay relevant and competitive (CB Insights, 2019). These developments have led some observers to talk of a “great new era of FinTech partnerships” between incumbents and more innovative start-ups (Tweddle, 2018). According to 42
The future of finance
this line of thinking, banks need to become consumers of innovative FinTech because they cannot deliver such services by themselves. On the other hand, FinTech start-ups are often narrowly focusing on a specific problem or issue and the development of one particular solution. This creates a potential “win-win” in which larger banks benefit from the specific new product or service developed by the start-up, and the start-up can benefit from the reach, network, and infrastructure of the incumbent bank. Nevertheless, for such partnering-for-innovation to work effectively, incumbent banks need to re-evaluate existing practices. For example, they need to strengthen their mechanisms for assessing their current internal capabilities, develop robust systems for evaluating potential partners, devise mutually acceptable financial arrangements, and ensure adequate testing capabilities for deploying new technologies (both, initially, on a small, experimental scale and later, when full-scaled implementation is planned, on a larger scale). This is not always easy for incumbents, and there are skeptical voices about the feasibility of such an approach (Shevlin, 2019). The enormous cultural differences between incumbents and start-ups lead some observers to conclude that it is difficult for incumbent banks to partner-for-innovation in this way. According to this line of thinking, the challenges of the external environment are too significant, the expectations of consumers are too high, and the banks do not – as things stand – have the internal capacities or resources to implement this new partnering-for-innovation and ecosystem style organization (Beaumont, 2018). Nevertheless, such skepticism about the kind of partnering outlined here seems to push against the approach of more and more incumbents in the sector. More and “better” partnering, rather than less, appears to be inevitable, given the trend towards the “unbundling” of banking that we will discuss below. Of course, this raises profound challenges for incumbent financial service providers that have become accustomed to working entirely “in-house” using settled internal procedures. And, for sure, such partnering in an open ecosystem means giving up a certain degree of control. However, the benefits in the long term of such partnering, sharing of know-how, and co-creation seems to justify any risks. Moreover, it is hard to see a better alternative for incumbent financial service providers than operating as a more open and inclusive ecosystem. Most obviously, it maximizes opportunities for incumbents to innovate and ensures that such innovation is “hard-wired” into the organization and, over time, its culture.
Re-imagining the role of government Traditional accounts of the role of, and justification for, regulation in financial services have focused on risk management. A crucial difference between financial service providers and other businesses is the regulatory environment in which they now operate. Moreover, the level of regulation is much higher for banks, particularly post-2008. This complicates efforts to borrow some of the FinTech “genie’s magic,” and a shift in approach by regulators may also be required, particularly if the goal is delivering more innovative products and services. Two considerations have dominated the post-2008 regulatory environment for financial service providers: first, ensuring a higher degree of consumer protection, particularly for retail clients, investors, and depositors (i.e., the micro-prudential aspect of regulation) and second, ensuring financial stability by minimizing systemic risk (the macro-prudential part of regulation). The 2008–9 Financial Crisis exposed shortcomings across both dimensions, and these failures triggered a significant process of regulatory reform and the imposition of stricter regulatory requirements. 43
Mark Fenwick, Erik P.M. Vermeulen
Moreover, a legacy of the 2008 crisis has been a shift in perceptions of innovation, at least on the side of regulators. Before 2008, innovations in financial products or services were generally perceived in favorable terms. This perception resulted in a “light-touch” approach to the regulation of innovations in financial services. However, since the crisis came to be blamed, in large part, on such innovation (so-called “financial weapons of mass destruction”), the regulatory trend shifted in the opposite direction (Tett, 2010). Innovation came to be seen in more negative terms by policymakers (not to mention the public), who were keen to avoid any repetition of the disruption of 2008–10. The timing of the emergence of FinTech has, therefore, proven enormously challenging for regulators (Zetzsche, 2017). Regulators have been placed in the awkward position of having to balance the post-2008 regulatory objectives of consumer protection and managing systemic risk with the promotion of innovation. From the perspective of regulators, it is easy to conclude that FinTech creates both micro- and macro-prudential risks or, at least, uncertainties, and managing these uncertainties is highly complex. However, as memories of the last financial crisis fade, the relationship between banks and regulators has shifted to a new stage. As discussed above, most financial institutions already engage in the proactive management of regulatory risk through expanded compliance departments (Fenwick, 2016). Banks have better integrated the two post-crisis objectives of regulation into their daily operations, and, in consequence, the regulatory agenda has shifted. Against the background of these changes, new regulatory approaches are now possible and desirable.
Open banking One clear example of a shift in regulatory thinking is the post-2016 UK experience. Traditionally, five big banks – Barclays, HSBC, Lloyds, Santander, and Royal Bank of Scotland – controlled over 80 percent of the retail current account market, offering near-identical products that remained unchanged for several decades. People would pick a bank, typically when they entered the labor market and would stay with them for life. However, in August 2016, the UK Competition and Markets Authority issued a ruling ordering to the nine biggest UK banks to allow licensed start-ups direct access to their data (see www.openbanking.org.uk). Account-holders needed to provide consent, but if they did, then all data in their current bank accounts – for example, utility bills, mortgage payments, etc. – could be made available to FinTech start-ups, who could then utilize that data to deliver innovative new financial products and services. To do this, Open Banking Limited, a non-profit organization, was set up to develop application program interfaces (APIs). These protocols transfer data automatically from one piece of software to another (Camerinelli, 2017). What makes these APIs potentially game-changing is that they can take the current account data and let software developers create new products that use this data in new ways. A simple example would be an app that collected an individual’s financial information together from several sources – several different bank accounts, for instance, and allowed that individual to manage their financial affairs from one app on their phone. The ability to access data on multiple bank accounts might not seem immediately game-changing. However, the thought behind Open Banking is that start-ups will take the data and develop innovative new services that no one has yet thought of. The expectation of the Open Banking movement is that innovative entrepreneur-founders will leverage the potential of this data to deliver more innovation. 44
The future of finance
The European Union introduced a similar set of financial services reforms in the Payment Services Directive 2 (PSD2) in 2015 and which will come into full effect in late 2020 (Price Waterhouse Cooper, 2016). The aim of the PSD2 was to develop the European single market in banking by forcing European banks to open up their data via APIs. PSD2 created two new types of licensed entities that can use this data, either for payments or other services. This created a unique opportunity for non-bank organizations to provide payment initiation and account information services, creating greater competition, and more choice for consumers. Security is ensured under PSD2 by the introduction of Strong Customer Authentication (SCA), which requires that when customers access their payment accounts online, two of three mandatory authentication measures must be used, so-called two-factor authentication (or 2FA). These SCA measures are: Knowledge, something only the user knows (e.g., a password or a PIN); Possession, something only the user possesses (e.g., a token, code, or key); Inherence, something the user “is” (e.g., a fingerprint, biometric, or voice.) Other countries are monitoring these trends towards open banking. Japan’s Banking Act, for example, was amended in June 2018 to promote open banking. Around 130 chartered banks in Japan have plans to open-up APIs by the end of 2020 (Creehan and Tierno, 2019). Incumbent banks were initially skeptical of these developments. For example, it was reported that slightly over 40% of European banks failed to meet the PSD2 deadline in March 2019 – the banks were supposed to provide a testing environment to the third-party providers. However, the smarter banks recognized the value of partnering with FinTech firms in the way described in the previous section to minimize the business and regulatory risk of the new world of open banking. As such, the PSD2 represents a critical case study in how regulatory interventions might “nudge” incumbents into partnering-for-innovation and creating the open and inclusive financial service ecosystem of the future.
Co-creation A related option – often preferred by the FinTech companies themselves – is for the government to give much higher weight to those actors that are driving technological innovation and its dissemination, namely the technology companies, in designing the regulatory framework. Stated slightly differently, the technology companies believe that to make a partneringfor-innovation strategy work, regulators need to become more active players in the open ecosystems described above. However, is this a sensible strategy? Or is it a case of putting the animals in charge of the zoo? There is something to the idea that governments should outsource to companies the task of designing regulatory policies suitable for a digital age (Kaal and Vermeulen, 2017; Malan, 2018). Disruption has become one of the main issues for any business, markets are changing fast, and new competitors enter the stage all the time. Under such circumstances, business models must continuously evolve. As a result, companies are obliged to take emerging technology seriously to remain relevant. One effect of such an environment is that technology companies have greater access to better information about the impact of technology. Increasingly, companies are better equipped than states to take a leading role in the design of regulation. Moreover, new digital technologies empower the customers and employees of such companies in new ways. The voice of these stakeholders must be considered for such companies to survive. For instance, in many cases, employees are no longer satisfied with being “cogs” in a corporate machine but want to be treated as active stakeholders. In the context of the Gig Economy, for instance, such employees have become entrepreneurs themselves. 45
Mark Fenwick, Erik P.M. Vermeulen
They will speak up or, in a worse case, “exit” when they do not support a company’s policies or actions (Kessler, 2018). Advocates of this type of outsourcing see such stakeholders as an essential check on how technology companies behave and how they approach the issue of designing a regulatory framework. However, what is perhaps even more important is that the consumers of tech products and services have become much more critical, at least compared to an earlier industrial phase of capitalism. In technology-driven markets, consumers are not just consumers anymore. They have become an essential stakeholder in the ecosystem of the firm and its governance. This functions as a constraint on the behavior of larger firms. It is becoming more dangerous for them to abuse their market power, as such abuse will risk user migration to rivals and, in the medium-long term, damage to the brand and a decline in a firm’s fortunes (Fenwick and Vermeulen, 2019a). Such risks are particularly acute for firms that operate a platform as a crucial part of their business model (think Amazon, Airbnb, Facebook, or Uber) because platforms are dependent on the network effects created by having as many users as possible, and their business will suffer if users desert the platform (Reddy, 2018). Of course, there are risks. Even if technology companies have good intentions, they may have difficulties proposing effective regulatory schemes because their interests are not wellaligned within the company. A recent example is Google’s failed attempt to set up an ethics council to examine developments in artificial intelligence (Levin, 2019). So, what is the role of government in a digital age? If the delegation of policymaking and regulation to the private sector might lead to the “capturing” of the policy process and policy by established tech companies, what would be a better alternative? The government still has an essential role to play. However, a bureaucrat-led approach to policymaking has had its time. A new, more dynamic approach that is responsive to the need for innovation must be implemented.
Policy experimentation One option that is being utilized in the context of FinTech regulation is to place greater emphasis on policy experimentation. Here, we do not refer to (rather traditional) consultation models in which the market is invited to respond to and provide limited input to policy and regulatory proposals, but more radical approaches that emphasize the testing of innovation in real-world settings in order to collect data that can then inform regulatory design (Bennet-Moses, 2013; Black, 2012). For example, in April 2016, the UK Financial Conduct Authority (FCA) announced the introduction of a “regulatory sandbox,” which allows both start-ups and established companies to roll out and test new ideas, products, and business models in the area of FinTech (Financial Conduct Authority, 2018). This model has proved very influential, especially in an Asian context where a number of countries have implemented similar schemes (Fenwick et al., 2020). The aim of this sandbox is to create a “safe space” in which businesses can test innovative products, services, business models, and delivery mechanisms without immediately incurring the normal regulatory consequences. In practice, this means that relevant rules and regulations are temporarily suspended and do not apply to a particular firm. With the sandbox, the regulator aims to foster innovation by lowering regulatory barriers for testing disruptive innovative technologies, while ensuring that consumers will not be negatively affected. In exchange, regulators are given access to the most cutting-edge data, thus closing the informational asymmetries discussed above. 46
The future of finance
What makes the regulatory sandbox attractive is that, insofar as technology has consequences that flow into the everyday lives of citizens, such technology will be open to discussion, supervision, and control. In this way, public involvement in regulatory debates can help to create a better sense of legitimacy. Arguably, the most significant tangible benefits to sandbox firms are the contacts formed with the regulator and the market credibility that participation in the regulator’s sandbox gives them vis-à-vis customers and financiers. Skeptics argue that regulatory sandboxes create a “two-tier” system of start-ups, where those that are selected in the sandbox are given an unfair advantage over rivals, including incumbents. The credibility gains, as well as the lighter regulatory requirements, could undoubtedly prove advantageous. Moreover, one could also question if regulators have the necessary capacities to determine whether a business should be included within a sandbox scheme or not. Indeed, such judgments would seem to presuppose that regulators engage in a more open dialogue and engagement with a broader group of start-ups operating in the FinTech space. This could be achieved through more regulatory dialogue, for example, “innovation hubs” set up by competent authorities to enable firms to engage with the authorities on FinTech-related issues and seek clarification on licensing and regulatory requirements. Regulatory sandboxes have been adopted or considered in many other jurisdictions and data-driven regulatory design, in a broader sense, is an increasingly popular approach. As different countries compete to attract innovative start-ups, the issue of the regulatory environment becomes increasingly important. After all, the regulatory situation will be a crucial consideration for any firm when deciding its base. In a technology-driven, global economy, those jurisdictions that fail to engage with new technologies and don’t put in place rules and regulations that are attractive to founder-innovators risk being left behind. As such, we must also introduce ecosystem thinking into regulation, and companies, banks, start-ups, and governments must all work together in partnership with the various stakeholders to ensure that vital interests are protected while facilitating innovation. There is already some evidence of such a shift. Regulators acknowledge their informational disadvantages and are participating in more events – training courses or hackathons – to close this gap. Moreover, regulators are building new collaborative relationships with actors in the private sector to understand and develop technologies. There is greater outsourcing of legal work and cooperation with public–private partnering to create new technologies, such as blockchain. Finally, the introduction of sandboxes and recognition of the importance of innovation suggests a growing awareness of the need for a different kind of approach. Of course, national governments and other regulators must set “smart” boundaries for the risk they are willing to take that are agreed with regulated entities. However, within these boundaries, they must allow and encourage freedom and innovation. This does not mean that within these boundaries, a free-for-all should be allowed. Instead, within carefully negotiated limits, it is all about building and maintaining trust amongst all participants via constant dialogue and sharing of information. In this respect, trust must be earned from all the stakeholders that are involved and affected by new technologies. As such, community-driven regulatory design is a version of policy experimentation. The crucial factor here is the changing context. In the context of the digital revolution and the new pressures it has created, there is a unique degree of openness and visibility both within society in general and within the emerging ecosystems. The check on regulatory capture is the new visibility that digital technologies have created and the dependency that 47
Mark Fenwick, Erik P.M. Vermeulen
open ecosystems have on remaining committed to the values of a free, open digital culture. If the check on power is visibility, transparency, and the demand for authenticity, then the key to ensuring these values are maintained are those infrastructures that facilitate speech, for example, social media.
Conclusion The government needs to take on a vital role in the development of the financial service ecosystems of the future. As such, they can help establish the trust that is necessary for such ecosystems to flourish. However, this means that everyone in the government needs to embrace “going digital.” Regulators and other policymakers need to think more about the meaning of technologies, what they can do for us, and how they can help us to build a better future. Doing nothing or restricting innovation are worse options. This goal of fostering innovation often means rejecting and replacing old, formalized ways of doing things, such as hierarchies, legacy processes, and settled procedures. Instead, “going digital” will lead to looser connections and relationships, and more flexible forms of organization and operation. As such, regulators must re-learn what it means to interact, transact, and become visible in a digital environment. They must build their brand, and government officials must learn how to think more like entrepreneurs. Being creative and innovative in this way will ensure that “going digital” creates more opportunities than it eliminates for everyone. And this includes incumbent financial institutions and the new FinTech firms that are innovating in the sector. After all, there is no turning back. There are good reasons to believe that the impact of next-generation technologies – particularly developments in artificial intelligence and automation – will be more significant than what we have experienced already. Digital technologies will clearly play a central role in the financial services of the future, but for these innovations to be successful requires government to become more “tech-savvy” and do more to enable everyone to engage with technology in a socially responsible manner.
References Arner, D.W., Zetzsche, D.A., Buckley, R.P., and Barberis, J.N. (2017). Fintech and regtech: enabling innovation while preserving financial stability. Georgetown Journal of International Affairs, 18: 47–58. Beaumont, L. (2018). Banks can’t partner themselves into digital relevancy. American Banker (26 June). https://www.americanbanker.com/opinion/banks-cant-partner-themselves-into-digitalrelevancy. Accessed 20 August 2020. Bennett Moses, L. (2013). How to think about law, regulation and technology: problems with ‘technology’ as a regulatory target. Law, Innovation and Technology, 5: 1–20. Black, J. (2012). Paradoxes and failures: new governance techniques and the financial crisis. Modern Law Review, 75: 1037–63. Blakstad, S (2018). Fintech Revolution: Universal Inclusion in the New Financial Ecosystem. London: Palgrave Macmillan. Buckley, R.P. and Webster, S. (2016). FinTech in developing countries: charting new customer journeys. Journal of Financial Transformation, 44: 151–9. Camerinelli, E. (2017). Open bank, APIs, and financial services ecosystems: the future of banking. https://www.softwareag.com/corporate/images/Aite_Group_Open_bank_api_and_financial_ services_Research_ Jul17_tcm389-167453.pdf. Accessed 20 August 2020. CBInsights. (2019). Where top U.S. banks are betting on fintech. CBInsights https://www.cbinsights. com/research/fintech-investments-top-us-banks/. Accessed 20 February 2020. Chishti, S, and Barberis, J.N. (2016). The FINTECH Book: The Financial Technology Handbook for Investors, Entrepreneurs and Visionaries. Chichester: Wiley. Christensen, C. (1997). The Innovator’s Dilemma. Cambridge, MA: Harvard Business Review Press.
48
The future of finance Creehan, S. and Tierno, P. (2019). The slow introduction of open banking and APIs in Japan. https:// www.frbsf.org/banking/asia-program/pacif ic-exchanges-podcast/open-banking-apis-japan/. Accessed 20 August 2020. Fenwick, M. (2016). The new corporate criminal law and transnational legal risk. In Flexibility in Modern Business Law: A Comparative Assessment. Singapore: Springer. 149–71. Fenwick, M. and Vermeulen, E.P.M. (2015). The ‘New Firm’. European Business Organization Law Review, 16: 595–623. Fenwick, M. and Vermeulen, E.P.M. (2016). Seven corporate venturing strategies to foster innovation. TILEC Discussion Paper No. 2016–23. https://papers.ssrn.com/sol3/papers.cfm?abstract_ id=2823008. Accessed 20 August 2020. Fenwick, M. and Vermeulen, E.P.M. (2019a). A sustainable platform economy and the future of corporate governance. 441 European Corporate Governance Institute – Law Working Paper. https://ssrn. com/abstract=3331508. Accessed 20 August 2020. Fenwick, M. and Vermeulen, E.P.M. (2019b). The end of the corporation. European Corporate Governance Institute – Law Working Paper. https://ssrn.com/abstract=3472601. Accessed 20 August 2020. Fenwick, M., Kaal, W.A., and Vermeulen E.P.M. (2017). Regulation tomorrow: what happens when technology is faster than the law. American University Business Law Review, 6: 561–94. Fenwick, M., Van Uytsel, S. and Bi Ying. (2020). Regulating Fintech in Asia: Global Context, Local Perspectives. Singapore: Springer. Financial Conduct Authority. (2018). Crypto Assets Taskforce: Final Report. London: FCA. Haycock, J. and Richmond, S. (2015). Bye Bye Banks?: How Retail Banks are Being Displaced, Diminished and Disintermediated by Tech Startups and What They Can Do to Survive. London: Adaptive Lab. Kaal, W.A. and Vermeulen, E.P.M. (2017). How to regulate disruptive innovation: from facts to data. Jurimetrics: The Journal of Law, Science, and Technology, 57: 169–209. Kessler, K. (2018). Gigged: The End of the Job and the Future of Work. New York: St. Martin’s Press. King, B. (2018). Bank 4.0: Banking Everywhere, Never at a Bank. Chichester: Wiley. Levin, S. (2019). Google scraps AI ethics council after backlash: back to the drawing board. The Guardian. (4 April) https://www.theguardian.com/technology/2019/apr/04/google-ai-ethics-councilbacklash. Accessed 20 August 2020. Macheel, T. (2019). Banks are turning branches into coworking spaces. Tearsheet. https://tearsheet. co/modern-banking-experience/banks-are-turning-branches-into-coworking-spaces/. Accessed 20 August 2020. Malan, D. (2018). The law can’t keep up with new tech. Here’s how to close the gap. World Economic Forum. https://www.weforum.org/agenda/2018/06/law-too-slow-for-new-tech-how-keep-up/. Accessed 20 August 2020. McMillan, J. (2014). The End of Banking: Money, Credit, and the Digital Revolution. Zurich: Zero/One Economics. Newton, R. (2015). How to borrow some of the start-up genie’s magic. Financial Times (28 July). https://www.ft.com/content/6b5e4ba6-ffcc-11e4-bc30–00144feabdc0. Accessed 20 August 2020. Pratskevich, A. (2019). New! 2019 Financial Apps Report. Liftoff. https://liftoff.io/blog/financeapps-2019/. Accessed 20 August 2020. Price Waterhouse Cooper (2014). Retail Banking 2020: Evolution or Revolution. https://www.pwc. com/gx/en/banking-capital-markets/banking-2020/assets/pwc-retail-banking-2020-evolutionor-revolution.pdf. Accessed 20 August 2020. Price Waterhouse Cooper (2016). PSD2 in a Nutshell. https://www.pwc.com/cz/en/bankovnictvi/ assets/psd2-nutshell-n01-en.pdf. Accessed 20 August 2020. Realini, C. and Mehta, K. (2015). Financial Inclusion at the Bottom of the Pyramid. Victoria: Friesen Press. Reddy, N. (2018). How to harness the power of network effects. Forbes (2 January) https://www. forbes.com/sites/forbescoachescouncil/2018/01/02/how-to-harness-the-power-of-networkeffects/#2b41823462e8. Accessed 20 August 2020. Shevlin, R. (2019). Bank/fintech partnerships will be a huge disappointment in 2019. Forbes (7 January). https://www.forbes.com/sites/ronshevlin/2019/01/07/ bankfintech-partnershipswill-be-a-huge-disappointment-in-2019/#3dc33dc837ba. Accessed 20 August 2020. Sironi, A. (2018). The evolution of banking regulation since the financial crisis: a critical assessment. https://ssrn.com/abstract=3304672. Accessed 20 August 2020. Sironi, P. (2016). FinTech Innovation: From Robo-Advisors to Goal Based Investing and Gamification. Chichester: Wiley.
49
Mark Fenwick, Erik P.M. Vermeulen Teubner, G. (1986). Dilemmas of Law in the Welfare State. London: De Gruyter. Tett, G. (2010). Fool’s Gold. New York: Free Press. Tweddle, K. (2018). Doomed? Fintech partnerships are vital. American Banker (10 July). https://www. americanbanker.com/opinion/doomed-fintech-partnerships-are-vital. Accessed 20 August 2020. West, S. (2018). Meeting Millennial expectations in these four areas of technology. Forbes. https:// www.forbes.com/sites/forbestechcouncil/2018/06/28/meeting-millennial-expectationsin-these-four-areas-of-technology/#3e7c9b244ffc. Accessed 20 August 2020. Zetzsche, D.A. (2017). Regulating a revolution: from regulatory sandboxes to smart regulation. Fordham Journal of Corporate and Financial Law, 23: 31–103.
50
4 DIGITAL CURRENCIES What role in our financial system? Grégory Claeys and Maria Demertzis
1. Introduction Under the Bretton Woods monetary arrangement put in place in 1947, the main global currencies were anchored to the US dollar (through a fixed exchange rate) and were, at least partially, convertible with gold. This system broke down in 1971 when US President Richard Nixon declared a temporary suspension of the dollar’s convertibility into gold. Since then, monetary systems in most developed countries have been based on fiat currencies, in other words, currencies that are not backed by physical assets but that rely on the ability of monetary authorities to ensure the currency’s stability. These currencies are issued by central banks in the form of (physical) coins and banknotes and (dematerialised) reserves, combined with highly regulated (dematerialised) bank deposits convertible at par with central bank money. The fiat-based monetary system has functioned in this form since the demise of Bretton Woods, with only minor innovations. However, there have been four major developments in the last decade that have challenged and continue to challenge the status quo and have reopened the debate on the forms that money will take in the future. 1.
2.
3.
The share of transactions in cash in developed countries has fallen. In countries such as Sweden, coins and banknotes have become so marginalised as a means of payment that there is even talk of abandoning them completely. The Riksbank, the Swedish Central Bank, has opted against the total elimination of cash, but there is unequivocally a trend towards less cash usage in some countries. The emergence of distributed-ledger technology, or blockchain (i.e. a decentralised, secure and unalterable record of financial transactions), has enabled the appearance of thousands of cryptocurrencies, such as Bitcoin, which launched in 2009. This technology has since given rise to many private forms of digital money. While the first generations of digital coins proved not to be stable means of payment and storing value, more recent versions have explicitly aimed to provide stability. A number of so-called ‘stablecoins’ have been issued in recent years, but the idea became more prominent with global tech giant Facebook announcing on 18 July 2019 its intention to issue its own fiat-currency-backed stablecoin: the Libra. Given its potential to reach 51
Grégory Claeys, Maria Demertzis
4.
millions, if not billions, of users across the world, authorities have taken a significant interest in how this might challenge official currencies and the need for regulation. Finally, if cash is scarce or even disappears, citizens risk losing direct access to sovereign money, the ultimate safe asset. Should cash altogether disappear, citizens would only have access to bank deposits, which are not as tangible as cash and not as safe to store value. Given the existence of digital technology, central banks are now contemplating the idea of creating central bank digital currencies (CBDCs). These could replace coins and banknotes and potentially make central banks’ digital reserves available to all economic agents and not only to banks.
We will review the emergence of these different forms of digital currencies and whether they fulfil a good role as currencies. To this end, we will also examine how and whether they could be able to challenge traditional currencies as well as the role guardians of money, price and financial stability, namely central banks, play. Last, we will review the need for central banks to also adapt by becoming more digital and to what effect.
2. The first generation: cryptocurrencies Money is a social convention that facilitates trade when there is a lack of a double coincidence of wants, by solving the problem of a lack of trust in exchanges. In practice, money tends to be defined by the three functions it traditionally performs: first, a unit of account, as it serves as a common measure of values for goods and services traded in an economy; second, a medium of exchange, as an item accepted for the payment of goods and services, and for the repayment of debts; and third, a store of value, a way to store wealth to transfer purchasing power from the present to the future. To perform these three functions, money can take various forms, including non-perishable goods and non-financial and financial assets. Historically, various goods and assets have been used and even co-existed as money, but some have been very successful while others have led to monetary instability and have been replaced. Therefore, money can vary both with respect to its characteristics and its relative success in performing its three main functions.
2.1 A taxonomy of money: where do cryptocurrencies fit in? We classify the various types of money to understand how cryptocurrencies differ from other forms of money based on three main criteria (among others discussed by Bech and Garratt, 2017): a) the issuer – government or private; b) the form it takes – physical or digital; and c) how transactions are settled – centralised or decentralised. Cryptocurrencies represent a form of money that has not previously been available, as a particular combination in the money taxonomy. Specifically, they are: •
• •
Privately issued. This is not new per se. Privately issued currencies have been used and have sometimes performed well in the past. However, unlike bank deposits for instance, cryptocurrencies are not a liability and cannot be redeemed. Digital. This is also not new per se; it is similar to electronic money issued by central and commercial banks. Cryptocurrencies are also fiat money in that they have no intrinsic value. Decentralised settlement of transactions. Exchanges via cryptocurrencies are peer-topeer, based on the distributed-ledger technology (DLT). Such technology is used to avoid the so called ‘double spending problem’, which is traditionally solved through 52
Digital currencies
Issuer
Form
Transaction
Centralised
N/A
Decentralised
Coins and banknotes / pre-paid cards
Centralised
CB reserves / CBDC
Decentralised
Crypto-CBDC / mobile wallets
Centralised
N/A
Decentralised
Commodity money
Centralised
Bank deposits / Libra-style Stablecoins
Decentralised
Cryptocurrencies / USDTrust-style Stablecoins
Physical Gov’t Digital Money Physical Private
Examples
Digital
Figure 4.1
A taxonomy of money
Source: Authors based partly on the typology proposed by Bech and Garratt (2017)
record-keeping by a trusted central agent. This means that with a DLT there is no central authority needed for the settlement of digital transactions between counterparties.1 In fact, no single entity is responsible for operating cryptocurrencies, though a number of intermediaries are needed to provide technical services (a digital wallet is needed to use the cryptocurrencies, and intermediaries are involved when exchanging them with other currencies, etc.). Essentially, the novelty of cryptocurrencies is the feasibility of peer-to-peer digital transactions (see Figure 4.1). What could be the main advantages of cryptocurrencies given these main characteristics? •
•
•
Anonymity. Decentralisation would ensure (almost) anonymity of transactions, which is good for privacy, although it could also mean that cryptocurrencies can facilitate transactions related to illegal activities or tax evasion. Arguably cryptocurrencies are even more prone to such activities than cash given the enhanced possibility to handle large transactions. The DLT is also in principle less vulnerable to malicious attacks compared to centralised systems and therefore should allow a reliable ledger of past transactions to be maintained. Free of political backing. Private issuance is decided not by a political institution but by an algorithm which is seen by supporters of cryptocurrencies as a way to avoid discretionary decisions that can lead to excessive inflation. The automatic issuance of cryptocurrencies would also increase transparency (for anyone able to read the algorithm at least) and the predictability of their ‘monetary policy’. As we will later discuss, this is also a disadvantage because discretionary decision making allows for flexibility to deal with shocks. Global reach. The digital form of cryptocurrencies and the absence of a link to a particular jurisdiction allow for a truly global and easily accessible currency that could facilitate global trade. 53
Grégory Claeys, Maria Demertzis
2.2. Cryptocurrencies: where do we stand? There are over 2000 cryptocurrencies2 but a vast majority of transactions is done in just a few of them. The total market capitalisation is $221bn (as of 8 October 2019).3 The 10 most important cryptocurrencies represent just below 80 percent of the total market, while the two most important currencies, Bitcoin and Etherium, represent around 65 percent of the market value. As Claeys et al. (2018) have shown, the first generation of cryptocurrencies traded in small volumes and therefore never threatened to challenge traditional global currencies, like the euro or the dollar. The main reason behind this is that they were not able to fulfil the three main functions of money. First, currencies like Bitcoin have been very volatile and therefore unable to store value properly. This is the direct result of its supply protocol. In the case of Bitcoin, the quantity supplied follows a predictable, near-predetermined path towards a fixed upper limit (21 million of bitcoins). Importantly this means that supply does not move to match the quantity demanded. The inelastic nature of the supply embedded in the protocol rules led to huge volatility of its purchasing power of goods and services which prevented it from functioning as a good store of value. This, in turn, also limited its adoption and kept the network of users relatively small, thus reducing its role as a medium of exchange and as a unit of account.4 These two problems reinforced each other because the high volatility in price is partly the result of their limited use and the fact that the networks of users consisted mainly of speculators. Cryptocurrencies have so far therefore been more akin to speculative assets, expected to yield returns only as a result of capital gains. The second reason why first-generation cryptocurrencies have not been a good medium of exchange is the time they took to be verified in the decentralised ledger. While this democratised the verification process, it came at the expense of time and energy efficiency (Krause and Tolaymat, 2018). The amount of computing power needed to validate cryptocurrency transactions in order to avoid any falsification of the ledger has been very energy inefficient and has represented a significant waste of resources. Bitcoin (BTC) Ethereum (ETH) 25%
XRP (XRP) Bitcoin Cash (BCH)
1% 1% 1%
Tether (USDT) 1%
Litecoin (LTC)
2% 2% 2%
67%
5%
EOS (EOS) Binance Coin (BNB) Bitcoin SV (BSV)
9%
Stellar (XLM) Other
Figure 4.2
M arket shares of the 10 most popular cryptocurrencies (market capitalisation, October 2019, in %)
Source: Bruegel based on Yahoo Finance.
54
Digital currencies
Third, the borderless nature of cryptocurrencies runs against the national nature of protecting price stability. From a monetary policy perspective, a global cryptocurrency area is unlikely to be an optimal currency area, as this would lead to an inability to adjust exchange rates within the ‘area’. The result would thus be a crypto-monetary policy (i.e. its supply protocol) that would be consistently too tight and too accommodative for different countries at different times. Last, the first generation of cryptocurrencies is also prone to other major risks: the market concentration which could lead to the falsification of the ledger and to ‘double spend’ issues; the manipulation of the value of the currency via insider trading; or the lack of regulation for intermediaries necessary to use cryptocurrencies.5
3. The second generation: stablecoins The second generation of private digital currencies aimed at resolving some of the first generation’s shortcomings. A number of so-called ‘stablecoins’ have been issued in recent years, including Tether in 2014, which was intended (originally at least) to be fully backed by US dollar reserves, TrueUSD with a similar model in 2018, and Basis in 2017, which promised to create an algorithmic stablecoin.6 However, the stablecoin idea has recently become more prominent, when global tech giant Facebook announced the creation of its own stablecoin in 2019 – the Libra – in the form of a private digital currency, run on a (more) centralised network and backed by fiatcurrency reserves to ensure stability. Even if the initiative is currently in limbo, it is worth discussing how it tried to resolve some of the shortcomings of the first generation of cryptocurrencies. First, the Libra intended to solve the problems of energy and time inefficiency by being a centralised system (permissioned blockchain), at least in the initial phase, in which the founding members – the so-called ‘Association’ – would be in charge of the validator nodes. This would have the aim of ensuring an optimal balance between effectiveness and security of the system. However, a centralised system in the hands of a limited number of association members posed the risk of collusion (Abadi and Brunnermeier, 2019) to the detriment of users. The ability of the Libra association to be an unequivocally trustworthy custodian of the ledger was immediately challenged. Second, Facebook wanted to ensure the stability of the Libra by backing it with reserves composed of a basket of liquid and stable assets (themselves in official credible fiat currencies),7 not unlike a currency board8 or a simple investment fund guaranteeing redemption at par. As long as the Libra Association would back each Libra coin with an identical pool of safe and liquid assets, its value should be stable. However, as explained in Claeys and Demertzis (2019) and Chang and Velasco (2019) such an arrangement is not incentive compatible. Motivated by the profit motive, the Libra Association might be tempted to renege on its promise and to not back each coin fully, or to change the composition of the pool of assets. The problem arises because there might be a conflict between maintaining a stable price for Libra (which implies the issuer honouring the initial pledge at any price) and profit maximisation (which gives the issuer the incentive to deviate from full collateralisation and a stable basket). The value of the Libra would depend crucially on the Association’s commitment to keep it stable. But unlike central banks that have a public function and are accountable to citizens to fulfil their stability mandate, the Association is not bound by a similar commitment. Stability of the basket and profit maximisation are not necessarily aligned objectives. 55
Grégory Claeys, Maria Demertzis
But perhaps the biggest issue with Libra was that it was going to be issued by a global tech giant that would immediately grant it scale and accessibility, two features that are pivotal for critical uptake of currencies. This immediately implied that it might be a more credible competitor to traditional forms of money. But what were the risks for central banks and governments? The first risk was that of scale. By means of a comparison, almost ten years after its creation, Bitcoin was estimated in 2017 to have 7.1 million owners worldwide.9 The active number of users of Facebook (and hence Libra’s immediate potential network size) is, at around 2.5 billion, much larger than the number of people using international currencies such as the euro or even the dollar on a daily basis. As former Bank of England Governor Mark Carney pointed out,10 Libra could become ‘instantly systemic’ on launch day and should therefore be put under tight regulatory scrutiny. Similarly, Financial Stability Board Chair Randal K. Quarles highlighted the need to contain the risks that arise from financial innovation and particularly the ‘wider use of new types of cryptoassets for retail payment purposes would warrant close scrutiny by authorities to ensure that that they are subject to high standards of regulation’.11 Second there were concerns beyond financial stability arising from complex trade-offs between competition and data protection (BIS, 2019). Calibra, the digital wallet on which Libra would be stored, was going to be bundled with Facebook’s ecosystem and made available to all its users. This would give Facebook the power to push its customers to use its own digital wallet, just like Amazon had the power to push their Kindle e-book reader to many of its customers that used its other services. The potential for a massive user base can lead to monopoly power for the issuer, which in turn can lead to severe financial vulnerabilities from system failures (either deliberate and fraudulent or simply erroneous). Third, the most important concern voiced since the Libra announcement in mid-2019 has been distrust about the way Facebook operates, particularly in relation to data privacy and Facebook’s global dominance. Wolf (2019), for instance, was very critical of Facebook’s continuing failure to appreciate the way it is affecting modern democracies. Libra therefore would start with a sizable trust deficit that could hinder its promised popularity. Following these concerns, the European Council and Commission declared in November 2019 that ‘no global stablecoin arrangement should begin operation in the European Union until the legal, regulatory and oversight challenges and risks have been adequately identified and addressed’.12 Facebook has since then significantly scaled back from its original plans.13 The possibility of a big-scale stablecoin is therefore for the moment not imminent. However, the possibility for future stablecoins to challenge established currencies remains real. We discuss next whether private digital currencies could possibly dominate over traditional currencies.
4. Private and official currencies: a ‘peaceful’ coexistence? At the moment, private digital currencies operate alongside official currencies. The current volumes are small and do not challenge the position of official money as the main currency. But is that always going to be the case, and if not, would that entail risks for central bank monetary policy? Could central banks lose their grip on respective economies as a result?
4.1 The control of money The interaction between private currencies and central bank monetary policy is treated in detail by Fernandes-Villaverde and Sanches (2018). Their theoretical model predicts that the 56
Digital currencies
coexistence of central bank and private money depends on the type of monetary policy the former follows. In particular, privately issued currencies would be used if the official currencies do not ensure price stability, but would lose their value as a medium of exchange when the central bank credibly guarantees the real value of money balances. The ramifications are two-fold. First, the coexistence of government money and private currencies that are valued as mediums of exchange is not a theoretical impossibility. Second, central banks have an advantage in that by choosing a specific type of monetary policy they can prevent private currencies from being valued as a medium of exchange (but they could still be valued for other reasons, for instance as a pure speculative asset). From this perspective, rather than posing a threat, the coexistence of government money and cryptocurrencies can have a positive effect by acting as a disciplining device on central banks. This is a partial vindication of Hayek (1976), who argued in favour of breaking the state monopoly on money as a way to ensure the stability of the official currency. Nevertheless, from a more practical standpoint, central banks’ monopoly position could be threatened by the emergence of cryptocurrencies as relevant mediums of exchange with stable purchasing power. First, the extent of the substitution of cash and bank deposits for cryptocurrencies by economic agents will determine the effectiveness of monetary policy. Extensive substitution of bank deposits would shrink the amount of broad money in the economy and therefore reduce control over monetary conditions. At the extreme, the provision of base money and the resulting influence over interest rates would be rendered ineffective. However, as Stevens (2017) points out, as long as money issued by central banks retains the role of unit of account, the switch to private currencies as a medium of exchange would be limited and thus the associated threat to monetary control would also be limited. Second, the shrinking role of central bank money creates a possible fiscal risk in the form of reduced seigniorage revenue. The response could be higher distortionary taxes that would hurt growth. That said, such risks appear to be exaggerated given that seigniorage revenues make up an insignificant fraction of total government revenue. The last, but probably most pertinent threat does not emanate from the potential use of cryptocurrencies as money, but from their attractiveness as investment assets. As a speculative investment – an investment made in expectation of a return from capital gains only – private currencies will be prone to bubbles. The collapse of such a bubble could reverberate into wider financial instability if households, corporates and financial institutions hold unhedged debt positions. Central banks would then face a double risk: first to the stability of financial institutions they supervise, from the potentially unregulated cryptocurrency debt markets, and, second, to price stability, from the effects on the real economy of deleveraging and defaulting by economic agents.
4.2 Financial stability implications of a potential private currency takeover Given the natural monopoly enjoyed by central bank-controlled currencies, it would take a deep crisis of trust for a private currency to substitute an established currency in full. An episode of very high inflation could be such a shock, but even then, agents might wish to switch to other established safe-haven currencies (such as the US dollar or the Swiss franc) before resorting to private digital currencies. However, as argued on p. 53, the broad accessibility of these currencies, compared to traditional currencies, might offer an easily accessible alternative. What would that mean for the financial system and the broader economy? 57
Grégory Claeys, Maria Demertzis
Focusing on the case of cryptocurrencies, can a fractional reserve banking system, as we have today, function in a cryptocurrency world and how would the cryptocurrency protocol influence it? In a fractional reserve banking system, bank deposits are matched by currency (bills, coins and central bank reserves) only up to a fraction. In such a system, bank deposits are the result of the provision of loans by commercial banks to companies and households, and, therefore, money and credit creation are closely intertwined. In theory, there is nothing that prevents the creation of a fractional reserve banking in a full cryptocurrency form. However, money creation by private banks would reduce the level of control the cryptocurrency protocol exerts over the money supply, placing additional complexity on the supply algorithm. In fact, central banks that have tried to target the total stock of money in the past renounced it because they found it difficult to achieve price stability with that strategy. Today, the money stock that is created by private banks is ultimately influenced, but not fully controlled, by the central bank. Monetary policy operates mainly through the interest rate at which the central bank provides currency to private banks. Successful influence over monetary conditions in the presence of a fractional reserve banking system would, thus, require an algorithm that manages to affect the lending behaviour of banks. Even if this were to be achieved, these banks would still be vulnerable to bank runs. Under a fractional reserve system, banks generate profits by engaging in maturity transformation: using short-term, money-like deposits as funding for illiquid, long-term loans. This leaves them vulnerable to the possibility of bank runs. When there is such a general flight to liquidity, the central bank acts as a lender of last resort that restores confidence in the banks and ensures financial stability. Arguably cryptocurrencies would not be able to provide liquidity readily in times of crisis as that is not part of their ‘mandate’. This is not unlike the gold standard, where new currency could not be mined in real time and made available to absorb excessive demand. Similarly, deposit guarantees would not be available as a solution in a crypto-financial system. A third ex-post solution would be for the banks themselves to suspend the convertibility of their deposits into the cryptocurrency. However, the existence of such ex-post risk would translate to an ex-ante discount of each bank’s respective IOUs. The uniqueness of money would then break down, as private-bank issued money would fragment into assets that are not traded at par with the predominant cryptocurrency (which is in some way similar to what happened during the free-banking era in the US between 1836 and 186414). Therefore, the ex-ante absence of credible solutions to bank runs would increase their likelihood and lead to instability in the system. Does that mean, at the opposite end of the spectrum, that we could see the emergence of a financial system similar to full reserve banking? In such a system, the bank’s IOUs that serve as money (e.g. bank deposits) are fully backed by a government fiat currency or by a commodity. Here, the cryptocurrency could play the role played traditionally by official fiat currencies. This would have two main advantages: first, money supply would be decoupled from credit and would thus only depend on the cryptocurrency algorithm; and, second, there would be no bank runs. However, one has to ask what forces would give rise to banking in such a cryptocurrency world. In the case of today’s fiat government currency, the possibility for users to hold and store its physical form (i.e. bills and coins) is fraught with security risks and inconvenience. Full-reserve banks (which do not provide lending) would at least provide a solution to this problem, by serving their clients’ needs to make payments. In a cryptocurrency world by contrast, full-reserve banks would be irrelevant: as payments would be done directly in the decentralised ledger, there would be no need to resort to an intermediary to complete a payment. 58
Digital currencies
In summary, in a full crypto-financial system, savers would have to choose between holding IOUs labelled in a cryptocurrency unit of account issued by unstable banks (not benefiting from a lender of last resort) or sticking to cryptocurrencies that stay idle in the ledger. In that case, who would provide lending to the rest of the economy? One possibility is direct peer-to-peer lending but this would force individuals to screen, monitor and diversify their investments themselves. Or individuals could pool their wealth to share risks and lend to other agents. However, these entities, say ‘cryptobanks’, that would provide loans to the economy would look more like investment funds than banks, as their funding sources (in cryptocurrencies) would not be deposits but equity. Although liquidity risk would be less of a concern for the holders of equity, they would also be more exposed to credit risk than bank depositors, because they would not benefit from the seniority that bank depositors enjoy in case of default. This risk could thus disincentivise savers from lending and lead to a system that is inherently unstable by being prone to severe credit squeeze that would clearly be detrimental for the economy.
4.3 Money and power: ensuring a system of checks and balances The potential of private currencies to credibly challenge official currencies cannot only be based on the intelligence of underlying algorithms. Technology and our understanding of underlying economic mechanisms can always help improve those algorithms just as they currently inform monetary policy decisions. But is that all we need, or is there something more intrinsic to the power of the money issuer? Can a private currency ever be a credible money? Currency management has a societal value – effectively the societal value of monetary policy. The value and stability of money is what enables societies to function well and is not separate from broader choices governments make when they run policy. It is therefore also a part of what constitutes the social contract (Collard, 2017) between the principal (the citizen) and the agent (the government). Manipulating a currency has historically been a powerful means of enabling the sovereign to pursue certain objectives, including financing wars. In other words, this power of controlling money can be used and abused. This is why in modern democracies currency management goes through appropriate layers of legitimacy and accountability. A modern authority that controls the currency will be evaluated according to how well it sticks to the implicit social contract agreed through democratic procedures. This means that it can be released for deviations from agreements made. In the case of cryptocurrencies, how could an intelligent algorithm that is automatic and anonymous ever be held responsible for failing to deliver agreements? The complexity of currency management implies that the system will fail sometimes, just like financial crises periodically happen. No algorithm, no matter how intelligent (and indeed benevolent), will remove the possibility of crisis. The automation of monetary policy would remove it also from the system of checks and balances. This type of ‘independence’ of monetary policy effectively also removes the possibility of accountability, and makes monetary policy exogenous to the process that identifies, monitors and evaluates agreements. But what about known associations that are no longer neither anonymous nor automatic. Can private money challenge the sovereign character of traditional money? The private nature of such endeavours and therefore their private objectives, be it profit or other things like accessibility, make them unlikely to take on the ‘externality of monitoring prices’. Price stability, a key feature of money, is based on a welfare criterion that alludes to a public good and are therefore best served by public institutions. 59
Grégory Claeys, Maria Demertzis
Last, it is only the existence of this system of checks and balances that allows modern lenders of last resort to create money out of nothing and provide ample liquidity in times of crisis. As soon as this system breaks down and trust in authorities goes, the currency ceases to be an acceptable means of payment or even a unit of account. The currency is only as strong as its lender of last resort, and the lender of last resort is only as strong as the backing that it has from its constituents. Constituents in turn, build trust depending on how well social contracts are adhered to.
5. Should central banks issue their own digital currencies? Interest in central bank digital currencies (CBDCs) on the part of authorities is partly motivated by the popularity of private digital currencies that could challenge the role of official currencies. Providing digital currencies issued by the central bank could possibly make private digital currencies less attractive and slow down their adoption. There is no universally agreed definition of what constitutes a CBDC, but the term has become commonly used15 to designate any form of central bank digital fiat liability that is accessible to all economic agents. We discuss here what we believe is the most promising version of CBDC: deposit accounts at the central bank available to all.
5.1 What would be the purpose of CBDCs? A first reason why CBDCs might be useful is that allowing households and companies to open accounts at the central bank would give them direct access to efficient and instantaneous retail payment systems that banks already use to exchange reserves between them. This would remove one reason for switching to a private digital currency with a better payment system. But there are other reasons. Replicate the ‘safety’ of cash. If cash were to become scarce or even disappear, citizens would lose direct access to sovereign money, the ultimate safe asset. Should cash disappear, citizens would only have access to bank deposits, which are not as safe. Deposits above a certain threshold (EUR 100,000 in the euro area) are uninsured, and even below this threshold, there is the possibility of losing access to savings even for a few days or weeks. For this reason, they cannot replace the safety that cash provides. Also, the possibility of re-directing bank deposits to central bank deposits provides incentives for bank discipline. As argued by Brunnermeier et al. (2019), in the absence of cash, banks would not ‘fear’ convertibility of their deposits into central bank currency, and could lose some of the incentives (even though regulation would still be a major disciplining device) to reduce their solvency and liquidity risks. In extremis, if deposits do not have to be converted into a common currency, deposits from different commercial banks could at some point become imperfect substitutes for one another. In that case, the trustworthiness of each issuer, would lead to the creation of ‘exchange rates’ between them, as was the case during the US free banking era in the nineteenth century. CBDCs would solve this problem by allowing households to access central bank currency in a new form, and thus restore the convertibility threat for banks. Strengthen the transmission of monetary policy. The introduction of CBDCs could also strengthen monetary policy by transmitting it directly to the general public. Changes in policy rates would be transmitted directly to CBDC depositors, in contrast to what happens today, where interest paid by commercial banks on deposits are relatively sticky.16 This also means that CBDCs would make unconventional policies easier to implement. 60
Digital currencies
First, as long as the CBDC is interest-bearing, it could help relax further the zero lower bound constraint because interest rates applied to the CBDC could be negative (unlike for banknotes). Abolishing cash altogether would strengthen this effect although it must be acknowledged that this might not be desirable for other reasons. Cash could still be useful at least as a back-up for a CBDC in case of a technical failure or cyberattack, and for privacy reasons. But even if cash continues to exist, as long as its use is inconvenient (which would be even more the case if CBDC were introduced) and its storage is costly, implementing negative rates on CBDC holdings would be possible. Second, CBDCs could reduce one of the potential side effects of quantitative easing (QE), namely excessive reserves. Currently, central bank bond purchases from non-bank institutions create additional reserves that are inevitably held by the commercial banks in the deposit facility of the central bank. This is because non-banks cannot hold reserves directly at the central bank. On aggregate, this means that banks cannot control fully the quantity of reserves they want to hold. When rates are negative, this becomes costly for banks and might result in potential side effects such as increased rates for lending to the real economy. If non-banks could hold CBDCs directly, QE would not affect the banking sector negatively. Finally, provided the concept of helicopter money is a politically acceptable monetary policy tool, it would be easier to implement if all citizens had accounts at the central bank, because the central bank would be able to credit their accounts with CBDC units directly.
5.2 The potential risks of CBDCs The introduction of CBDCs is sufficiently disruptive that it could pose a number of risks. Cyclical bank runs. First, one the main fears of policymakers (see, for example, Coeuré, 2018) is that CBDCs will lead to cyclical bank runs. If households and companies have access to central-bank reserves, there is a risk of a flight-to-safety from commercial-bank deposits to CBDCs in each economic downturn. This type of run from banks to the central bank happened in the 1930s during the Great Depression in France, when it was possible for nonbanks to maintain accounts at the Banque de France (Baubeau et al., 2018). Bank runs are already possible today by withdrawing cash or transferring deposits between banks, 17 but the main concern is that digital bank runs towards CBDCs would be easier and happen more rapidly than traditional bank runs. Altering and possibly removing the need for intermediation. In addition to this cyclical financial stability risk, CBDCs could interfere and distort the objective of financial intermediation. Banks would compete with the central bank to hold deposits. It is very difficult to predict what would happen, because it would depend on the particular properties of the CBDC introduced and on the behaviour of the central bank after its introduction, but this could lead to different outcomes (Stevens, 2017). A first possible outcome could be an evolution towards a financial system characterised by narrow(er) banks that are less reliant on deposits. Banks could either offer higher returns to depositors to try to retain their deposit base, or they could rely on other sources of financing. This would have profound implications, positive as well as negative. The extra competition from CBDCs would reduce the monopoly power of the banking sector and allow depositors to obtain higher returns on their deposits. For banks, by definition, the effect would be the opposite because they could be forced to rely on more expensive and potentially less stable sources of funding, such as the wholesale market.18 This new banking model would make banks look more like investment funds, which could be less stable thus requiring an 61
Grégory Claeys, Maria Demertzis
adjustment of the financial safety net. The need for traditional deposit insurance would be reduced because deposits could be kept safely in the form of CBDCs. However, if we consider that the maturity transformation provided by banks is a valuable service, then it needs to be protected from liquidity risk. Either insurance cover for banks’ short-term liabilities would have to be broadened to include wholesale funding, with all the risks that this would entail (but the alternative would be frequent ‘wholesale runs’ such as those that happened during the last financial crisis). Or regulation would have to be significantly stricter to avoid any maturity mismatch on banks’ balance sheets, for example by forcing them to be financed mainly through equity and long-term debt. Another possibility would be a tightening of credit conditions by banks if they are unable to retain depositors or attract new sources of funding. This tightening would lead to less lending and/or at a higher price, which would, all else being equal, result in a significant drag on investment and ultimately on economic activity.
5.3 How could central banks minimise these risks? How can policymakers reduce bank runs? Policymakers have several tools at their disposal, should bank runs become more frequent as a result of the introduction of CBDCs. First, deposit insurance offsets the risk of runs when deposits are within the guaranteed amount. Second, the central bank should play its crucial role of lender of last resort by providing liquidity through loans to the banks that suffer runs, as long as they are solvent. The financial instability episodes in France in the 1930s discussed in Baubeau et al. (2018) showed that the main problem was not the bank runs (towards the central bank or towards other saving institutions) per se but rather the strong ‘gold standard mentality’ prevailing at the Banque de France at the time. This mentality prevented the central bank from playing its role as lender of last resort and replacing the shortfall in deposits held at commercial banks with central bank loans to avoid a strong credit crunch. How can central banks discourage disintermediation? Central banks would also have various instruments to counter the risk of structural financial disintermediation if that were to endanger price or financial stability. The central bank’s reaction could vary depending on the magnitude of the problem. In moderate cases, such as if the quantity of credit provided by banks was not significantly affected, but banks asked for higher lending interest rates (for example because they needed to increase the returns paid to depositors to retain them), the central bank would have to lower its policy rates structurally to offset this effect and maintain financial conditions at the same (presumably adequate) level, all things being equal. In normal times this should not be a particular problem, but at a time when the effective lower bound is binding, it might be problematic, and might involve the increased use of unconventional monetary policies. What if disintermediation were to become a more significant issue and there was a clear downward pressure on bank credit availability? The main way for the central bank to offset this trend would be to provide structurally more funding to the commercial banks to replace the lost deposits, so that they could maintain the same level of financing to the economy. This means the central bank balance sheet would have to become structurally much bigger19 and also more exposed to the banking sector than otherwise. The debate on the optimal size of central banks’ balance sheets has not been settled.20 However, the two main risks for central banks in increasing massively their refinancing operations would be the following. 62
Digital currencies
•
•
First, the central bank would take more risks onto its balance sheet because it would be more exposed to the risks faced by banks: in a way, the central bank would become itself a financial intermediary between depositors that would hold CBDCs and the commercial banks. Second, this means that the central bank would be involved more directly in the credit allocation process. In order to be able to provide a much greater amount of refinancing to the banks, the central bank might have to adjust significantly its collateral framework so that banks are able to access its operations at a sufficient scale. Central banks’ decisions on collateral eligibility and haircuts are often perceived as purely technical decisions, but they are not always as neutral as they seem (Claeys and Goncalves Raposo, 2018). In particular, deciding to include new asset classes as eligible collateral (in order to increase the pool so that banks can obtain more refinancing) could have some powerful effects on credit allocation by the banks. The main advantage is that this would give the central bank greater control over the macroeconomic situation. However, the drawback would be that it could potentially make the overall allocation of resources in the economy less efficient, and could also have some distributional effects that central banks cannot and should not control.
The central bank could also try to carefully calibrate the properties of CBDCs in order to reduce ex ante the incentive to use a CBDC as a main store of value. This should avoid the extreme situation in which deposit accounts held at the central bank would fully crowd out bank deposits. The simplest way to do this would be through its remuneration system. CBDC accounts should benefit from lower than other policy rates in order to reduce both the structural disintermediation risk and the frequency of bank runs. But these returns should not be so much lower that they make CBDC unattractive mediums of exchange. In particular, when policy rates are negative, a portion of CBDC holdings could be exempted to avoid the negative impact on small savers and prevent households from switching back to holding cash. Bindseil (2019) proposed a very practical system to put that in place with a two-tier remuneration system for CBDCs: below a threshold of €3,500, CBDC holdings would be remunerated at the maximum level between the deposit rate and 0, and above that threshold CBDC holdings would be remunerated at the deposit rate minus 200 basis points. These numbers are indicative and the central bank would need to experiment to find the right balance. Such a balance would incentivise the use of CBDCs as a medium of exchange, and gives access to everyone to the ultimate safe asset when necessary (especially if cash disappears), but disincentivises the use of these accounts as a main store of value in normal times.
5.4 Who else could provide an equivalent to CBDCs? Finally, an alternative solution to give the general public access to digital central bank liabilities would be not to provide it directly through a CBDC, but to do it indirectly through what could be considered ‘full-reserve banks’ (sometimes also referred to as ‘narrow banks’21). The idea, as described for example by Adrian and Mancini-Griffoli (2019), would be to allow new entities to hold reserve balances at the central bank, subject to some specific conditions. These entities – which actually would not be so different from some form of stablecoin – would have a very particular balance sheet with only central bank reserves as assets (they would not give credit, nor buy any other type of asset) and only simple deposits as liabilities (they probably would not need to hold much capital, if any at all, given the absence 63
Grégory Claeys, Maria Demertzis
of risk from their portfolios). The entity would pass the remuneration of central bank reserves to depositors and earn a small fee for the service provided. This system would allow households and companies to hold indirectly the central bank currency and would have two additional advantages. First, it would allow central banks to focus on their mandates and not use their resources to provide direct services to their new customers (which could also have some negative reputational consequences for central banks if not handled properly). If all households and companies of the euro area opened a CBDC account at the European Central Bank, the number of accounts in the Eurosystem would grow from around 10,000 to more than 500 million (Bindseil, 2019). Second, as argued by Bordo and Levin (2019), this would help prevent a conflict of interest for the central bank. Competition from a CBDC could be considered unfair by banks given the crucial role central banks play in the organisation of the banking sector (for instance as a supervisor, among other functions). For all these reasons, privately managed alternatives to CBDCs should not be discarded by central banks and, on the contrary, should be considered as one way to provide CBDCs to the general public.
6. Conclusion Privately issued digital currencies have the capacity to change global payment systems and possibly challenge some official currencies. But their issuers are unlikely to play the role of managing price stability, safeguarding financial stability and being the ultimate guarantor of value. This is a task for public authorities serving the greater public good that cannot be met when operating for profit, which is the case for these privately issued coins. But their potential for scale, global reach and ultimately also convenience has an appeal that has forced public authorities to reconsider their role. One potential effect will be the need for creating Central Bank Digital Currencies, to first meet the rapid reduction and possible elimination of cash. This is not simple and it could imply central banks taking more of a financial intermediation role in the economy. This is not without problems but it is something that needs to be considered as the whole monetary system becomes digital.
Notes 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
See BIS (2018) for details on how a DLT works in practice. See https://capital.com/top-cryptocurrencies-to-invest-in-spring-2020 Yahoo Finance (8 October 2019). See Koning (2015). See for instance Griffin and Shams (2018) showing how the cryptocurrency Tether might be used to provide price support and manipulate other cryptocurrency prices. Despite an original and potentially promising model, Basis shut down its operations in December 2018. Although important details are not described in the White Paper (Libra Association, 2019). Anderson and Papadia (2020). See https://www.bitcoinmarketjournal.com/how-many-people-use-bitcoin/. https://www.ft.com/content/189c1c66-91dd-11e9-aea1-2b1d33ac3271 https://www.fsb.org/2019/06/fsb-chairs-letter-to-g20-leaders-meeting-in-osaka/ https://data.consilium.europa.eu/doc/document/ST-13571-2019-INIT/en/pdf https://www.nytimes.com/2020/04/16/technology/facebook-libra-cryptocurrency.html See details in Frieden (2016). See for instance Meaning et al. (2018) for a detailed discussion on the definition of a CBDC. In the euro area, before the crisis when policy rates were high, interest on bank deposits was significantly lower, while now the opposite is true, as shown by Bindseil (2019).
64
Digital currencies
References Abadi, J. and M. Brunnermeier (2019) ‘Blockchain Economics’, manuscript, 31 August 2019. Available at: https://scholar.princeton.edu/sites/default/files/markus/files/blockchain_paper_v7a.pdf Adrian, T. and T. Mancini-Griffoli (2019) ‘The rise of digital money’. IMF Fintech Notes No. 19/001. Available at: https://www.imf.org/en/Publications/fintech-notes/Issues/2019/07/12/ The-Rise-of-Digital-Money-47097 Anderson, J. and F. Papadia (2020) ‘Libra as a currency board: are the risks too great?’, Bruegel, Blog post, 20 January. Available at: https://www.bruegel.org/2020/01/libra-as-a-currency-boardare-the-risks-too-great/ Baubeau, P., E. Monnet, A. Riva and S. Ungaro (2018) ‘Flight-to-safety and the credit crunch: a new history of the banking crisis in france during the Great Depression’, Banque de France Working Paper No. 698. Available at: https://ssrn.com/abstract=3285119 Bech, M. and R. Garratt (2017) ‘Central bank cryptocurrencies’, BIS Quarterly Review September 2017, Bank for International Settlements Bindseil, U. (2019) ‘Central Bank Digital Currency – financial system implications and control’, manuscript, 30 July 2019. Available at: https://ssrn.com/abstract=3385283 BIS (2018) Cryptocurrencies: looking beyond the hype. BIS Annual Economic Report, Bank for International Settlements BIS (2019) Big tech in finance: opportunities and risks. BIS Annual Economic Report, 23 June. Available at: https://www.bis.org/publ/arpdf/ar2019e3.htm Bordo, M. and A.T. Levin (2019) ‘Improving the monetary regime: The case for U.S. digital cash’. Cato Journal, 39(1), 383–405. Available at: https://www.cato.org/sites/cato.org/files/serials/files/ cato-journal/2019/5/cj-v39n2-9.pdf Brunnermeier, M.K., H. James, and J.P. Landau (2019) ‘The digitalization of money’ (No. w26300). National Bureau of Economic Research Working Paper. Available at: https://www.nber.org/ papers/w26300 Chang, R. and A. Velasco (2019) ‘Will Facebook’s Libra turn into a cancer?’ Project Syndicate. 16 July 2019. Available at: https://www.project-syndicate.org/commentary/facebook-libra-becomescancer-by-andres-velasco-and-roberto-chang-2019-07 Claeys, G. and M. Demertzis (2017) ‘How should the European Central Bank “normalise” its monetary policy?’, Bruegel Policy Contribution No 2017/31, prepared for the Economic and Monetary Affairs Committee (ECON) of the European Parliament. Available at: https://bruegel.org/2017/11/ how-should-the-european-central-bank-normalise-its-monetary-policy/ Claeys, G. and M. Demertzis (2019) ‘The next generation of digital currencies: in search of stability’, Policy Contribution Issue n˚15, Bruegel, December. Available at: https://www.bruegel.org/ wp-content/uploads/2019/12/PC-15_2019.pdf Claeys, G., M. Demertzis and K. Efstathiou (2018) ‘Cryptocurrencies and Monetary Policy’, Bruegel Policy Contribution No 2018/10, prepared for the Economic and Monetary Affairs
65
Grégory Claeys, Maria Demertzis Committee (ECON) of the European Parliament. Available at: https://bruegel.org/2018/06/ cryptocurrencies-and-monetary-policy/ Claeys, G. and I. Goncalves Raposo (2018) ‘Is the ECB collateral framework compromising the safe-asset status of euro-area sovereign bonds?’ Bruegel blog, 8 June 2018. Available at: https:// bruegel.org/2018/06/is-the-ecb-collateral-framework-compromising-the-safe-asset-status-ofeuro-area-sovereign-bonds/ Cochrane, J. (2018) ‘Basecoin’, The Grumpy Economist – John Cochrane’s blog, 22 April. Available at: https://johnhcochrane.blogspot.com/2018/04/basecoin.html Cœuré, B. (2018) ‘The future of central bank money’, Speech at the International Center for Monetary and Banking Studies, Geneva, 14 May 2018. Available at: https://www.ecb.europa.eu/press/key/ date/2018/html/ecb.sp180514_4.en.html Collard, B. (2017) ‘Money is the real social contract’, Foundation for Economic Education Blog, 25 August. Available at: https://fee.org/articles/money-is-the-real-social-contract/ Fernandes-Villaverde, J. and D.R. Sanches (2018) ‘On the economics of digital currencies’, Working Papers 18-7, Federal Reserve Bank of Philadelphia Frieden, J. (2016) Lessons for the euro from early American monetary and financial history, Essay and Lecture Series, Bruegel Griffin, J.M. and A. Shams (2018) ‘Is Bitcoin really un-tethered?’, mimeo. Available at: https://papers. ssrn.com/sol3/papers.cfm?abstract_id=3195066 Hayek, F.A. (1976) ‘Choice in currency: a way to stop inflation’, Occasional Papers, The Institute of Economic Affairs Koning, J.P. (2015) ‘The dollarization of bitcoin’, Moneyness Blog, 12 June. Available at: http:// jpkoning.blogspot.com/2015/06/the-dollarization-of-bitcoin.html Libra Association (2019) ‘Official Libra White Paper’. Available at: https://libra.org/en-US/white-paper Krause, M.J. and T. Tolaymat (2018) ‘Quantification of energy and carbon costs for mining cryptocurrencies’. Nature Sustain 1, 711–718. Available at: https://www.nature.com/articles/ s41893-018-0152-7 Meaning, J., B. Dyson, J. Barker, and E. Clayton (2018) ‘Broadening narrow money: monetary policy with a central bank digital currency’, BoE Staff Working Paper No. 724. Available at: https:// www.bankofengland.co.uk/-/media/boe/files/working-paper/2018/broadening-narrow-moneymonetary-policy-with-a-central-bank-digital-currency.pdf Stevens, A. (2017) ‘Digital currencies: threats and opportunities for monetary policy’, NBB Economic Review, 1 ( June 2017), 79–92, National Bank of Belgium. Available at: https://www.nbb.be/doc/ ts/publications/economicreview/2017/ecorevi2017_h5.pdf Wolf, M. (2019) ‘Facebook enters dangerous waters with Libra cryptocurrency’, Financial Times, 25 June. Available at: https://www.ft.com/content/07c05f ba-b1c4-11e7-a398-73d59db9e399
66
PART II
Blockchain and cryptocurrencies
.
(
Taylor & Francis Taylor & Francis Group http://taylorandfrancis.com
5 DECENTRALIZED AUTONOMOUS RISK TRANSFER ON THE BLOCKCHAIN Alexander Braun, Niklas Haeusle and Stephan Karpischek
1. Introduction Traditional insurance is plagued by high operating costs, information asymmetries and in-transparency. However, the rise of modern technologies, such as blockchain and smart contracts, brings about new options to transfer risks. This chapter introduces the concept of a decentralized autonomous organization (DAO) on the public Ethereum blockchain, which decomposes the traditional insurance value chain and disintermediates the associated risk transfer transactions. A decomposition and decentralization of the insurance value chain introduces competition for the different insurance company functions. Similarly, it can help to increase fairness and to tackle conflicts of interest between customers and firms. In traditional insurance markets, for example, claims managers are employed by the risk carriers, implying that the payout to the customer is not overseen by an independent third party. In contrast, a decentralized autonomous risk transfer system allows for the payout to be guaranteed under well-defined conditions. Other benefits for customers are a higher transparency, control of their data, and a more resilient system. Since the record- keeping of a DAO relies on a public blockchain, all transactions are provable at any time and data is not lost if one node fails. In combination with smart contracts, the classical boundaries between markets and companies blur and hybrid institutional arrangements become viable. This effect is especially relevant for industries with virtual products, such as insurance. In this chapter, we describe the mechanics of a decentralized risk transfer process on the blockchain. We begin with a review of the technological innovations that allow for this new institutional arrangement in the first place. Next, we discuss the structure and dynamics of a decentralized risk transfer platform, including all the necessary entities for a fully functioning system. Subsequently, we look at insurance products that may be suitable for decentralization and explain how participants in a DAO can be incentivized using system-specific tokens. Finally, based on three use cases, which are currently being piloted or developed, we discuss the potential of decentralized insurance in practice.
69
A. Braun, N. Haeusle, S. Karpischek
2. Essential innovations Blockchain technology Decentralized autonomous risk transfer can only be realized due to a few enablers. One of them is blockchain technology, a form of a distributed ledger technology (DLT), which combines several pre-existing technical innovations such as cryptography, peer-to-peer networks and distributed consensus in order to create a robust and reliable record, without the need for a trusted central party (World Economic Forum 2018). The blockchain itself is a ledger or data base, in which all transactions are timestamped and sequentially arranged in blocks, starting with the so-called genesis block. The ledger itself is stored in a distributed fashion across a network of computers. Writers take turns in recording information and are selected from the network participants based on algorithmic principles. Apart from past transactions, plenty of other data, such as contract details, can be stored in a distributed ledger, too. The first blockchain was created to serve as a ledger for Bitcoin, the cryptocurrency invented by an unknown individual or group called Satoshi Nakamoto in 2008 (Nakamoto 2008). Today, many kinds of blockchain designs exist. They mostly differ in the way writers are selected to add new blocks and how participants achieve consensus regarding the validity of transactions. If concurrent blocks are being added to an existing blockchain, writers must be able to identify the legitimate one. Similarly, all other network participants, who read the information stored in the blockchain, must be able to select the true version from two or more competing chains. This validation requires a consensus mechanism such as Proof of Work (PoW), which is employed by Bitcoin and Ethereum, the two largest blockchain networks. Other mechanisms such as Proof of Stake (PoS) are also possible (Saleh 2018). The latter is, for example, employed by the cryptocurrencies Dash and Neo. The general benefits emphasized by supporters of blockchain technology are anonymity, self-sovereignty, fairness, transparency and irreversibility of the recorded transactions. The latter two characteristics are particularly important for the transfer of insurance risk through a DAO.
Smart contracts Another innovation which is required for decentralized risk transfer is the so-called smart contract (Buterin 2013). Smart contracts are not legal contracts, but pieces of software that include predefined rule sets and are deployed on a blockchain. A very simple smart contract could have the following form: if a node contributes 20 units of computing power to the network, then reward it with five newly generated coins. The programming code contains all the necessary information: the receiver of the payout, the source and the amount of the reward, as well as the trigger used to determine when the transaction should happen. While technically all different kinds of smart contracts are possible, most are currently simple payment devices, exchanging a workload against a specified number of tokens. Cryptocurrencies based on a PoW consensus mechanism, for example, trade computing power against monetary rewards. Fully decentralized value chains require more complex smart contracts that can account for tasks processed by humans. The advantages of smart contracts are numerous (Sheth und Subramanian 2019). They permit trusted transactions and agreements to be carried out among disparate, anonymous parties without the need for a central authority, legal system, or external enforcement mechanism. The most important contribution of smart contracts for decentralized risk transfer is the ease of enforceability. 70
Decentralized autonomous risk transfer on the blockchain
Decentralized autonomous organizations The third enabler is the DAO, a new organizational form which became feasible through the aforementioned technological advancements. Figure 5.1 shows the structure of a very general DAO. Customers buy the product, for example an insurance contract. The product is produced by a nexus of workers, who are linked by smart contracts on the blockchain. The exact relationships between workers may differ from industry to industry, but usually they need to hold network-specific tokens to participate in the DAO. Oliver Williamson’s famous transaction cost theory (Williamson 1973, 1987) identifies uncertainty and asset specificity as major drivers of transaction costs. Asset specificity describes how well an asset can be used across different situations. A factory that is customized to produce special equipment for hospitals, for example, exhibits a high degree of asset specificity. In contrast, a generic plastic manufacturing plant can build various products and therefore has a lower asset specificity. Once a highly specific asset has been purchased, it cannot be easily sold or redeployed for another purpose. Therefore, it creates vulnerabilities to opportunistic behavior, since counterparties obtain leverage and may drive prices down to the marginal costs. Such behavior can be prevented by contracts. However, due to incomplete information, contractibility becomes increasingly harder for higher asset specificities, and therefore also more cost intense. At some point, the total transaction costs are lowest if a firm is established to orchestrate the production process through its hierarchy. In contrast, market-based transactions dominate if asset specificity is low (Williamson 2010). In this case, opportunistic behavior can be easily avoided by turning to competitors of the counterparty. A DAO is a hybrid institutional arrangement between markets and hierarchies. The ease of enforceability of smart contracts and the transparency of blockchains help Holder
C
W
Customer
C
W
W
W
W
W
W
W
P
W
T = Token P = Product W = Worker C = normal currency, like USD
Figure 5.1
T
W
Smart Contracts
Generic structure of a decentralized autonomous organization
71
A. Braun, N. Haeusle, S. Karpischek
to significantly reduce the total transaction costs for higher asset specificities. Goods which were most efficiently produced in traditional companies before can now be provided by a nexus of workers in a decentralized network. The two industries which are most affected by this potential shift in transaction costs are the financial sector and the insurance industry. The first DAO in the financial sector (‘The DAO’) was a venture capital fund, which got disbanded shortly after creation due to a vulnerability exploit that allowed hackers to steal most of its funds. More recent examples of DAOs are ‘Augur’, which specializes in prediction markets and ‘MolochDAO’, which focuses on community funding. In the insurance sector, technology startups are working on platforms such as the decentralized insurance protocol (DIP) that will allow for a decentralization of insurance products with more complex value chains.
3. Decentralized autonomous risk transfer Structure The aforementioned technologies are essential for the advent of decentralized autonomous risk transfer. Figure 5.2 shows the basic functions that a DAO needs to comprise for an insurance product (Braun, Häusle and Karpischek, 2021). The boxes represent different entities, consisting of multiple workers, or individuals, each of which executes a specific part of the production process. The relayer is comparable to an insurance broker. S/he connects customers with the decentralized risk transfer system, for example through a website on which users can select their insurance coverage. Relayers are not mutually exclusive. Multiple relayers can offer the same insurance product and charge different prices for their services, because they have different customer acquisition costs. Another role is the risk pool keeper, who manages the portfolio of insurance policies and carries out administrative tasks, such as developing risk models and defining rules and License provider
Relayer
Risk Pool Keeper
Customer Data Oracle
Data Broker
Figure 5.2
Decentralized insurance on the blockchain
72
Decentralized autonomous risk transfer on the blockchain
conditions for other auxiliary workers, which are needed in the decentralized risk transfer system. Each risk pool is linked to exactly one product and contains the contracts and their premiums. Furthermore, the DAO requires a license provider. Since insurance markets are heavily regulated, it would not be feasible to offer coverage without an insurance license or compliance with regulatory authorities. License providers make sure that a decentralized system can operate in such an environment. Finally, the data provider and the data oracle are responsible for supplying the decentralized insurance system with raw data and converting said data into actionable information. Necessary data could, for example, be weather statistics for parametric catastrophe insurance (Martin 2018), which are needed to calculate the premium and determine whether a payout is due. These roles are crucial in every decentralized insurance system. Depending on the product and setting, however, further steps may be needed. For example, if traditional loss-based policies are sold instead of parametric products, underwriters and claims handlers will be required. Similarly, an additional payment agent can ensure that the customers get their indemnification without a delay and in their preferred currency.
Feasible products Some insurance products are more viable for a decentralized system than others (Calcaterra et al. 2019). Network incubators, protocol designers and investors need to assess the feasibility of the decentralization before it can be implemented. The higher the standardization of the insurance product and the automation of the value creation process, the easier it is to implement a DAO. There are several reasons why parametric insurance is particularly well suited for decentralization. In a traditional insurance contract, the customer is indemnified if he or she experiences a loss. Often this loss must be confirmed by a claim adjuster. The payoff of a parametric insurance policy, on the other hand, depends on a predefined trigger event. Consider hurricane insurance, for example. Under a traditional contract, the indemnity is paid after an investigator has evaluated the loss of the policyholder. In case of parametric insurance, in contrast, the payment is made if a physical parameter value, such as wind speed, exceeds a preset threshold level. The parameter may be highly correlated to the actual losses, but the insured could also face a situation in which the wind speed was just not high enough to trigger the payout of his policy. This is known as basis risk. A significant advantage of parametric insurance are its lower costs. Instead of having to maintain an expensive claims management, the insurance company can instantly and without much effort verify whether the trigger level was hit, even for a very large portfolio of policies. Further factors that determine a product’s suitability for decentralization are the regulatory complexity and contract length. Health insurance, for example, has very specific regulatory requirements and exhibits long contract terms. Both characteristics are difficult to reconcile with a decentralized value chain, in which competing market forces can cause high fluctuations. Given these considerations, the best use cases for decentralized insurance are parametric property and casualty products with relatively short contract durations.
Quality assurance A DAO needs rules to work properly. In addition, workers must be convinced to participate in the organization and to complete tasks. In a traditional company, they receive a 73
A. Braun, N. Haeusle, S. Karpischek
salary from their employer. In the decentralized setting on the blockchain, in contrast, they get a reward or transaction fee for their contribution, which is determined by market prices and paid in cryptocurrency. It turns out that a key problem in a DAO is the incentivization of workers to put in enough effort for a high-quality result. In traditional firms, quality is typically assured through managerial oversight, which does not exist in a DAO. Unfortunately, smart contracts alone are an imperfect means for quality assurance. Often essential tasks in a DAO are too complex to be evaluated by a self-executing piece of computer code and thus the trigger for a payment is hard to verify. Although it is easy to design a smart contract for a data provider, which triggers a payout when a file has been submitted, imperfect contractibility leaves room for opportunistic behavior and malfeasance. A verification by checksums or similar monitoring methods can only address simple errors but may miss those that can cause severe damage to the whole network. In the decentralized insurance case, it would lead to enormous disutility if the customer did not receive his indemnity due to poor-quality work by one of the service providers in the DAO. In addition to the financial loss for the policyholder, the network itself would probably lose the trust of its customer base. Customers of a traditional insurance firm have the option to sue if a product is faulty. In developed countries, this certainly helps to assure a certain level of the quality for the customer. Since DAOs are essentially a network instead of a legal entity, they cannot be held accountable. The workers themselves, on the other hand, are generally liable for their behavior. However, enforceability is heavily restricted due to their anonymity and their global distribution. Hence, quality assurance in a DAO must exclusively rely on economic incentivization mechanisms instead of managerial control and the legal system. One option in this regard is staking, a digital form of collateralization that requires network participants to acquire cryptographic tokens and deposit them in a smart contract or bound wallet. In case of worker malfeasance, the collateral can be withheld and reallocated to the customers to compensate them for their losses. Staking is economically akin to performance bonds (Martimort et al. 2016; White 1992; Joseph A. Ritter und Lowell J. Taylor 1994).
Bootstrapping the network At launch, a DAO must quickly attract enough participants to operate the network and offer products of interest to customers. If several relayers want to join, for example, but no license provider is interested, the system will not be usable. To address this issue, early movers as well as the programmers of the protocol and the smart contracts have to be disproportionally compensated for the high business risk they are taking (Karpischek and Mussenbrock, 2018)(. Just as in the case of regular startups, non-recoverable investments must be made right at the beginning, when only few customers buy the product and the success of the business is highly uncertain. Especially license providers are affected by this issue, since their capital outlay is completely front-loaded. One solution is the issuance of system-specific staking tokens, particularly to early network participants. To start with, these tokens do not have much value, since they are not backed by any real asset. Economically, they represent a share in the network rather than a currency and can therefore be expected to appreciate with increasing throughput. If the decentralized business ultimately becomes successful, early participants can sell their tokens at a much higher price.
74
Decentralized autonomous risk transfer on the blockchain
4. Tokenomics Due to the complexity of the different transactions involved in a DAO for insurance and risk transfer, three types of tokens are required. In the following, we explain their differences and highlight which specific purpose each one of them serves.
Payment The first category of tokens are payment tokens. In principle, the payments for the insurance product offered by the DAO could also be made in fiat money such as USD or EUR. Cryptocurrencies instantiated on the blockchain, however, are a better fit for digital networks, because they enable instantaneous worldwide payments and can be processed by smart contracts. Customers pay their premiums and receive their indemnities in the form of payment tokens. The latter are also used to compensate workers for their effort. Since both customers and workers are assumed to be risk averse, volatility in the exchange rate with fiat currencies is not desired and should be avoided as far as possible. This can be achieved by resorting to stable coins such as Tether(USDT) or DAI. USDT for example is backed by the US Dollar and tries to mimic a 1:1 exchange rate with the USD.
Staking Staking serves as an incentivization mechanism to prevent malfeasant and opportunistic behavior by network participants. The system-specific tokens issued to secure early-stage funding for a DAO can be used for this purpose. In contrast to the payment token, which needs to be stable, the exchange rate between the staking token and the fiat currency should ideally exhibit an upward trend. The latter is of key interest to investors and workers, particularly those that joined early on, and reflects the increasing value of the decentralized insurance network or token economy. The token supply is often fixed at the outset, while the token demand is driven by speculators and workers. The demand of speculators does not contribute to the intrinsic value of the network and thus drives the exchange rate volatility. The demand of workers equals the total amount of tokens required for their stakes. Of course, workers will only participate in the network if the reward compensates them for their effort and for the risk of losing the stake. By expanding the network, more workers will participate and therefore the token demand rises. Although, at first glance, the token demand can also be increased by setting the stake size above its optimum, this would exert the same effect as taxes and therefore decrease the overall value of the network.
Risk transfer If enough customers are insured by the DAO, the law of large numbers allows for a reliable estimation of the average portfolio loss. In some years, however, pure randomness can cause claims to exceed their statistically expected value. Furthermore, extreme events such as natural disasters may lead to a correlation of risks in the portfolio and cause aggregate losses to be much higher than in average years. Such larger-than-expected payouts are called tail risks and, in traditional insurance companies, are buffered by equity capital. Since the latter is expensive, however, insurers frequently also use substitutes such as reinsurance. A modern
75
A. Braun, N. Haeusle, S. Karpischek Table 5.1 Comparison of token types Token type
Payment token
Staking token
Risk token
used byt used for
customers premium and indemnity payments USDT, DAI
workers Quality assurance, incentives, early network funding DIP
investors transfer of risk to the capital market type: ERC-20
Examples
alternative is direct risk transfer to the capital markets through insurance-linked securities such as catastrophe (cat) bonds. By buying these instruments, investors provide risk capital to the ceding insurance company. In case of a trigger event before maturity, this capital is used to indemnify the cedent. Otherwise, investors are repaid their principal plus coupons. An insurance DAO has no balance sheet and therefore needs equity substitutes to handle tail risk. To this end, it can issue digital securities called risk tokens that work the same way as a cat bond. They allow for risks to be compartmentalized and transferred directly to the capital markets. Table 5.1 summarizes the three different types of tokens: The payment token, the staking token and the risk token.
5. Use cases The five biggest decentralized finance networks grew over 40% in less than two months and the total sum invested passed 600 million USD in 2019 (Sandner 2019). Decentralized autonomous risk transfer can be expected to follow with a delay, but ultimately develop a similar momentum. The yearly event D1Conf brings together blockchain insurance projects from around the world. Regarding the volume of risks transferred, Nexus Mutual is currently one of the leading projects in the decentralized insurance space. They insure users against smart contract risks and had an active cover amount of about 3.5 mn USD in Spring 2020. Below we present three more applications of the decentralized autonomous risk transfer process which are currently under development or in use.
Flight delay One application which has been deployed by the blockchain startup Etherisc in cooperation with Atlas Insurance PCC Limited is flight-delay coverage. This product pays an indemnity in case an insured flight is delayed or cancelled (Peterson et al. 2013). The trigger is fully parametric: the payout solely depends on the duration of the delay instead of the actual damage for the policyholder, such as losses due to a missed business meeting. In Europe, consumers are protected by the Air Passenger Rights Regulation (European Union 2004), which grants a legal right for compensation of up to EUR 600 in case of a delay or a cancellation. However, many other countries do not exhibit such customer protection laws. In the United States, for example, a federal act for flight delay reimbursement does not exist. This means that any form of compensation is entirely subject to the goodwill of the providers. Especially cheaper airlines do not offer repayments at all. Flight Delay insurance implies protection against this issue. If this type of insurance is run on the blockchain and uses stable coins for payments, a smart contract ensures that customers receive their payout immediately. While the advantages of a decentralized insurance are not as enormous as in other use cases, this product is comparably easy to implement in practice due to the rich availability of data and simple structure of the underwriting process. 76
Decentralized autonomous risk transfer on the blockchain
Hurricane guard Another use case for the decentralized risk transfer process is insurance against hurricanes. These extreme storms can entail huge property losses for individuals and families (Insurance Information Institute 2017). Parametric hurricane insurance offers payouts depending on the wind speed recorded at specific measurement stations. Due to climate change, the importance of coverage against natural disasters is going to increase significantly. However, especially in developing countries, insurance penetration is still very low. Providers of coverage against natural disasters do either not exist or demand high premiums to cover their inefficient cost structures. To address this problem, the Insurtech firm Raincoat is currently developing a decentralized insurance product for Puerto Rico, where homeowners are highly exposed to hurricane risk. By relying on a parametric trigger and complete automation on the blockchain, costs can be greatly reduced compared to traditional insurance, since claims handling becomes redundant. In addition, it is possible to pay indemnities much faster, helping residents to recover and rebuild as soon as possible. Excess losses are covered through tokenization. This is especially important for hurricane insurance, which exhibits a significant tail risk. Through a direct transfer of tail risk to investors, large amounts of solvency capital or reinsurance coverages can be avoided. Through tokenization, the risk becomes more accessible, liquid and diversifiable.
Crop insurance Finally, Etherisc, AON, and Oxfam are currently testing decentralized crop insurance in cooperation with farmer organizations and local insurers in Sri Lanka. The income of farmers is volatile and strongly dependent on the weather. Natural catastrophes such as floods or droughts can wipe out large amounts of crops and thus threaten the existence of whole families. This is a particular problem in developing countries, where a large share of the population directly or indirectly depends on agriculture, and farmers have little private wealth to absorb losses. Decentralized crop insurance on the blockchain is an efficient form of coverage in this case. The advantages are threefold: first, costs are much lower. Through automation, smart contracts and the use of existing distribution infrastructures, coverage can be sold at a price close to the actuarial fair premium. Second, parametric triggers in combination with smart contracts allow for rapid payouts. Third, transparency, automatic execution, and irreversibility of blockchain transactions preclude opportunistic behavior by insurers, thus earning the trust of farmers who may doubt the enforceability of regular insurance claims due to the corrupt legal systems in many developing countries.
References Braun, Alexander and Häusle, Niklas and Karpischek, Stephan, Incentivization in Decentralized Autonomous Organizations (2021) Available at SSRN: https://ssrn.com/abstract=3760531 Buterin, Vitalik (2013) A Next-Generation Smart Contract and Decentralized Application Platform. In: Ethereum White Paper ethereum.org. Available online at: https://cryptorating.eu/whitepapers/ Ethereum/Ethereum_white_paper.pdf. Calcaterra, Craig; Kaal, Wulf A.; Rao, Vadhindran K. (2019) Decentralized Underwriting. In: SSRN Journal. DOI: 10.2139/ssrn.3396542. Etherisc (2017) White Paper. Available online at: https://etherisc.com/files/etherisc_whitepaper_ 1.01_en.pdf. European Union (2004) Passengers Rights Nr. 261/2004. Available online at: https://europa.eu/ youreurope/citizens/travel/passenger-rights/air/index_de.htm.
77
A. Braun, N. Haeusle, S. Karpischek Insurance Information Institute (2017) Homeowners and Renters Insurance. Available online at: https://www.iii.org/fact-statistic/facts-statistics-homeowners-and-renters-insurance. Karpischek, Stephan; Mussenbrock, Christoph (2018) Etherisc White Paper. Available online at: https://etherisc.com/files/etherisc_whitepaper_1.01_en.pdf. Martimort, David; Semenov, Aggey; Stole, Lars (2016) A Theory of Contracts with Limited Enforcement. In: Review of Economic Studies 88, rdw024. DOI: 10.1093/restud/rdw024. Martin, Andre (2018) What is Parametric Insurance? Available online at: https://corporatesolutions. swissre.com/insights/knowledge/what_is_parametric_insurance.html. Nakamoto, Satoshi (2008) Bitcoin: A Peer-to-Peer Electronic Cash System. Available online at: https://bitcoin.org/bitcoin.pdf. Peterson, Everett; Neels, Kevin; Barczi, Nathan; Graham, Thea (2013) The Economic Cost of Airline Flight Delay. In: Journal of Transport Economics and Policy 47 (1), pp. 107–121. Ritter, Joseph A.; Taylor, Lowell J. (1994) Workers as Creditors: Performance Bonds and Efficiency Wages. In: American Economic Review 84 (3), pp. 694–704. Saleh, Fahad (2018) Blockchain Without Waste: Proof-of-Stake. In: SSRN Journal. DOI: 10.2139/ ssrn.3183935. Sandner, Phillipp (2019) Decentralized Finance (DeFi). Available online at: https://medium.com/ @philippsandner/decentralized-finance-defi-what-do-you-need-to-know-9cd5e8c2a48. Sheth, Alpen; Subramanian, Hemang (2019) Blockchain and Contract Theory: Modeling Smart Contracts Using Insurance Markets. In: Managerial Finance 46 (6), pp.803–814. White, William D. (1992) Information and the Control of Agents. In: Journal of Economic Behavior and Organisation 18 (1), pp. 111–117. Williamson, Oliver E. (1973) Markets and Hierarchies: Some Elementary Considerations. In: American Economic Review 63 (2), pp. 316–325. Williamson, Oliver E. (1987) The Economic Institutions of Capitalism. Firms, Markets, Relational Contracting. New York, NY: Free Press. Available online at: http://www.loc.gov/catdir/bios/ simon051/87011901.html. Williamson, Oliver E. (2010) Transaction Cost Economics: The Natural Progression. In: American Economic Review 100 (3), pp. 673–690. World Economic Forum (2018) Blockchain Beyond the Hype: A Practical Framework for Business Leaders. Available online at: http://www3.weforum.org/docs/48423_Whether_Blockchain_WP.pdf.
78
6 DISTRIBUTED LEDGER TECHNOLOGIES AND BLOCKCHAIN FOR FINTECH Principles and applications Raghava Rao Mukkamala and Ravi Vatrapu 1 Introduction Historically, technology advancements and financial innovations have been interlinked. In the recent past, technological innovations have played a crucial role in financial innovations that led to not only new forms of products and services but also disruption and disintermediation in the wider financial sector. One of the enduring impacts of the global Financial Crisis (GFC) has been the trust deficit between consumers and companies in the financial sector in general and the banking sector in particular. This erosion of institutional trust extended from the established players such as retail and investment banks to other centralized organizations such as central banks. One consequence of this has been the proliferation of FinTech entities that offer decentralized trust mechanisms and alternatives to fiat currencies. Due to their unique technological features such as the lack of centralized control and high level of anonymity, distributed ledger technologies (DLT) underpin much of the FinTech innovation and entrepreneurship and empower the evolution of decentralized applications in multiple domains such as finance, health care, supply chains etc. In this chapter, we present the underlying technical principles of DLT and blockchain technology and outline their practical applications to FinTech. According to a World Bank report (Natarajan et al., 2019), with the rapid development and spread of new technological advancements, the finance sector is undergoing a significant transformation to embrace these innovations for the betterment of existing and innovation of new financial products and services. Blockchain technology came into the limelight when Bitcoin, a decentralized digital cash system was introduced as a peer- to-peer cryptocurrency in 2009 (Nakamoto, 2008) and as of 2020, Bitcoin is the largest cryptocurrency with a market capitalization of approximately more than 100 billion USD.1 Moreover, an important feature of Bitcoin is maintainability of its currency value without any central authority or governmental administration but purely based on the transactions that are stored in the public distributed ledger (datastore) using blockchain technology. When Bitcoin was making its initial buzz, many institutions and people thought that it would not make any significant impact on the global economy; this was supported by several reports (European Central Bank, 2012; European Central Bank, 2015). However, such a view has changed drastically 79
R.R. Mukkamala, R. Vatrapu
in the last few years, especially with regards to cryptocurrencies; many financial institutions like banks started to explore and test the technologies behind blockchain and DLTs. Apart from Bitcoin, several hundred more cryptocurrencies were introduced, with a market cap of more than a couple of hundred billion dollars.2 However the interest shown towards DLT and blockchain technologies is not only limited to the domain of finance; it has also garnered attention from many sectors like power generation, health, education, government, supply chain and logistics, transportation and others. The new technology has also inspired many people and organizations, resulting in fortunes for some entities and people and bankruptcy for others, for example cryptocurrency exchanges (Szostek, 2019). Due to their innovative and disruptive nature, blockchain technologies have captured the attention of many governmental organizations, the academic community, enterprises and financial institutions. For example, the European Union has taken several initiatives to promote and harness the innovative technology. One such is the EU Blockchain Observatory & Forum,3 started in 2018 by the European Union to promote collaboration between various initiatives on blockchain technologies among the member countries and also to highlight the important progress made in these technologies and at the same time promote education and awareness among the population and organizations. At the same time, since the cryptocurrencies provide high-level anonymity for their users, they became go-to currencies for illicit activities such as money laundering and cybercriminal activities (Sun Yin et al., 2019). Therefore, there is a strong desire on the part of regulatory authorities to initiate more regulatory guidance concerning DLTs and blockchain-based applications. To understand the kind of impact DLT and blockchain technologies can have on FinTech, it is important to understand the technological foundations of the DLTs and blockchain, and the main principles and applications behind these technologies. It is also necessary to understand the current legal and regulatory initiatives concerning DLT and blockchain to foresee what kind of impact those will bring into the FinTech domain in terms of legal regulatory frameworks. Therefore, in this chapter we focus on the principles and the concepts behind DLT and blockchain in sec. 2. In the next section (sec. 3) various FinTech applications and their suitability concerning blockchain will be discussed. Then we turn our focus on legal and regulatory aspects of DLT and blockchain technologies in sec. 4. Final we conclude in sec. 5 with few comments about the market adoption of these technologies in FinTech.
2 Principles behind distributed ledger technologies and blockchain The disruptive and innovative nature of blockchain technology resulted in the evolution of many decentralized applications such as cryptocurrencies and smart contracts. Bitcoin, a decentralized cryptocurrency based on blockchain technology was introduced in 2009 (Nakamoto, 2008) and in 2020 Bitcoin is the largest cryptocurrency with a market capitalization of approximately more than 200 billion USD.4 Even though bitcoin is considered an innovative and disruptive technology, its underlying technical foundations and those of other cryptocurrencies actually originated back in the 1980–1990s (Narayanan and Clark, 2017b). The following is a brief description of various concepts and underlying technical components of DLT and blockchain technologies.
2.1 DLT and blockchain According to Rauchs et al. (2018, p. 15), “Distributed ledger technology (DLT) has established itself as an umbrella term to designate multi-party systems that operate in an 80
Distributed ledger technologies
environment with no central operator or authority, despite parties who may be unreliable or malicious (‘adversarial environment’)”. The notion of DLT systems evolved in the 1980s in distributed computing and in that context a DLT system can be considered as a distributed datastore using state machine replication, where multiple parties operate in a decentralized environment without any central authority, communicate the atomic and incremental changes to the global state. Blockchain is a subset of DLT technologies. Even though the notion of blockchain evolved around the 1990s, it only became popular after 2009 when bitcoin cryptocurrency (Nakamoto, 2008) was introduced. Blockchain is the decentralized distributed data structure that is combined with guarantees against the tamper-resistance of transactions/records using cryptographic methods. By using the time-stamping of its transactions and messages, blockchain provides universally verifiable proofs for the existence or absence of a transaction in the distributed database, and the underlying cryptographic primitives, using hash functions and digital signatures, provide a guarantee that these proofs are computationally secure and verifiable at any point in time. Blockchain is decentralized, jointly maintained by a plurality of independent parties/nodes, and achieves consistency of transactions among distributed nodes by using distributed consensus protocols (such as Byzantine fault tolerance algorithm (Lamport et al., 1982)) without the need of having a central authority. In this chapter, we confine our discussion mostly to blockchain applications which will have practical relevance to FinTech rather than focusing DLTs which are mostly confined to the distributed computing domain.
2.2 Distributed ledger The central idea of blockchain and DLT is the distributed and decentralized ledger maintained by several participating entities. The main difference between a centralized and decentralized ledger is the way it is maintained and how consensus is achieved. In the case of a centralized ledger, since it is maintained by a trusted central authority (such as a bank or a financial institution) there is no need for consensus. However, a distributed ledger is a global data structure that is collectively maintained by the participants who may not trust each other in decentralized environments on the Internet. Therefore, distributed ledgers need certain characteristics to maintain the integrity and consistency of the ledger. First of all, the ledger must be immutable in the sense that it only allows for new data to be appended, i.e. it neither allows deletion nor modifications to the ledger. Secondly, it should be possible to compute a succinct cryptographic digest to the state of the ledger for verification purposes, so that rather than storing the entire state of the ledger, the digest can be used to verify the state of the ledger, for example, to make sure that the ledger has not been tampered with. Built on the concept of peer-to-peer networks and distributed storage (Xu, 1999), distributed ledgers can be considered as a distributed data store with state machine replication using a peer-to-peer protocol, where transactions are atomic changes to the data store which are grouped into blocks (Mamoshina et al., 2018).
2.3 Hash functions Hashing is known as a one-way function used to ensure the integrity of data. A hash function is an input independent average linear time algorithm that takes a set of variables or data and transforms it into a fixed size hash digest (Carter and Wegman, 1979). A successful hash function has the following characteristics. It is deterministic – the same input always creates the same output, efficient – output is computed in a timely manner, distributed – evenly spread 81
R.R. Mukkamala, R. Vatrapu
across the output range, meaning that similar data should not correlate to similar hashes, preimage-resistant – it will be impossible to find the input document x, based on the hash value (h(x)) and nearly collision resistant – no two different inputs x and y create the same hash h(x) = h(y) =⇒ x ≡ y. Furthermore, hash functions are used for organizing and linking data together in blockchains. Today’s cryptographic hash functions are built on certain standards, a particularly popular one is SHA-2 (Penard and van Werkhoven, 2008), a version of it (SHA-256) being used in the bitcoin blockchain. SHA-2 was developed partly by the United States National Security Agency (NSA) and it builds on two concepts: Merkle-Damg˚ard construction (Merkle, 1989; Damg˚ard, 1989) and Davies-Meyer compression (Simmons, 1994). Other common cryptographic standards include SHA-3 (Dworkin, 2015), Blake, and MD5. Another key concept of hash functions in the blockchain is that of organizing and linking data together. This is done through the hashing of various elements in the block header containing a hash of the previous block, Merkle root of transactions, time, and nonce. The concept of the Merkle Tree (Merkle, 1980) is that each transaction is hashed, then the resulting hash of each transaction is hashed to build a tree structure until the top node known as the Merkle root is obtained. This type of organizing of data allows secure and efficient verification of contents of a block and summarizes all the transactions in a block (Antonopoulos, 2014).
2.4 Digital signatures One of the main goals of blockchain technology is to be able to verify the authenticity and non-repudiation of data/transactions. The digital signature is a cryptographic scheme that guarantees two properties: authenticity, that the data/message created or owned by the known sender and the non-repudiation property guarantees that the data is not altered, using a pair of keys with an asymmetric cryptographic algorithm like Rivest–Shamir–Adleman or RSA algorithm (Rivest et al., 1978). In the asymmetric cryptographic algorithm, two corresponding keys (e.g. public and private) will be generated and the data encrypted with one key can only be decrypted with the other corresponding key. Participants in the blockchain network use a public/private key pair, where the public key is used as an address to digital assets (such as coins in cryptocurrency) and the private key is used to claim the ownership over these digital assets (e.g. to spend coins in cryptocurrencies). Over the years, more secure versions of digital signatures have been developed. For instance, bitcoin uses the Elliptic Curve Digital Signature Algorithm (ECDSA) for key generation ( Johnson et al., 2001).
2.5 Digital timestamping The concept behind the distributed ledger data structure is adopted from digital timestamping and was proposed by Haber and Stornetta (1990) and Bayer et al. (1993) in the 1990s to address the challenges of a digital notary service that provides proof to establish that the documents were signed at a certain point in time and no later than that (Narayanan and Clark, 2017b). In the proposal made by Haber and Stornetta (1990), the documents were created and broadcasted continuously and the creator of a new document would sign the document digitally, attach the timestamp and then link the newly created document with the previously broadcasted document. Since the previous broadcasted document had been created and signed by someone else, this created long-chain documents authored by many participants
82
Distributed ledger technologies
in a collaborative manner. The link between the documents is created using hashing rather than digital signatures, as the hash values of the document are quick and efficient to compute. Moreover, it was also proposed that documents created around the same time can be grouped into some kind of block structure so that these documents will have the same timestamp. The documents inside a block are linked by a binary tree of hash pointers known as a Merkle Tree (Merkle, 1980), which is an efficient data structure for the storage of hash values of documents (Narayanan and Clark, 2017b).
2.6 Fault tolerance The requirements of a distributed ledger are much more stringent when compared to a centralized ledger due to the absence of a central authority. The distributed ledger has to deal with issues such as that participating nodes in the network may fail, or be malicious; it also needs to account for latency in the network and, therefore, achieving consensus in distributed ledgers is more challenging. There is a significant amount of research in fault-tolerant distributed computing (Lamport et al., 1982; Lamport, 2019; Lamport et al., 2001; Castro et al., 1999) where the problem of achieving state replication across different distributed nodes is explored. Several fault tolerance protocols like Paxos (Lamport et al., 2001), Practical Byzantine Fault Tolerance (PBFT) (Castro et al., 1999) and other protocols have contributed to some of the main ideas behind the fault-tolerance mechanisms blockchain and DLTs.
2.7 Consensus protocols In order to avoid having a central authority for enabling trust in the system, there needs to be some mechanism that establishes trust between the involved parties, which is achievable by distributed consensus of those parties. In a blockchain, trust is ensured through a distributed consensus protocol. Although the protocol can vary slightly from system to system, the idea of achieving trust with the consensus of those parties involved remains the same. The two most widespread concepts in the distributed consensus protocol are proof-of-work and proof-of-stake. Proof-of-work (PoW) refers to the idea that a service requester is required to solve a cryptographic puzzle (computational work) to participate in a network and it was initially proposed in hashcash (Back, 2002) as a countermeasure for denial of service attack using CPU (Central Processing Unit) cost-functions. In blockchain and especially in bitcoin (Nakamoto, 2008), it is used as a verification technique for finding the appropriate header for new blocks of data and to append them to the chain of blocks. To add a block, a node has to solve a cost-function (find the right nonce), that results in a pre-defined hash format with certain restrictions. At the same time, blocks can only be added to the longest chain (with the most proof-of-work invested), to avoid ‘dishonest’ attempts at altering the ledger. The concept of using a cost-function as a proof-of-work was first proposed by Adam Back in 2002. Proof-of-stake (PoS) is another method for verifying and adding blocks to the blockchain, where the node that creates the next block is chosen (Wang et al., 2019). Therefore, a node adds and verifies blocks according to how much stake they have in the system. Thereby, ownership will lead to actors behaving honestly, because they would lose their stake if they behaved dishonestly. There are a few other consensus protocols such as proof of burn, proof of luck, elapsed-time and ownership; for more details about these consensuses, protocols can be found in (Wang et al., 2019).
83
R.R. Mukkamala, R. Vatrapu
2.8 Smart contracts Using blockchain as a tamper-proof ledger would record the transfer and prove ownership of assets beyond any doubt. This enables smart contracts, an idea that was conceptualized already 20 years ago (Szabo, 1997) in the creation of computer programs that can securely enforce previously closed contracts in a decentralized manner. The idea of smart contracts is to take contractual clauses, translate them into code, thereby making them self-enforceable. Hence, intermediaries who are responsible for enforcing the contract are not needed, but instead, a trusted computer program is relied upon. Complex contractual and payment agreements can be included in standardized contracts and then be monitored and executed at low transactional costs, as they are managed digitally and immutably (Swan, 2015). Smart contracts will become powerful when combined with cryptocurrency platforms as they can be executed to handle money such as transferring money from one account to other (Narayanan and Clark, 2017b).
2.9 Public verses permissioned One of the major design choices in blockchain and DLT is whether the network runs in a permission-less mode or a permissioned mode. Public blockchain and DLT like bitcoin and Ethereum are permissionless in the sense that anybody can join the network and perform transactions without any prior approval. Alternatively, in the permissioned blockchains, like Ripple, Corda, Hyperledger, an entity, a consortium or a central administrator controls access to the participants and assigns privileges to various participants based on the role they play in the blockchain ecosystem. One of the best advantages of permissioned blockchains is that, since all the participants are prescreened/known ahead/trusted, there is no need to adopt very expensive and computing-intensive consensus mechanisms like proof-of-work; on the contrary, a very simple mechanism can be adopted to achieve consensus among the participants. The orchestration of all the above-described technologies lead to the following characteristics (Faber et al., 2019) in blockchain as shown in Table 6.1.
3 Applications of DLT in FinTech In this section, we focus our discussion on various applications in the FinTech domain where blockchain technology can make a significant impact in the coming years.
3.1 Cryptocurrencies There were several attempts since the 1990s to create a decentralized digital currency which would be resistant to censorship, and able to protect itself against outside attacks. Some of Table 6.1 Characteristics of the blockchain Immutability Decentralization Transparency Pseudonymity Chronology
Data written to database cannot be changed or deleted without consensus leading to data integrity No single point of failure/control achieved by decentralized architecture and a distributed database All data sent through the blockchain is visible to all network participants The identity of data senders and receivers is unknown Every transaction is time-stamped and can be traced back
84
Distributed ledger technologies
the notable attempts are the payment solutions such as b-money,5 Hashcash (Narayanan and Clark, 2017a), and Bitgold,6. However, these attempts ended in failure, even though they came a bit closer to the main motivation. On Halloween in 2008, the Cypherpunk member (or group of members) under the pseudonym of Satoshi Nakamoto released a whitepaper on bitcoin (Nakamoto, 2008); this was essentially a detailed document outlining the technical and logical workings of blockchain. One of the members of Cypherpunk, Hal Finney, started engaging in conversation with Satoshi, and in 2009 received the first prototype from him (Chohan, 2017). After running the software, Hal Finney mined the first bitcoins and performed the first peer-to-peer blockchain transactions. Although conceived in the perfect storm after the 2008 financial crisis, and promising liberation from banking, bitcoin’s first real use case was its manifestation as a black-market payment, secretive and essentially cost free – a reputation that it still carries to this day. As of 2020, the total market cap of cryptocurrencies is $335 billion,7 out of which the top 20 of the 2,500 live coins make up 90% of the value. It was not until 2014 when the banking world caught onto the underlying technology, that ‘blockchain’ as a term was even propagated. For the last few years, a unanimous belief in blockchain dealing with universal issues like corruption, inefficiency, and malfeasance has grown. The concept of blockchain has since merged with other areas of cryptography, peer-to-peer networking and economics, to form variants of the system more applicable to the financial sector (Treleaven et al., 2017). Obstacles to faster adoption remain within the areas of cyber risks, transaction delays due to scalability problems and high volatility, however, work is being done in solving these issues.
3.1.1 Cyber risk The nature of blockchain-based technologies is that they are hard to tamper with. Since the data is linked using hash pointers any modifications to data in a blockchain can be easily identified. However, hashing and the storage of private-keys can be a weak point. Massachusetts Institute of Technology’s Digital Currency Initiative (DCI) found that the IOTA project, whereby payments across interconnected devices happen through their cryptocurrency, was using an in-house created hash function Curl-P-27 that had some security lapses (Heilman et al., 2019), including that its function lacked a so-called seed-generator to help users generate keys for their wallets. Publishing of the cryptoanalysis of Curl-P-27 and other attacks on IOTA cryptocurrency (Heilman et al., 2019) have caused user losses amounting to 4 million USD worth of IOTA cryptocurrency. The DCI has pointed out that such architectural issues can be avoided by using well-established and scrutinized open-source functions available, and disadvise creating proprietary software which is not studied well enough or verified by cryptoanalysis. Crypto-dedicated security audits – such as SmartDec,8 OpenZeppelin,9 and Chainsecurity10 are also being used more readily.
3.1.2 Transaction delays and scalability The issue of scalability of cryptocurrencies comes from its consensus mechanisms. Since the cryptocurrencies operate in a decentralized fashion, they use a consensus mechanism to achieve an agreement between nodes. Bitcoin’s proof-of-work can only process seven transactions per second, and Native Ethereum can process four transactions per second (Dalvit, 2020). However, new proof algorithms and techniques such as Zero-Knowledge Proofs (Goldwasser et al., 1989) and ZK-STARKS (Ben-Sasson et al., 2019) are being developed which allow for faster transaction times through better consensus. StarkWare,11 a team made 85
R.R. Mukkamala, R. Vatrapu
up of developers from various groups such as ConsensSys, Coinbase, Intel Capital and researchers have developed a payment system on Ethereum called Stark allowing for 9,000 trades and 18,000 payments per second.
3.1.3 Volatility Volatility in cryptocurrencies is one of the biggest issues and to address this the development of stable coins has been ongoing for the last few years. A stable coin is a cryptocurrency designed to mitigate volatility by being tied to an asset or grouping of assets. This asset can be either another cryptocurrency, fiat money, or any other tradeable commodity: this form of tying is called ‘backed’. The relationship between the stablecoin and the backed commodity has to be pre-defined and pegged either on-chain via smart contracts, or off-chain through banks or financial institutions where the currency is located. By market capitalization, Tether is the largest stablecoin with a market-cap of surpassing 18 billion USD12 as of 2020. It has been put under scrutiny multiple times for not providing sufficient documentation of auditing, bringing into question the stated 1:1 backing ratio. A more complicated system, which has not yet been found at fault is Maker’s DAI. The DAI is backed by collateral on the Maker platform which runs on Ethereum, executed by smart contracts, and attempts to maintain a value of 1 USD which, so far it, has successfully managed. It is founded on decentralized margin trading, facilitated by a Collateralized Debt Position (CDP): users deposit assets into smart contracts as collateral for loans. Once held by the CDP, the user can generate an equivalent amount of USD value in DAI and borrow it. This means the platform and thus the ratio-fixture is run by the users themselves and is highly transparent.
3.2 Corporate finance and governance Public interest in blockchain technology is wider than in creating alternative monetary systems. The 2019 study on industry trends by Cambridge Centre for Alternative Finance (CCAF) has shown that 43% of enterprise blockchain networks used or being tested are in the Financial Services sector (Rauchs et al., 2019). This particularly includes already existent service providers experimenting with the technology; use cases extend to not only accounting but broader networks of interaction such as supply chain tracking, trading infrastructure, and document certification (Rauchs et al., 2019, p.10). The remaining percentage of blockchain systems known are spread across a wide range of sectors including government, media, manufacturing, energy, and health. According to a report by the OECD (Akgiray, 2019), blockchain is meaningful for financial applications within the areas of capital markets, where the entire ecosystem can be modelled – payment systems, both cross-border and intra-national; OTC trading, a full trading cycle of bonds derivates, commodities, and other illiquid assets; and trade finance, where processes currently take several weeks and can be cut by circumventing intermediary steps, extra tasks and paperwork (Akgiray, 2019, p. 15). It is the capacity and ambition of the separate projects which set the limit for how far they can go, whether they end up using elements of blockchain technology or build entire infrastructures. The main choice lays with what sort of consensus the system will be based on, and whether this system will be multi-party or as a catalyst for process transformation. The first variant is where control across the network is shared and not centralized, allowing for collective agreement over data, whereas the second does not work with that aim, and instead use components of DLT technologies in order 86
Distributed ledger technologies
to solve or potentially improve a business case. Unless a system is truly distributed, both in terms of data and authority, it remains an internal enterprise blockchain, with a pragmatic purpose rather than an idealistic one. A study of 67 enterprise blockchain systems revealed that only 3% were multi-party consensus networks, 20% had plans of becoming such, and the remaining 77% were using the system to reconfigure internal operations (Rauchs et al., 2019, p.19). This is not to say that internal systems have no impact; most new services and business models will come from such projects.
3.3 Financial accounting Blockchain and DLT are ledger-based technologies, where distributed ledger constructs are being used to record transactions, and as such, they are by default an accounting methodology. Due to their decentralized and distributed nature, the creation, storage and updating of financial records is entirely changed. Through universal entry bookkeeping of sorts, every entry is shared identically and permanently with everyone participating. The key features which enable this are propagation, meaning live updating of the system, shared with everyone; permanence, determined by the consensus mechanism and allowing only for addition not editing; and programmability, essentially smart contract allowing for program code to be stored alongside transaction entries. Much of what accounting is concerned with is the measurement, sharing, and analysis of financial information. The goal is to ascertain or define ownership and duty or plan for the most beneficial way of allocating resources. Blockchain and DLT can be deployed to improve the entirety of the accounting profession by lowering the costs of keeping up and auditing ledgers, as the technology provides complete assurance of factuality. Instead of focusing on the recording process, accountants can use the improved clarity to concentrate on strategic decision-making and value estimation: blockchain would never take away the need for subjective decision-making. The scope can also be expanded by bringing in value from the measurement of previously unaccounted for data, such as proof of ownership. It is not a new phenomenon that there is an effort to bring more transparency to accounting information. The organization International Financial Reporting Standards (IFRS, 2020) has been promoting trust and transparency to global financial markets and legislation against unlawful auditing has increased. Since the 1980s, triple-entry bookkeeping has been heralded as a new means for openness towards external users (Ijiri, 1986), whereby a third party can read a shared cryptographic receipt of a transaction between two parties. Since a 2014 article in Bitcoin Magazine (Tyra, 2014), the term has become synonymous with blockchain, and several projects have been created. The most dominant actors in the accounting space are The Big Four accounting firms – Deloitte, PwC, EY, and KPMG – and all have been very active in R&D in the blockchain space the last couple of years, setting up teams like Deloitte’s Rubix, FinTech strategy-consulting platforms such as PwC’s Denovo, business applications like EY’s Ops Chain, and partnerships with tech-providers such as KMPG’s collaboration with Guardtime. Startups, smaller initiatives and research projects have also been developing enterprise blockchain solutions for the financial sector, and some noteworthy projects are zkLedger,13 Pacio,14 Request Network,15 and Ledgerium.16 To explain how blockchain and DLT based financial accounting works, we take the example of Ledgerium. Ledgerium is an Australian company founded in 2018, building a ledger Luca, which is a cloud-based platform recording payment transactions between parties via blockchain. The product is being partner tested, which means that some companies have cheap access to try out 87
R.R. Mukkamala, R. Vatrapu
the technology for their external auditing. How it works in practice is that company A needs to pay company B for a service done, and uses its accounting service, for example Xero, popular with Australians. The invoice is transmitted to Luca, with a hash and its details, and encrypted with the public key of company B. Company B will be notified of a request, and it can then verify and accept the transaction by requesting a hash. This is then added to the common ledger of the two parties, and Luca will automatically ping the bank to process the financial transaction, and add the payment after it has been completed. Auditors can then use the hashes of transactions or each party’s actions to verify their truthfulness.
3.4 Financial reporting and compliance When it comes to reporting and compliance there is a heightened focus on the legal quality of the accounting where the blockchain’s aspect of immutability has always struck a promising chord. The two parts of the capital market confidence are transparency and monitoring. The transparency happens through the standards and regulations companies have to adhere to, and the monitoring happens by the regulating agencies which reinforce fairness. Blockchain helps both these processes, helping companies be more transparent and making monitoring easier. For example, in 2018 a US government guidance (ASC 606) stated exactly how revenue is recognized when it includes the use of contract, and companies which had been exploring smart contracts found it easier to comply as they had done exploratory work on products and services (Lewtan et al., 2018). In light of regulations such as ASC 606, companies that use novel, and often more complex, business models will be able to more easily keep track of their revenue stream using blockchain-based services. As an example, video game publishers no longer focus on their point-of-sale/marketplace but rather on their end-users when it comes to creating long-term relationships: their games have deluxe versions in which extra content can be downloaded, currency and add-ons which can be used in-game, and continuous bug-fixes are all included. Users can also buy these things separately, at a higher price each. In a blockchain-setup, the end-user has a unique serial number, and the reseller can attach one upon sales. Sales will be linked back to the customer, but a token will be released by the smart contract with value price assigned which, after encryption onto the ledger, will show as revenue for the company. With downloaded updates, new content, or in-game currency, the tokens will be added, and even though the end-income shows as ‘total sum’, each new component will be traceable. However, Molina-Jimenez et al. (2018b), in their exploratory paper on hybrid architectures for smart contract components, argue that both online and off-line systems have drawbacks: purely permission-less and truly democratic blockchain platforms lack scalability, speed, and are costly, among other limitations, whereas off-line systems require corruptible third parties. The smart contract acts as a service determining trust, transparency, throughput, response time, and cost, all depending on the type of operation taking place (e.g. rental or selling). No single type of uniform scenario exists, and each blockchain platform (e.g. Ethereum or Hyperledger) needs to be paired to the right off-line processing. In a Cambridge Computer Lab test, the team mimicked a hybrid model by using Rinkeby tenet of the Ethereum Blockchain for the on-blockchain component, and the Contract Compliance Checker tool for the off-line part (Molina-Jimenez et al., 2018a). These can read each other, and the result showed the architecture is implementable in case the off-blockchain component uses standard APIs able to communicate with the ones that blockchains use. 88
Distributed ledger technologies
3.5 Crowdfunding, peer-to-peer lending and ICOs In 2017, the digital equivalent of an Initial Public Offering (IPO), Initial Coin Offerings (ICOs) exploded, raising more than five billion USD in investments for crypto-based projects that year.17 ICOs are decentralized fundraising initiatives for crypto-based projects. ICOs launch their tokens at a set price and the investors who are interested in these projects can buy these tokens using cryptocurrencies like Bitcoin (BTC) or Ether (ETH). Once sufficient tokens have been sold, the project is funded, can be started, built, and opened to holders of tokens to exchange these for services on the platform. Some platforms have a split between different tokens; some purely act as a currency, eligible for public exchange; others can be used for different services, or represent the value of different assets.18 As with any class of investment, particularly novel ones, most ICOs are not only very ambitious (“changing the game”), but actual scams, with no intent to realize project goals (Dowlat and Hodapp, 2018). ICOs obviously also have their benefits, as highly regulated crowdfunding-platforms can be avoided, smaller-scale initiatives can gather trusted support, creating value together, and sometimes very quickly. Some ICOs were suspended due to an overwhelming level of interest, such as Filecoin, Tezoz, and Brave (Adeyanju, 2017). The Brave browser was started by the former CEO of Mozilla Firefox, and is a decentralized web browser that gives users heightened data privacy and the opportunity to pay their most frequently used websites, encouraging an add-free internet; it raised 35 million USD in 30 seconds.19 After China banned ICOs in late 2017, the market froze globally as heavy regulation was expected to follow, but recent offerings have slowly been picking up. Any financially risk-averse entity rightfully awaits a reliable legal framework, and the European Securities and Markets Authority (ESMA) has only recently posted its advice (ESMA, 2019), concluding for instance “Because the range of crypto assets are diverse and many have hybrid features, ESMA believes that there is not a ‘one size fits all’ solution when it comes to legal qualification” (ESMA, 2019, p. 9). Tokens being judged on a case-by-case basis means that even after a launch, time has to be devoted to certification and approval. Today the cost for raising funds for a regulatory compliant ICO can be very high; an analysis by OECD20 showed that conducting third-party security audits, communicating the benefits, and navigating other forms of requirements has costs ranging from between 50,000 and 500,000 USD. Despite this, projects that have emerged and continue to do so prove that the technology can help significantly support financial services through better workflow and management, contractual relationships and security (Weber, 2019). Crowdfunding mechanisms have also grown from ICOs and funnel investment into larger projects or a group of people, through their online platform systems. The project may be tied to financial gain (equity or interest) or non-financial rewards, or pure social impact. Examples of DLT representation in crowdfunding can be found in donation-based platforms (BlockBonds), match-funding (GOTEO), equity crowdfunding ecosystems (RealMarket), and more. New versions keep developing, like the 2018-founded platform WHIRL which is based on a “pay it forward” principle. Here a new project can only be initiated following active support of other projects by the founder(s), inspiring a perpetual loop of generosity and good reputation (Baber, 2019).
3.6 Derivative markets and smart contracts As author Shermin Voshgmir puts it in the book Token Economics (Locklin, 2019) tokens are the most promising application of crypto-based technologies, and beyond ICOs there exist 89
R.R. Mukkamala, R. Vatrapu
security tokens. These represent security, a type of asset which is reliant upon or derived from, an underlying asset or group of assets such as stocks, bonds, interest rates, currencies etc. The asset and token are then interlinked via smart contracts, and these are tradeable financial instruments which can represent fractions of the total value of an asset. According to the European Securities and Markets Authority, security tokens can qualify as transferable securities under the MiFID directive, which has the role of protecting investors, but that the classification of each case is the role of the individual member states (ESMA, 2019, p. 4). The report highlights no disruption risk to traditional securities trading services, and even the European Central Bank has been exploring possible use-cases (ESMA, 2019). Due to the way in which DLTs simplify the process of issuing, and reduce the duration of clearing and settlement, such processes can become simplified also on an institutional level. This is probably also where the largest success potential lies, in financial institutions taking up the securities to create liquidity in a decentralized way, skipping the step of fourth-party governance. It could lead to a significant lowering of cost for both investors and issuers, and therefore not just influence the transaction process but also the types of trading and exchanges that exist (EU Commission, 2019). Places where security exchanges and clearing houses have actively been testing blockchain-based models are national stock exchanges. In 2017 the Australian Stock Exchange (ASX) started planning the replacement of its widely recognized platform CHESS by blockchain technology from a startup named Digital Asset Holdings. The project is built on the open-source smart contract language DAML and was projected to be in deployment by 2019 but is still ongoing and with more collaboration partners. National stock exchange projects in Japan, Korea and India have been testing the technology for trading in low liquidity, shares of startups, and know-your-customer data protocols respectively (OECD, 2018). There are new experiments happening across many banks and exchanges, experimenting with things such as ways to make digital central bank-issued money available in trade and transaction of tokenized assets amongst different actors in the legacy systems (Swiss Stock Exchange and Swiss National Bank). These are currently still happening on an exploratory level, as the lack of stability and scalability of cryptocurrencies make it hard to reach large adoption. Further, formal regulation and legal adjustments in accordance with securities law are still catching up and leaving the situation in a precarious state.
4 Legal and regulatory perspectives of DLT Blockchain-based technologies and DLTs have attracted significant attention from researchers in the law discipline, especially on aspects of regulation and governance of cryptocurrencies and blockchain-based applications such as smart contracts etc. The relevant extant literature research on the legal aspects of Blockchain regulation, compliance and governance is summarized in detail in Table 6.2. As shown in Table 6.2, we can characterize the current research discussion on blockchain regulation into two distinct and opposing research streams: stringent regulation vs. openminded regulation. First, the stream of research studies (Kleiman, 2013; Ajello, 2014; Lin, 2016) that is primarily concerned with money laundering and digital crime using cryptocurrencies and their economic and social impacts on societies, argues for establishing clear and stringent regulations, compliance protocols and guidance frameworks for the cryptocurrency industry. On the other hand, a significant amount of research (McLeod, 2014; Turpin, 2014; Kiviat, 2015; Sonderegger, 2015; Tsukerman, 2015; Colombo, 2016; Morgan, 2018; Priem, 2020) considers that blockchain is a highly disruptive and innovative technology and 90
Distributed ledger technologies Table 6.2 Research on regulatory initiatives of blockchain and DLTs Article
Primary theme
Main arguments and recommendations
Kleiman (2013) Regulation
Cryptocurrencies can be potential security and economic threats. Argues for establishing clear jurisdictional lines and regulations for the virtual currency industry. Turpin (2014) Regulation Argues for regulation, but is in favor of embracing the new (favorable) technology. Recommends that governments should further study and regulate Bitcoin, but without attempting to stop or slow the growth of the currency itself and without attacking otherwise lawabiding citizens who transact in Bitcoins. Ajello (2014) Regulation, Is concerned with Bitcoin’s money laundering and its economic money and social consequences. Advocates for more stringent regulations laundering and argues that cryptocurrencies deserve greater attention from regulators and law enforcement officials. McLeod (2014) Regulation Argues for regulation, but suggests amending the existing legal provisions in an amicable way to create a workable regulatory model for cryptocurrencies. Sonderegger Regulation Suggests that given Bitcoin’s ideological and technological (2015) (favorable) underpinnings, it requires a degree of regulatory freedom to succeed. Argues that proper regulation will not stifle innovation but will allow it to self-regulate within a vaguely defined regulatory framework. Kiviat (2015) Digital Assets, Argues that the true value of technology lies in its potential to Regulation facilitate more efficient digital-asset transfers and advocates that (favorable) policymakers carefully define the specific activities that they seek to regulate. Tsukerman Regulation Claims that unmasking actors on the blockchain will help Bitcoin (2015) (favorable) shed its infamous reputation and that Bitcoins must be brought into the light and seen as a useful currency, and not simply as the refuge of dark web inhabitants. Colombo (2016) Regulation Argues that responsibility of regulators and lawmakers is to (favorable) establish rules that safeguard consumers and markets without hindering growth and innovation. Opines that it will be difficult to get right when dealing with something as new and alien (to fiat currency regulatory apparatus) as virtual currency. Lin (2016) Cybersecurity, Focuses on the challenges of financial cybersecurity, technology compliance integration, compliance, and the role of humans in the future of modern finance. Lee (2016) Cyber Argues that blockchain will create crypto securities that will securities and allow the public to verify transactions if they want, which will stock markets remove some of the hidden secrecy surrounding much of the high frequency and dark pool trading occurring today. Gabison (2016) Regulation, Argues for the need for policymakers to reinvestigate several laws accountability and rights for blockchain. Fears that lack of accountability to a policing authority exposes users to attacks and allows potential transfers that finance criminal activities. Christopher Enforcement Argues that Bitcoin requires more trust than is generally understood (2016) and trust and both currency and payment systems benefit from the involvement of trusted intermediaries in response to problems and crises. (Continued)
91
R.R. Mukkamala, R. Vatrapu Article
Primary theme
Main arguments and recommendations
Rosner and Kang (2016)
Flexible regulation
Argues for a flexible and principles-based approach to amend current regulatory frameworks to account for modern technological realities. Claims that the cryptocurrency Ripple’s advantages suggest that users will increasingly use these systems in place of traditional payment. processes. Examines blockchains through the lens of polycentric governance to ascertain what could be done to build trust in distributed systems and ultimately promote cyber peace. Assess that it will take many years to build sustainable blockchain system by involving numerous stakeholders and policymakers. Discusses whether trade secret or copyright protection should apply to protect the claims and uses of blockchain technology and states that only time will tell if blockchain technology can be claimed as intellectual property or be used in court. Advocates for usage of blockchain technologies for legal systems (crypto law) and argues that crypto law could offer a legal discourse to serve more rapidly, more efficiently, more transparently, and in creative ways, that may encourage increased civic engagement. Claims that regulatory emphasis on the threat posed by cryptocurrencies has created a hostile environment to innovation. Advises establishing a national charter for FinTech to absorb the full potential of blockchain technology. Advocates that blockchains with self-executing smart-contracts provide compelling opportunities in derivatives markets and that they can reduce dependency on central counterparties that are exposed to large amounts of credit risk. Argues that semantic contracts offer two forms of flexibility: linguistic ambiguity and enforcement discretion, which are important in the contracting process and therefore smart contracting will impose more costs than semantic contracts. Argues that intermediaries’ role is crucial in the processes of firms’ strategies and contracting and therefore cannot be replaced by blockchain but argues for private blockchains for archiving purposes within standard registration systems. Argues for the regulation of cryptocurrencies with a more openminded approach to promote truly informed policy decisions as opposed to irrational and poor investment decisions. Advocates for a smart constitution, to make the government operate transparently and within its mandate; also argues that “When something is codified, and connected to the blockchain, code is law. When the code is the law, any entity tied to it is powerless to act outside of the code. This will ensure that governments stay within their expressed powers.” Young (p.71, 2018). Explores the initiatives for the regulation of crypto-based businesses such as ICOs in the UK, Estonia and Switzerland; posits that these are friendly companies when it comes to doing digital businesses.
Shackelford and Cyber security, Myers (2017) regulation
Guo (2017)
Patents for blockchain technology
Reyes (2017)
Crypto-legal structures
Ross (2017)
Regulation, National charter
Surujnath (2017)
Derivatives markets
Sklaroff (2017)
Smart vs semantic contracts
Arrun˜ada (2018)
Smart contracts for property rights
Morgan (2018)
Open-minded regulation
Young (2018)
Smart constitution, smart social contract
Suciu et al. (2019)
Regulation for crypto-based businesses
92
Distributed ledger technologies Article
Primary theme
Main arguments and recommendations
(Franks, 2020)
Records management Policy debate on regulation of DLTs
Presents several opportunities and challenges in terms of using DLT for records management and information governance. Explores which regulatory barriers need to be removed for full adoption of DLTs; identifies the challenges and risks related to DLT systems; provides a good summary of the regulatory initiatives in EU for DLTs.
(Priem, 2020)
argues for a more open-minded regulation without attempting to stop or slow its growth with suggestions to amend the existing legal provisions if necessary to create a workable regulatory model. Moreover, it is argued that a proper regulatory model that does not constrain the innovation of cryptocurrencies will allow them to self-regulate within a vaguely defined regulatory framework; at the same time revealing the identities of the actors in case of necessities (e.g. money laundering) will help the cryptocurrencies to get rid of their infamous reputation and potentially revolutionize organizations. Apart from cryptocurrencies such as Bitcoin, prior research also focused on the regulation and compliance in terms of using blockchain and DLTs for various applications such as digital-asset transfers (Kiviat, 2015), property rights (Arrun˜ada, 2018), crypto securities (Lee, 2016), derivatives markets (Surujnath, 2017), smart contracts (Sklaroff, 2017), records management (Franks, 2020) and so on for financial, accounting and other administrative domains. Unlike the case of cryptocurrencies, the research from the law disciplines has argued for the usage of blockchain technology for developing applications in these areas, as it would enhance transparency by removing hidden secrecy and provide a way for more efficient document and authorship verification, title transfers, and contract enforcement. Finally, just to provide an example of the scope of research regarding blockchain regulation and governance Young (2018) advocated for a smart constitution, a blockchain-based implementation for governance, which would make government operate in compliance with smart constitution laws visibly and would also prohibit it from operating outside of its mandate. Given that this extensive discussion on regulation and compliance is important for trusted third-party providers, it is apparent that governmental agencies or regulators need to implement flexible regulatory and compliance measures for cryptocurrencies without burdening law-abiding citizens who transact with cryptocurrencies and DLTs within the legal framework.
5 Conclusion In this book chapter, we discussed various principles and concepts behind the DLT and blockchain technologies such as hash functions, digital signatures and digital timestamping which form the foundations of the DLTs and blockchains. We have also introduced how various consensus protocols and fault tolerance mechanisms can help to achieve consensus in a distributed and decentralized environment like blockchain and DLTs. In addition to that, a detailed description of how DLT and blockchain applications can impact the FinTech applications is also provided. We also summarized the existing research and initiatives in terms of legal and regulatory perspectives on DLT and blockchain. 93
R.R. Mukkamala, R. Vatrapu
5.1 Regulation For the last few years, there has been a growing demand for regulatory measures on blockchain applications, especially for cryptocurrencies. Since cryptocurrencies employ a high level of anonymity (Sun Yin et al., 2019), they have been labelled as the go-to currencies for illicit activity. The shutdown of the drug market Silk Road provides the most well-known example (Christin, 2013). Moreover, recent articles and reports (Hout and Bingham, 2013; Martin, 2014a; Martin, 2014b) have stated that Bitcoin has been used for terror financing, thefts, scams, and ransomware. Financial regulators, law enforcement, intelligence services, and companies who transact on the Bitcoin blockchain have become wary observers of technical developments in Bitcoin, economic issues with it, and the societal implications of its adoption (Ali et al., 2015; B¨ohme et al., 2015). In light of these developments, there has been a great demand for regulatory measures to be put in place to control the illicit activities associated with cryptocurrencies and other blockchain applications. Although there has been no specific regulation adopted for cryptocurrencies and blockchain applications, there have been several initiatives (Financial Conduct Authority, 2017; Financial Stability Board, 2018) to examine the need for new regulation or whether to modify existing regulation in order to harness the full potential of blockchain technology. However, many regulatory authorities are still monitoring the rapid growth of these technologies and their influence on FinTech applications. As discussed in the previous section, many regulators and legal authorities still think that blockchain technology is in its infancy stage although evolving at a fast rate, and that it will be necessary to study challenges and barriers in a much more thorough way before putting in place a regulatory framework for DLT and blockchain applications.
5.2 Market adoption Technological advancements are a necessary but not sufficient condition for the widespread market adoption of DLT-based FinTech products and services. Advancements in DLT are necessary to address the known critical issues of low transaction volume, energy inefficiencies, and regulatory challenges in terms of cybersecurity vulnerabilities and cybercriminal activities. That said, technological advancements in DLT are not sufficient as established players and entrenched interests still experience favorable market conditions such as highly standardized terms and inter-institutional operations. In conclusion, technological advancements in DLT intertwined with service innovations will continue to play an influential role in the evolution of FinTech in the near future.
Notes
94
Distributed ledger technologies
References Adeyanju, Craig (2017) How a loyalty program on blockchain works, explained, December. https:// cointelegraph.com/explained/how-a-loyalty-program-on-blockchain-works-explained Ajello, Nicholas J. (2014) Fitting a square peg in a round hole: Bitcoin, money laundering, and the fifth amendment privilege against self-incrimination. Brooklyn Law Review, 80(2): 435. Akgiray, Vedat (2019) The potential for blockchain technology in corporate governance. Technical report, OECD Corporate Governance Working Papers. Ali, Syed Taha, Dylan Clarke and Patrick McCorry (2015) Bitcoin: Perils of an unregulated global p2p currency. In Cambridge International Workshop on Security Protocols. Springer. Antonopoulos, Andreas M. (2014) Mastering Bitcoin: unlocking digital cryptocurrencies. O’Reilly Media, Inc. Arruñada, Benito (2018) Blockchain’s struggle to deliver impersonal exchange. Minnesota Journal of Law, Science & Technology, 19(1): 55. Baber, Hasnan (2019) Blockchain-based crowdfunding: A ‘pay-it-forward’model of whirl. International Journal of Recent Technology and Engineering (IJRTE), 8(3): 2277–3878. Back, Adam (2002) Hashcash-a denial of service counter-measure. http://www.hashcash.org/papers/ hashcash.pdf. Bayer, Dave, Stuart Haber and W. Scott Stornetta (1993) Improving the efficiency and reliability of digital time-stamping. In Sequences Ii. Springer, pp. 329–334. Ben-Sasson, Eli, Iddo Bentov, Yinon Horesh and Michael Riabzev (2019) Scalable zero knowledge with no trusted setup. In Annual International Cryptology Conference. Springer, pp. 701–732. BÖhme, Rainer, Nicolas Christin, Benjamin Edelman and Tyler Moore (2015) Bitcoin: Economics, technology, and governance. The Journal of Economic Perspectives, 29(2): 213–238. Carter, J. Lawrence and Mark N. Wegman (1979) Universal classes of hash functions. Journal of Computer and System Sciences, 18(2): 143–154. ISSN 0022-0000. http://www.sciencedirect.com/science/ article/pii/0022000079900448. Castro, Miguel Barbara Liskov, et al. (1999) Practical byzantine fault tolerance. OSDI, 99: 173–186. Chohan, Usman W. (2017) A history of bitcoin. https://papers.ssrn.com/sol3/papers.cfm?abstract_ id=3047875. Christin, Nicolas (2013) Traveling the silk road: A measurement analysis of a large anonymous online marketplace. In Proceedings of the 22nd International Conference on World Wide Web, WWW. New York, NY, USA, pp. 213–224. ACM. ISBN 978-1-4503-2035-1. Christopher, Catherine Martin (2016) The bridging model: Exploring the roles of trust and enforcement in banking, bitcoin, and the blockchain. Nevada Law Journal, 17: 139. Colombo, Ronald J. (2016) Bitcoin: Hype or harbinger. Journal of International Business and Law, 16(1): 1. Dalvit, Lorenzo (2020) Ethereum: A new record of transactions per second thanks to Istanbul, January. https://en.cryptonomist.ch/2020/01/07/ethereum-transactions-per-second-istanbul/. Damård, Ivan Bjerre (1989) A design principle for hash functions. In Conference on the Theory and Application of Cryptology. Springer, pp. 416–427. Dowlat, Sherwin and Michael Hodapp (2018) Ico quality: Development and trading, March. https:// www.bitcoininsider.org/article/21901/ico-quality-development-trading. Dworkin, Morris J. (2015) Sha-3 standard: Permutation-based hash and extendable-output functions. Technical report, National Institute of Standards and Technology (NIST). ESMA (2019) Advice, initial coin offerings and crypto-assets. Technical report, European Securities and Markets Authority.
95
R.R. Mukkamala, R. Vatrapu European Central Bank (2012) Virtual currency schemes. Frankfurt am Main, Germany. European Central Bank (2015) Virtual currency schemes: A further analysis. Technical report, Frankfurt am Main, Germany. EU Commission (2019) Potential use cases for innovative technologies in securities post-trading. Technical report, European Central Bank. Faber, Benedict, Georg Cappelen Michelet, Niklas Weidmann, Raghava Rao Mukkamala, and Ravi Vatrapu (2019) Bpdims: A blockchain-based personal data and identity management system. In Proceedings of the 52nd Hawaii International Conference on System Sciences. Financial Conduct Authority (2017) Discussion paper on distributed ledger technology. Technical report, Financial Conduct Authority. Financial Stability Board (2018) Crypto-assets: Report to the g20 on the work of the fsb and standardsetting bodies. Technical report, Financial Stability Board, Basel, Switzerland. Franks, Patricia C. (2020) Implications of blockchain distributed ledger technology for records management and information governance programs. Records Management Journal. Gabison, Garry (2016) Policy considerations for the blockchain technology public and private applications. Science and Technology Law Review, 19: 327. Goldwasser, Shafi, Silvio Micali, and Charles Rackoff (1989) The knowledge complexity of interactive proof systems. SIAM Journal on Computing, 18(1): 186–208. Guo, Angela (2017) Blockchain receipts: Patentability and admissibility in court. Chicago–Kent Journal of Intellectual Property, 16(2): 440. Haber, Stuart and W. Scott Stornetta (1990) How to time-stamp a digital document. In Conference on the Theory and Application of Cryptography. Springer, pp. 437–455. Heilman, Ethan, Neha Narula, Garrett Tanzer, James Lovejoy, Michael Colavita, Madars Virza, and Tadge Dryja (2019) Cryptanalysis of Curl-P and other attacks on the IOTA cryptocurrency. IACR Cryptol. ePrint Arch.: 344. https://eprint.iacr.org/2019/344.pdf. Hout, Marie Claire Van and Tim Bingham (2013) Silk road, the virtual drug marketplace: A single case study of user experiences. International Journal of Drug Policy, 24 (5): 385–391. ISSN 0955–3959. http://www.sciencedirect.com/ science/article/pii/S0955395913000066. IFRS (2020) List of IFRS Standards, September. https://www.ifrs.org/issued-standards/list-ofstandards/. Ijiri, Yuji (1986) A framework for triple-entry bookkeeping. Accounting Review, 745–759. Johnson, Don, Alfred Menezes and Scott Vanstone (2001) The elliptic curve digital signature algorithm (ECDSA). International Journal of Information Security, 1(1): 36–63. Kiviat, Trevor I. (2015) Beyond bitcoin: Issues in regulating blockchain tranactions. Duke Law Journal, 65: 569. Kleiman, Jared A. (2013) Beyond the silk road: Unregulated decentralized virtual currencies continue to endanger US national security and welfare. American University National Security Law Brief, 4(1): 59. Lamport, Leslie (2019) The part-time parliament. In Concurrency: the Works of Leslie Lamport, pp. 277– 317. ACM Books. Lamport, Leslie, Robert Shostak and Marshall Pease (1982) The byzantine generals problem. ACM Transactions on Programming Languages and Systems (TOPLAS), 4(3): 382–401. Lamport, Leslie et al. (2001) Paxos made simple. ACM Sigact News, 32(4): 18–25. Lee, Larissa (2016) New kids on the blockchain: How bitcoin’s technology could reinvent the stock market. Hastings Business Law Journal, 12(2): 81. Lewtan, Jacob, Joseph McManus and Saeed Roohani (2018) Blockchain: Opportunity to improve financial reporting and corporate governance. In Hawai’i Accounting Research Conference (HARC). Lin, Tom C.W. (2016) Compliance, technology, and modern finance. Brooklyn Journal of Corporate, Financial and Commercial Law, 11(1): 159. Locklin, Scott (2019) Token economics. Technical report, Brave Software. Mamoshina, Polina, Lucy Ojomoko, Yury Yanovich, Alex Ostrovski, Alex Botezatu, Pavel Prikhodko, Eugene Izumchenko, Alexander Aliper, Konstantin Romantsov, Alexander Zhebrak, et al. (2018) Converging blockchain and next-generation artificial intelligence technologies to decentralize and accelerate biomedical research and healthcare. Oncotarget, 9(5): 5665. Martin, James (2014a) Drugs on the dark net: How cryptomarkets are transforming the global trade in illicit drugs. Springer. Martin, James (2014b) Lost on the silk road: Online drug distribution and the cryptomarket. Criminology & Criminal Justice, 14(3): 351–367. http://dx.doi.org/10.1177/1748895813505234.
96
Distributed ledger technologies McLeod, Patrick (2014) Taxing and regulating bitcoin: The government’s game of catch up. CommLaw Conspectus, 22: 379. Merkle, Ralph C. (1980) Protocols for public key cryptosystems. In Security and Privacy, 1980 IEEE Symposium on, pp. 122–122. IEEE. Merkle, Ralph C. (1989) One way hash functions and DES. In Conference on the Theory and Application of Cryptology, pp. 428–446. Springer. Molina-Jimenez, Carlos, Ioannis Sfyrakis, Ellis Solaiman, Irene Ng, Meng Weng Wong, Alexis Chun and Jon Crowcroft (2018a) Implementation of smart contracts using hybrid architectures with on and off–blockchain components. In 2018 IEEE 8th International Symposium on Cloud and Service Computing (SC2), pp. 83–90. IEEE. Molina-Jimenez, Carlos, Ellis Solaiman, Ioannis Sfyrakis, Irene Ng, and Jon Crowcroft (2018b) On and off-blockchain enforcement of smart contracts. In European Conference on Parallel Processing, pp. 342–354. Springer. Morgan, Joshua S. (2018) What I learned trading cryptocurrencies while studying the law. University of Miami International and Comparative Law Review, 25:159. Nakamoto. Satoshi (2008) Bitcoin: A peer-to-peer electronic cash system. [White Paper] https://bitcoin. org/bitcoin.pdf. Narayanan, Arvind and Jeremy Clark (2017a) Bitcoin’s academic pedigree. Communications of the ACM, 60(12):36–45, November. https://doi.org/10.1145% 2F3132259. Narayanan, Arvind and Jeremy Clark (2017b) Bitcoin’s academic pedigree. Communications of the ACM, 60(12): 36–45. Natarajan, Harish, Solvej Karla Krause and Helen Luskin Gradstein (2019) Distributed ledger technology (DLT) and blockchain. Fintech note; no. 1. Washington, DC: World Bank Group. OECD (2018) The potential for blockchain technology in public equity markets in Asia. Technical report, OECD Capital Market Series. Penard, Wouter and Tim van Werkhoven (2008) On the secure hash algorithm family. Cryptography in Context, pp. 1–18. Priem, Randy (2020) Distributed ledger technology for securities clearing and settlement: Benefits, risks, and regulatory implications. Financial Innovation, 6(1): 1–25. Rauchs, Michel, Andrew Glidden, Brian Gordon, Gina C Pieters, Martino Recanatini, Francois Rostand, Kathryn Vagneur, and Bryan Zheng Zhang (2018) Distributed ledger technology systems: A conceptual framework. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3230013. Rauchs, Michel, Apolline Blandin, Keith Bear, and Stephen B. McKeon (2019) Second global enterprise blockchain benchmarking study. Technical report, Cambridge Centre for Alternative Finance (CCAF). Reyes, Carla L. (2017) Conceptualizing cryptolaw. Nebraska Law Review, 96(2): 384. Rivest, Ronald L., Adi Shamir, and Len Adleman (1978) A method for obtaining digital signatures and public-key cryptosystems. Communications of the ACM, 21(2): 120–126. Rosner, Marcel T. and Andrew Kang (2016) Understanding and regulating twenty-first century payment systems: The ripple case study. Michigan Law Review, 114(4): 649. Ross, Elizabeth Sara (2017) Nobody puts blockchain in a corner: The disruptive role of blockchain technology in the financial services industry and current regulatory is- sues. Catholic University Journal of Law and Technology, 25(2): 7. Shackelford, Scott J. and Steve Myers (2017) Block-by-block: Leveraging the power of blockchain technology to build trust and promote cyber peace. Yale Journal of Law and Technology, 19(1): 334. Simmons, Gustavus J. (1994) Contemporary cryptology: The science of information integrity. IEEE Press. Sklaroff, Jeremy M. (2017) Smart contracts and the cost of inflexibility. University of Pennsylvania Law Review, 166: 263. Sonderegger, Daniela (2015) A regulatory and economic perplexity: Bitcoin needs just a bit of regulation. Washington University Journal of Law & Policy, 47: 175. Suciu, Christina Marta, Christian Nasulea and Diana Florentina Nasulea (2019) To- wards developing a friendlier regulatory framework for blockchain-based businesses. Economic Convergence in European Union, p. 15. Sun Yin, Hao Hua, Klaus Langenheldt, Mikkel Harlev, Raghava Rao Mukkamala and Ravi Vatrapu (2019) Regulating cryptocurrencies: A supervised machine learning approach to de-anonymizing the bitcoin blockchain. Journal of Management Information Systems, 36(1): 37–73. Surujnath, Ryan (2017) Off the chain: A guide to blockchain derivatives markets and the implications on systemic risk. Fordham Journal of Corporate & Financial Law, 22 (2): 257.
97
R.R. Mukkamala, R. Vatrapu Swan, Melanie (2015) Blockchain: Blueprint for a new economy. O’Reilly Media, Inc. Szabo, Nick (1997) Formalizing and securing relationships on public networks. First Monday, 2(9), Sept. ISSN 1396–0466. http://dx.doi.org/10.5210/fm.v2i9.548. Szostek, Dariusz (2019) Blockchain and the Law. Nomos Verlag. Treleaven, Philip, Richard Gendal Brown, and Danny Yang (2017) Blockchain technology in finance. Computer, 50(9): 14–17. Tsukerman, Misha (2015) The block is hot: A survey of the state of bitcoin regulation and suggestions for the future. Berkeley Technology Law Journal, 30(4): 1127. Turpin, Jonathan B. (2014) Bitcoin: The economic case for a global, virtual currency operating in an unexplored legal framework. Indiana Journal of Global Legal Studies, 21(1): 335–368. Tyra, Jason M. (2014) Triple entry bookkeeping with bitcoin, February. https://bitcoinmagazine. com/articles/triple-entry-bookkeeping-bitcoin-1392069656. Wang, Wenbo, Dinh Thai Hoang, Peizhao Hu, Zehui Xiong, Dusit Niyato, Ping Wang, Yonggang Wen, and Dong In Kim (2019) A survey on consensus mechanisms and mining strategy management in blockchain networks. IEEE Access, 7: 22328–22370. Weber, Conny (2019) Exploring DLT and blockchain for alternative finance: A collection of case studies. Technical report, European Crowdfunding Network, November. Xu, Lihao (1999) Highly available distributed storage systems. PhD thesis, California Institute of Technology. Young, Steve (2018) Enforcing constitutional rights through computer code. Catholic University Journal of Law and Technology, 26(1): 1.
98
7 INITIAL COIN OFFERINGS A statistical analysis of the main characteristics Paola Cerchiello and Anca Mirela Toma
1 Initial coin offerings Initial coin offerings (ICOs) became famous as the finance model of cryptocurrencies. They are a digital way of public capital funding for entrepreneurial use through the issue of an own virtual token [1]. A token is a ‘cryptographically’ secured digital asset’ ([2], p.2). For companies whose business model stands in connection with blockchain technology, ICOs have surpassed the traditional venture capital financing in the shortest of time [3]. This means that ICOs are a new way to raise capital for young and unestablished ventures. The first ICO was held in July 2013 by Mastercoin, which is a digital currency built on the Bitcoin blockchain [4]. Since then, over 1,000 ICOs have followed. CoinSchedule, a leading website monitoring current ICOs, reports that 366 ICOs took place in 2017, raising a combined amount of USD 6.2 bn. According to Fisch [5], the aggregated 2017 funding volume was surpassed in the first three months of 2018 alone; 254 ICOs raised USD 7.8 bn in this period. The premier crowdfunding platform Kickstarter in contrast has raised a total of USD 4.6 bn since its inception in 2009 [6]. In the month of June 2018 alone, the ICO funds raised the amount to over USD 5 bn with 91 ICOs ending in that period. However, since then, monthly funds raised remained more or less distinctly under the mark of USD 1 bn except for May 2019. The number of token sales has also declined from the staggering number of 146 ended ICOs at the maximum in April 2018 to four ended token sales in September 2019 [7]. It is to be noted that an officially recognized definition of ICOs does not exist [1]. The name initial coin offering is a reference to the well-established concept of initial public offerings (IPOs). However, at first sight, ICOs have relatively few things in common with traditional public offerings. Through an ICO, a new firm offers a token to a crowd of investors for the first time. In IPOs the company is most often already established and has had rather a successful past. In an IPO, shares of the company are sold. In an ICO, the sold token is created by the firm offering it using distributed ledger technology (DLT) and can be bought in exchange for at money or other cryptocurrencies. The functions of the token may equal classical shares but are manifold [8]. The spike in the occurrence of ICOs followed the development of the Blockchain by Nakamoto in 2008 and the subsequent development of cryptocurrencies such as Ethereum (short: ether) [9]. A blockchain enables the direct, secure transfer of value over the Internet 99
Paola Cerchiello, Anca Mirela Toma
between parties that do not trust each other [2]. It consists of a sequential list of transactions in a unit of value that is native to the blockchain. For example, the Bitcoin blockchain uses the cryptocurrency bitcoin and the Ethereum blockchain uses ether. Additional text, such as contingent terms of contracts, can be appended to a transaction. Bitcoin permits simple and limited additional text, whilst other blockchains such as Ethereum essentially permit any code to be executed as part of a transaction. Blockchains are so-called distributed ledgers, providing decentralized record-keeping that cannot be retroactively edited. Cryptography enables rapid verification and prevents hacking [2]. Distributed ledger technology (DLT) describes a decentralized database stored on a set of individual nodes. The records are synchronized through a consensus algorithm, which allows peer-to-peer transactions without the need for an intermediary. This is the technical basis of an ICO in various forms [1]. The token sales themselves are conducted via complex, self-enforcing and state-contingent smart contracts, which are pieces of code embedded in a blockchain. This enables the exchange of money, property or other assets without an intermediate party. Smart contracts guarantee the fulfilment of the transaction and regulate for example the conditions of sale for the tokens [1]. Therefore, due to low transaction costs similar to crowdfunding, ICOs might become a significant driver for financial inclusion by democratizing access to investments and capital [10]. Recently, there has been a growing literature studying the ICOs drivers aiming to predict their future outcome. A study by [1] offers an exploratory empiric classification of ICOs and the dynamics of voluntary disclosures. It examines to what extent the availability and quality of the information disclosed can explain the characteristics of success and failure among ICOs and the corresponding projects. Another important research focuses on the effectiveness of signalling ventures and ICO projects’ technological capabilities to attract higher amounts of funding [5]. Other streams of research concentrate on the impact of managers quality on the ICOs. Momtaz (2019) studies the impact of CEOs’ loyalty disposition and the magnitude of asymmetry of information between managers and investors on ICOs’ performance [11]. Moreover, to remain in the management area, an interesting spark comes from research specifically directed at CEOs’ roles and effects on ICO results [12]. Finally, another area of study focuses on the driving factors impacting the liquidity and trading volume of crypto-tokens listed after the ICOs. Identified among these factors have been the quality level of disclosed documentation (source code public on GitHub white paper published, an intended budget published for use of proceeds), the community engagement (measured by the number of Telegram group members), the level of preparation of the management (using as proxy the entrepreneurial professional background of the lead founder or CEO), and other outcomes of interest (i.e., the amount raised in the ICO, outright failure – delisting or disappearance, abnormal returns, and volatility) [2]. Despite the interest created by ICOs and the constantly growing trends, it is worth mentioning that almost half of ICOs sold in 2017 had failed by February 2018. What should drive more attention to ICOs is the consistent presence of scam activities only devoted to fraudulently raising money. According to Cointelegraph, the Ethereum network (the prevalent blockchain platform for ICOs) has experienced considerable phishing, Ponzi schemes, and other scam events, accounting for about 10% of ICOs. On the other hand, it is interesting to assess which factors affect the probability of success of an ICO. Adhami et al. in 2018, based on the analysis of 253 ICOs, showed that the following characteristics contribute: the availability of the code source, the organization of a token pre-sale and the possibility for contributors to access to a specific service (or to share profits)[13]. 100
Initial coin offerings: statistical analysis
The main source of information about blockchain, tokens and ICOs is the Web. Here we can find sites enabling exploration of the various blockchains associated with the main cryptocurrencies, including Ethereum’s. We can also find websites giving extensive financial information on prices of all the main cryptocurrencies and tokens, sites specialized in listing the existing ICOs and giving information about them. Often, these sites evaluate the soundness and likelihood of success of the listed ICOs. One of the most popular among these sites is icobench.com, which evaluates all the listed ICOs and provides an API (Application Programming Interface) to automatically gather information on them. ICOs are usually characterized by the following features: a business idea, most of the time explained in a white paper, a proposed team, a target sum to be collected, a given number of tokens to be given to subscribers according to a predetermined exchange rate with one or more existing cryptocurrencies. Nowadays, a high percentage of ICOs are managed through Smart Contracts running on Ethereum blockchain, and in particular through ERC-20 Token Standard Contract [14]. On top of all the characteristics explained so far, there is a further and not yet explored point of interest: the Telegram chats. Telegram is a cloud-based instant messaging and voice over IP service developed by Telegram Messenger founded by the Russian entrepreneur Pavel Durov. In March 2018, Telegram stated that it has 200 million monthly active users – “This is an insane number by any standards. If Telegram were a country, it would have been the sixth-largest country in the world” (Telegram, 2018). Telegram is completely free and has no ads, users can send any kind of media or documents and can program messages to self-destruct after a certain period. Some characteristics are imposing Telegram among the first social networks; indeed it intentionally does not collect data about where its clients live and what they use the platform for. This is one of the main reasons why, according to AppAnnie rankings, Telegram is particularly popular in countries like Uzbekistan, Ukraine, and Russia, where Internet access may be limited or closely monitored by the government. As of October 2017, Telegram was by far the most popular offcial discussion platform for current and upcoming ICOs, with 75%+ of these projects employing it. This means that retrieving Telegram discussions associated with each and every ICO would produce a huge amount of textual information potentially useful for understanding the chance of success and more interestingly possible signs of fraudulent activities. In this chapter, we explain how to leverage two kinds of information: structured and unstructured ones. Regarding the former, we take advantage of classical statistical classification models to distinguish the status of an ICO that is made up of two classes, intended as follows: • •
Success = 1: the ICO collects the predefined hard cap within the time horizon of the campaign; Failure/scam = 0: the ICO does not collect the predefined hard cap within the time horizon of the campaign.
Logistic regression aims at classifying the dependent variable into two groups characterized by a different status [1=scam vs 0=success or 1=success vs 0=failure] according to the following model: p ln i = α + β j xij (1) 1− pi j
∑
where pi is the probability of the event of interest, for ICO i, x i = (x i1,…, x ij,…, x iJ) is a vector of ICOs specific explanatory variables, the intercept parameter, as well as the regression 101
Paola Cerchiello, Anca Mirela Toma
coefficientsj, for j = 1; :::; J, are to be estimated from the available data. It follows that the probability of success (or scam) can be obtained as: pi =
1 1 + exp(α +
∑ β x ) (2) j
j
ij
Since the target variable is naturally categorized according to three classes, success, failure and scam we extend the aforementioned binary logistic regression to a multinomial one. Such model assesses all the categories of interest at the same time as follows: p ln k = αk + 1− pk
∑β x
k ij
(3)
j
where pk is the probability of kth class for k = 1;…… K given the constraint P that
∑p = 1. k
k
Considering the textual analysis of Telegram chats, we take advantage of the quantitative analysis of human languages to discover common features of written text. In particular, the analysis of relatively short text messages like those appearing on the micro-blogging platform presents several challenges. Some of these are the informal conversation (e.g. slang words, repeated letters, emoticons) and the level of implied knowledge necessary to understand the topics of discussion. Moreover, it is important to consider the high level of noise contained in the chats, witnessed by the fact that only a fraction of them with respect to the total number available is employed in our sentiment analysis. We have applied a Bag of Word (BoW) approach, according to which a text is represented as an unordered collection of words, considering only their counts in each comment of the chat. The word and document vectorization has been carried out by collecting all the word frequencies in a Term Document Matrix (TDM). Afterwards, such matrix has been weighted by employing the popular TF-IDF (Term Frequency Inverse Document Frequency) algorithm. Classical text cleaning procedures have been put in places like stop-words, punctuation, unnecessary symbols and space removal, specific topic words addition. For descriptive purposes, we have used wordclouds for every Telegram chat according to the general content and to specific subcategories like sentiments and expressed moods. The most critical part of the analysis relies on sentiment classification. In general, two different approaches can be used: •
•
Score dictionary-based: the sentiment score is based on the number of matches between a predefined list of positive and negative words and terms contained in each text source (a tweet, a sentence, a whole paragraph); Score classifier-based: a proper statistical classifier is trained on a large enough dataset of pre-labelled examples and then used to predict the sentiment class of a new example.
However, the second option is rarely feasible because to be a good classifier, a large amount of pre-classified examples are needed and this represents a particularly complicated task when dealing with short and extremely non-conventional text like micro-blogging chats. Therefore, we decided to focus on a dictionary-based approach, adapting appropriate lists of positive and negative words relevant to ICO topics in the English language. The lexicons used are based on unigrams, i.e., single words; they contain many English words and the words are labelled with scores for positive/negative sentiment and also possibly emotions like joy, anger, sadness, and so forth. Among the possible vocabularies, we consider the following ones: NRC, BING and AFINN. The NRC lexicon categorizes words in a binary fashion (yes/no) into categories of positive, negative, anger, anticipation, 102
Initial coin offerings: statistical analysis
disgust, fear, joy, sadness, surprise, and trust. The BING lexicon categorizes words into a binary manner into positive and negative categories. The AFINN lexicon assigns words with a score that runs between -5 and +5, with negative scores indicating negative sentiment and positive scores indicating positive sentiment. By applying the above lexicons, we produce for every ICO a sentiment score as well as counts for positive and negative words. All these indexes are used as additional predictors within the logistic models.
2 Data As already mentioned, we exploit structured and unstructured information and empirically examine 195 ICOs starting from January 2017 till November 2018. The first step in collecting data about each project is to gather information from the most used ICO related platforms such as Icobench, TokenData, Coinschedule or similar. During this phase, we look for general characteristics such as the name, the token symbol, start and end dates of the crowdfunding, the country of origin, financial data such as the total number of tokens issued, the initial price of the token, the platform used, data on the team proposing the ICO, data on the advisory board, data on the availability of the website, availability of white paper, white-paper characteristics, and social channels. Some of these data, such as short and long description, and milestones are textual descriptions. Others are categorical variables, such as the country, the platform, the category (which can assume many values), and variables related to the team members (name, role, group). The remaining variables are numeric, with different degrees of discretization. As concerns the unstructured data, insightful information can be derived from the white papers in terms of quality of the technical report and specific content. A white paper is a summary report that provides detailed information about the project, its originality and the benefits available to investors and users, the technological features, the team behind the project, the project’s background and plans. The dimensions captured through the white paper features consist in checking the availability of the document for the ICO in question and additional information about its scope (pages, sections, appendix). Furthermore, data on the disclosure of information regarding the issuers were collected (names and/or photographs of issuing and advising team members). Social channels are more personal than every database, rating platform or website, so they are a way to reach a wide range of users, to update them constantly about the evolution of the project and in the end to create a trusted environment that can finalize in successful crowdfunding activity. To conduct the textual analysis, we enrich our database with the social channels data, such as the presence of a channel, the numbers of users as a proxy of the community engagement and as mentioned in the introduction the textual chat, retrieved backwards till the creation of the chat, used to produce a sentiment-based score for every ICO. In Table 7.1 we report the complete list of collected and employed variables.
3 Empirical evidence In this section, we report our main results obtained from classification and textual analysis. In Table 7.2 and Table 7.4 we report results respectively for logistic regression on Success/ Failure (class 2 variable) and multi logit regression estimated on failure (f ) and scam (sc) compared to success as the baseline. Regarding the first model, in Table 7.2 we report the final configuration after several stepwise selection steps. The reader can see that the only two relevant dummy variables are: the presence of a white paper (Paper du) and a Twitter account 103
Paola Cerchiello, Anca Mirela Toma Table 7.1 Explanatory variables Class0 Class1 Class2 W_Site Tm W_Paper Usd Tw Fb Ln Yt Gith Slack Reddit Btalk Mm Nr_Team Nr Adv Adv Project Nr Tm Tot Token Pos Bing Standardized Neg Bing Standardized Pos NRC Standardized Neg NRC Standardized Sent NRC Standardized
f = failed, sc= scam, su= success 0 = scam 1 = failed+ success 0 = failed 1 = success Website (dummy) Telegram White paper (dummy) Presale price in USD Twitter (dummy) Facebook (dummy) LinkedIn (dummy) YouTube (dummy) GitHub (dummy) Slack(dummy) Reddit (dummy) Bitcoin talk (dummy) Medium (dummy) Number of Team members (quantitative) Number of advisors (quantitative) Existence of advisors (dummy) name of the ICO (categorical) Number of users in Telegram (quantitative) Number of Total Tokens (quantitative) no. of positive words for BL list (quantitative) no. of negative words for BL list (quantitative no. of positive words for NRC list (quantitative) no. of negative words for NRC list (quantitative) sentiment for NRC list (quantitative)
(tw). Both present positive coefficients showing their impact on increasing the probability of success of an ICO. It should be stressed that the influence of Twitter channel is much higher than the presence of a white paper, indeed if we calculate the associated odds ratio we would get respectively 11.94 and 3.85. In other words, if the ICO has a Twitter account the probability of success is almost 12 times higher (almost four times higher for the white paper). Regarding the three continuous variables, number of elements of the team (Nr team), number of advisors (Nr adv) and scaled sentiment score based on NRC lexicon (Sent NRC sc), they are all highly significant and again positive suggesting that increasing people and advisors in the team has a positive impact. Regarding the sentiment, we notice a particularly high positive value, stressing the importance of the perception of possible investors who interact with the ICO proposer by means of a social media, namely Telegram. To further evaluate such configuration, we have explored the VIF index (Table 7.3) that accounts for the level of multicollinearity brought by every variable. The VIF results for the two model configurations are reported in Table 7.3 (logistic) and 7.5 (multinomial), with useful insights in defining the lack of multicollinearity. Therefore in Table 7.3, we can see low values for the VIF index associated with the estimated logistic model (given in Table 7.2). The reader can easily notice that there is not any multicollinearity effect, making the model robust. Moreover, reported performance indexes, namely AIC and pseudo R 2, present good values above 50%. 104
Initial coin offerings: statistical analysis Table 7.2 Logistic regression results on success/failure ICOs Dependent variable Class 2 Tw
2.481∗ (1.381) 1.351∗∗ (0.635) 0.461∗∗∗ (0.135) 0.233* (0.088) 0.233* (0.088) 0.233* (0.088) 196 89.41 0.63 0.57 0.49
Paper_du Nr_adv Nr_team Sent_NRC_sc Constant Observations Akaike Inf Crit McFadden pseudo R2 McFadden Adj. pseudo R2 Cox & Snell pseudo R2 Note *p