Too Big to Fail III: Structural Reform Proposals: Should We Break Up the Banks? 9783110421231, 9783110426052

Citation preview

Too Big to Fail III: Structural Reform Proposals ILFS

Institute for Law and Finance Series

Edited by Theodor Baums Andreas Cahn

Volume 16

Too Big to Fail III: Structural Reform Proposals Should We Break Up the Banks? Edited by Andreas Dombret Patrick S. Kenadjian

DE GRUYTER

ISBN 978-3-11-042605-2 e-ISBN (PDF) 978-3-11-042123-1 e-ISBN (EPUB) 978-3-11-042125-5 Library of Congress Cataloging-in-Publication Data A CIP catalog record for this book has been applied for at the Library of Congress. Bibliographic information published by the Deutschen Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the Internet at http://dnb.dnb.de. © 2015 Walter de Gruyter GmbH, Berlin/Munich/Boston Cover Image: Medioimages/Photodisc Printing and binding: Hubert & Co. GmbH & Co. KG, Göttingen ♾ Printed on acid-free paper Printed in Germany www.degruyter.com

Table of Contents Andreas Dombret, Patrick S. Kenadjian Introduction 1 Dr. Andreas Dombret Cutting the Gordian Knot or splitting hairs – The debate about breaking up the banks 5 Dr. Paul Achleitner “What kind of financial system do we want – A global private sector perspective” 13 Jan Pieter Krahnen Rescue by Regulation? Key Points of the Liikanen Report Miguel de la Mano Bank Structural Reform and Too-big-to-fail Debra Stone The Volcker Rule

23

47

95

Adam S. Posen Confronting the Reality of Structurally Unprofitable Safe Banking Douglas J. Elliott Ten Arguments Against Breaking Up the Big Banks

105

115

Randall D. Guynn and Patrick S. Kenadjian Structural Solutions: Blinded by Volcker, Vickers, Liikanen, Glass Steagall and Narrow Banking 125 Simon Gleeson Structural Separation and Bank Resolution

145

The Authors Dr. Paul Achleitner Born 1956 in Linz, Austria, Paul Achleitner was educated at the University of St. Gallen (HSG) and Harvard Business School (HBS). He holds a Ph.D. from HSG and an honorary professorship from WHU – Otto Beisheim School of Management. After four years at Bain & Co. in Boston he joined Goldman Sachs in 1988 where he served in New York, London and Frankfurt in various capacities, since 1994 as a partner of the firm. Between 2000 and 2012 Paul Achleitner was CFO of Allianz in Munich. Besides chairing the Supervisory Board of Deutsche Bank Paul Achleitner serves on the Supervisory Board of Bayer and Daimler as well as on the Shareholders’ Committee of Henkel. He also chairs the German Government Commission of Capital Markets Experts and is a member of the International Advisory Board of Allianz, the Board of Trustees of the Brookings Institution, the European Advisory Board of the Harvard Business School and the Advisory Council of the Munich Security Conference. Since 2013, he is Co-Chairman of the Hong Kong/Europe Business Council. Paul Achleitner is married to Professor Ann-Kristin Achleitner. They have three children.

Prof. Dr. Andreas Cahn Andreas Cahn studied law at the Johann Wolfgang Goethe-University Frankfurt/ Main and at the University of California at Berkeley, where he earned an LL.M. After his Second State Examination in Frankfurt he worked for 6 years as a research assistant at the University of Frankfurt. During this period of time he wrote his doctoral thesis on problems of managers’ liability (published in 1996) as well as his post-doctoral thesis on legal aspects of intra-group financing (published in 1998). In 1996 he took up the Chair of Civil Law, Commerce Law and Corporate Law at the University of Mannheim. Since October 2002 he is Director of the Institute for Law and Finance at Goethe-University in Frankfurt. He has published extensively on corporate law, capital markets law, the law of products liability, general civil law as well as on civil procedure. He is co-publisher of “Der Konzern”, a law journal focusing on company law, taxation and accounting of corporate groups, of “Corporate Finance law”, a journal with a focus on current legal issues corporate finance, co-editor of the Institute for Law and Finance Series and member of the editorial board of the law journal “European Company Law”.

VIII

The Authors

Miguel de la Mano Miguel de la Mano is currently Head of Economic Analysis of Financial Markets at DG Internal Market since mid-2012. His team of specialised economists provides support in the analysis of complex economic or financial issues related to regulatory initiatives in the areas of financial markets, free movement of capital, corporate governance, banking, asset management and insurance. He actively engages with external stakeholders, including the academic community, practitioners and related regulatory institutions or agencies with the goal of identifying fault lines, monitoring early warning indicators, and facilitating a targeted and timely regulatory response to endogenous and exogenous risks to the EU financial system. Prior to his current assignment Miguel was a member of the Chief Economist Team at DG Competition since its inception in 2003. He was appointed Deputy Chief Economist in early 2009. From October 2011 to May 2012 he was Acting Chief Economist at the UK Competition Commission. He has co-drafted various guidelines setting out the European Commission’s analytical framework in competition enforcement. As a competition policy enforcer he was closely involved in dozens of in-depth merger and antitrust investigations, both during the administrative and court proceedings and has co-drafted multiple prohibition decisions. He has written extensively in particular for internal policy development. He completed graduate studies in economics at the Institute for World Economics in Kiel, Germany, and the European Institute at Saarbrucken University, Germany. He conducted his PhD research at Oxford University, UK. He joined the European Commission in 2000.

Dr. Andreas Dombret Dr. Andreas Dombret was born in the USA to German parents. He studied business management at the Westfälische Wilhelms University in Münster and was awarded his PhD by the Friedrich-Alexander University in Erlangen-Nuremberg. From 1987 to 1991, he worked at Deutsche Bank’s Head Office in Frankfurt, from 1992 to 2002 at JP Morgan in Frankfurt and London, from 2002 to 2005 as the CoHead of Rothschild Germany located in Frankfurt and London, before serving Bank of America as Vice Chairman for Europe and Head for Germany, Austria and Switzerland between 2005 and 2009. He was awarded an honorary professorship from the European Business School in Oestrich-Winkel in 2009. Since May 2010, he has been a member of the Executive Board of the Deutsche Bundesbank with currently responsibility for Banking and Financial Supervision, Risk Control and the Bundesbank’s Representative Offices abroad. He is also responsible for G7, G20 and IMF (Deputy of the Bundesbank), Supervisory Board

The Authors

IX

(SSM) (Member), Basel Committee on Banking Supervision (BCBS) (Member of the Bundesbank) and Bank for International Settlements, Basel (Board of Directors).

Douglas Elliott Mr. Elliott is a Fellow in Economic Studies at The Brookings Institution. A financial institutions investment banker for two decades, principally at J.P. Morgan, he was the founder and principal researcher for the Center On Federal Financial Institutions, a think tank devoted to the analysis of federal lending and insurance activities. At Brookings, he focuses primarily on financial institutions and markets and their regulation. He has written extensively on bank regulation and on international coordination of financial regulation (see http://www. brookings.edu/experts/elliottd.aspx) Mr. Elliott’s work as a financial institutions investment banker over two decades has given him a wide-ranging and deep understanding of the industry. He has researched financial institutions or worked directly with them as clients in a range of capacities, including as: an equities analyst, a credit analyst, a mergers & acquisitions specialist, a relationship officer, and a specialist in securitizations. His work encompassed banks, insurers, funds management firms, and other financial institutions. In addition to 14 years at J.P. Morgan, Mr. Elliott worked as an investment banker with Sanford Bernstein, Sandler O’Neill, and ABN AMRO. A deep interest in public policy led him to found the Center On Federal Financial Institutions (COFFI) in 2003. COFFI focused on providing objective analyses of the federal government’s 100 %-owned financial institutions. He has testified before both houses of Congress and participated in numerous speaking engagements, as well as appearing widely in the major media outlets. The New York Times referred to his analyses at COFFI as “refreshingly understandable” and “without a hint of dogma or advocacy”. Mr. Elliott graduated from Harvard College magna cum laude with an A.B. in Sociology in 1981. In 1984, he graduated from Duke University with an M.A. in Computer Science.

Simon Gleeson Simon Gleeson joined Clifford Chance in 2007 as a partner in the firm’s Financial Regulation group, where he specialises in financial markets law and regulation. He has advised Governments, regulators and public bodies as well as banks, investment firms, fund managers and other financial institutions on a wide range of regulatory issues. He advised the World Economic Forum on their report on

X

The Authors

their 2009 Report on The New Global Financial Architecture, and has worked with regulators and governments around the world on the establishment of regulatory regimes. He has been a member of the Financial Markets Law Committee, chairs the Institute of International Finance’s Committee on Cross-Border Bank Resolution, has written numerous books and articles on financial regulation, and is the author of “International Regulation of Banking”, recently published by Oxford University Press.

Randall Guynn Mr. Guynn is the head of Davis Polk & Wardwell’s Financial Institutions Group. His practice focuses on bank M&A, bank regulation and enforcement, and bank capital markets transactions. He is widely recognized as a thought leader on financial regulatory reform, and was recently named one of the ten most innovative lawyers in the United States by the Financial Times for his work in the area. He has taken a leading role in helping the FDIC develop its single-point-of-entry recapitalization approach for solving the too-big-to-fail problem, and has represented five out of the six largest U.S. banks on their recovery and resolution planning. He has been ranked in Band 1 in both Financial Institutions M&A and Bank Regulation by Chambers USA for many years, and as one of the 500 leading lawyers in America by Lawdragon Magazine. The group he heads at Davis Polk was named the 2013 Financial Regulation Team of the Year by IFLR. Mr. Guynn is editor of the 7th edition of Regulation of Foreign Banks & Affiliates in the United States, the leading treatise in the area, and a co-author of the chapters on bank M&A and the expanded powers of financial holding companies. He is the Co-Chair of the Bipartisan Policy Center’s Working Group on Failure Resolution and co-authored the BPC’s white paper on solving the too-big-to-fail problem. He is the author of numerous articles and chapters in books, has been a guest lecturer at the Harvard, Yale, Pennsylvania and Virginia law schools and frequently speaks on panels at bank M&A and bank regulatory conferences. He holds a B.A. from Brigham Young University and his J.D. from the University of Virginia School of Law. Mr. Guynn was a law clerk to former Justice William H. Rehnquist of the U.S. Supreme Court.

Andy Haldane Andrew Haldane is Executive Director, Financial Stability at the Bank of England. In this role, he has responsibility for developing Bank policy on financial stability issues and the management of the Financial Stability Area. Andrew is a member of the Bank’s Financial Stability Executive Board and the new Financial

The Authors

XI

Policy Committee. He is a member of various international public policy committees, economics associations, editorial boards and academic advisory committees. Andrew has written extensively on domestic and international monetary and is co-founder of ‘Pro Bono Economics’, which aims to broker economists into projects in the charitable sector.

Patrick Kenadjian Patrick S. Kenadjian is currently an Adjunct Professor at the Goethe University in Frankfurt am Main, Germany, where he teaches courses on the financial crisis and financial reform and comparative public mergers and acquisitions at the Institute for Law and Finance. He speaks frequently on topics related to financial reform, including too big to fail, the architecture of financial supervision and the new regulatory environment in the US and the EU. Mr. Kenadjian is also Senior Counsel at Davis Polk & Wardwell, LLP in their London office. He was a partner of the firm from 1994 to 2010, during which time he opened the firm’s Tokyo and Frankfurt offices in 1987 and 1991, respectively and spent over 25 years in their European and Asian offices. His practice includes cross-border securities offerings, especially for financial institutions, mergers and acquisitions, privatizations and international investments and joint ventures, as well as general corporate advice, with an emphasis on representing European clients. He has been active in securities transactions for issuers in Asia and Europe, particularly on initial public offerings and privatizations in Germany, Austria, Italy and Switzerland. He has represented bidders and targets in cross-border acquisitions throughout Europe, in particular in France, Germany, Italy, Switzerland, the United Kingdom and the United States. Mr. Kenadjian has also represented European and Asian issuers in U.S. debt private placements. He speaks French, German and Italian.

Lowri Khan Lowri Khan is Director of Financial Stability at the UK Treasury. The Financial Stability Group has the aim of securing the stability of the UK financial sector for the benefit of economy. It does this through identifying and analysing the emerging risks to the financial stability of the UK and the global financial system, and preparing and responding appropriately. As part of this, it is responsible for intervening directly to manage the provision and withdrawal of public support for the financial services sector. It is also responsible for delivering the Government’s response to the Independent Commission on Banking, and the UK Government’s response to the Parliamentary Commission on Banking

XII

The Authors

Standards, including through the Banking Reform Act 2013. She returned to the Treasury from a secondment to the Bank of England in 2012, where she worked on financial stability. Her previous Treasury career includes financial services and government debt, cash and reserves management.

Jan Krahnen Jan Pieter Krahnen is a Professor of Finance at Goethe University’s House of Finance. He is a Director of the Center for Financial Studies (CFS) and the Center of Excellence SAFE. His current research interests focus on the implications of the 2007– 2010 financial turmoil for banking, systemic risk, and financial market regulation. His publications appeared, among others, in the Review of Economic Studies, the Journal of Financial Intermediation, the Journal of Banking and Finance, and Experimental Economics. Krahnen is a CEPR research fellow, and was President of the European Finance Association in 2011. Krahnen has been involved in policy advisory on issues of financial market regulation, most recently as a member of the High Level Expert Group on Structural Reforms of the EU Banking Sector (“Liikanen Commission”), implemented by EU Commissioner Michel Barnier. From 2008 until 2012 he was a member of the Issing-Commission, advising the German government on the G-20 meetings. He is also a member of the Group of Economic Advisors (GEA) at the European Securities and Markets Agency (ESMA), Paris, and a member of the Academic Advisory Board of Germany’s Federal Ministry of Finance. Policy briefs and opinion pieces: http://safe-frankfurt.de/en/policy-center/ policy-publications.html.

Adam Posen Adam Posen is President of the Peterson Institute for International Economics, the world’s leading independent non-partisan research institute on globalization. From 2009 – 2012, he was a member of the Monetary Policy Committee of the Bank of England, and is globally recognized as an influential advocate of central bank activism in response to the financial crisis. His policy and research work focuses on macroeconomic policy and forecasting, European and Japanese political economy, central banking issues, and the resolution of financial crises. He is a member of the Panel of Economic Advisers to the US Congressional Budget Office, and, prior to joining the MPC, was a consultant to several U.S. government agencies, the UK Cabinet Office, the European Commission, and to the IMF, as well as a visiting scholar and consultant to a number of central banks

The Authors

XIII

in the Asia-Pacific, Europe, and North America. He is the author of Restoring Japan’s Economic Growth , the co-author with Bernanke, et al, of Inflation Targeting: Lessons from the International Experience, as well as of over forty research articles, and the editor of four conference volumes on financial crises and on the future of the euro. He received his Ph.D. and his A.B. from Harvard University, and is a member of the Council on Foreign Relations, the Trilateral Commission, and the Bellagio Group of international finance officials and scholars.

Debra F. Stone Ms. Stone is a Counsel and Vice President in the Legal Department of the Federal Reserve Bank of New York. Prior to joining the Federal Reserve Bank of New York, Ms. Stone had over 20 years of experience as a lawyer in financial services, primarily as an in-house counsel for global financial firms. She began her legal career as an associate at Skadden, Arps, Slate, Meagher & Flom. Ms. Stone graduated cum laude from the University of Pennsylvania and received her law degree from New York University School of Law.

Andreas Dombret, Patrick S. Kenadjian

Introduction

On January 21, 2014, the Institute for Law and Finance at the Johann Wolfgang Goethe University in Frankfurt am Main held its third full day symposium on the topic of “too big to fail”. The first session, held on November 5, 2010, was devoted to helping build consensus in Germany on the need for special bank resolution laws to deal with the phenomenon of “too big to fail”. The second, held on May 3, 2012, was devoted to evaluating the progress made in developing bank resolution laws as the European Commission prepared to publish the first draft of its directive on bank recovery and resolution. The third session was planned around a different approach to dealing with “too big to fail”, the so-called “structural solutions” pursuant to which the problem would be solved by breaking up the banks. We waited until 2014 to present this program because in the immediate aftermath of the crisis, the idea of breaking up the banks, which went under the dismissive label of “a return to Glass-Steagall”, the US Depression era legislation that separated commercial and investment banking in 1933, was viewed in policy-making circles as unrealistic, a result of a naive nostalgia for simpler times which would not return and unsuited for a complex interconnected commercial and financial world. This attitude was perhaps best summed up by Lord Turner, Chairman of the UK Financial Services Authority, who branded it, in his famous review, as appropriate for “an era of fixed exchange rates and exchange controls, with far more limited capital flows and trade flows as a percentage of GDP, and a much smaller role played by cross‐border corporations”.¹ He probably best summed up the immediate post-crisis consensus when he wrote that “serving the financial needs of today’s complex globally interconnected economy… requires the existence of large complex banking institutions which can only be delivered off the platform of extensive market making activities, which inevitably involve at least some position taking”. He went on to observe that “[m]any activities which before the lifting of Glass-Steagall were in the US conducted by investment banks … are core elements within an integrated service to corporate partners in a world where a significant element of debt is securi-

 Financial Services Authority, The Turner Review, A regulatory response to the global banking crisis. March 2009, at 94.

2

Andreas Dombret, Patrick S. Kenadjian

tized”² and that “‘narrow banks’ focusing almost entirely on classic commercial or retail banking activities can be extremely risky. Northern Rock, Washington Mutual and IndyMac were all ‘narrow banks’”³. He might have added that Bear Stearns and Lehman Brothers who failed and Merrill Lynch which almost failed before it was bought by Bank of America, were classical investment banks while JP Morgan, which comes as close to a universal bank as US regulation allows, did not fail or need to be rescued. That was, and largely remains, the core of the case against structural reform and initially it carried the day. The Volcker Rule, prohibiting proprietary trading and the ownership of certain kinds of investment funds by US banks, was not originally part either of the US Treasury’s White Paper for reform of the financial sector or of either the U.S. House of Representatives’ or the U.S. Senate’s drafts of what ended up becoming the Dodd-Frank Wall Street Reform and Consumer Protection Act of 2010. Its last minute inclusion was widely seen as a political accident rather than as the result of serious policy deliberations and as an American peculiarity which, like Glass-Steagall or the overall limitation on the size of banks, would find few takers internationally. Then the Independent Banking Commission in the UK, better known as the Vickers Commission after its chairman, also recommended a different but equally radical separation between domestic retail banking services and global wholesale and investment banking operations, albeit under the umbrella of a single holding company. Those recommendations were endorsed by a conservative-liberal UK government and have been enacted into law. Some observers dismissed this as an attempt to get the UK taxpayer off the hook for the liabilities incurred by their banks abroad in an industry where domestic operations account for only one third of banks’ balance sheets, so as a result of the particular situation of the UK. However, when the High-level Experts Group appointed by EU Commissioner Michel Barnier under the Chairmanship of Governor Erkki Liikanen of the Finnish Central Bank to examine bank structural reform also recommended a structural solution for European Union banks generally and the Commission set about transforming these recommendations into an EU directive, it became clear that an important change had occurred in the public and policy-makers’ attitude towards structural solutions and that the tide in favor of such solutions was rising. However, the fact that a tide is rising does not mean that it has the better arguments. The objections raised by Lord Turner five years ago must still be answered. Hence, the idea to gather in Frankfurt on January 21, 2014 rep-

 Id.  Id, at 95.

Introduction

3

resentatives of the public and private sectors to discuss the issues involved, and whether breaking up the banks is a sensible approach to the “too big to fail” dilemma. At the end of the day, although we had started off wanting to discuss the pros and cons of structural reform, most of the speakers ended up advocating or conceding that without a credible bank resolution system, structural proposals alone would not suffice to solve “too big to fail”. The following speeches and papers trace the route taken by the participants. Frankfurt, August 2014

Dr. Andreas Dombret

Cutting the Gordian Knot or splitting hairs – The debate about breaking up the banks Content  Introduction  Breaking up the Banks – the Objectives  Breaking up the Banks – Would it Work?  Breaking up the Banks – The Alternatives  Conclusion

1 Introduction Ladies and gentlemen Thank you for inviting me to speak at today’s conference. It is a pleasure and an honour to be here. The story goes that at some point in history the kingdom of Phrygia was without a king. The priests therefore consulted an oracle to determine who should fill the vacant position. The oracle decreed that the next man to enter the city in an ox cart should become the new king. That man was Gordias, a peasant. Gordias became the king of Phrygia and, out of gratitude, his son dedicated the ox cart to the gods. The gods in turn tied the cart to a pole with a complicated knot – the famous Gordian Knot. This knot was so complex that for centuries no one was able to untie it. Hence, the cart was still tied to its pole when Alexander the Great entered the scene in 400 BC. Meanwhile, it had been prophesied that whoever untied the knot would become king of Asia. Alexander, full of ambition, took on the challenge. Unable to find the loose end of the knot, he drew his sword and sliced the knot in two, thereby undoing it. Since then, the Gordian Knot has become a symbol for solving seemingly insoluble problems by thinking outside the box and then taking bold action. The insoluble problem in our case is how to secure financial stability. And after six years of crisis, more and more people are, understandably, looking for a Gordian solution. To some, the solution would be to break up the banks – with one bold stroke of the sword, as it were. Broadly speaking, the dividing

6

Dr. Andreas Dombret

line would run between those parts of a bank engaged in commercial banking and those parts engaged in investment banking. In my speech today, I will address the question of whether this is a sensible approach. As I speak first at this conference, my objective is to give an overview of the issue, discuss the pros and cons of the proposal and try to draw a tentative conclusion. Let us begin by looking at the arguments put forward by the proponents of breaking up the banks.

2 Breaking up the Banks – the Objectives The underlying notion is, of course, that there are two different worlds. There is the world of commercial banking, which is populated by conservative and “prudent” bankers who support the real economy and provide a haven for savings. And then there is the world of investment banking populated by the Gordon Gekkos of this world, who pose a constant risk to financial stability and to taxpayers’ money. The combination of these two worlds in a universal bank is seen as a source of systemic risk. It is therefore being proposed that they be separated: instead of universal banks there would just be pure commercial and pure investment banks. But why is the combination of commercial and investment banking perceived as a source of systemic risk? And what specifically are the objectives of those who want to separate them? The general objective when breaking up the banks is to shield those parts that are deemed vital for the real economy from those parts which have little or no connection to the real economy – to shield commercial banking from investment banking. The underlying assumption is that investment banking is risky and thus more prone to losses and failure. If the investment banking branch of a universal bank were to fail, it would drag the commercial banking branch down with it. Both parts would go down together, disrupting the real economy. It is therefore argued that separating the two parts would block this channel of contagion, shield the real economy and protect savers and taxpayers alike. But there is more to the argument. Two aspects are usually emphasised. The first aspect is the existence of deposit insurance schemes. These schemes render deposits risk-free – at least up to a certain amount. Consequently, depositors demand lower risk premiums. This holds even if the bank in question is engaged in risky investment banking activities. Universal banks can therefore draw on a subsidised source of funds to finance their investment banking activities. It is argued that separating commercial and investment banking would abolish this subsidy,

Cutting the Gordian Knot or splitting hairs – The debate about breaking up the banks

7

align incentives for investment banks and force them to reduce the size of their business. The second aspect is the existence of implicit government guarantees for certain banks. If a large and interconnected universal bank fails, the whole financial system might be disrupted, as might the real economy. The taxpayer might therefore be forced to step in and save the bank to prevent an even worse outcome. Banks that enjoy such a status are termed “too big to fail” or “too interconnected to fail”. Just like deposit insurance, such an implicit government guarantee subsidises investment banking activities. Proponents of breaking up the banks argue that pure investment banks with no connection to the real economy would be treated differently. They would be excluded from deposit insurance schemes and also from implicit government guarantees. Consequently, if things went wrong, they would not be rescued by the government at the taxpayers’ expense – moral hazard would be reduced. Moreover, banks would become less complex and thereby easier to resolve. These are – in a nutshell – the central arguments brought forward by the proponents of breaking up the banks. They claim that breaking up the banks would block channels of contagion, would remove subsidies for risky investment banking, would lower the risk of systemic crises, would make banks easier to resolve and would eventually save taxpayers’ money.

3 Breaking up the Banks – Would it Work? Now, could one stroke of the sword really take us all the way to financial stability? To answer this question, we have to take a step back. To me, financial stability cannot be achieved as long as banks are too big, too interconnected or too complex to fail. Finding a solution to this problem would eliminate implicit government guarantees, would align incentives and would increase financial stability. What we need at a very basic conceptual level are two lines of defence. First, we have to make banks safer to reduce the likelihood of failure. Second, if a bank fails, it must be able to do so without disrupting the entire financial system. Erecting these two lines of defence is the basic challenge we are facing. And at the same time, it should be the benchmark for evaluating the benefits of breaking up the banks. The first question is therefore: would breaking up the banks make them safer? It is true that commercial banking would benefit if it had some degree of insulation from the perils of investment banking. But is commercial banking in itself really safer than investment banking? Looking back, we have to admit

8

Dr. Andreas Dombret

that pure commercial banks were at the centre of the recent crisis: Washington Mutual, Countrywide, Hypo Real Estate and the Spanish savings banks, to name just a few. The question of stability ultimately depends on the sustainability of the business model. And a commercial bank that is highly leveraged and has an unsustainable business model can be as risky as any investment bank. In addition, breaking up the banks would reduce their potential to diversify. This, in turn, could increase their exposure to individual shocks – they might even become less safe. The second question is: would breaking up the banks limit the size of investment banks by removing funding subsidies? Well, it certainly would remove at least those subsidies related to deposit insurance. But there are two things we should bear in mind. First, if deprived of deposits as a source of funding, an increasing number of pure investment banks would have to revert to less stable sources of funding – they would become less safe. Second, what about that part of the subsidy that is due to implicit government guarantees? This point relates to our second line of defence – to the question of whether a bank can fail without disrupting the whole system. And frankly, I am not fully convinced that breaking up the banks would block the relevant channels of contagion. True, it would block channels of contagion within banks. But wouldn’t they just be replaced by channels of contagion between banks? I think that even after breaking up the banks, the financial system would be characterised by a large degree of interconnectedness. And again, we could take a look into history. Lehman Brothers was a pure investment bank, yet when it failed in 2008 it brought the financial system to the brink of collapse – because it was so interconnected. Against this backdrop, the question is: what would happen if another pure investment bank were to fail? Would the government really just stand on the side-lines and watch the potential fallout? Or would it be compelled to step in, as it has done before, to save the failing bank – regardless of whether it is attached to a commercial bank or not? And what about the problem of shadow banking? I see the danger that breaking up the banks might provide incentives to export more and more risky activities to the realm of non-bank banking. However, the risks could easily be re-imported to the regular banking system. In the end, we would have gained nothing. This is why our efforts to control the shadow banking system have to continue. Regarding the proposal to break up the banks, a lot of questions on the effects and side-effects remain unanswered. And this takes me directly to the last issue I wish to raise. We all know that the worst enemy of good ideas is the need to implement them. In the present case, this refers to the challenge of designing

Cutting the Gordian Knot or splitting hairs – The debate about breaking up the banks

9

an actual system of separated banking functions. In the theoretical debate, an imaginary line is drawn between investment banking and commercial banking. But where exactly would that line be drawn in reality? If commercial banking is supposed to support the real economy, it should comprise more than just lending money. Take the example of Germany, which has a bank-based system of corporate finance. German companies procure a wide range of services from their banks. The boundaries between customer business, hedging transactions, market making and traditional proprietary trading are consequently fluid. Finding the right dividing line in the grey area between those activities would be difficult and prone to lobbyism. The fact that current legislation and legislative proposals are so different in content and scope partly reflects these problems. In the US, the Volcker Rule seeks to prohibit proprietary trading by banks and places severe restrictions on certain forms of investment. In the UK, the Vickers proposal seeks to ringfence deposit-taking and the provision of credit facilities in legal, organisational and operational terms. In Europe, the Liikanen Commission suggests yet another model which would maintain the universal bank within a holding structure but ring-fence deposit-taking units.

4 Breaking up the Banks – The Alternatives Based on the arguments I have laid out, it seems unclear to me whether breaking up the banks would take us all the way to financial stability. It could make banks somewhat easier to resolve – that much is clear. But there are three crucial points on which I have doubts and which should be discussed in more detail: – I doubt that breaking up the banks would make them safe enough; – I doubt that breaking up the banks would ensure that they can fail without disrupting the system; – and I doubt that the proposal of breaking up the banks could be implemented in a suitable manner. I also doubt whether it is the responsibility of the state to determine what business model works best. In my view, we do not need the same business model all over the world. A lot of countries, including Germany, fared very well with universal banks. Personally, I therefore do not see a reason to abandon the concept of universal banks. And as a matter of fact, legislative proposals such as the Liikanen-proposal also do not seek to abolish universal banks as a general concept. Why not let the market determine which business models work and which don’t? And please do not misunderstand me: the state certainly has to set boun-

10

Dr. Andreas Dombret

daries to guide this selection process and to protect the financial system, the real economy, depositors and the taxpayer. In my view, other things are at least equally important if we want to secure financial stability. To make banks safer, we have to increase capital requirements. Banks need more capital and they need better capital. This is exactly what the new Basel rules prescribe. To enable banks to fail without disrupting the financial system, we need resolution mechanisms. At the international level, the Financial Stability Board has published relevant principles. At the European level, a Single Resolution Mechanism is under construction. This mechanism will be a central pillar of the banking union and should be established as soon as possible. In my view, these measures are more important and bring us closer to financial stability than just breaking up the banks. And I would like to bring up a final point: we must be aware that we cannot solve all our problems through regulation. In addition to regulatory reform, there also has to be a change of culture within the world of banking. And here, investment banks might, on balance, have to change more than commercial banks. Just take the example of compensation schemes. Compensation schemes that reward excessive risk-taking need to be replaced with more sustainable solutions. Relevant regulation exists but to be really effective these rules must become part of the culture.

5 Conclusion Ladies and gentlemen, I began my speech with an account of how Alexander the Great solved the seemingly insoluble problem of the Gordian Knot. Instead of attempting to untie it as everyone else had done, he just split the knot with his sword. This is certainly the most famous version of the story, but it is not the only one. There are alternative accounts of how Alexander untied the cart from its pole. According to Aristobulus, Alexander unfastened the cart by removing the pin which secured the yoke to the pole of the cart, then pulled out the yoke itself. This exemplifies that there is never only one solution to a given problem. And the more complex a problem is, the less likely that it can be solved with a single stroke of the sword. To the contrary, it could make things even worse. Just recall the story of how Heracles cut off the head of the Hydra, just to have two grow back. To ensure financial stability, we have to apply a wide range of measures. We have to adjust capital requirements, we have to implement liquidity requirements, we have to make banks resolvable to name just a few.

Cutting the Gordian Knot or splitting hairs – The debate about breaking up the banks

11

Breaking up the banks might shield commercial banking and thereby the real economy to a certain degree. But it cannot, on its own, ensure financial stability. And when it comes to implementing a system of segregated banking functions, attempts to cut the Gordian Knot might mutate into a tedious exercise of splitting hairs. Thank you very much.

Dr. Paul Achleitner

“What kind of financial system do we want – A global private sector perspective” Dear Ladies and Gentlemen Thank you very much for the invitation and thank you very much for your flexibility to move my slot to the first session at short notice. Apologies for any inconveniences this may have caused. The timing of this year’s conference could not be more topical and telling: – A month ago in Brussels, on 19 of December, the European heads of State and Government agreed on the roadmap to a European Banking Union, calling to adopt the Single Resolution Mechanism within the current legislative period of the Parliament (i. e. until May 2014) – A week ago in Basel, on 12 of January, central bankers and supervisors approved an international standard for the leverage ratio – And as we speak, the European Commission is finalizing its proposal on separating retail and investment banking. 2014 remains another challenging and exciting year for the industry. Today, I have been asked to present to you my thoughts on the global financial system from a private sector perspective. And as you know, Adam Posen will speak from a public policy perspective later in the afternoon. Let me start out by saying that I do not see any major contradictions between a global public policy perspective and a global private sector perspective. Both ultimately have the same goal: a better financial system. Point taken that the term better needs further color: I see three key qualities that we should prioritize. In my view a financial system should be: 1) Stable 2) Functional 3) Competitive Today, we are far from that ideal condition – particularly in Europe. Above all, the nexus between the sovereign sector on the one hand and the banking sector on the other has not yet been cut through: When a crisis occurs in either of the two sectors, it quickly spills over into the other part of the nexus. A banking crisis necessitating fiscal support will cause sovereign debt problems, which in turn undermine the asset quality of banks, leading to further injections… You are all familiar with the sequence.

14

Dr. Paul Achleitner

The key lever to cut this vicious cycle is to ensure that regulatory rules are effectively designed to actually tackle existing problems while avoiding unintended consequences. A cornerstone of such regulation is a credible Bank Resolution Mechanism – Simon Gleeson will address this particularly important issue in detail later this afternoon and I will also briefly comment on it. However, other areas of the regulatory debate are of similar importance when we want to achieve our prioritized qualities: (1) stable, (2) functional and (3) competitive. Before I will comment on them separately and address associated risks, let me suggest that we back away from the ambition to create the optimum financial system. Remember, we are not starting from a blank page – financial market participants and regulations are in constant motion. A “blueprint type of approach” would certainly be over-ambitious. With a multitude of half-baked proposals in the air, we are well advised to concentrate on those concepts that represent material improvements to stability. In parallel, we need to question all proposals critically that may lead to unintended consequences or foster an unlevel playing field. So allow me to illustrate – by examples of the current regulatory debate – the three imminent priorities that need to be achieved to align the private and public sector perspective for an improved financial system.

1st Priority: Stability – simple is not always better The first priority is to keep our financial system secure and stable on a regional and global scale. A crucial condition for stability is adequate capital requirements for financial institutes. To come to global rules, we have spent over 10 years in the Basel process with a focus on a detailed RWA approach of looking at bank capital. Within a few months this risk-based approach has come under fire following a US push that RWA are too prone to model assumptions leading to uncertainties. Instead our counterparts west of the Atlantic proclaim that one should put much more emphasis on using a simple leverage ratio (which takes your exposure, ie. your assets, at face value without risk adjustment and compares it with your Tier 1 capital). Understandable from a US perspective, difficult to see why so many Europeans support that approach. We all know the general reservations: within the boundaries of liquidity requirements, a rational bank manager would be immediately incentivized to re-

“What kind of financial system do we want – A global private sector perspective”

15

duce his cash positions, on which he earns nothing but they cost him the same as a high risk-high return asset. We should also recall that there are important differences between the US and the European banking system: the former builds on fewer but riskier assets with high returns (for example consumer loans that are volatile and therefore pay 15 – 18 % interest or mortgages with up to 100 % loan-to-value and a legal framework that allows you to default on your house without any implications for your personal credit standing – you just give the keys to the bank and ‘go on with your life’). In addition the two Federal mortgage agencies act de facto as a state-sponsored SPV into which every bank can unload its mortgage assets routinely rather than keeping them on its balance sheet. In Europe, we have safer consumer loans and mortgages, which based on 60 % loan-to-value and a comprehensive personal liability in case of default constitute from a banking perspective much lower risk but also much lower return assets. To generate similar returns (as we compete for the same investors) European banks need to use much more of these (lower risk) assets – and there is no government sponsored vehicle to offload mortgages to. By definition therefore European banks will have in general more, but lower risk assets on their balance sheets than US banks. So what works for US banks may not work for European banks – not because the later are riskier but because of different circumstances. Ladies and Gentlemen, simple is not always better for financial market regulation. We need to consider the underlying circumstances or we will generate unintended consequences. Indeed and to be fair, last week brought partial relief as the Basel Committee on Banking Supervision left the leverage ratio of 3 % unchanged but revised the scope of application by allowing a more favourable treatment of products such as credit derivatives. Rightfully so because this is an area where accounting standards differ significantly as you all know. For Europe, however, the story is far from over. It will now be up to the European Commission and the European Parliament to decide how it will take these proposed equity requirements into consideration when it comes to the amendment of the EU leverage ratio rules during the first half of 2014. It remains to be seen whether the awareness that ‘one size does not fit all’ will also gain acceptance in Brussels’ decision making bodies. Against that background, the financial sector must clarify that adequate equity requirements and the necessity to keep the financial system secure and stable are not a contradiction. It remains open whether the voice of the financial industry will be heard in Brussels. The problem is: The banking sector has lost so

16

Dr. Paul Achleitner

much credibility that whenever it puts forward an argument it is automatically seen as self-serving. When rules are made, those who understand their implications because they have to execute them do not have a reserved seat at the table anymore. Ladies and Gentlemen, the loss of credibility of the banking sector has involuntarily forced another – even more worrying – regulatory debate in Europe, which I would like to use as an example for the functionality of financial system.

2nd Priority: Functionality – the ‘Too-Big-To-Fail’ Conundrum This topic is central to this conference and has seen a huge ongoing debate. It is common consensus that banks that are ‘Too-Big-To Fail’ are the core of the problem of a non-working financial system. There is also broad consensus that we will need to fix that problem through enhanced rules to ensure that the taxpayer shall never again have to bail-out a bank. So far so good – but what conclusions do we draw from that? And what is the solution? Should banks become all smaller? And what implications would that have for the second quality of a better financial system – i. e., being functional? Ladies and Gentlemen, any concept behind the term ‘Too-Big-To Fail’ is meaningless and will be subject to political sentiment and misinterpretation unless somebody defines what is ‘Small Enough To Fail’. Apparently neither IKB nor Northern Rock fell into that category. Let me, against this background, dare the following question: In how many pieces would you like to carve up Deutsche Bank then – pre-emptively? My provocative question shows that the ongoing debate focuses too much on the ‘Too Big’ instead of concentrating on the ‘To Fail’ – with well known negative side effects and unintended consequences. Only a credible and independent banking resolution mechanism can decide on the ‘To Fail’ issue. Mr. Gleeson will address this issue later in detail, but please allow me the following remarks from a banker’s perspective. While most people agree on the need for an insolvency process for banks, there are still discussions ongoing on how that process should be designed. The case of General Motors demonstrates that not only banks are ‘Too-BigTo-Fail’. Also, it shows the benefits of an established process when industrial companies are troubled: the Chapter 11 Filing.

“What kind of financial system do we want – A global private sector perspective”

17

As you know Chapter 11 proceedings mean that a formal administrator takes control away from the management, the shareholders lose most of their money, the creditors lose some of their money – as do suppliers, employees – even taxpayers – particularly if the government decides to inject capital as the US government did in the case of GM. The administrator then decides which parts of the company need to be shut down, which are to be sold, which will continue – if necessary under new leadership … Now what is the difference with a large global bank? Mainly three things: 1) the immediate need to stabilize the short-term liquidity issues, 2) the multifaceted funding structure including depositors and 3) the complexity of the global activities that make it very difficult for any administrator coming in to know where to cut without creating new problems. The liquidity issue can only be solved by the central banks – very much at the core of their responsibility of ensuring a stable banking system. In terms of credit priorities, i. e. who loses how much in which sequence we are making good progress based on tough experiences (in particular Cyprus) and are coming to a clear hierarchy of capital. Finally, the agreed Resolution Regime – the so-called ‘living wills’ are designed to serve as a blueprint for a potential administrator to know what to close down or sell in which sequence … I am therefore convinced that the adequate solution for the TBTF Conundrum is a de facto Chapter 11 process for banks. The recent political agreement and the preparatory works for a Single European Resolution Mechanism are very important steps into the right direction. Still, other issues are far from being solved. While a cross-national resolution mechanism should be fully independent in order to prevent vested interests, the solution for the Euro area currently provides for an involvement of national resolution authorities and the Council of Ministers at all critical stages. It remains to be seen whether this strong political momentum will enable the mechanism to take decisions quickly and consistently for any bank in the banking union – and whether it will be able in the end to fully solve the TBTF conundrum in order to make our financial system more functional. Time is pressing as the US with their Federal Deposit and Insurance Commission (FDIC) is already able to intervene at short notice. Competition never sleeps – also in regulatory terms. Together with enhanced capital requirements and enhanced supervision and regulatory cooperation a common recovery and resolution mechanism could make European financial systems more competitive in the future.

18

Dr. Paul Achleitner

This leads me to the third quality that a good financial system should fulfill and priority No. 3: Competitiveness.

3rd Priority: Competitiveness – The Banking Separation Allure As you all know, the idea to split the commercial and the investment banking activities of financial institutes is everything but new. The Glass-Steagall Act of the 1930s was designed to deal with the pre-depression abuse of replacing shaky corporate loans with newly issued bonds stuffed into depositors’ accounts … Against this background, it is difficult to see why this is suddenly seen by some as a panacea to deal with today’s very different challenges – interestingly the enthusiasm comes mostly from Europeans not from those who have the actual experience – i. e. the Americans. The UK, EU, France and Germany have all proposed some form of separation of investment banking activities from deposit-taking. These all stem from a concern that risks in the investment bank are ‘underwritten’ by implicit government guarantees for deposits. Empirical evidence has shown that the ability to withstand crisis situations in the past did not depend on the structure of the financial institute. – Narrow retail banks collapsed as well as pure investment banks; cooperatives suffered alongside listed banks. – The common themes around failure were poor risk management and overall interconnectedness of the system. – Neither aspect would be directly addressed by the separation principle alone. – As a matter of fact universal banks tend to be more stable thanks to the diversification of their operations and offer client benefits through economies of scale and scope. It goes without saying that universal banks only offer these benefits if they have a proper risk framework, the right governance and lived values. My concern is that the debate over the business model, while politically popular, will distract decision-makers from the risks that any proposals for bank separation bear for financial stability and the economy as a whole. By severing even client serving capital markets activity from retail deposits we are de facto negating the very essence of banking, moving money from savers

“What kind of financial system do we want – A global private sector perspective”

19

to those who require funding. This happens today in the capital markets and not only on the bank balance sheets. The only living example of such a construct that we know of here in Germany was the division of labour between the savings banks who took in the deposits and their partly owned Landesbanken who invested the ‘excess cash’. The crisis has relentlessly revealed risks associated with that model. Let’s take a step back: What is the underlying intention? Understandably politicians want to insure that grandmother’s savings are not used by some young London trader engaged in speculative market activity … If we talk about dedicated proprietary trading – ok, agreed – but then why not simply forbid dedicated proprietary trading? If the ‘Liikanen Proposal’ were adopted as suggested it would mean that when a company comes to a bank like Deutsche Bank and asks to buy e. g. Turkish Lira for a trade we would have to respond “thank you for your inquiry – we will on a best efforts basis look for somebody who has that amount of Turkish lira to sell and inform you of his price expectations.” Why? Because today when we simply sell you those Lira at a fixed price it means we either have a highly liquid market (where we could immediately cover the request) or an inventory of Turkish Lira. Such inventory – regardless if we hedge the position overnight or simply keep it unprotected – would constitute ‘under Liikanen’ a proprietary position which would not be allowed. The German ‘Trennbanken Law’ appropriately therefore explicitly allows ‘market making’ on behalf of clients. Unfortunately – from all what you can read in the press – the draft regulation by the European Commission does not. I will not waste your time on this but who has thought through the complexity of dividing a balance sheet of a big universal bank – I am sure if you are a counterparty you would willingly argue to transfer your claim into the entity that is legally designated to fail. Let me be clear, it is essential that any concerns about the TBTF issue are addressed and that all banks are resolvable. There is no question that financial market participants need to be regulated more stringently and more effectively than during pre crisis times. And it is also clear that especially the banking industry is obliged to demonstrate it has drawn its lessons from previous mistakes. Also, there is nothing to say against drastic reforms that are well thought through and implemented on a global basis. The European financial sector is standing at historic cross roads, where a single decision can lead to negative developments of historic dimensions. Therefore, in their common interest, financial and public sector need to cooperate in order to achieve the aspired system.

20

Dr. Paul Achleitner

As mentioned before, one can only hope that the industry gets the chance to bring its practitioner expertise and perspective to the table when new rules are being designed. If a constructive dialogue between regulators, politicians and the financial sector participants is not fostered, I see in particular two dangers:

1st Danger: The Real Economy Illusion The first danger that I see is that politics and society regard the real economy and the financial system as two autonomous systems with little interdependencies. The linkage which is obvious for today’s audience is not self-evident for the broader public and is often disregarded in public debate. It is a dangerous illusion to assume that we could do in a modern society without capital markets. On the contrary the increasing regulatory demands on bank capital and bank balance sheets will require a much higher proportion of debt financing to be done via capital markets in the future – I am sure you are familiar with the fact that in the US roughly 70 % of corporate debt is provided through capital markets and only 30 % via bank balance sheets – while in Europe it is exactly the reverse. Given the rising regulatory capital demands it is clear that we Europeans will have to move more into the direction of the US. In this context, the mantra that the financial system has detached itself harmfully from the so-called real economy and that it needs to be shrunk to the size of the real economy is outright erroneous and dangerous. Argumentations fail to recognize an important function of capital markets: risk provision and risk transfers. I do not want to waste your time on the topic of the demographic time bomb – suffice it to say “tomorrow’s pensioners are today’s speculators”. Bottom line is: Without a financial system there is no real economy – just as there is no real economy today without IT. Underestimating the role of the financial sector for the real economy and their mutual complementarities could indeed turn out to be a competitive disadvantage for Europe in the long run. Vivid and well functioning financial markets are a pre-requisite of economic success. In other domains – such as the Internet or Energy supply – other regions and above all the US have competitive advantages; be it due to their natural endowments, due to strategic investments into infrastructure or due to their ability to define global standards at an earlier stage. Europe should take care that it does not jeopardize its current – may I say tenuous – position in the capital markets. And we should be aware of the risk that by jeopardizing that position we would also undermine the industrial base of our continent in the long run.

“What kind of financial system do we want – A global private sector perspective”

21

And this is not the only danger that I see if mutual dialogue does not take place:

2nd Danger: The Trend towards Renationalization When I use the term Renationalization, I do not refer to the risk of having stateowned institutions, but rather to risks associated with a crisis-driven reorientation to allegedly national agendas. I see a real danger that intentions to safeguard national interests will manifest themselves at the expenses of hard-earned achievements of a free global market economy. It is insofar alarming that the design of regional regulatory rules is currently perceived as the most decisive competitive advantage in the banking industry. A risk to our aspired level playing field and the prerequisite for the financial system we discussed earlier. I am most concerned of the strong European efforts towards domestic solutions. Just one example in the interest of time: Since May 2008 international exposure of Japanese banks doubled, those of Canada and Australia tripled and those of Brazilian banks quadrupled… while those of the European banks declined by a third. For good reasons a clear sign that others try to avoid and counteract a trend towards renationalization. Dear Ladies and Gentlemen, Let me stop here and reiterate that I am a firm believer in stringent and effective regulation – there is a good reason why banking is a regulated industry i. e. why society, or if you want, the taxpayers have specialists control what is going on because at the end taxpayer funds will be at risk in case of failure. The challenge is to find regulation that avoids the abuses of the past but does not stifle the positive aspects of an effective financial system or is onesidedly disadvantageous to European institutions. The old continent is currently at a turning point where it needs to decide if it still wants to be represented in the ‘Champions League’ of financial markets – at least with some teams. To pause for some years and then go back on the field is not an option. We should not underestimate the risk: The NSA spy affair unveiled very bluntly what happens when others are so far ahead and you fully depend on foreign goodwill… and capital markets are in that regard no different than cyber space. I would go so far to hazard a rather grand analogy:

22

Dr. Paul Achleitner

In the 19th century we had an industrial revolution in Europe. In hindsight the benefits were coupled with unbelievable abuses and side-effects – from child labour to environmental pollution to murdered union activists. How did we deal with that – not by ‘dis-inventing’ the technical accomplishments as the Chinese did in the 17th century when they destroyed all their seafaring vessels with more than one mast and also not by trying to completely change human behaviour as the Soviet Union tried and failed. In the West we successfully established tougher laws, changed societal acceptance and relied on enlightened individuals to move from Manchester capitalism to Social Market Economy. Let me submit to you that the 20th century saw a financial revolution – one that was extremely powerful in its accomplishments – without securitization, risk transfer and capital mobility based on deregulation we would have never been able to finance German Unification, Eastern European Expansion or globalization in general – or is there anybody here who believes that the investments of German or US industry in China or other so called ‘Emerging Markets’, which were so instrumental in their development could have been financed on the balance sheets of a few banks. There is no question, though, that this financial revolution was accompanied by totally unacceptable behaviour, by exploitation of the system as well as the weak and that we need to do everything to avoid such excesses in the future – not by unwinding and reversing financial market integration but by further enhancing and improving the regulatory environment. It will require cool heads though to find the right way to achieve this goal, cool heads that challenge conventional wisdom, cool heads that look at facts – and side-effects, cool heads that put future effectiveness and fairness before revenge or ‘Schadenfreude’. Good to have events like the ILF conference – good for mutual dialogue and (and this is more important) for mutual understanding. Thank you for your attention.

Jan Pieter Krahnen*, **

Rescue by Regulation? Key Points of the Liikanen Report***

Summary This paper summarizes the key proposals of the report by the Liikanen Commission. It starts with an explanation of a crisis narrative underlying the Report and its proposals. The proposals aim for a revitalization of market discipline in financial markets. The two main structural proposals of the Liikanen Report are: first, for large banks, the separation of the trading business from other parts of the banking business (the “Separation Proposal”), and the mandatory issuing of subordinated bank debt thought to be liable (the strict “Bail-in Proposal”). The credibility of this commitment to private liability is achieved by strict holding restrictions. The anticipated consequences of the introduction of these structural regulations for the financial industry and markets are addressed in a concluding part.

Content . Introduction . Liikanen Report: Interpretation of the crisis . Liikanen Report: Key reform proposals a. Separation of trading from universal banking: facilitating resolution b. Subordinated debt for non-banks: reducing contagion . Consequences for banks, investors and supervisory authorities a. Consequential considerations regarding the separation proposal i. Reduced profitability? ii. Regulatory arbitrage?

* Department of Economics, House of Finance, Goethe University Frankfurt [www.finance.unifrankfurt.de], Center for Financial Studies (CFS) and Center of Excellence-SAFE [www.safefrankfurt.de], E-mail: [email protected], Tel.: +49-69-798-33699. ** The author was a member of the High Level Expert Group on reforming the structure of the EU banking sector (Liikanen Commission 2012). The author alone is responsible for the selection of the topics discussed in this paper and for their further interpretation; this does not necessarily represent the opinion of the HLEG. I should like to thank Horst Gischer, Frank Heinemann, Thomas Mayer, Karl-Heinz Paqué, and Hermann Remsperger for their helpful comments during a Symposium organized by Verein für Socialpolitik, September 16, 2013 in Frankfurt. *** This paper was previously published, in German, in Perspektiven der Wirtschaftspolitik, Volume 14, Issue 3 – 4, pp 167– 185.

24

Jan Pieter Krahnen

iii. Reduced liquidity? b. Consequential considerations on the creation of bail-in capital i. Who will bear the risk for bail-in debt? ii. Will this create a sufficiently large market? iii. What are the requirements placed on supervisory authorities? . Outlook and unanswered questions . Literature

1. Introduction The organizers of this conference have asked me to analyze the existing concepts for a European Banking Union from an economic (and political) point of view. The fundamental arguments for the proposals introduced in the Liikanen Report, set out over the following pages, represent a personal interpretation of the “key” recommendations of the report – irrespective of my membership in the team of authors. The term “key” already represents an individual judgmental assessment, which is neither immediately evident, nor directly derived from reading the Report.¹ The Liikanen Report, and the proposals for structural reforms in the banking sector it sets out, was published in early October 2012. The official response from the European Commission, to whom the report was addressed, is expected toward the end of 2013. This conference is therefore taking place at a date (i. e. January 2014) that enables us to take a balanced view of these proposals within the context of the Banking Union as a whole. Several aspects of the bigger picture of an emerging Banking Union (BU) are still in limbo. Of the four elements that define the BU – a cohesive restructuring and organizational regime (Recovery and Resolution Directive, RRD), a unified microprudential banking supervision (Single Supervisory Mechansim, SSM), the Single Resolution Agency, SRA, including an adequate funding backstop, and the joint Deposit Guarantee Scheme, DGS – only the first two are reasonably well defined and are at the stage of final agreement with the European Parliament (RRD) and/or at a stage of implementation (SSM). In both cases, the

 This article draws on a thread presented at a workshop held at the European University Institute in Florence in April 2013 entitled Krahnen (2013) “Banking Union in the Eurozone? A Panel contribution”, in: Political, Fiscal and Banking Union in the Eurozone?, edited by F. Allen, E. Carletti and J. Gray, Wharton FIC Press, pages 29 – 40.

Rescue by Regulation? Key Points of the Liikanen Report

25

legal robustness of a shift of competence from national to European institutions continues to be a controversial issue.² A draft for a European Single Resolution Authority (SRA) was recently put forward by the Commission. The instrumental and process-related details of this relate to the Recovery and Resolution Directive (RRD) while a Resolution Authority equipped with far-reaching intervention rights is still controversial. In contrast, no such specific plans exist for a European Deposit Insurance. The main reason for the slow progress with these last two issues is due to the widespread fear, particularly in Northern European countries, of the tacit socialization of claims that have originated due to local political flaws.³ This is also one of the reasons for a review of the intrinsic value of all banking assets in preparation for the European Supervisory System (SSM). A loss of value accrued in past years and not yet realized would therefore be fully accounted for either by recapitalization or by bail-in – as a necessary prerequisite for a start of the Banking Union. Under the heading of “Asset Quality Review”, discussions and negotiations are already under way about the rules of play when it comes to splitting up old and new losses. Note that a proper AQR carries significant risks for financial stability if not implemented properly.⁴ In addition to the pronounced concern, particularly in Germany, about a Europe-wide collectivization of liability, there are also competition-based concerns about a transfer of decision-making power and resolution discretion from a national level to a European (or transnational) level. For example, key players (e. g., association representatives) in Germany regard this as a threat to the so-called three-pillar banking system. It is now feared, European supervisory bodies will

 Thus, there are very differing assessments regarding anchoring supervision with the European Central Bank and resolution with the European Commission, say, and its respective conformity with the European Treaties.  A corresponding advertising campaign by the German Association of Savings Banks and of the Cooperative Union on 12.09. 2012 [http://www.dsgv.de/de/presse/pressemitteilungen/ 120912_PE_DSGV-BVR_EU-Einlagensicherung.html] highlighted the high level of sensitivity.  To see the perils, consider a tough AQR revealing significant asset write-downs for some banks. Investors in bank liabilities would then probably avoid banks with lots of writedowns, in fear of possible bail-ins, unless these banks have ample back-stop funding assured from third parties. Note that “avoiding a bank” is a euphemism for a rush of unsecured depositors to the exit, i. e. a run on bank assets. Thus, an improperly designed AQR may bring down the banking system and may induce the supervisor not to be as tough with asset valuation as announced. A classical Catch 22 – which can only be avoided if there are sufficient funds available for an orderly, non-contagious wind down of weak institutions.

26

Jan Pieter Krahnen

have no respect for the inherited banking structure as a specific national characteristic, as would be the case with a primarily national supervisory body.⁵ Taken together, both concerns – legacy assets of current banking system and distinctive characteristics of the national financial architecture – suggest that it will still take a major political effort of epic proportions to bring the issue of European deposit insurance, as well as the problem of processing, to a positive outcome, with the Banking Union as currently envisioned. There is a need for a convincing institutional solution to the issue of limiting national liability, for which, to date, only isolated proposals have been made.⁶ As the Banking Union (BU) project will hardly advance further without support from its biggest stakeholders, notably Germany, the sustainability of the project continues to be in question. A BU project is only workable if all four institutional features mentioned above – SSM, RRD, SRA and DGS – come into force at about the same time. These four features are complementary to one another and mutually effective. If individual elements are not implemented, the much-hoped-for restoration of the functional market-based rules in the banking sector will likely not take place. Therefore, particularly without an adequate resolution mechanism including the requisite financial reserves or commitment, it will be hard to solve the too-big-to-fail problem and return government bank rescues (and therefore also government bail-outs) to what they were always meant to be: a radical exceptional measure and not a permanent state of affairs. The following pages lay out an interpretation of the crisis – nowadays often referred to as a crisis narrative in section 2, followed by the key proposals of the Liikanen Report in section 3. These include the creation of a separate banking system for large institutions and the obligation on the part of all banks to issue subordinate debt capital held by non-banks. Possible implications for banks, investors and supervisory authorities are discussed in section 4, including adapted business models for banks, the reaction of the investment sector to stricter liability regulations for private investors, and extended powers for banking supervisory authorities. A consideration of remaining unanswered questions brings the paper to a close in section 5.

 Private banks, two cooperative banking groups and the state-owned/community-owned savings banks (Sparkassen and Landesbanken) make up roughly 30 %, 15 % and 35 % of the 2007 market (numbers are averaged over a) deposits from non-banks and b) loans to nonbanks (excluding specialized banks).  Krahnen (2012), Deposit insurance suitable for Europe: Proposal for a three-stage european deposit guarantee scheme with limited european liability, Policy Letter No. 8, Center of Excellence SAFE.

Rescue by Regulation? Key Points of the Liikanen Report

27

2. Liikanen Report: Interpretation of the crisis The proposals contained in the Liikanen Report need to be viewed against the background of the crisis narrative laid out in the first part of the report. This describes the growth of the banking sector in the European Union (EU-27) since 2000 based on balance sheet figures. The banking sector, (meaning the cumulative balance sheet figures of all EU-27 banks), as related to the gross domestic product (GDP), is considerably larger than in the USA. The corresponding figures for 2010 are 350 % for Europe and 80 % for the USA. These figures are inversely related to the size of the market for corporate bonds, which is large in the United States, as compared with a relatively much smaller market in Europe (although there is a large market for European bank bonds). Even if, on closer inspection, the comparability of the absolute figures – partly due to different accounting policies for derivatives – is questionable, the fundamental issue nevertheless remains that the European banking sector has a dominant role in the national economies’ savings and investment processes – and therefore the restoration of a properly functioning banking market in Europe is not just more critical than in the USA, but is also potentially harder to achieve. The Liikanen Report’s analysis of the financial crisis since 2007 diagnoses a substantial and sustained breakdown of the operation of the banking market in Europe. Fundamental to this breakdown, expressed in terms of low (average and marginal) borrowing costs and high bank debt levels, is the (persistent) state of acute systemic risk, which was only fully recognized for the first time following the collapse brought on by Lehman Brothers. We understand systemic risk as meaning the risk of a simultaneous insolvency of a number of financial institutions, causing a serious impact on the proper functioning of the so-called real economy. Almost imperceptibly,⁷ the risk of systemic banking crises grew in the years preceding the outbreak of the crisis. The nature of this systemic risk in the financial system, which had not been experienced in the past, can be attributed to the substantially increased level of interconnection between financial institutions, as compared with banking in earlier decades. Over the past 25 years, the rapid growth of derivatives markets, the increasing role of secured and unsecured interbank lending and the heightened dependency on refinancing funds borrowed short-term on the capital market, significantly increased interdependencies

 However, refer to the early warnings from the Bank for International Settlements, under the authorship of Claudio Borio and William White (Borio/White 2003), as well as Rajan (2005), in a widely recognized critical paper at the Federal Reserve conference in Jackson Hole.

28

Jan Pieter Krahnen

amongst financial institutions. Indirect relationships, such as the correlation of the mark to market valued assets on bank balance sheets and liquidity-sensitive market prices in the event of the fire sale of exactly these assets, further reinforce these inter-dependencies. Direct and indirect bank inter-dependencies, in turn, fuel the risk that difficulties encountered by individual institutions or price slumps on individual product markets will “infect” other markets and be transferred to other institutions. Large sections of the banking system may be affected by a type of knock-on effect. The risk therefore increases that basic financial services – specifically payment transactions and loan provision – will no longer be available to the real economy across the board, resulting in high expected economic (welfare) costs. The anticipation of a collapse of this nature is referred to as systemic risk. Should this occur, government intervention and creditor bailout is almost always required to avoid the anticipated significant costs to the real economy. Given the current circumstances, this unavoidable (’compulsory’) government rescue of individual institutions or entire groups of institutions in the event of a crisis, has a significant ex-ante impact on the behavior of the key players in the market and on the market prices observed. Indeed, it is a fundamental matter of fact that players on the financial markets – investors, borrowers, creditors – can reasonably anticipate the forced rescue by the state (i. e. taxpayers) and price this in when valuing securities – and more generally adjust their behavior accordingly. In terms of the experiences of 2007– 2012, it was primarily creditors, including hybrid investors, who expected a rescue by taxpayer funding in the event of a crisis.⁸ On the other hand, equity investors expected that they would, at the very least partially and often completely, lose their capital. Completely in line with rational expectations, this experience has led to a far-reaching adjustment of market prices with these apparently inevitable and therefore predictable rescue operations. In a widely recognized study that included data from the US, UK and the EU, Schweikhard/Tsesmelidakis showed as early as 2011 that the ratio of (risk neutral) estimations of default risk based, on the one hand, on CDS prices and, on the other hand, on share yields related to the same respective companies follow a standard pattern. More specifically, the two default risk estimators mentioned above have systematically deviated from each other since mid-2007, with stock markets signaling a significantly

 See the extensive collection of case studies on the practice of bail-outs in Europe during the financial crisis in Dübel (2013).

Rescue by Regulation? Key Points of the Liikanen Report

29

higher risk than the credit markets. Interestingly, these discrepancies in assessing risks essentially only apply to banks and hardly ever to non-banks.⁹ Taken together, the observations of Schweikhart/Tsesmelidakis (2011) can be explained as the expression of an efficient but dysfunctional market. The bank refinancing market is dysfunctional because market interest rates do not reflect the actual default risk, due to the expected ’rescue’ of the debt in the event of a systemic risk, as well as due to the general expectation that there is a high probability of a case like this occurring. Instead, interest rates for bank refinancing reflect the much lower, remaining default risk, after taking into account likelihood of the government rescue operation. As the same argument does not apply to equity capital, or at least to a much lesser extent, both prices and thus the two risk signals systematically fall apart. As a consequence, conflicting risk information from the price signals in the two markets for equity and unsecured debt can arise. While the equity market is reporting high default risks and correspondingly claiming high expected returns, the default risks appear low through the lens of the debt market – an incorrect assessment, resulting from the dysfunctionality of the market. In spite of this dysfunctionality, the market is nevertheless efficient – at least in the conventional sense used in financial economics, according to which efficiency is synonymous with rational expectations, i. e. implying the speedy processing of all available information. However, the dysfunctional, efficient market now provides the basic informational input for banks’ risk management. Thus, the implicit risk-information inherent in prices for bonds and swaps, such as CDS, forms part of the monitoring data for banks’ internal risk models. It is therefore conceivable that, before and during the financial crisis, the banks’ risk managers acted in good faith in the signals received from their markets. As a consequence they took – in a sense clueless – increasingly greater risks.¹⁰ The line of argument thus ends in a self-reinforcing cycle (a ’vicious circle’), in which existing systemic risks lead to expectations of rescue, and they, in turn, to indirect subsidies to the owners of banks (via subsidized refinancing costs).

 One of the exceptions is General Motors, an industrial operation of substantial importance for the US economy which, at the same time, has played a prominent role as a reference value in the American CDS market. In 2009 GM experienced a bail-out as part of the TARP program.  The quintessence of this statement is that it does not require greedy (etc. ) bankers to explain a cycle of increasingly high risk-taking followed by public bail-outs – it merely requires the externality (of the systemic risk) on debt market prices, and rational expectations on the part of those involved.

30

Jan Pieter Krahnen

This type of vicious circle reasoning summarizes the crisis diagnosis underlying the Liikanen Report.¹¹ Against the background of this explanation of the crisis, the Liikanen Proposal now attempts – not surprisingly – to break through the vicious circle described above. The core idea here is to reinforce the processing capability of individual institutions and, at the same time, eliminate the contagion effects between banks, at least at the level of subordinated debt. In this context, we should remind ourselves of the repeated experience of bank rescuers – in Germany as well as in other countries – that can be described as the “weekend effect”. This experience relates to the progress of a banking crisis, which typically starts on a Friday with a call to a banking inspector, in which the institutions’ difficulties are disclosed. Then slightly over 48 hours remain, i. e. up to the opening of the Tokyo Stock Exchange at midnight on Sunday/Monday CET (and the possibility of international investors reallocating their capital), to find a solution comprehended by the markets for the bank’s precarious situation. The institution, in its restructured form – to the extent that it is dependent on market financing and thus is exposed to a run by uninsured depositors – then has to regain a state of solvency visible from outside. The sobering experience of several “weekend restructurings” like this then culminated on the Sunday in a tax-financed rescue, i. e. in a bail-out.¹² The sheer complexity of business transactions and the high level of penetration with frequent short-term market-related transactions played a key role here. The complexity of the business portfolio of individual banks is often extremely high, impeding the ability of an institution to resolve matters within the narrow window of a weekend action. The two proposed measures – strengthening resolution capability, reducing contagion effects – will be discussed in more detail below. Both, taken together, have the same aim: comprehensive replacement of taxpayer liability by liability on the part of bank creditors. If we succeed in making the resolution of banks credible with the involvement of creditors, market prices for bank capital will once again reflect the actual cost of the risk, so that banks will also choose

 It should be noted here, only for the sake of completeness, that this analysis differs fundamentally from the crisis analyses on which the recommendations made prior to the Liikanen Report for the USA (Volcker-Rule as part of the Dodd-Frank Act, 2010) and Great Britain (VickersReport , 2011) were based.  This statement can be verified with the descriptions of the IKB bank, LB matters and on the HypoReal Estate cases. Cf. Spiegel-Online , 7.7. 2009 , “HRE Rescue: Top bankers feared the death of their institutions,”, http://www.spiegel.de/wirtschaft/hre-rettung-top-banker-fuerchtetenden-tod-ihrer-institute-a-634716.html

Rescue by Regulation? Key Points of the Liikanen Report

31

less risky strategies (i. e. business models). As a result of reduced intensity of contagion in particular, ultimately the systemic risk will again fall to the inevitable minimum level inherent in every functional banking system.¹³ The two key structural recommendations of the Liikanen Report will be presented below and discussed in more detail.¹⁴ The main thinking behind the two proposals follows the logic of the crisis narrative presented above. The restoration of private liability in the banking sector is the prerequisite for a revival of market discipline – relating here mainly to the price mechanism for debt instruments. Note, however, the argument can also be applied to the interests of bank creditors (thus the unsecured depositors) to the management of the bank and its owners. The two proposals should therefore be understood as being complementary to the four pillars of the Banking Union.¹⁵

3. Liikanen Report: Key reform proposals a. Separation of trading from universal banking: facilitating resolution The proposal for the separation of the banking system with a generous de-minimis rule, which limits application to large institutions, took up the majority of the space in the Liikanen Report and in the related press reports. It remains to be discussed below, whether the two proposals do justice to this implied assessment. I will present the sequence as listed in the Report regardless of my own emphasis of this question. The first proposal sets out the breakup of large and complex institutions into two entities, in order to facilitate a continuation of its solvent parts, and a winddown of the rest. The background for the proposal is the growing interconnection of traditional information-driven banking (relationship/commercial banking) and modern, capital market-oriented trading business (trading/investment bank-

 For the causes of and limits of prevention of systemic risk, refer to Bluhm/Krahnen (2011).  Additionally, there are diverse recommendations, among them larger capital buffers for trading assets and a strengthening of corporate governance.  At the time of drafting of the Liikanen Report, it was impossible to foresee the extent to which the ’bold design’ of a Banking Union could succeed; the HLEG (Liikanen Commission) has therefore basically assumed the success of this major political project with its four building blocks (RRD, SSM, SRM, DGS) and has effectively concentrated on the two capstone provisions: Facilitating resolution, and reducing contagion.

32

Jan Pieter Krahnen

ing). The – ultimately untested – theory behind this proposal is that at least a significant part of the complexity of banking, making dismantling and resolution impossible over a crisis weekend, derives from the interaction of banking and commercial transactions.¹⁶ Given this background, the Liikanen Report proposes separating trading from the banking activities of large financial institutions into two legally independent entities (’trading bank’ and ’commercial bank’), both being managed under a single holding company.¹⁷ Capitalization and refinancing of both units must be separate: there can only be a guarantee from the trading bank towards the commercial bank, but not vice versa. This would ensure that the implicit government guarantee that exists for a commercial bank, thanks to its deposits,¹⁸ cannot readily be transferred to the trading bank. This thereby prevents the cross-subsidization of the trading banking arm by the commercial trading arm, with the corresponding distortion of the investment and risk allocation. Trading banks and commercial banks should each provide their own funding, and thus be confronted by the market with their own respective risk costs. Based on the assumption that in the past there has been substantial cross-subsidization, one would expect the separation of both banking arms to cause the refinancing costs of the trading bank to rise and those of the commercial bank to fall. This pricing adjustment would result in the tendency for trading on the overall market to be reduced.¹⁹ The main objective of this measure is not to reduce trading activities per se or to call into question the economic benefits of trading – there is no concrete evidence on this in academic studies. The Liikanen

 The Liikanen Commission interviewed a number of European institutions during its work stream about their business models, cf. European Commission (2012)- Liikanen Report. Nevertheless, independently of this off-hand sample of bank business models, serious empirical research remains a key task for supervisory authorities, and an interesting field of study for academics.  The economic details of a break down like this are not discussed in the Report, as this probably lies beyond the time capacity and technical possibilities of a high-level expert group.  Admittedly a state guarantee can be ruled out legally – but it is doubtful that the players (Central Bank, supervisory authorities, Ministry of Finance) can adhere to such a requirement in an acute case of systemic risk. This explains the use of the term “credible” in the last sentence.  However, predictions about the impact of such a rule change, such as the separation model described here, are difficult because it is not a partial behavioral change, but rather a change of the entire system leading to a new general equilibrium. This can, for example, result in new prices for trading products on the markets, which may compensate for the assumed pricing effect, either in whole or in part.

Rescue by Regulation? Key Points of the Liikanen Report

33

Proposal is solely about avoiding cross-subsidization between the trading bank and the commercial bank. The proposed separation also provides for a threshold (de minimis rule) relating to the absolute size of the bank (€ 100 billion trading assets) or the relative volume of these assets (15 – 20 % of total assets). Perhaps the most important detail in the separation proposal put forward in the Liikanen Report is the intention not to differentiate between proprietary trading and client-related business or between proprietary trading and market-making. Proprietary trading, so the argument goes, can scarcely be separated from client business, as market-making in less-than-perfectly liquid markets consists essentially of a sequences of trades that end up on the bank’s own books.²⁰ The separation of market-making and proprietary trading (or client-related and proprietary trading) in large institutions has encountered considerable opposition from the industry. They argue that neither were trading activities at the center of the crisis, and nor were banks with large trading books any more affected by the crisis than other institutions. Why, then, should trading be singled out for separation, thereby rendering universal banking, including trading activities, less attractive as a business model? Moreover, competition on the international market would be much more difficult, unless there was a level playing field. As outlined above, the reason the Liikanen Report is pushing for the separation of trading from universal banking is to increase the chances for resolution and creditor bail-in. International universal banks have become extremely complex internally, and their trading activities have played a major role in this development. Any attempt to rescue a failing institution over the weekend, the infamous “flash” emergency event, is doomed to fail if trading activities and traditional banking are closely interrelated. Over the past two decades, major international banks have become specialists in risk management and hedging services. Thanks to the massive order flows and a strong position in many derivatives and securities markets, they are in a position to offer quick execution of clients’ orders, offering liquidity to their clients. By extensively using netting possibilities and innovative ways to internalize transaction flows and manage risk in terms of portfolio models, a bank is able to

 In the US, the Volcker rule as part of the Dodd-Frank Act, prescribes that banks should fully separate their proprietary trading activities from their other business, while market-making and client-related business remain untouched. The regulatory changes in Germany and France follow this American model. According to the ideas of the Liikanen Report, however, the separation of proprietary trading activities would be difficult to enforce, as modern forms of market-marking are hard to differentiate from proprietary trading.

34

Jan Pieter Krahnen

significantly reduce the required number and volume of hedging transactions with third parties. The savings of embedded spreads and risk premiums translate into bank earnings. It is precisely this integrated commercial/trading bank with its complex internal portfolio of risks that renders quick restructuring difficult. On the other hand, a separation of trading from banking will create two distinct institutions: a trading house (or broker-dealer) and a remaining (universal) bank. Both institutions will have their own separate equity capital, possibly provided by a mutual bank holding company. While the universal bank will be refinanced, as before, by deposits, bonds and unsecured credit, the trading house will have its own funding now, most probably from bond or inter-bank wholesale markets. It will not have access to the deposit market and therefore will not benefit from an implicit government guarantee.

b. Subordinated debt for non-banks: reducing contagion Will a separated banking system be more immune to systemic risks than an integrated banking system?²¹ The answer to this question is not an easy one: when the occurrence of a risk has been identified (such as a run on the deposits of one or more banks, triggered by a sudden shock to the bank’s asset value), it is impossible to make a clear diagnosis of the effect on systemic risk. The risk of a simultaneous collapse of various institutions may be smaller or greater in a separated banking system than in an integrated system. The trading bank’s level of equity capital plays a key role here, as does the level of unsecured debt vis-a-vis commercial banks. However, if the occurrence of the risk case is not predetermined but endogenous in its operations, i. e. resulting from the normal business conduct of the institutions involved, it is possible to venture a diagnosis. As a separated banking system is believed to have improved resolution capabilities in individual institutions, the financial markets will consider (and price in) the reduced government guarantee and thus create an incentive to reduce bank risks and mutual

 The term ’universal’ is avoided at this point – because, according to the Liikanen Report, the bank left after separation of trading activities will also continue – and should continue – to be known as a universal bank in the separated bank model. An institution like this will include investment banking (IPO, SEO, syndicated lending business, M&A, Private Equity) alongside typical commercial banking activities – just not trading. External contracts will be awarded for this, partly to the proprietary holding “sister” company and partly to other broker-dealers in the market.

Rescue by Regulation? Key Points of the Liikanen Report

35

dependencies. Both will lead to reduced systemic risk, because chain reactions will become less probable across the banking system. A simple example illustrates this. Let us assume that the unsecured liabilities of a bank (also) consist of outstanding accounts from other banks, as was suspected, for example, in 2008 with the German Industrie Kredit Bank (IKB). The aforementioned ’weekend effect’ is expressed here in the fear that a writedown of outstanding debts will trigger a refinancing crisis the next day – possibly starting a systemic banking crisis. In view of this danger, the supervisory authorities are basically forced to call upon taxpayers for help and initiate a bailout. This measure will successfully stabilize the banking system this weekend – but at the cost of increased systemic risks in the future. If you look at the example in more detail, you will recognize that the real problem, from the viewpoint of the supervisory authorities, lies in identifying the depositors. As they may be banks themselves, the supervisor is barred from bailing-in private debt, as would normally be the case in the usual bankruptcy proceedings for industrial firms. If the supervisory authorities knew with certainty that the creditors of the unsecured bank debt were non-bank investors, rather than banks, nothing would stand in the way of a bail-in. The simple yet fundamental contribution of the second structural proposal laid out in the Liikanen Report asks for holding restrictions for bail-in bonds. These are intended to limit the identity of the buyers of subordinated bank capital. The restriction called for is a simple one: banks should be prohibited to hold bail-in bonds on their balance sheet. The justification for this follows from the above: if the supervisory authorities can assume that the owners of bail-in bonds are not themselves banks, then they can initiate a write down (or a debt-equity swap) without having to fear contagion or immediate systemic consequences. And as this is the case, their announcement of a bail-in will also be credible ex-ante. However, it is precisely this credibility that is called into question, if, at a moment of crisis, there is uncertainty on the part of the supervisory authorities concerning the identity of the holders. The following story, taken from the Financial Times, may underpin this consideration. The newspaper reported on the TPG (Texas Pacific Group) investment fund. The Group, under the direction of its founder David Bonderman, invested around 7 billion dollars in the ailing bank Washington Mutual in 2008. As the newspaper reported, the fund manager based his decision to invest on a comprehensive bailout by the U.S. government (i. e. FDIC). His conviction was based on the devastating experience of a bail-in at Lehman Brothers, a short time earlier. To TPG’s great surprise, however, the FDIC refused the bailout and instead practiced a bail-in. TPG lost its entire investment, $1 billion of its own money and another $6 billion of clients’ money. It is to be assumed, as an explanation for the conduct

36

Jan Pieter Krahnen

of the FDIC – although the reader can only speculate here – that the financial supervisory authorities could only risk the bail-in because the identity of the investors, i. e. TPG as a Private Equity Fund, was known and so they estimated that it was unlikely that this would have a retroactive effect on other financial institutions: “Sheila Blair, the FDIC’s Chair at the time, described in her memoirs how shocked she had been by Mr. Bonderman’s “combative way” in pressing for access to Fed lending. TPG lost all its money (more than $1bn) as well as that of its investors and co-investors, a group that included major sovereign wealth funds and state pension funds. But Mr. Bonderman’s admirers point out that if the government had decided to bail out the bank instead, it would have been a fantastic investment.”” (Private Equity: Last of the risk-takers? in: Financial Times, European Edition, 19. 8. 2013, page 5).

This second proposal in the Liikanen Report therefore addresses the core problem of the dysfunctionality of the bank refinancing market. It directly addresses the reason for the ubiquity of systemic risk in today’s banking sector. The proposal requires banks – all banks – to issue a minimum amount of unsecured bonds with specific holding restrictions. Forcing banks to issue subordinate debts is, by itself, not unusual, as most financial institutions already have a host of junior or hybrid tier-2 instruments outstanding.²² It should be stressed, to avoid misunderstandings, that even after the creation of special bail-in bonds, all items on the liabilities side of a bank will fundamentally continue to be subject to private liability, that is to say will be bail-in-able.²³ Thus the special feature of this new class of subordinated bonds is not in the loss absorption capacity but rather in the credibility of the liability announcement. The essential innovation in the Liikanen Proposal is therefore an explicit holding restriction as a bond covenant. It proposes that no institution within the banking system, and thus an institution potentially linked to the outbreak of a systemic risk, may hold the aforesaid subordinated debt at any time. The institutions outside of the banking sector that may hold subordinated debt will be discussed below. Importantly, a retrocession i. e. the transfer of the default risk of this subordinated debt back into the banking system, through the purchase of

 Much has been written about the role of junior debt to incentivize banks, cf. Calomiris 2000, Calomirirs/Herring (2010) – although the issue of a holding restriction has been ignored up to now in all the literature. The reason for this lies in the point of view of the individual institutions dominant in the literature. In contrast, the Liikanen Report takes into account market-wide systemic risk, and the severe consequences of contagion for financial stability.  Paul Tucker, then Deputy Director of Financial Stability at the Bank of England, has repeatedly cautioned against the term “bail-in debt”used in the Liikanen Report, as implicitly attributing a bail-out characteristic to other and more senior layers of bank debt, like bank bonds and unsecured deposits.

Rescue by Regulation? Key Points of the Liikanen Report

37

credit default swaps (CDS), where a bank acts as protection seller, needs to be systematically ruled out. By allocating junior debt (bail-in debt) outside of the banking system, losses exceeding the equity of the affected institution will be borne, up to the amount of such debt, by these outside investors, without a direct contagious feedback effect. A supervisory authority considering a bail-in can be confident that its decision will not trigger the next systemic crisis in the banking sector: this situation can and should encourage it to involve creditors in the rescue. In turn, creditors will know from the very start, i. e. from the date of issue, that the threat of losing part or all their capital is a very real possibility and that taxpayers will not jump in – or will do so much less predictably. Accordingly, the actual default risk of any particular bank – subject to market efficiency – will be reflected properly in market prices of debt as well, and the disciplinary effect of these markets will be resurrected. According to the Liikanen Report, the strict holding restriction can be replaced by a provision permitting banks to invest in subordinated debt of banks provided that it fully backed by tier-1 equity (risk weight = 1250 %). Both formulations lead to essentially the same thing: subordinated or junior (“bail-in”) bank debt will be held outside of the banking system, minimizing the systemic feedback effects of bail-in operations on the stability of the financial system. The holding restrictions make the affected capital “bail-in-able”.

4. Consequences for banks, investors and supervisory authorities Much of the regulatory reform project of the last couple of years has been welcomed by the financial sector: there was a positive response to the Basel III innovations relating to the strengthening of equity capital and the introduction of a leverage ratio. The Banking Union project with its focus on a level playing field for supervision and resolution within the European Union has also met with broad support. However there is strong opposition on specific issues of concern for national institutions and their business models. There are also concerns about the future role of national supervisory authorities. As a result, there is a general lack of willingness to transfer authority from a national to a supranational level.²⁴

 In terms of joint banking supervision, it continues to be unclear what powers the ECB (as the Supervisory Authority) has over the largest banks and what decision-making or veto rights it has over national authorities. This point is particularly important for decisions with fiscal conse-

38

Jan Pieter Krahnen

The submission of the Liikanen Report raised a number of questions relating to the possible organization of regulation. I now wish to address some of these questions, without any claim to completeness. The structure of the existing organizations will firstly be based on three frequently occurring questions on the separation model: what will be the effects of a separated banking model on the profitability of a bank, on competitiveness between banks and the liquidity of the securities market? In conclusion, I will address possible effects of the introduction of an issuing and holding regulation for subordinated debt (bail-in capital). I will address three aspects here: Who should hold the junior bonds, if banks are prevented from doing so? Will there be a sufficiently large group of investors? Should the holding regulation be supervised? If yes, by whom?

a. Consequential considerations regarding the separation proposal Overall, the Liikanen proposals on separation have largely met with opposition from the industry. In Germany, for example, the associations representing all three pillars of the banking sector, the public savings banks (Sparkassen and Landesbanken), the private banks and the cooperative banks, have argued that a separation of trading activities would threaten their proven model of universal banking. Each of the three aforementioned parties offers up somewhat different reasons for their opposition, a factor that can be explained by their own very different organizational and business models. It is impossible at this juncture to provide a comprehensive study of the anticipated effects. The following comments are purely speculative and should be understood as a prelude to a more academic debate involving testable hypotheses.

i. Reduced profitability? One commonly held argument refers to the profitability of a stand-alone securities firm, or broker-dealer. Some commentators have argued that stand-alone funding, as is implied in the separation concept, will lead to higher refinancing costs for the broker-dealer, lowering the profitability of market-making and

quences, such as with bank resolution. No decision has been made to date about which decision-making powers should be transferred to a designated unified resolution mechanism.

Rescue by Regulation? Key Points of the Liikanen Report

39

therefore reducing the provision of liquidity in markets dominated by marketmakers. An impact study carried out in the United States on behalf of the US Chamber of Commerce analyzed the consequences of implementing the Volcker Rule. It came to the conclusion that implementation will have a negative influence on market liquidity (cf. Thakor 2012). However, the latter study is not entirely convincing, as the influence of rising costs was only analyzed for a single financial institution, while the effects of a general rule change on the entire equilibrium were disregarded. In other words, the separation of the trading activities (and most importantly marketmaking) of all institutions in the market would be likely to change other prices as well, not just broker-dealer refinancing costs. In particular, one would expect that after such a rule change the price of market-making services might go up as well. The overall impact on the profitability of broker-dealers has to encompass three adjustments: refinancing costs (up), the price for market-making services (presumably up) and the volume of transactions (presumably down). The resulting spread, market turnover and, hence, the profitability of the broker-dealer is thus difficult to predict. It is even conceivable that a new market environment, defined by legislation as separating market-making and proprietary trading from universal banking, might reduce the number of institutions offering these services and increase profitability from trading for the remaining institutions.²⁵

ii. Regulatory arbitrage? Another argument raised against the separation rule challenges the Report’s threshold value, the so-called de-minimis rule, according to which only banks with significant trading books have to transfer their trading activities over to a separate entity. The Report has suggested drawing the line at an absolute size of the trading book of € 100 billion Euro (or when the trading book exceeds 15 – 25 % of total assets). Would such a generous de minimis rule invite mid-sized banks to expand their trading activities up to a level just below the critical threshold? And wouldn’t a market development like this take away significant business volume from larger institutions and make their separate broker-dealer business ineffective or unprofitable? The concentration of trading activities observed in ever-fewer large players makes the aforementioned scenario appear less likely, assuming there are indeed

 This would then be the case if it produced positive economies of scale for investment firms.

40

Jan Pieter Krahnen

real economies of scale in the broker-dealer industry. Without these economies of scale, there would be a decentralized trading architecture already in place and the separation requirement would entail no (further) negative effects on bank resolvability – as the complexity of universal banks with relatively small trading business is believed to be manageable anyway. Conversely, however, if a few large broker-dealer institutions prevailed in the market in future, these institutions could present a significant systemic risk given their high degree of interconnection with many non-broker-dealer banking institutions across Europe. The Liikanen Report therefore recommends the further tightening of equity capital requirements on trading institutions

iii. Reduced liquidity? Major providers of market liquidity in today’s financial markets are above all large international universal banks with significant trading books (assets held for active trading or available for sale). To what extent is their business model dependent on unrestricted access to customer order flow? Put differently, is it possible that a bank like this can offer its customers the usual range of services, if its trading business has been separated into a broker-dealer institution? It seems at least conceivable that a newly separated broker-dealer institution could continue to advise the universal bank with respect to financial strategy and risk management, leaving merely the execution of trading orders to external broker-dealers in the market. While this may lower overall the profitability of broker-dealer services, as competition among them is rising, this would not necessarily mean a decline in the profitability of the banking institutions. In principle, it would also be possible that profits remain at current levels, for example if execution prices rise or if a reasonable price is paid for the advisory services provided by the “old” broker-dealer. These compensatory price effects are difficult to estimate overall, as there may well be new players in the broker-dealer markets putting a certain pressure on their profitability.

b. Consequential considerations on the creation of bail-in capital Who should hold subordinated bonds if other banks are prevented from doing so? Will there be a sufficiently large group of investors? Should the holding regulation be supervised – and by whom?

Rescue by Regulation? Key Points of the Liikanen Report

41

The absence of a real risk transfer outside of the banking system was at the root of the early stage of the financial crisis, as was evidenced during the first stages of this financial crisis by the allocation of Asset Backed Securities tranches (see Franke/Krahnen 2009 for an assessment of why securitization failed so dramatically). Rather than transferring the risk to investors, as was claimed by issuers and confirmed by some supervisors, risk was in fact at least partially shifted to and fro between financial institutions, resulting in increased interconnection, contagion and systemic risk. The Liikanen Proposal prohibits banks from holding subordinated bonds from other banks and, at the same time, forces them to issue these bonds – a factor that will, by construction, contribute to a significant transfer of risk outside the banking system. This is actually the main objective of the (really inaccurately named) “bail-in-instruments”. Creating a layer of subordinate bank debt, from which there can be no contagion amongst banks, represents, by its very nature, a form of non-systemic risk. The associated risk transfer is therefore credible, as a rescue of the creditors by the state should not be expected. Other requirements relating to credible bail-in target the assurance of a lasting cost effect for banks. It should be ensured that the bail-in bonds issued have finite terms (i. e. are not consols) and have staggered maturities. The former is intended to prevent the need for further issues when an adequate bail-in volume has been achieved, by making the bond coupon payments directly dependent on a bank’s actual (current) risk. The staggering of debt maturities ensures that there is no one-time refinancing cost effect in the event of a high bail-in volume, thereby jeopardizing the solvency of the bank and triggering a run, but rather a smoothing out over an extended period of several years.²⁶

i. Who will bear the risk for bail-in debt? Conditions are to be set on the future holders of non-systemic subordinate debt: they may not hedge themselves in the banking sector and they should not themselves experience refinancing difficulties, when suffering losses on their investments caused by a bail-in. Diversified investment firms with long term maturities on their funding side, restrictive termination rights and variable payment promises appear to be suitable investors. Life insurance companies, as well as sovereign wealth funds, pension funds, high-net-worth individuals and specialized

 I am grateful to Frank Heinemann for pointing out this implication

42

Jan Pieter Krahnen

hedge funds seem to fit this profile in particular. In all cases, the attractiveness of an investment is based on the expected high interest coupon. If the market is efficient in terms of information aggregation and the announced bail-in is credible, the interest coupon to be paid by the issuer will be at a level that fully covers the expected loss from the possession of a bailin bond and will include a risk premium on top. For an insurance company, the high interest coupon will therefore only pay a distributable profit after many years of setting aside accruals – the revenue previously accrued will be used as provision for anticipated losses from outstanding bail-in bonds. If investors, particularly life insurance companies are aware of this fact – and do not prematurely pay out their profits to clients, the investment world will now have in place a robust shock absorption mechanism, thereby stabilizing the system at large. If this is the case, the financial system can then speak of a successful transfer of bank risks, free from negative feedback.

ii. Will this create a sufficiently large market? One argument, raised many times in discussions, questions the likelihood of a sufficiently large market to absorb the subordinate bonds called for in the Liikanen Report. In order to assess this argument, it is worth estimating the assumed size of a market for junior bank bonds. Estimates based on a target of 5 % of total banking assets in 2012 indicate a required issue volume in the order of € 2 trillion. This figure is around 5 to 6 times as high as the European junk bond market in the same year.²⁷ Against the background of this figure, the intention to issue 5 % of the total assets of European banks in the form of bonds with a holding restriction appears already to be an ambitious goal. Achieving this goal will require regulators to think ahead of time about the supposed exigency of investors when designing the instruments. A new class of asset known as ’bail-in bonds’ should therefore feature standardized bond issuance conditions to promote the emergence of a large and liquid secondary market. In particular, requirements will relate to the bail-in process itself, defining a standardized and effective triggering of the liability provision, and equally standardized form of risk-sharing. The triggering of the credit event, initiating the loss sharing, therefore needs to be defined in a transparent way, in advance, to facilitate the evaluation by capital markets. My own discussions with institutional investors point to the

 The aggregate balance-sheet total of the EU-27 banks is approximately € 13 trillion.

Rescue by Regulation? Key Points of the Liikanen Report

43

role of so-called exogenous triggers. These are pre-defined triggers for liability that are removed as far as possible from discretionary decisions by individuals, or by the supervisory authorities. An exogenous trigger would be, for example, the definition of a minimum capital ratio, below which subordinate debt would be triggered. Other conceivable exogenous triggers would be market prices, like the respective bank’s share price or the price of credit insurance (CDS). Both cases produce further questions: To what extent is an exogenous trigger actually ’exogenous’? That is, to what extent is it vulnerable to specific influence by the credit institution per se, or by a third party – and to what extent does this affect the role of private investors in preventing systemic contagion? There is also the question of the extent to which the demand for an exogenous trigger could lead to an incorrect (type-II error) credit event. In conclusion, the optimum combination of exogenous and discretionary triggers that would meet the interests of investors in terms of predictability needs to be defined. A further requirement of issuing conditions relates to the possible form of risk-sharing induced by a potential bail-in. In recent years, creditors were occasionally held liable in the form debt write-downs. Although the current day value loss appears to be identical between debt write down and debt conversion into equity, the implicit risk-sharing differs significantly in both cases. In the second case, of a debt-equity swap, the creditor maintains involvement in the company and, even after the bail-in. In the event of a future recovery, and he will participate in a possible future recovery in value of the equity. Involvement in potential subsequent appreciation is a call option which is limiting the cost of a false alarm (in the sense of a type-II error). The resulting ownership structure in the event of a crisis also favors the conversion of debt into equity, instead of a write-off. A gradually rising dilution would be conceivable, starting with the highest degree of dilution for former owners, followed by a lesser dilution (compared to the original nominal claim) for junior creditors. In the event of further reductions in the institution’s assets, successive senior creditors could become subject to the bail-in in the same form (conversion and dilution). The basic rule of private liability therefore applies to all liabilities, including deposits.²⁸ This procedure would ensure that the original ownership and creditor structure would largely be retained by the institutions, at least in terms of its composition. By contrast, the total write-off of owners’ and junior creditors’ positions, be it gradual over time or not, would result in erratic and, if applicable, temporary ownership structures, depending on the size of the

 In the case of deposits, the existing deposit insurance scheme will limit the loss to retail depositors.

44

Jan Pieter Krahnen

conversion steps. Due to the required membership of supervisory boards, this would not only make corporate governance more difficult in a crisis, but could also burden the valuation of these instruments with an additional risk premium.

iii. What are the requirements placed on supervisory authorities? The credible announcement of a bail-in, executed under transparent conditions without any margin for manipulation, is a prerequisite for a well-functioning market in bail-in instruments. This, in turn, signals properly the risk perceived by the market about the bank’s risk management mechanisms, creating the toehold for a stronger role of market discipline. Thus, by overseeing the bank bail-in debt market, the supervisory authorities can contribute to their own, and the market’s, trust in a proper bank resolution practice. The fear of initiating a systemic risk shock, which had been the cause of many bank rescues in the past, are now largely eliminated. If this last condition is met, it will be possible to speak of the credibility of the bail-in announcement. Ensuring this condition will present a new task which would probably be handled preferably by the banking supervisory authorities.²⁹ Monitoring “bail-in-ability” is a core element of the new regulatory framework for the European financial markets laid out in the Liikanen Report. It requires several individual steps, in the course of which new requirements will be set for data held by the supervisory authorities, including information about the identity of the holders of bail-in bonds at all times, and transparency for the market. This is associated with the possible narrowing down of the circle of potential holders. Furthermore, it is necessary to ensure that there is no transfer of the default risks of bonds into the banking sector, either by credit default swaps (CDS) or by other instruments. Overall, the supervisory authorities will be responsible alone for the necessary (self‐) confidence needed to provide prompt corrective action relating to bail-in procedures. It should also be emphasized (emphasized!) that the term “bail-in instrument”, introduced in the Liikanen Report, does not imply that all other sources of (bank) debt are in the form of bail-out debt. The opposite is the case. According to the Report, all bank debt is bail-in-able, depending on its position in the ranking order of debt instruments, and can therefore be called upon for a rescue.  Under certain circumstances, the market supervisory authorities can also be mandated with guaranteeing bail-in credibility; in Europe it would then be the role of the ESMA in conjunction with the ECB to provide uniform supervision, possibly with the national banking supervisory authorities, if applicable.

Rescue by Regulation? Key Points of the Liikanen Report

45

Only the lowest tranche, referred to here as bail-in capital, has to be of a minimum volume (e. g. 5 % of total assets) and is subject to the aforementioned holding restrictions.

5. Outlook and unanswered questions Seen from today, the future of the European banking sector largely depends on the will of Europe’s policy makers to execute a list of new laws and implement regulations, which, taken together, will form a coherent overall picture. There is a need for courage on the part of politicians to run away from the shadow of national interest and accept supranational European institutions, partly replacing national sovereignty. All this, in turn, will be needed to restore the regulatory framework that will enforce market discipline. Starting points for this include the prices for banks’ subordinated debt in particular. The Liikanen Proposal can thus make a significant contribution to the stabilization of financial intermediation in Europe and the repair of the regulatory framework of the market economy. Economic stringency is, however, not of paramount importance when it comes to executing and implementing this. A capacity to generate political majority then takes on more importance than this. The order in which the individual issues of the Banking Union are addressed, and the quality of safeguards preventing premature European-level ’community’ liability, will play a key role. Therefore, it is important to emphasize that the proposals put forward by the Liikanen Commission and discussed in this paper are particularly suitable for restoring credibility among the public of the validity of private liability rules with banks, even large banks. It remains to be seen whether a sensible, targeted institutional formation such as this is possible within the EU-27 or EU-28 without the catharsis of a renewed banking crisis.

6. References Bluhm, M. and J. Krahnen (2011). Default risk in an interconnected banking system with endogenous assets, CFS Working Paper No. 19, Center for Financial Studies, Frankfurt. Borio, C. and W. White (2003). Whither monetary and financial stability? The implications of evolving policy regimes. Proceedings of the Federal Reserve Bank of Kansas City Symposium at Jackson Hole. Calomiris, C. and C. Kahn (1991). The role of demandable debt in structuring optimal banking arrangements. American Economic Review, Vol. 81, No. 3, pages 497 – 513.

46

Jan Pieter Krahnen

Calomiris, C. and R. Herring (2011). Why and how to design a contingent capital debt requirement. Working Paper Columbia University und Wharton School, U Penn. Dübel, A. (2013). The capital structure of banks and the practice of bank restructuring. Case Studies on current bank restructurings in the Eurozone and conclusions for the reform of resolution and restructuring and regulation of funding Instruments. CFS Working Paper, Center for Financial Studies, Frankfurt. European Commission (2013). Proposal for a DIRECTIVE OF THE EUROPEAN PARLIAMENT AND THE COUNCIL for the establishment of a framework for the reorganization and resolution of credit institutions and investment firms and to amend Council Directives 77/91/EEC and 82/891/EC, Directives 2001/24/EC, 2002/47/EC, 2004/25/EC, 2005/56/EC, 2007/36/EC and 2011/35/EC and Regulation (EU) No 1093/2010, COM (2012) 280. Brussels, June. European Commission (2012). Report by the High Level Expert Group on Structural Reforms of the EU Banking Sector. Brussels, October. Franke, G. and J. Krahnen (2009). The future of securitization, in: Prudent lending restored: Securitization after the mortgage meltdown (Eds. Fuchita Y., Herring J., Litan E.; Brookings Institution), pages 105 – 161 Krahnen, J. (2012). Europataugliche Einlagenversicherung: Vorschlag für eine dreistufige Einlagensicherung mit begrenzter europäischer Haftung. Policy Letter No. 16, Policy Platform at the House of Finance, Goethe University Frankfurt. Rajan, R. (2005). Has financial development made the world riskier? Proceedings of the Federal Reserve Bank of Kansas City Symposium at Jackson Hole. Schweikhard, F. and Z. Tsesmelidakis (2012). The impact of government interventions on CDS and equity markets, SSRN Working Paper. Thakor, A (2012). The economic consequences of the Volcker rule. Report by the US Chamber’s Center for Capital Market Competitiveness, Washington D.C.

Miguel de la Mano

Bank Structural Reform and Too-big-to-fail Contents I. Introduction II. Why structural reform can eliminate the too-big-to-fail problem . Public safety nets are indispensable to resolve market failures . Public safety nets create moral hazard and force the state to protect banks from insolvency a. Without subsidies (explicit or implicit) banks will not resume lending to the real economy b. The time inconsistency problem . Banks can endogenously extend the public safety net to cover activities that do not deserve to be subsidised but enhance the bank’s systemic importance . Modern banks increasingly engage in scalable and short-term oriented activities backed by the public safety nets III. Complementarity of structural reform with the existing reform agenda . Structural reform is complementary to the enhanced EU capital and liquidity requirements framework (CRD IV/CRR) . Structural reform is complementary to enhanced bank resolution . Structural reform is complementary to Banking Union IV. Bank structural reform enjoys broad public support . Structural reform is appealing to politicians across the political spectrum and appeals to people’s common sense . Structural reform has been introduced in the past . Structural reform addresses the overbanking of Europe and the crowding out of EU capital markets V. Responding to some of the concerns raised against structural reform . Will structural reform give rise to important foregone societal benefits? . What is the impact of structural measures on market liquidity? . Would structural reform have avoided bank failures? . Do structural measures undermine the benefits associated with universal banking? VI. Concluding remarks References Annex

48

Miguel de la Mano

Stan Maes, Dimitrios Magos, Miguel de la Mano¹ [The too-big-to-fail issue] is not solved and gone; it’s still here … it’s a real problem and needs to be addressed if at all possible. … Too-big-to-fail was a major part of the source of the crisis. And we will not have successfully responded to the crisis if we don’t address that problem successfully. –Ben S. Bernanke, Chairman, Federal Reserve Board, March 20, 2013

I. Introduction Several large EU banking groups have weathered the crisis well. However, without extraordinary taxpayer support in the dawn of the financial crisis (European Commission (2011, 2012)) the EU financial system would have likely imploded due to a system-wide cascade of banking failures. State aid enjoyed by banks to date amounts to 40 % of EU GDP (€ 5.1 trillion) and has undermined the solidity of several Member States’ public finances turning a banking crisis into a sovereign crisis. State aid is a transfer from taxpayers to shareholders and creditors of financial institutions, so are not pure waste. However, to raise the funds needed to support the beneficiary banks, governments need to raise taxes, borrow from capital markets at a premium, or reallocate funds from other public programs and activities. On the margin all of these actions introduce severe distortions and the opportunity social costs of these funds is high, especially in times of crisis where targeted and expansive fiscal stimuli might be most needed. Hence, beyond the moral considerations associated with the redistribution of money from taxpayers to relatively more wealthy bank creditors, the associated total welfare costs of bank bailouts, are likely very large. But an even deeper concern is that bailout is liable to create an expectation of future bailouts, further confirming and reinforcing moral hazard and the perception that beneficiary banks are too-big-to-fail (“TBTF”)². This TBTF status further enables beneficiary banks to borrow at a lower rate than their less reckless or better managed rivals, distorting competition even further, allowing beneficiary banks to take unjustified excess risks, focus on short-term profits, and retain inadequate risk management and

 The views expressed in this paper are those of the authors and are not necessarily those of the European Commission. The paper does not prejudge the position of the European Commission in any way.  Too-big-to-fail is meant to cover too-interconnected-to-fail (TITF), too-complex-to-fail (TCTF), and too-systemically-important-to-fail (TSITF). See also European Commission (2013).

Bank Structural Reform and Too-big-to-fail

49

compensation practices at the expense of customers and ultimately taxpayers. This is self-reinforcing since in the absence of market discipline, negligent conduct in turn, makes future bailouts more likely. In response to the financial crisis and the role played therein by TBTF banks, an ambitious and broad financial reform agenda has been put in place in Europe. The consistency, expected benefits and costs of this fundamental overhaul of bank regulation and supervision are discussed in the Commission’s “Economic Review of the Financial Regulation Agenda” (European Commission (2014a)). The EU has raised minimum capital requirements and has introduced liquidity obligations to increase the resilience of banks against solvency and liquidity shocks. Tighter prudential rules and supervision will reduce the probability of bank default. The EU has also adopted legislation to introduce bank recovery and resolution frameworks to deal with failing banks. This includes the ability to bail-in creditors to force them to absorb loses or accept equity conversions. Moreover, a genuine Banking Union has been set up among Eurozone countries in which both bank supervision and bank recovery and resolution are being dealt with at the supranational level, rather than at the national level, thereby reinforcing the resilience of the Eurozone to systemic shocks and fostering cross-border activity and integration. However, the EU bank reform agenda is not complete without bank structural reform. Structural measures differ from other regulatory measures as they impose restrictions on the corporate structure of banking groups, limiting either their ability or incentive to engage in certain activities. Structural reform essentially entails (i) restrictions on the activities a deposit-issuing entity is allowed to perform within a large and complex banking group, (ii) restrictions on the economic, operational and legal links between entities within a large banking group, and (iii) restrictions on the connections within and between banking groups. Bank structural reform is being discussed and/or implemented at the highest political level in both Europe and the US, but also elsewhere and in international fora such as G20, FSB, IMF, BIS and OECD.³ In September 2013 G20 Countries gathered in St-Petersburg jointly stated:

 The European Commission tabled a proposal on bank structural reform on 29 January 2014. This proposal is inspired in the recommendations of the High-Level Expert Group (HLEG) report chaired by Erkki Liikanen (Liikanen (2012)). See also European Commission (2013, 2014a, 2014b). In the UK, bank structure reform is implemented following the recommendations of the Independent Commission on Banking chaired by Sir John Vickers (ICB (2011)). In the US, the Dodd-Frank Act added the Volcker Rule and Swaps Push Out rules to the already existing Bank Holding Company Act. The FSB is consulting on the impact of different bank structural re-

50

Miguel de la Mano

“We recognize that structural banking reforms can facilitate resolvability and call on the FSB, in collaboration with the IMF and the OECD, to assess cross-border consistencies and global financial stability implications, taking into account country-specific circumstances, and report to our next Summit.”

Several EU Member States (Germany, France, Belgium and the UK) as well as third countries (US) have introduced or are currently advancing structural reform at national level.⁴ These measures all have in common that they mandate the separation of selected banking activities from a deposit taking entity (for banks above certain thresholds) but differ in several respects, notably as regards the activities subject to separation and the rules on the strength of separation. This means that under national structural reform, banks operating in the same national market would potentially be subject to different rules depending on whether they are subsidiaries (subject to domestic reform) or branches (not subject to domestic reform). These national reform efforts may thus be ineffective, if locally incorporated banks were to relocate and branch back in (for domestic banks subject to reform) or switch from subsidiary to branch status (for banks from another country).⁵ In sum, all these national reforms are vulnerable to regulatory arbitrage. Regulatory arbitrage, as the crisis has proven, is particularly acute within the European Union. Divergent national legislation may undermine the functioning of the Internal Market and the level playing field. National measures have the potential to improve domestic financial stability but if other Member States fail to adopt similar rules, or worse, actively adopt conflicting measures, then banks will likely remain TBTF. Divergent national structural reform legislation will also tend to undermine the single rulebook applicable throughout the Internal Market. Importantly, this would also weaken the effectiveness of the Banking Union since the SSM would have to supervise banks subject to different bank legislation. Bank efficiency would deteriorate since management of cross-border institutions would become more difficult and costly, notably in terms of ensuring compliance with divergent and possibly inconsistent rules. In itself this would

form initiatives across the globe (FSB (2013)). The BIS, OECD and IMF have written extensively about bank structural reform in the recent years (Arcand et al. (2012), Blundell-Wignall et al. (2012, 2013, 2009), Boot and Ratnovski (2012), Cecchetti and Kharroubi (2012), Chow and Surti (2011), Gambacorta and Van Rixtel (2013), Laeven et al. (2014), Lumpkin (2011)).  Structural reform measures are also under consideration in the Netherlands.  The effects could also be more pervasive in the sense that banks, rather than relocating by legal means, could relocate in “economic” terms and thus avoid national rules by moving particular activities (by e. g. booking certain transactions in another part of the banking group located in another Member State).

Bank Structural Reform and Too-big-to-fail

51

discourage cross-border banking activity and contribute to market fragmentation. In sum, addressing TBTF banks in an effective manner requires a coordinated EU approach. There are different possible paths that one could pursue with respect to bank structural reform. Firstly, structural reforms differ on which activities should be undertaken by separate entities (the location of the fence) and the legal, organisational and economic links allowed between such separate entities (the height of the fence). Secondly, structural reform may also differ in what triggers structural separation or related measures. Targeted banking groups either need to implement the foreseen activity restrictions or subsidiarisation when and if certain necessary and sufficient conditions are met (to be ascertained by the competent authority) or, alternatively, only after an institution-specific assessment and discretionary decision by such authority⁶. An intermediate possibility is that prespecified conditions triggering separation are established, but the bank can advance evidence to the requisite legal standard, supporting the claim that due to specific and idiosyncratic circumstances, structural measures affecting that bank would likely increase financial instability or the costs of separation would disproportionately outweigh the benefits. In that case, the competent authority may decide to waive the requirement to separate activities into distinct legal subsidiaries either temporarily or until conditions materially change (in other words, the presumption that imposing structural measures on the bank in question would lead to the net benefits pursued by the regulation can be rebutted by the bank itself). Issues of policy design and scope are obviously of critical importance in practice. However we do not discuss them here. We also do not discuss the merits of the proposal tabled by the European Commission, or for that matter, any of the many alternative approaches to structural reform pursued by other jurisdictions and extensively debated in the financial economics literature. Our goal is more modest: to explain, intuitively, why imposing restrictions on bank activities can contribute to eliminate the existence of too-big-to-fail banks, and consequently, reduce both the incidence and severity of future systemic crises. Section 2 is the core of the paper and explores the important link between public safety nets in banking and bank structural reform and explains how structural reform can address the TBTF problem. Section 3 discusses the complementarity of structural reform with the existing regulatory reform agenda. Sec Both approaches will have advantages and disadvantages conditional on their design. Whereas the former approach may for example provide more clarity and legal certainty, the latter approach may help in ensuring that the “right” banks are subject to bank structural reform (i. e. reducing so-called Type I and Type II errors).

52

Miguel de la Mano

tion 4 discusses the many reasons behind the broad public support for bank structural reform. Section 5 questions the merits of industry and other interested parties’ concerns against structural reform. Section 6 concludes.

II. Why structural reform can eliminate the too-big-to-fail problem 1. Public safety nets are indispensable to resolve market failures Who should bear the financial burden when a firm is threatened with insolvency? In the case of non-financial companies, the obvious answer is that the firms’ stakeholders—its shareholders and creditors—should absorb the losses, according to well-established and crisis-tested bankruptcy rules. After all, shareholders and creditors willingly financed the firm, were responsible for managing its investments, and enjoyed the upside in the good times. However, unlike for a normal manufacturing business, the assets of a traditional bank are not equipment or machines, but loans. When a bank makes a loan to a home owner or a SME, it is listed as an asset on its balance sheet. So the value of a bank’s assets is determined by the borrowers’ ability to make (re)payments. What makes a bank further unique is how it gets funding on the liability side of its balance sheet. The primary liability of most banks is deposits. Most depositors do not think of their money as an investment in a bank, but that is exactly what it is. It is a “loan” to a bank that can be withdrawn on demand. The bank does not keep a depositor’s money in its vault. It makes loans on the asset side of the balance sheet using depositors’ money. These loans will be paid back over years, even though deposits can be demanded back immediately. The main justification for protecting the banking system is based on the role of deposits in the payment system. Deposits are not just a bank liability; they are the means of settling transactions in the economy. Further, depositors can pull out their money at any time they want. Depositors may all demand their money at the same time when they sense the bank is in trouble – a bank run. Bank runs can lead to even solvent banks going under. For example, even a depositor in a healthy bank will “run” if he believes that other depositors are withdrawing their funds in a panic. Bank runs threaten financial stability. A run forces banks to sell assets in a fire-sale, i. e. at prices below their intrinsic value reflecting cash flows generated and valued in normal times, thereby fuel-

Bank Structural Reform and Too-big-to-fail

53

ling yet more asset losses and reinforcing the required asset sales and ensuing panic. Depositor runs can easily become self-fulfilling and a liquidity shock can degenerate into insolvency in a matter of days or even hours.⁷ Runs can also damage the payment system of a country, which relies on bank deposits: when someone makes a transaction, it is cleared by shifting deposits from one bank to another. Businesses often pay their workers using deposit accounts. If the value of bank deposits is called into question, the entire payment system of a country may break down, crippling its economy. It is this concern, above all, that requires the government to set up a public safety net and justifies regulation and close supervision of the financial sector. The public safety net consists of two elements. First, the central bank should act as a “lender of last resort”, to solvent banks, against good collateral, and at a penalty rate. Second, governments explicitly insure bank deposits up to a certain level through deposit guarantee schemes, also to prevent bank runs and maintain confidence in the payments system. Thus, if a solvent bank faces a run, it can get financing from the central bank to meet the deposit withdrawals. If the bank is insolvent and the value of its assets would be even lower than the value of its deposits, then the regulator can step in and take over the bank, guarantee its deposits and prevent disruptions to the payment system (as the FDIC has done in the US for hundreds of banks in the recent crisis alone). The mere presence of a lender of last resort and deposit insurance can prevent runs from happening in the first place. Thus, a public safety net consistent of these policies is indispensable to ensure financial stability. They avoid self-fulfilling confidence crises and various forms of contagion, prevent wide-scale collapse of the intermediation services of the banking sector, and facilitate the ability of banks to engage in effective maturity transformation that may affect even solvent banks.⁸ However, public safety nets may also incentivise banks to expand their balance sheets and take excessive risks with the funds made available to them

 Diamond and Dybvig (1983) is the seminal article on the liquidity insurance provision role of banks and its policy implications. Adrian and Shin (2010a) elaborate on the fire-sale dynamics and endogenous risk in banking.  Deposit-taking banks are vulnerable to bank runs given their asset maturity transformation role. Their mix of illiquid assets and liquid liabilities (deposits that may be withdrawn at any time) may give rise to self-fulfilling confidence crises and force banks to liquidate illiquid long term assets at a loss even when they are, in reality, solvent (as explained in the seminal paper by Diamond and Dybvig (1983)).

54

Miguel de la Mano

(“moral hazard”).⁹ Safety nets take away disciplining incentives of depositors and/or bank creditors and artificially lower the bank’s cost of capital (funding cost), which allows banks to expand.¹⁰ Hence, in the absence of adequate regulation and supervision, safety nets indirectly allow banks to leverage up more easily than would be possible otherwise. Furthermore, holders of banks’ debt may have the expectation that they would not face the (full) risk of loss in the event of bank failure but would instead be bailed out (“implicit subsidy”). And, this expectation is stronger for larger banks, which are more likely to benefit from the implicit public safety net (as the impact of their uncontrolled failure would be, ceteris paribus, greater). This setting incentivises risk-taking by banks, as upside gains are being privatised, whilst downside losses may end up being socialised and distorts the level playing field across large and small banks (see Annex 4, European Commission (2014b)). There has been significant interest by academics and policymakers on “putting a figure” on the size of the implicit government guarantee to the banking sector. These implicit subsidies represent a transfer of resources to the banking sector from taxpayers and are found to be significant, with subsidies reaching levels of several billion Euros annually that represent a significant share of banks’ profitability. However, the precise estimate of the level of the implicit subsidies is challenging and dependent on the exact methodology used, as well as on the sample period and countries under consideration. The total implicit subsidy has been estimated by the European Commission (2014b) to be in the range of EUR 72– 95 billion in 2011 and EUR 59 – 82 billion in 2012, based on a sample of 112 EU banks. This amounts to 0.5 % to 0.8 % of annual EU GDP and between one third and one half of the banks’ profits.¹¹

 Ultimately, the net effect of safety nets on bank risk taking is theoretically ambiguous and depends on the relative empirical importance of the two channels. Gropp et al. (2010) find that government guarantees are on balance associated with strong moral hazard effects. Dam and Koetter (2012) use pre-crisis German banking data to show that significant increases in expectations of bailouts for banks lead to significant increases in risk-taking by banks.  It is implicitly assumed that an adequate pricing of the deposit insurance is not feasible, given the complexity and fluctuating riskiness of a bank’s activities. Demirgüc-Kunt et al. (2005) find that deposit insurance underpricing seems to be the rule rather than the exception. See Admati and Hellwig (2013) for a good review of why banks chose to become big through increased leverage.  Last year, using a study (Ueda and Di Mauro, 2013) that relied on credit-rating data from 2009, Bloomberg View put the value of the subsidy at $83 billion in a typical year for the 10 largest U.S. banks. See http://www.bloombergview.com/articles/2013 – 02– 20/why-should-taxpayers-give-big-banks-83-billion-a-year-. In March, using various methodologies, the International

Bank Structural Reform and Too-big-to-fail

55

Preventing runs should not be viewed as a bailout. For solvent banks, the money will be paid back with interest. For insolvent banks, if the government is acting appropriately, long-term creditors and shareholders of the bank will be wiped out. But note that, to prevent runs, protect depositors and preserve the payment system, there is absolutely no reason for the government to protect long-term creditors and shareholders of banks. It is possible to completely wipe out shareholders and long-term creditors while preserving the integrity of the payment system. The FDIC has done this many times. With the adoption of the BRRD, EU countries have equipped themselves with a modern resolution regime, with instruments to resolve and restructure insolvent banks, shielding depositors from losses and ensuring the continuity of essential financial services.

2. Public safety nets create moral hazard and force the state to protect banks from insolvency In the wake of the current financial crisis, both the US and multiple EU authorities went far beyond protecting depositors and the payment system. Some banks and other financial institutions were perceived too-big-to-fail, or so important to the financial system that the government can’t let them collapse. Because creditors assume they’ll get rescued at taxpayer expense, such banks can borrow for less than they otherwise would. This implicit subsidy, a sort of free insurance, makes disasters more likely by encouraging reckless behaviour. Indeed, government policies took money from taxpayers and gave it directly to banks’ creditors and shareholders. Why did they do this?¹²

a. Without subsidies (explicit or implicit) banks will not resume lending to the real economy One explanation is that banks’ long term creditors and shareholders must be protected because the only way to recover from the crisis is to resume lending.

Monetary Fund (IMF, 2014) put it at somewhere between $16 billion and $70 billion annually in 2011 and 2012 for eight systemically important U.S. banks (see also footnote 20 below).  Mian and Sufi (2014) argue that “When a financial crisis erupts, lawmakers and regulators must address problems in the banking system. They must work to prevent runs and preserve liquidity”. They also argue and provide extensive evidence in support of the claim that “policy makers have gone much further, behaving as if the preservation of bank creditor and shareholder value is the only policy goal.”

56

Miguel de la Mano

Banks, particularly large banks, the argument goes, have the unique ability, incentive and expertise to lend. To perform this function they require access to cheap funding and liquidity and should obtain a satisfactory return on equity. In sum, bank bailouts (followed by regulatory restraint) are needed to kick start lending to the real economy. If the argument that redistributing funds from innocent bystanders (taxpayers) to those partly responsible for causing the financial crisis (bank shareholders and creditors) seems flawed, that is, because it is. The resilience of this justification for sustaining implicit state support six years after such a devastating financial crisis is a testament to the greater lobbying power of the banking industry relative to the influence and persuasive abilities of academic economists and career regulators. It is true that the social cost of a bank’s bankruptcy is larger than its private cost, for at least, three reasons. First, it affects uninformed depositors who do not have the incentives or the means to assess the risk they face. Second, the bank’s proprietary knowledge of its customers, and especially of small corporations, is a valuable asset that would be lost in a bankruptcy. Finally, the bankruptcy of one bank may generate a negative externality for all other banks through a contagion effect and the ensuing financial panic (deposits would be frozen in bankruptcy and hence are no longer deemed perfectly liquid and accessible). Yet bank executives maximising shareholder value will take into account only the private cost of their bankruptcy, so the market will be characterised by excessive risk taking. As explained above, lender of last resort facilities and deposit guarantees are required to maintain financial stability but this safety net should not extend to shield bank creditors and shareholders from failure. The fundamental business of a bank is lending, just as the fundamental purpose of a law firm is to offer legal advice or that of a bakery to produce and sell bread. All these activities are valuable if not essential in any well functioning economy. But few believe that taxpayers should compensate the partners in a law firm that loses a big case or the owner of a bakery that goes bankrupt after making bad business decisions. So if banks get in the business of producing bad loans, why should the government step in to protect incompetent bank managers and their creditors and shareholders? It should not. “Implicit” state guarantees, essentially statecontingent subsidies, not only create significant moral hazard – the seed of TBTF banks – but they also distort financial markets by giving banks an artificial advantage over capital markets in credit intermediation. It is also not clear, as seen in Japan, that the economy would heal if we could just get the banks to lend again immediately. The ongoing severe recession originates in a systemic banking crisis caused by a massive pullback in household

Bank Structural Reform and Too-big-to-fail

57

and firm spending, after a long period of uncontrolled credit expansion. “This is like trying to cure a hangover with another binge-drinking episode” (Mian and Sufi, 2014). More debt is unlikely to be the key to escape a recession caused by excessive and unsustainable private debt.

b. The time inconsistency problem A more compelling reason for implicit state support beyond the explicit safety net is that when a systemic bank fails the government faces a dismal choice between rescuing the bank and ensuring its business continuity or risking contagion, massive domino-like failures and ultimately financial catastrophe¹³. It is not surprising that – quite irrespective of any assurances that bailouts would not take place – our leaders would prefer to err on the side of increasing moral hazard in exchange for minimizing the risk of financial meltdown. No one doubts policymakers would like to commit to not bailing out any bank to avoid the negative incentive effects ex ante (moral hazard) and the large costs of crises ex post. But it has proven extremely difficult to make such commitment credible. More so, as banks have become in just decades, large, complex, heavily interconnected systemic institutions, in some cases with balance sheets as large as or larger than their domestic economy. In effect, the TBTF problem stems from the lack of credibility of any government’s commitment not to bail-out large systemic banks. This lack of credibility is a manifestation what economists Kydland and Prescott (1977) have labeled the “time inconsistency problem”. Even if policymakers recognize that bail-out policies lead to long run moral hazard and distortions,

 Systemic instability is defined in a variety of ways, but in general arises when financial distress in one financial institution is communicated to other institutions. Such contagious distress may occur when problems in one institution trigger a crisis of customer confidence in other institutions. Alternatively, the failure of one institution to settle its obligations may cause the failure of other, fundamentally sound, institutions. Short-term debt provides valuable discipline inside financial firms, but it can also create systemic problems. Specifically, the need to repay the debt may force banks to dump assets and reduce lending during a financial crisis. And because each bears only a tiny slice of the systemic costs it creates, banks issue more than the socially optimal amount of short-term debt. Moreover, this systemic cost is in addition to concerns one might have about the mismatch between the maturities of a bank’s assets and liabilities. Whether the bank’s assets mature in two years or twenty, the risk that it will be forced to sell illiquid assets in a financial crisis increases with its use of short-term debt. Thus, it is not sufficient to make capital requirements increase in relation to the maturity mismatch between assets and liabilities.

58

Miguel de la Mano

they face a commitment problem with respect to their pledge that they will not intervene in cases of bank crises. When a large bank is at the brink of failure, policymakers will want to renege on their pledge in order to avoid systemic risk. In turn unsecured creditors, anticipating such policymakers’ incentives, will not monitor large banks sufficiently in the run-up to a crisis, allowing the TBTF problem to grow. This time inconsistency problem cannot be avoided simply by eliminating the explicit public safety net, since, as explained above, in the absence of deposit guarantee schemes and lender of last resort facilities, financial stability cannot be preserved. The commitment not to bail-out systemic banks lacks credibility because in certain circumstances, the authorities may have neither the ability nor the incentive to do otherwise. First, the authorities may have no ability (i. e. tools and instruments) to orderly resolve a large systemic bank, avoid contagion and shield depositors. In that case a bailout is virtually inevitable and a commitment not to rescue a failing systemic bank lacks credibility simply because a bail-out is the only course of action available to avert a systemic crisis. The recent crisis exposed the fact that many countries did not have an adequate resolution regime for any bank, systemically important or not, and commitments to coordinate among authorities to resolve cross border banks also proved to be time inconsistent. It follows that introducing adequate resolution regimes is a necessary condition to eliminate the moral hazard problem and hence the implicit subsidies that TBTF banks enjoy. However, a resolution regime alone may not suffice to prevent the emergence or to eliminate the TBTF status in the case of systemic banks. Even countries with decade-old, well established resolution authorities, such as the US, could not prevent the buildup of massive implicit subsidies for their largest banks. The crucial problem is that the ability to resolve banks does not in itself eliminate the time inconsistency problem. In other words, the incentives to completely avoid bail-outs are weak in the event a large, systemic bank is failing or likely to fail (even one subject to orderly resolution) if, even with low probability, such failure risks triggering a systemic banking crisis. Even the best designed resolution regime suffers from mild to severe time inconsistency problems depending on the size, degree of interconnectedness, complexity and cross border activity of the bank that falls under resolution. When a large systemic bank is failing or likely to fail, the incentives of a resolution authority are, above all, to preserve financial stability and protect depositors. If the failing bank is systemically important, by definition, it cannot be wound down without inflicting severe collateral damage upon the financial sys-

Bank Structural Reform and Too-big-to-fail

59

tem. In that case, experience and logic suggests that the failing bank will benefit from lender of last resort facilities irrespective of whether the losses originated in activities that merit or not coverage by the public safety net (e. g. proprietary trading or off-balance sheet investment vehicles). This is the case, in particular, when different activities are comingled together and not operated in stand alone and separately capitalized entities or subsidiaries. In addition, in order to protect depositors, run-prone wholesale creditors may need to be bailed out. Even shareholders of the failing systemic bank are unlikely to be wiped out if this may spook “existing or will be” shareholders of similar or interconnected banks in the path of a contagion storm – yet another manifestation of the time inconsistency dilemma. Bail-in of certain creditors can and should contain (maybe even eliminate), ceteris paribus, the extent of the bail-out. The problem is that the excess risk taken by any bank seeking to remain or become TBTF is endogenous – that is to say, the ceteris paribus assumption does not typically hold. Shareholders of TBTF banks are aware that in the event of excess (tail) risks leading to failure they may be wiped out. However, in the upside, they immediately benefit from excess returns. But in the downside their financial liability is limited (by law), and incentives by senior creditors to impose discipline are subdued by the fact that the larger the losses, the larger the risk of contagion, and the more likely that the authorities will be compelled to rescue the bank’s senior creditors and ensure the continuity of the bank’s core activities. And yet, the same economic mechanism (and logic) that explains why shareholders would allow the managers of their banks to take risks that could wipe them out applies to the holders of bailinable debt. The goalpost triggering time inconsistency is shifted but so is the excess risk the bank is willing to take. Bailinable creditors benefit from higher coupon payments in the good times (paid from the higher returns derived from even greater financial risk) to compensate for a wipe out (bail-in) if the bank′s excess risk taking turns sour. But just like in the case of shareholders, the financial liability of bailinable bondholders is limited, and any excess loses would need to be absorbed, in the event of liquidation, by non-bailinable creditors. ¹⁴ Some of these non-bailinable

 Moral hazard, is a situation in which a party is more likely to take risks because the costs that could result will not be borne by the party taking the risk. But with limited liability, bailing in certain creditors in addition to shareholders, that is, spreading the losses does not in itself eliminate the moral hazard. It will bring more market discipline but excess risk taking would still be rational if the excess returns benefit shareholders and bailinable creditors and the losses are in excess of their limited liability – and hence part of the losses are borne by non-bailinable creditors and/or taxpayers.

60

Miguel de la Mano

creditors may seek to run, triggering panic, and the government will be called in to absorb excess loses and avoid contagion threatening financial stability. Others, such as retail depositors may start to run, even if belatedly (as in the case of Northern Rock). Retail depositors can exert little ex-ante discipline but remain vulnerable, and hence covered by the public safety net, when loses are in excess of a bank′s loss absorption capacity (including bailinable debt). The point is that the size of the contingent loses is not exogenous but depends on the banks′ risk taking activity. Moreover, the bank can seek to increase its leverage and balance sheet and unrestrained, engage in activities that make it more interconnected, complex and thereby systemically important. This will further enhance the risk of contagion in the event of failure, and hence reinforce the time inconsistency affecting the ex-post actions of the resolution authority. By its very nature, systemic importance is endogenous to the actions of the bank, and like moral hazard it thrives on time inconsistency. Orderly resolution authority mitigates the consequences of a failing systemic bank but it does not eliminate its systemic importance, and the associated implicit subsidies. ¹⁵ Implicit subsidies are like a hard drug; the more resistant the body becomes to its influence, the greater the intake necessary to trigger a pleasurable effect. Similarly, the greater the loss absorption capacity the banks are forced to have, the greater the losses (and hence the willingness to take on excess risk) necessary to trigger the time inconsistency and hence to benefit from the implicit state support. The painful Lehman collapse, provides further indirect evidence of how difficult it is to contain the expectation (i. e. resolve the time inconsistency problem) that state support would be extended, in a systemic crisis, to activities that do not (ex-ante) and should not fall under the protection of the explicit safety net. Only a few days after the collapse of Lehman, government emergency measures explicitly extended the public safety net to money market funds and ultimately also to pure investment banks, thereby validating the implicit state support that these entities had enjoyed in the run-up to the crisis. The reason, quite simply, was that these financial institutions (not least Lehman itself) were heavily interconnected with commercial and deposit taking banks, and in order to

 At the same House Financial Services Committee hearing in June 26 2013, two Federal Reserve district bank presidents, Richmond’s Jeffrey Lacker and Richard Fisher of Dallas, said they believe Dodd-Frank didn’t end the perception that some banks are too big to fail. “I don’t think we have prevented taxpayer bailouts by Dodd-Frank,” Fisher told lawmakers. Lacker said the law failed to end the too-big-to-fail perception because it gives a “tremendous amount of discretion” to regulators in deciding how to take down banks, leaving government officials too vulnerable to lobbying and outside influence.

Bank Structural Reform and Too-big-to-fail

61

protect the latter from failure, market players reasoned, the regulator would need to protect the former – a rational expectation indeed. In sum, an orderly resolution regime can mitigate the harm caused by the failure of a large systemic bank. However, there remains a (residual) risk that large, complex and interconnected banks, remain systemic, in the absence of structural measures, and can engage in activities that allow them to retain and further entrench their too-big-to-fail status, in (a) the knowledge that their shareholders and creditors enjoy limited liability and (b) the expectation that, whatever the ex-ante commitment to the contrary, a resolution authority would do “whatever it takes” to avoid a systemic crisis, including, if need be, partial bailouts. This is a view also held by many practitioners and commentators (see annex), including some of those entrusted with running the new orderly resolution mechanism that gives the FDIC more power to impose losses on bondholders when banks do fail. For example, Thomas Hoenig, FDIC’s vice chairman, has expressed doubt that the mechanism could handle multiple large bank failures simultaneously. At a House Financial Services Committee hearing as recently as June 26, 2013, he said that, while his agency can handle the wind-down of one big failing bank at a time “if you have a systematic meltdown, as we did last time, I feel pretty confident that the Congress will be asked for another TARP”.¹⁶

 A recent study published in late July 2014 by the GAO (GAO, 2014) suggests that the largest banks’ funding advantage may have disappeared or even turned negative in 2012 and 2013. Unfortunately, like distant planets, that can only be detected and their size estimated indirectly by observing their gravitational influence on the path of other moving (and observable) celestial objects, implicit subsidies can only be indirectly estimated in the credit upswing, when banks start taking further credit and market risk. As explained in the editorial article in BloombergView on July 31, 2014, “in a period of low perceived credit risk and unusual market calm, you’d expect the subsidy to be low. At such times, as the GAO recognizes, investors aren’t concerned about banks going bust, so they might not give a funding advantage to the ones they think are too big to fail”. Indeed GAO researchers also asked banks what the big banks’ funding advantage would have been in 2013 if credit risk had been as high as in 2008. Bloomberg further reports that: “[GAO]′s models found an advantage ranging from about one to six percentage points. Applied to just the uninsured liabilities of the five largest U.S. banks, that would be worth $67 billion to $400 billion a year. In other words, the GAO study suggests that the subsidy will be there for the big banks if and when the credit cycle turns” (see http://www.bloombergview.com/articles/2014– 07– 31/too-big-to-fail-is-still-too-dangerous)

62

Miguel de la Mano

3. Banks can endogenously extend the public safety net to cover activities that do not deserve to be subsidised but enhance the bank’s systemic importance The systemic importance of a bank is closely related to the size of its balance sheet, the degree and nature of its interconnections across financial markets and with other banks, the absence of effective competition and its organizational complexity and governance structure. But the real concern is that there is no simple way for a (systemically important) bank to continue to provide essential banking functions whilst in insolvency, and in the case of a failure of a large bank, those functions could not simply shut down without significant systemic damage. This implies that, as far as systemically important banks are concerned, the public safety net cannot be restricted or contained to protect a solvent bank from a depositor run or to support the stability of a public good such as the payment system, or to promote the provision of loans to SMEs, or households where it leads to positive spillovers in the economy. The government will step in to rescue a systemically important bank irrespective of the source of its losses. In other words, de facto, all activities of the bank benefit from the implicit government subsidy, including proprietary trading in complex derivatives, the origination and distribution of complex and exotic securitized investments, investment banking services such as market making, etc. This happens even though these activities are not generally underprovided and other rival banks (including specialized banks) or even non-banks can step in or indeed are more efficient in providing such services (absent the safety net distortions). In any case, there is absolutely no justification for the government to subsidize privately profitably speculation, as in the case of proprietary trading. In theory, to reduce moral hazard, the government could commit publicly that the public safety net only protects a bank insofar as its losses do not originate in trading activities that do not merit state support or result from excessive or reckless risk taking. However, in practice, in the case of systemically important banks this claim is also time inconsistent. The source of the losses matters little when considering the costs of intervening against the costs of inaction. The solution is obvious. To avoid this time inconsistency and the moral hazard it fuels, the only way to ensure that the public safety net is contained, is to separate activities that can and should benefit from this safety net from those that do not. By addressing directly the mechanism through which banks implicitly extend the public safety net to cover all its activities, structural measures re-

Bank Structural Reform and Too-big-to-fail

63

strict, at the source, the ability of the banks to become too-big or too-systemically important to fail.¹⁷ Structural measures typically aim at the separation of activities that can be allowed to fail from those that are not easily substitutable and cannot be discontinued and hence should benefit from the public safety net. This makes the bank less systemic and thus enhances the credibility of the government commitment not to bail out the banking group in its entirety irrespective of the source of the losses. Structural reform should make clear which part of the bank, if any, deserves to be bailed out and for what reason. Placing activities that should not benefit from the public safety net (typically trading activities not subject to market failure) in a separately capitalised entity ex-ante, before failure materialises, eliminates the time inconsistency that otherwise undermines the commitment not to bail out these activities. As a consequence, the implicit subsidy for these (ring-fenced) trading entities is reduced (or even eliminated). The question that arises is which activities should benefit from the implicit subsidy. The subsidisation of an activity is justified if it directly addresses a genuine market failure, such as when the activity resolves an information asymmetry (e. g. funding SMEs), avoids negative externalities (e. g. avoiding financial instability), provides a public good (e. g. market liquidity or financial stability due to continuity of financial services) or addresses a market power concern (e. g. due to switching costs or network effects). Therefore, banking activities that are easily scalable and short-term in nature and which do not suffer from the above types of market failure should not be bailed out (and hence should be allocated in a separate legal entity within the group) – see also section 2.3 below. On the other hand, core banking activities that are relationship-oriented and which address a genuine market failure deserve to be protected and covered by a public safety net. In this way, structural reform helps avoiding that (i) desirable core banking activities are underprovided, (ii) scalable and short-term oriented banking activities are overprovided, and (iii) governments make com-

 Higher capital requirements similarly constrain the ability of banks to engage in excessive leverage. However, capital requirements are a blunt instrument compared to structural measures since systemic importance is path dependent and derives from multiple factors such as organizational complexity, interconnectedness, market power in the provision of certain activities, cross-border activities, similarity with other banks, herding behavior, etc, not only excess leverage. As explained in the next section, structural measures and higher capital requirements should be seen as complementary instruments in restricting the ability of banks to become systemically important and hence, ultimately, TBTF.

64

Miguel de la Mano

mitments they cannot credibly maintain in the presence of a systemic crisis (namely, not to bail out banks in their entirety whatever the source of losses).¹⁸

4. Modern banks increasingly engage in scalable and short-term oriented activities backed by the public safety nets The changing nature of financial intermediation and banking in the last decades has significantly increased the concerns about the distortionary effects of the public safety nets. Total asset growth has outpaced EU GDP growth, with total assets exceeding 350 % of EU GDP in 2012, up from 250 % in 2001. This trend contrasts with the decades prior to the 1980s, where assets growth went in tandem with real economic growth.¹⁹ The underlying drivers of the 1980s structural break that triggered banks’ expansion are globalisation, technological innovation (securitisation, IT, etc.), deregulation, and increased competition. Prior to the 1980s, commercial banks could be characterised by a “originate and hold” banking model, which generally refers to a long-term oriented, customer relationship-based banking model, where loans are granted and held until maturity, and where bank funding is mainly derived from insured deposits, rather than tradable wholesale market instruments. The relationship-oriented model encouraged banks to originate loans and to gather information and monitor ultimate borrower performance, as the interests of the bank and its customers were typically aligned (the bank does well if the borrower does well and is able to pay off his loan). As of the 1980s, commercial banks increasingly moved towards a so-called “originate and distribute” or transactions-oriented model (see Acharya et al. (2009), Buiter (2008)). The “originate and distribute” banking model refers to the banking model in which granted loans are pooled, securitized and sold to investors (trading and market-based activities). The shift in model is associated

 National structural reform initiatives in the US (Volcker rule) and the UK (Vickers) have indeed been shown to have significant impact on respective banks, especially for systemic and investment banks. By analysing CDS spreads in an event study analysis Schäfer et al. (2013) find that these reforms seem to have credibly reduced bail-out expectations for systemic banks.  Allessandri and Haldane (2009) document these long term trends for the UK banking sector. They first show that the aggregate UK balance sheet remains roughly stable at 50 % of GDP for the century between 1880 and 1980, after which it started to grow quickly to reach more than 500 % of UK GDP before the crisis struck.

Bank Structural Reform and Too-big-to-fail

65

with an increased reliance on capital markets for funding²⁰. More emphasis is put on non-interest income, as income is derived to a significant extent from fees and trading. Information and principal-agent problems become more important, as the interest of the bank and its clients are no longer necessarily aligned. Banks became part of a long intermediation chain, rather than linking ultimate savers directly to ultimate borrowers (Adrian and Shin (2010b)).²¹. This ability of structured finance to repackage risks and to create “safe” assets from otherwise risky collateral led to a dramatic expansion in the issuance of structured securities, most of which were viewed by investors to be virtually risk-free and certified as such by the rating agencies. At the core of the recent financial market crisis has been the discovery that these securities are actually far riskier than originally advertised. Excessive trading and market-based activity has been an important risk factor in this crisis.²² Market-based activities (trading in, or holding, securitised debt instruments) contributed to the failures of major banks in Europe (amongst others RBS and Fortis) and of both investment and commercial banks in the USA (amongst others Lehman Brothers, Merrill Lynch, Washington Mutual). The majority of the large and complex EU financial institutions that received state support in 2008 and 2009 had above average trading income to total revenue ratios. Chow and Surti (2011) analyse a sample of 46 large and complex EU banking groups. 25 banks had trading income to total revenue ratios that exceeded the average ratio plus one standard deviation. 18 of those 25 “vulnerable” banks were effectively part of the sample of 23 banks that received official support in 2008/2009.

 According to Shin (2012) the “originate and distribute” model facilitates bank expansion and risk-taking.  Rather than simply taking deposits and making mortgages, a long chain of interconnected institutions arises. The mortgages are kept on the asset side of a mortgage pool that issues mortgage backed securtities (MBS). This paper is bought by an issuer of asset backed securities (ABS), who issues tranches of collateralised debt obligations (CDOs) in order to finance it. Investment banks hold some of this profitable ABS paper and finance it through collateralized borrowing (repo). Commercial banks make reverse repos and secure their funding short term by issuing commercial paper (CP). Money market mutual funds buy the CP and issue shares to the households that have excess savings. Note that the intermediation chain can be much larger, as ABS can be repoed multiple times, for example. And investment or commercial banks can set up conduits and SIVs in order to finance the direct holding of CDOs and other ABSs.  Trading and lending are not entirely disconnected. The traditional originate-and-hold or relationship oriented model of banking has been shifting towards a originate-and-sell or transaction-oriented model of banking. Loans, previously illiquid, have been made more liquid through securitisation.

66

Miguel de la Mano

Structural reform would contribute to ensure that large banking groups do not scale and leverage up in an overly aggressive, complex and procyclical manner and in activities that do not merit being protected through the (implicit at least) public safety nets, as they have done in the past 30 years. Avoiding such an expansion would also make the banks more likely to be resolvable and thus address directly the TBTF problem and the corresponding distortive implicit subsidies. ²³ Two characteristics of certain non-core banking activities can help explain the rapid expansion of the bank balance sheet size of TBTF banks and give rise to adverse market failures that should be addressed not by subsidies but instead taxation: “scalability” and “short-termism”. Certain non-core banking activities, such as proprietary trading, market making, securitisation, and derivatives transactions are easily “scalable” as they do not (always) require costly ex ante information gathering, ex post monitoring and long term relationship building with clients, but can be expanded swiftly as long as wholesale funding markets are deep and liquid. “Short-term orientation” refers to the fact that large banking groups can and do build up large inventories and investment portfolios of securities (in the broad sense, i. e. including derivatives) to reap short-term benefits. This opportunistic behaviour is very different from banks’ traditional hold-to-maturity loan book activity, which is more long-term and relationshiporiented. The short-termism of modern banking creates a procyclical “financial accelerator”, because banks will need to sell some of the securities to deleverage if the securities drop in value, which may amplify the initial shock and give rise to further sales (i. e. it also gives rise to dynamic systemic effects). The “scalability” and “short-term orientation” of certain modern banking group trading activities (i) give rise to complexity and interconnectedness impeding the resolvability of large EU banking groups, (ii) give rise to conflicts of interests and flawed bank standards, and (iii) push core credit institutions away from a relationship-oriented towards a transactions-oriented model of banking. Bank structural reform should hence aim to separate “easily-scalable” trading activities from those core banking activities that are “less easily-scalable” as well as separate the balance sheet part of the bank that is more “oriented to-

 Structural bank reform objectives have been listed in the Commission proposal as follows: Reduce excessive risk-taking within credit institutions; Remove material conflicts of interest between the different parts of the credit institutions; Avoid misallocation of resources and encourage lending to the real economy; Contribute to undistorted conditions of competition for all credit institutions within the internal market; Reduce interconnectedness within the financial sector; Facilitate efficient management, monitoring and supervision of the credit institutions; Facilitate the orderly resolution and recovery of the group.

Bank Structural Reform and Too-big-to-fail

67

wards short-term profit taking” from the part that is “oriented towards making loans and keeping assets until maturity”. This logic and focus is aimed at reversing undesirable trends at the systemwide level such as rapidly expanding aggregate balance sheets and banking system leverage following globalisation, deregulation and financial innovation. The underlying systemic concern justifies a focus on scalable and short-term oriented trading activities, such as market making, securitisation, and derivatives trading. Scalability and short-termism hence both give rise to negative externalities (in particular systemic risk in distress) and push core credit institutions away from their core role and expertise. Addressing the concern about the scalability and short-term orientation of bank balance sheets implies that scalable and short-term oriented activities should not be artificially and implicitly promoted by enjoying the coverage of the (explicit or implicit) public safety nets. Putting these activities at a distance from the public safety net would give rise to more sustainable balance sheet growth and leverage.

III. Complementarity of structural reform with the existing reform agenda To address TBTF concerns, the European Union and its Member States have engaged in a fundamental overhaul of bank regulation and supervision (see European Commission (2014a)). In a recent report (ESRB (2014)), the Advisory Scientific Committee of the ESRB concludes that the reform agenda is necessary, but insufficient, and that more needs to be done to address the EU’s documented overbanking problem and the underlying causes. One of their recommendations is to implement aggressive bank structural reform, so as to reduce both the size of the largest banks and their risk-taking in securities markets. We will argue in this section that structural reform is not a substitute but a necessary complement to the other key EU financial reforms. We discuss below the complementarity of structural reform with the enhanced capital and liquidity requirements framework (CRD IV/CRR), the bank recovery and resolution framework (BRRD), and Banking Union in the EU.

68

Miguel de la Mano

1. Structural reform is complementary to the enhanced EU capital and liquidity requirements framework (CRD IV/CRR) A landmark in the financial regulatory reform has been the adoption of the CRD IV package in July 2013. This package transposes, via a Regulation and a Directive, into EU law the new global standards on bank capital (Basel III agreement). The new rules notably involve tougher capital requirements (in terms of quantity and quality) and introduce liquidity and leverage requirements. They also set rules on prudential supervision and corporate governance (amongst others). There are several reasons structural reform can complement this prudential framework. In effect, structural reform measures go beyond the fine-tuning of Basel type regulations. Schäfer et al (2013) argue that the announcement of the Volcker Rule in the US, had a significant effect since it signalled that national governments would employ additional instruments than Basel regulations to tackle the TBTF problem. The CRD IV package strengthens minimum capital (and liquidity) requirements. Capital requirements are increased, among others, for trading activities and for banks that are deemed systemically important. The rationale is that more capital makes a bank safer, since it increases the cushion of losses it can absorb before going bankrupt. In theory, if the loss absorbency of banks were so great as to cushion the loss that a bank may have to face this would remove all substantial divergence between the private interests of banks and the public interest. As a result, there would no longer be a need for public policy to regulate the structure of banks. However, a first difficulty is to estimate how much capital would be enough to absorb losses in the next crisis. Both Bear Stearns and Lehman were compliant with the Basel requirements. Five days before its bankruptcy, Lehman had a Tier 1 capital ratio of 11 %, but yet was overwhelmed by market fears about its solvency and viability. A month after the Lehman bankruptcy, the settlement auction for Lehman credit default swaps valued Lehman senior unsecured debt at only 9 cents on the dollar, meaning that liquidating the firm’s assets was only expected to yield 9 % of the money needed to repay unsecured creditors. How much more capital would have been needed to prevent panic and avoid creditors losses, and how much capital will our banks need next time? The leverage of Basel-compliant banks still far exceeds the leverage in most other sectors of the economy, and there seems to be no realistic prospect of sufficient capital absorbency (Miles et al. (2013), Admati and Hellwig (2013)). A second problem has to do with the gaming of risk-weighted assets (RWA). Banks are allowed to use internal risk-based models to calibrate their capital requirements. By this means, for example Barclays’s total assets of GBP 1 596 bil-

Bank Structural Reform and Too-big-to-fail

69

lion are reduced to GBP 398 billion on a risk-weighted basis and Deutsche Bank’s total assets of EUR 1 910 billion are reduced to EUR 314 billion (at the end of March 2013). Ratios on the basis of risk weighted assets are not always meaningful without detailed knowledge of how the risk weighting is achieved. During the crisis it has become clear that the size of a bank’s non-risk weighted capital buffer (known as the leverage ratio) is an important health parameter. The RWA figure may give some comfort day-to-day, but in times of stress it will not count for much (Haldane and Madouros (2012)). Furthermore, addressing TBTF problems by means of higher capital requirements would not address the fundamental inconsistency of discouraging a bank’s systemic importance and trading-book related activities through higher capital requirements while at the same time artificially promoting a bank’s systemic importance and trading activity by allowing scalable and short-term oriented banking activities to be performed by entities that enjoy explicit coverage of public safety nets. Structural reform is needed to ensure that systemic importance and excessive trading activity are being curtailed, rather than implicitly promoted, fully in line with the CRD IV package. Indeed, even if banks were structured transparently in a deposit entity and a separate trading entity, capital requirements would still be necessary because each of a bank’s constituting entities would still face incentives to leverage up and increase their expected return on equity. In any event, structural reform can complement the CRD IV package with respect to supervision and market transparency. Pillar 2 of the Basel framework is about the supervisory review and Pillar 3 of the Basel framework is about enhancing market transparency and disclosure requirements. Structural reform is expected to facilitate bank management, but is also expected to help simplifying supervision and enforcement of capital requirements. Structural reform ensures that creditors can exert greater market discipline through increased transparency and reduced complexity and is therefore consistent with the objectives of pillars 2 and 3 of the Basel framework. This is important given the significant expansion of rulemaking complexity in capital requirements regulation. Zingales (2012) states that “[t]he simpler a rule is, the fewer provisions there are and the less it costs to enforce them. The simpler it is, the easier it is for voters to understand and voice their opinions accordingly. Finally, the simpler it is, the more difficult it is for someone with vested interests to get away with distorting some obscure facet.” The legislative responses of the two largest financial crises of the past century, the Great Depression and today’s crisis (sometimes referred to as the Great Recession), differ markedly. The Great Depression spawned the Glass-Steagall Act (1933) that ran to a mere 37 pages. The current crisis has triggered the Dodd-Frank Act (2010)

70

Miguel de la Mano

and runs to 848 pages (or more than 20 Glass-Steagalls). Once implemented and completed, Dodd-Frank might run to 30 000 pages of rulemaking. The EU rulemaking response, aggregating primary and secondary legislation, is likely to exceed even this number.

2. Structural reform is complementary to enhanced bank resolution The BRRD and structural reform are also natural complements. Implementation of the BRRD will pave the way for the orderly resolution of the bulk of EU banks that are not systemically important and will thus significantly reduce the impact of failure of such banks on public finances. One may then argue that if creditors anticipate that they would be fully bailed-in in case of bank failure then they would request the rate of return that reflects the underlying risk of the bank activities. As a result there would be no implicit subsidy and no moral hazard. However, as explained in section 2, the extension of the public safety net to implicitly support a wider range of activities including short-term oriented and scalable trading activities, aggravates the associated moral hazard. The problem is that tail risk within financial systems is not determined by God but by man: it is not exogenous but endogenous. Finance theory tells us that risk is linked to return. So there are natural incentives for banks to generate tail risk to the point that losses are potentially large enough to affect depositors and trigger systemic (system-wide) instability. In other words banks have incentives to seek the higher returns that come from assuming tail risk. They do so safe in the knowledge that the state will assume some of this risk if it materialises, through the public safety net. Scalability of trading activities and the focus on activities that generate a short-run income stream allows this endogenous excess risk taking. This dynamic means it is hazardous to believe that in the case a systemic bank fails, an ex-ante commitment to wipe out shareholders and bail-in long-term creditors (who then become the new shareholders) is significantly more time consistent, than a similar commitment not to bail out shareholders absent resolution authority. Once there are doubts on the credibility of the commitment of the government not to intervene, this would be reflected in the rate of return requested by market participants, and thereby the very activities that make resolution difficult would be encouraged. This perverse cycle is sometimes labelled the “doom loop”. This can arise in several situations: First, the resolution powers will be challenging to exercise for TBTF banks, given their particularly large, complex and integrated balance sheets and corpo-

Bank Structural Reform and Too-big-to-fail

71

rate structures. At the same time, the implementation of the BRRD is expected to pave the way for the orderly resolution of small and medium-sized EU banks and thus significantly reduce the impact of failure of such banks on public finances. The FDIC crisis-tested experience of resolving 400+ small and medium-sized banks in the US suggests that resolution powers can be a powerful and effective tool. As a result, while the potential for eventual public support is certainly reduced, it may still not be eradicated if the powers cannot be fully applied in all instances. The impact of a failure of a large and complex bank may still be significant. Structural reform will increase the options available to authorities when dealing with failing banking groups. By increasing orderly resolution credibility, it will also improve market discipline and bank balance sheet dynamics ex ante. Second, bail-in of creditors may give rise to contagion, unless the intra-financial sector exposures are clearly mapped and controlled, as is intended through structural reform (depending on the strength of the rules of separation). When interconnectedness risks triggering contagion, resolution and creditor bail-in mechanisms may lose credibility. In circumstances of very extraordinary systemic stress, even the BRRD framework allows authorities to provide extraordinary public support, instead of imposing losses in full on private creditors, to solvent banks in the form of a guarantee or precautionary recapitalisation, subject to specific qualifications and to remedy a serious disturbance in the economy and preserve financial stability (see European Commission (2014a) for more details). Such support will also have to comply with the Union State aid framework. Structural reform should help reducing the probability of facing situations of very extraordinary systemic stress. Third, national resolution powers are also far from a magic bullet, especially in the global world of modern finance. When a bank with assets in different countries fails, it is in each country’s immediate interest to have the strictest rules on freezing assets to pay off domestic creditors. As explained in section 2.2, there are several studies on the quantification of implicit subsidies. Overall, these studies point out that implicit subsidies are present, sizeable and can represent a significant advantage for larger banks. More recent evidence suggests that TBTF perceptions may have been dented from the peak of the financial crisis, but are by far not eliminated. The IMF has estimated that the implicit subsidies enjoyed by TBTF banks are still large (IMF (2014)), despite regulatory reforms. They also find that subsidy estimates remain much higher in the euro area than in the United States, likely reflecting the different speed of banks’ balance sheet repair, as well as perceived differences in policy frameworks for dealing with the TITF issue.

72

Miguel de la Mano

Structural reform and BRRD are no substitutes. Even if banks were structured transparently in a deposit and trading entity, special powers would still be required to resolve large banking groups in an orderly manner. Banks cannot enter normal bankruptcy and suddenly stop performing their special and critical role in the payment system, nor can they freeze their deposits, because of the financial panic that would result and because their business model and raison d’être is to provide liquidity to their depositors. Structured universal banks will provide authorities with more options to orderly resolve large banking groups.

3. Structural reform is complementary to Banking Union Inter alia, Banking Union creates a Single Supervisory Mechanism (SSM) which effectively shifts bank supervision from the national to the supranational level for euro area Member States and those Member States that choose to opt in. Over time, the SSM will imply that bank failures will be dealt with at the same supranational level. Member States, however, will be reluctant to mutualise risks originating in the banking sector, in particular given the impact of the past financial crisis on national public finances (for example Irish debt was below 25 % of GDP before the crisis but exceeds 100 % of GDP after having bailed out the banks). Targeting the safety net through structural reform to those core banking activities that deserve protection because they address a genuine market failure reduces the scope of the public safety net and may be a catalyst for the willingness of Member States to push ahead with and support a deeper banking Union over time. In this spirit, structural reform could be part of the necessary preconditions for the introduction of a European deposit insurance scheme. Banking Union is meant to reduce the negative feedback loop between sovereigns and their banks. The nexus between banks and sovereigns originates from the fact that banks hold government bond paper on their asset side and benefit from explicit and implicit support on their liability side. Banking Union has the potential to break these undesirable links. However, by doing so, this leads to mutualisation of risk at the EU level, which may further increase implicit subsidies and their corresponding problems of moral hazard, aggressive balance sheet expansion, and competition distortions between TBTF banks and other banks. In this respect, structural reform can complement the Banking Union by addressing the implicit subsidies of TBTF banks.

Bank Structural Reform and Too-big-to-fail

73

IV. Bank structural reform enjoys broad public support There is strong and widespread support for structural reform from international organisations, financial experts, economists, and the public at large. The annex summarises the views of different stakeholders and lists several statements from proponents of structural reform. The strong support for structural reform can be explained by at least three reasons.

1. Structural reform is appealing to politicians across the political spectrum and appeals to people’s common sense Structural reform is aimed at reducing implicit subsidies. As such, structural reform should appeal to ‘right wing’ politicians who seek to reduce subsidies because of their alleged distortive nature. The presence of a safety net triggers moral hazard and artificially promotes risk-taking. It is intellectually inconsistent to worry about regulation to control bank risk-taking whilst bringing all banking activities under the cover of the public safety nets. However, structural reform is also aimed to redirect the remaining implicit subsidies to those activities that are characterised by genuine market failures. The latter objective aligns well with the alleged intentions of ‘left wing’ politicians. Most, if not all, banking activities are valuable and useful to the real economy. However, just like most activities in a market economy, many should not be promoted or subsidised by the taxpayer unless there is an appropriate justification.

2. Structural reform has been introduced in the past The introduction of bank structural measures is not something new. To curtail the excessive risk-taking and expansion of banks that may result from the existence of the public safety nets, banking activities have always been heavily regulated and supervised. In fact, when the US introduced the very first set of public safety nets, it paired it with a battery of regulation including (i) the prohibition of deposit-taking banks to underwrite or deal in securities, (ii) the limitation of access to deposit insurance and lender of last resort facilities to commercial banks, and (iii) the introduction of a saving deposit rate ceiling to avoid destabilising competition amongst banks (1933 Banking Act which is better known as the

74

Miguel de la Mano

Glass-Steagall Act). In the wake of the Great Depression, several EU Member States, amongst others Belgium, France, and Italy, introduced structural separation rules similar to the Glass-Steagall Act.²⁴ Although their impact cannot be estimated accurately, to our knowledge, they did not stop decades of historically high economic growth in the 1950s and 1960s. Only more recently were risk-based capital requirements (capital adequacy regulation) introduced.²⁵

3. Structural reform addresses the overbanking of Europe and the crowding out of EU capital markets The EU financial system is bank-intermediated, rather than market-intermediated. One driver of this bank dominance of the financial intermediation in Europe is likely to be the implicit subsidies that are “channelled” to the largest banks, but not to other non TBTF capital market players. One way of promoting the development of capital markets in Europe may be to gradually reduce the distortive implicit subsidies towards European banks. For example, the hedge fund industry experience provides valuable lessons. Despite several problems and the need for reform in the sector, the hedge fund

 The first structural rules introduced in Belgium date back to 1934 and 1935 (Royal Decree n°2 of August 22, 1934; Royal Decree n°180 of July 9, 1935). “Mixed” banks were required to separate their deposit taking activities from their investment banking activities. Banks were prohibited from holding shares of industrial and commercial companies. Bank managers were prohibited from holding concurrent executive functions in other companies (National Bank of Belgium (2012), Box 2). In France, the 1941 Banking Law established an impenetrable barrier between deposit banks and investment banks. Restrictions were imposed on bank lending and on the lines of business different types of banks were permitted to transact. In 1945, the Banque de France and the four leading deposit taking commercial banks were nationalised. The investment banks avoided nationalisation. The 1984 Banking Act recognised the principle of universal banking and eliminated many of the applicable restrictions, which allowed banks to significantly expand in balance sheet size relative to GDP. The separation principle in Italy is regulated in Article 19 of the Legislative Decree no. 385 of September 1, 1993 (Single Banking Act).Subsequent reforms removed restrictions on mixing bank and securities activities. In several Member States structural rules still apply, but often limited to specific activities such as housing finance and mortgage banks (Bausparen, covered bond issuance, etc.). Moreover, under current EU legislation, banking and insurance activities are being prohibited from being supported by the same pool of capital in all EU Member States.  In 1988 a first-ever, landmark, genuinely international prudential regulatory agreement “International Convergence of Capital Measurement and Capital Standard” was reached. Basel 1 has been amended and revised in 1996, 2004 (Basel 2), 2009 (Basel 2.5), and 2010 (Basel 3).

Bank Structural Reform and Too-big-to-fail

75

industry did not trigger a systemic crisis.²⁶ It is hence relevant to compare the market structure and business models of the hedge fund industry with those of the banking industry. The structure of the industry as well as its evolution differs markedly from the banking sector. Contrary to banking, the hedge fund sector does not comprise a small number of large players, but rather a large number of relatively small players, entry and exit rates from the hedge fund industry are both high; and the business models of hedge funds are typically specialised rather than diversified. Also, the structure of the hedge fund sector emerged in the absence of state regulation and state support, and the majority of hedge funds operate as partnerships with unlimited liability and most hedge funds today operate with leverage less than a tenth that of the largest global banks.

V. Responding to some of the concerns raised against structural reform When analysing the impact of structural reform, it is important to distinguish “private” (i. e. stakeholder-specific) benefits and costs from “social” benefits and costs (i. e. benefits and costs for society as a whole). Private costs and benefits typically receive more attention, focussing on the impact on banks, i. e. bank shareholders, bank creditors, bank employees, or the EU banking sector competitiveness. Social benefits and costs relate to total or aggregate welfare more generally by incorporating the impact on all stakeholders in society, including notably bank customers (e. g. depositors, borrowers and consumers of financial services) and taxpayers (i. e. the public finances of governments). Whereas private costs are concrete and visible in the short run (e. g. funding cost  Hedge funds amplified the boom and bust and had inadequate liquidity and capital (i. e. shock absorbers). In addition to adverse market conditions, many hedge fund managers were faced with increased redemption demands from investors and with tighter lending conditions from banks. Leveraged funds were forced to unwind positions. Faced with such pressures, hedge funds were often forced to sell assets into declining markets, thereby realising losses and adding further pressure on declining asset prices. This pro-cyclical behaviour may have undermined financial stability and contributed to a deepening of the crisis. Excessive reliance on counterparties and trend-following at the expense of sound risk management and due diligence were observed. In April 2009, the Commission put in place a comprehensive and effective regulatory and supervisory framework for managers of alternative investment funds in the EU (Alternative Investment Fund Managers Directive or AIFMD). Concretely, the AIFMD makes all alternative investment fund managers subject to appropriate authorisation and registration requirements, allows monitoring of macro and microprudential risks, and introduces several investor protection tools.

76

Miguel de la Mano

increases) and being felt by a limited number of banks, social benefits typically only materialise over the medium term, are less easy to pin down reliably, and are dispersed over a much greater number of stakeholder (e. g. taxpayers). Structural reform is for example expected to result in increased private funding costs for the trading entity which is being separated from the deposit entity and hence no longer benefits to the same extent from being implicitly linked to the public safety net. Those funding costs are recurrent and private costs. However, they do not necessarily result in any social cost, as the increased funding cost for trading entity merely reflects the shift of bank risk and contingent liability away from the taxpayer and towards the investors in unsecured debt bank who should, in principle, bear the risk and be properly remunerated (through higher returns) for being exposed to it. Put differently, the increased private funding cost for the bank’s trading entity reflects and is being offset by a decreased (implicit) public subsidy to the benefit of taxpayers. In sum, the increase in the private funding costs due to the removal of the implicit public subsidy should not be considered as a social cost. Notwithstanding the above, there are a number of legitimate concerns raised against structural reform that merit to being addressed and discussed.

1. Will structural reform give rise to important foregone societal benefits? Opponents of structural reform argue that big systemic banks provide benefits to the economy that cannot be provided by smaller banks. A common argument is that large corporations require financial services that only large banks can provide. Related to this is the idea that the global competitiveness of EU firms requires that European banks be at least as big as anyone else’s banks, and in particular as large as US banks. Another argument is that large financial institutions enjoy economies of scale and scope that make them more efficient, helping the economy as a whole. Finally, supporters argue that large, global banks are necessary to provide liquidity to far-flung capital markets, making them more efficient and benefiting companies that raise money in those markets. These arguments suffer from a shortage of empirical evidence. Large multinational corporations have large, global financing needs, but there are currently no banks that can supply those needs alone. Instead, corporations rely on syndicates of banks for major offerings of equity or debt. And even if there were a bank large enough to meet all of a large corporation’s financial needs, it would defy business logic for that corporation to restrict itself to a single source of financial services, instead of selecting banks based on their expertise in particular

Bank Structural Reform and Too-big-to-fail

77

markets or geographies. In addition, EU firms already benefit from competition between EU and foreign banks, which can provide identical financial products. There is no reason to believe that the global competitiveness of our nonfinancial sector depends on our having the world’s largest banks. There is little evidence that large banks gain economies of scale above a very low size threshold. A review of multiple empirical studies found that economies of scale vanish at some point below $10 billion in assets (European Commission (2014b) reviews the literature). The 2007 Geneva Report “International Financial Stability” also found that the unprecedented consolidation in the financial sector over the previous decade had led to no significant efficiency gains, no economies of scale beyond a low threshold, and no evident economies of scope. Finance professor Edward Kane has pointed out that since large banks exhibit constant returns to scale (i. e. they are no more or less efficient as they grow larger), and we know that large banks enjoy a subsidy due to being too big to fail, “offsetting diseconomies must exist in the operation of large institutions”. In other words, without the TBTF subsidy, large banks would actually be less efficient than midsize banks. There is an element of truth to the argument that large banks are necessary in certain types of trading businesses such as customized (over-the-counter) derivatives, where a corporate client may want a hedge that spans multiple markets (currencies, interest rates, and jet fuel, for example). To manufacture such a hedge cheaply, a derivatives dealer has to have significant trading volume in each of the underlying markets, which implies some minimum efficient scale. However, this alone cannot explain the enormous growth in leading investment banks over the last ten years. Goldman Sachs, for example, grew from $178 billion in assets in 1997 to over $1.1 trillion in 2007, while Morgan Stanley grew from $302 billion to over $1.0 trillion. Recent academic literature finds a negative impact of private credit creation on long run growth when it exceeds a certain threshold (see ESRB (2014)). One of the reasons for this link is that bloated banking systems marginally divert both financial and human capital away from more productive projects. ESRB (2014) also find that bank based systems (i. e. systems that try to exploit economies of scope) tend to record lower long-run growth, ceteris paribus. Bank credit supply is found to be more volatile than supply from debt capital markets.

78

Miguel de la Mano

2. What is the impact of structural measures on market liquidity? The discussion of whether market making should be amongst the activities to be separated has been one of the main issues on defining the boundaries of the separation in all recent structural reform proposals. The opponents of such inclusion argue that separating market making, into a separate subsidiary within a banking group will harm market liquidity and hence will be socially costly. The term “liquidity” is left undefined, but it typically is implicitly linked to the level of bid-ask spreads. The fear is that bid-ask spreads may increase, increasing the costs to trade at any scale. Likewise, the set of options to investors may be reduced, as they may no longer be able trade as much or as easily as before. Price discovery may be made more difficult. And price volatility may increase if professional position takers can no longer spot price divergences from rational levels and correct them through trading. There are a number of comments and counter-arguments to be made in this respect. First, the liquidity concern neglects the fact that structural separation aims to reduce the implicit subsidies that distort the proper market functioning and level of bank activity. Indeed, market prices are distorted when contaminated with implicit public subsidies and may in fact produce excess liquidity. The liquidity concern is built on the presumption that more liquidity is always and inherently positive, which is not the case²⁷. Richardson (2013) notes that the issue of liquidity is more relevant in times of crisis than in normal times when liquidity is typically not a pressing concern. How did banks perform in generating liquidity in the crisis? Banks have not performed a significant liquidity role during crisis period and central banks have stepped in to assume the role of Market Maker of Last Resort (in covered bond markets, government bond markets, etc.) next to their Lender of Last Resort role. If anything, the procyclical behaviour and excessive liquidity provision of the banks in the run-up to the crisis has sown the

 For example, benefits of market liquidity should become smaller with the degree of market liquidity. The additional benefits of the extra liquidity derived from high-frequency trading must be of negligible (or negative) value compared to the benefits from having a market which is reasonably liquid on a day-by-day basis. Moreover, ever greater market liquidity may give rise to destabilising momentum effects, such as cycles of undervaluation and overvaluation. In addition, voluntary market making may not occur when it is most needed, i. e. during troubled market conditions. Even dedicated market makers are typically only allowed to post quotes during 90 % of the trading period and of course they may decide to breach their contractual obligations if they deem that fulfilling them would threaten their solvency.

Bank Structural Reform and Too-big-to-fail

79

seeds of the crisis in which they want to deleverage excessively (Bruno and Shin 2014). Second, US experience has shown that concerns may be exaggerated. The US has decades of ongoing experience with subsidiarisation of investment banking activities, as deposit taking affiliates within a Bank Holding Company are not allowed to do anything other than core banking activities. There is however no evidence that suggests that US bond markets are less liquid than European ones. Also, medium-sized competitors or new entrants that are not subject to mandatory separation may gain market share from large banking groups if artificial competition distortions in favour of TBTF banking groups are being reduced. Hence, whereas some banking groups may face increased costs and may no longer serve certain customers, those activities may be picked up by smaller competitors that do not face structural separation requirements. Customers are accordingly not likely to be left unserved. Zingales (2012) argues that markets need a large number of independent traders, to function properly. A separation between deposit entities and trading entities deprives the latter of access to cheap and insured deposits, forcing them to limit their size and the size of their bets. Zingales argues that these limitations increase the number of market participants, making markets more liquid.²⁸ In the era when Glass-Steagall was in place, the US economy has on average been thriving, suggesting that the most extreme structural separation can go hand in hand with robust economic growth. Third, the increased funding cost for the trading entity that acts as market maker is unlikely to be passed on to the real economy and therefore harm economic growth. Households and SMEs that are clients of a banking group that needs to separate certain capital market activities are typically and mainly clients of the deposit entity. Hence, the increased funding cost for the entity not taking deposits would not necessarily affect borrowing conditions for house-

 Zingales (2012): “The third reason why I came to support Glass-Steagall was because I realised it was not simply a coincidence that we witnessed a prospering of securities markets and the blossoming of new ones (options and futures markets) while Glass-Steagall was in place, but since its repeal have seen a demise of public equity markets and an explosion of opaque over-the-counter ones. To function properly markets need a large number of independent traders. The separation between commercial and investment banking deprived investment banks of access to cheap funds (in the form of deposits), forcing them to limit their size and the size of their bets. These limitations increased the number of market participants, making markets more liquid. With the repeal of Glass-Steagall, investment banks exploded in size and so did their market power. As a result, the new financial instruments (such as credit default swaps) developed in an opaque over-the-counter market populated by a few powerful dealers, rather than in a well regulated and transparent public market.”

80

Miguel de la Mano

holds and SMEs. Market making entails significant risk.²⁹ These risks are important, given the size and importance of market making as a share of large banks’ trading activities. Separating market making from the deposit entity will reduce excessive risk-taking and artificial balance sheet expansion and hence may lower the funding cost for the deposit entity. To conclude, market making appears to be a financially viable activity on its own, as illustrated by the fact that several important market makers are not taking any deposits. Under a subsidiarisation model, market making is not prohibited within a banking group. It needs to be performed by a legally separate trading entity within the same group. The increased funding cost for the trading entity (that would perform market making) is part of the desired effects of the separation and reflects the underlying risks of such activities.

3. Would structural reform have avoided bank failures? Critics of structural reform often argue that structural reform would not have helped preventing specific bank failures. For example, it is often argued that Lehman Brothers did not take deposits and hence would not have been affected by structural reform or that several banking crisis were linked not to trading activities but rather to real estate bubbles. Depending on its design, structural reform has the potential to reduce systemic risk and TBTF banking. Some reform options go further than others, for example in limiting a deposit entity’s exposure towards leveraged financial institutions. By legally separating and discouraging excessive trading activity, the complexity and interconnectedness within and across banking groups is being reduced, facilitating the resolvability of trading entities and banking groups. Structural reform could have helped in several bank failures seen in the current crisis. The aggressive growth and mortgage lending practices of Northern Rock and Cajas have only been made possible by relying on wholesale funding

 When facilitating client business a bank is likely to try and hedge most of its risks. Hence, genuine market making should entail limited market risk. However, the actual exposure to risk may vary across time depending on the liquidity of the instruments, on changes in market volatility and on significant variation in the sizes of positions that market making clients may wish to acquire or liquidate. Moreover, there may be a mismatch between the position and the hedge (basis risk) and the hedge will need to be rebalanced over time as market moves alter risk profiles. Furthermore, market makers are still exposed to high counterparty risk and the concrete functioning of market making can vary in relation to different financial instruments and market models.

Bank Structural Reform and Too-big-to-fail

81

and securitisation practices, which in turn reflected the expansion of the largest European banks (Bruno and Shin (2014)). To the extent that the growth of the latter is reined in and not artificially promoted by the inappropriate linkage to the public safety net, real estate bubbles may not have arisen to the same extent. The whole idea of structural reform is to refocus the deposit entities of large banks towards a sustainable relationship-oriented model of banking and away from a transaction-oriented fee-based and short term oriented business model. The separation into distinct legal entities will re-focus banks’ attention to serving their clients’ needs. The real estate crisis, unlike most before it, imperilled sovereigns because states were obliged to bail out not only deposit entities that perform traditional core banking activities such as deposit taking and loan making, but also trading entities that do not deserve to be covered and artificially promoted and protected by a public safety net. More generally, excessive trading and market-based activity has been an important risk factor in this crisis. Market-based activities (trading in and holding securitised debt instruments) contributed to the failures of major banks in Europe (amongst others RBS and Fortis) and of both investment and commercial banks in the USA (amongst others Lehman Brothers, Merrill Lynch, Washington Mutual). The majority of the large and complex EU financial institutions that received state support in 2008 and 2009 had above average trading income to total revenue ratios. Chow and Surti (2011) analyse a sample of 46 large and complex EU banking groups. 25 banks had trading income to total revenue ratios that exceeded the average ratio by one standard deviation. 18 of those 25 “vulnerable” banks were effectively part of the sample of 23 banks that received official support in 2008/2009. Boot and Ratnovski (2012) argue that the deepening of financial markets in the last 10 to 15 years has fundamentally destabilised banks by introducing a trading culture in large, complex and interconnected banking groups. Specifically, such banks face incentives to use their franchise value and undrawn credit lines to trade on an excessive scale to make short term profits. More analysis is needed to confirm or invalidate such claims. European banks have been equally important as US commercial banks in channelling US savings to US borrowers (Shin (2012)). They did so by tapping USD MMFs, holding and investing in enormous volumes of US securitised assets, and sponsoring USD ABCPs (Maes (2014)). European banks are far more important as sponsors for USD ABCP entities than US banks have been (they sponsor twice the volume of ABCP conduits as US banks do). One wonders whether European banks have a genuine comparative advantage to intermediate US funds from savers to borrowers or whether European banks have expanded aggressively on the back of inadequate public safety net backing, in combination with inadequate regulation and supervision.

82

Miguel de la Mano

In addition, structural reform would also have helped to avoid failures of investment banks, such as Lehman Brothers. If Lehman Brothers would have been less connected to deposit taking banks, the scope for aggressive growth pre-crisis and contagion upon failure would have been reduced. Moreover, if the connectedness towards deposit entities is sufficiently curtailed, the failure of Lehman Brothers would not have endangered the deposit entities of EU universal banks, which leaves scope for alternative solutions ex post and which would inject more market discipline ex ante. By reducing the implicit support towards the trading entity, market discipline should strengthen and trading activity may not expand as rapidly as it has done on the back of an implicit state guarantee. Structural reform has the potential to directly reduce the interconnectedness and to indirectly reduce the size of banks and investment banks. The repeal of Glass-Steagall significantly increased the competitive pressure felt by pure investment banks like Lehman Brothers and Bear Stearns, given that they faced commercial banks that were allowed to enter the investment banking area after 1999 and which could do so at artificially low funding cost.³⁰

4. Do structural measures undermine the benefits associated with universal banking? Implicit subsidy estimates suggest that the EU universal bank profitability may be artificially increased by the indirect and implicit sponsoring by their sovereigns. Implicit subsidies primarily benefit the largest EU banks and represent a sizeable part of their profitability. Implicit subsidies are also significantly greater in Europe, compared to the United States (IMF (2014)). The universal banks today are very different from the universal banks that have served the economy well in the past. The current EU financial system is dominated by relatively few large, interconnected and diversified universal banking groups. Whereas several of those large EU universal banking groups have weathered the crisis well, the EU financial system as a whole would have likely imploded due to a system-wide cascade of banking failures without the extraordinary

 Form 10-K for fiscal year ended November 5 2005: “We face increased competition due to a trend toward consolidation. In recent years, there has been substantial consolidation and convergence among companies in the financial service industry. In particular, a number of large commercial banks … have established or acquired broker-dealers or have merged with other financial institutions. Many of these firms … have the ability to support investment banking and securities products with commercial banking, insurance and other financial services revenues in an effort to gain market shares.”

Bank Structural Reform and Too-big-to-fail

83

and on-going taxpayer, government and central bank support (European Commission (2011, 2012)). The (contingent) taxpayer support to date amounts to 40% of EU GDP (€ 5.1 trillion parliamentary committed aid measures) and has undermined the solidity of several Member States’ public finances. In the case of some Member States it has contributed to turn a banking crisis into a sovereign crisis. This has had the effect of further increasing the fragility of the banking system since banks hold large volumes of sovereign bonds on their balance sheet – and hence confidence on these banks depends on the robustness of the public safety nets). ESRB (2014) finds that in the last decade universal banks were exposed to substantially greater systemic risk than more narrowly focused banks. Despite the above, most structural reform proposals are not calling for the break-up of banking groups. Without exception, structural reform proposals simply want to disentangle the activities that are considered long-term and relationship-oriented from those that are short-term, transaction-oriented, and the most prone to rapid change. At the same time, the proposals under consideration aim to maintain banks’ ability to efficiently provide a comprehensive range of financial services to their customers. The proposals simply aim to introduce more structure in existing banking groups, not to break them up. Structural reform through subsidiarisation merely aims to inject clarity into the structure of universal banks. The nature of the activities is different and hence the culture will be different. So the aim is to keep the fundamentally different banking cultures apart. Commercial banking typically involves long-term lending relationships, whereas trading typically involves a short-term perspective. The separation of the two distinct cultures reduces the risk that a short-term oriented, deal- and fee-based trading culture negatively influences the long-term relationshipbased culture of the deposit and lending entity. Given the TBTF focus, structural reform would target a limited set of banking groups that are large, complex and engage in significant trading activity. Hence, structural reform would not affect the large majority of the 8000+ banks in the EU, and would in particular exclude small cooperative and savings banks that focus on serving the financing needs of local communities and small businesses. It would therefore preserve the diversity of the EU banking sector, e. g. in terms of different business models. Given the reduced reliance on implicit public subsidies, competition on the merits would become more important, which should lead to inefficient competitors exiting the market and allowing more efficient competitors, including new entrants, to gain market share and revenues. Each part of the group would be subject to its own profitability and resource constraints. To the extent that structural reform facilitates and enhances the effectiveness of bank resolution, exit barriers would be removed, giving more oppor-

84

Miguel de la Mano

tunities to banks that have a sound and prudent business model (European Commission (2011)).

VI. Concluding remarks The fragile and run-prone nature is a defining feature of deposit-taking banks. Public safety nets have been introduced several decades ago to avoid costly runs. However, public safety nets also give rise to excessive risk-taking (i. e. moral hazard), rapid balance sheet expansion, and competition distortions. Structural reform is about imposing certain activity restrictions within banking groups. It wants to provide clarity as to which activities benefit from funding covered by an explicit safety net and which do not. In the absence of structural reform, and supported by the public safety nets, banks have shifted from a relationship-oriented model of banking towards a transaction-oriented model of banking, sharply amplifying the potential risks and costs for society. The underlying drivers of this shift have been globalisation, technological innovation, deregulation, and increased competition. Member States are pushing ahead with structural bank reforms that are not harmonised and that risk fragmenting the internal market and imposing significant costs on EU banks that are active at European and global level. There has been support for structural reform, despite strong opposition from banks. The HLEG recommended structural bank reform measures, appropriately phased in. International institutions and the G20 call for a global debate on bank structure. The September 2013 G20 St-Petersburg statement says “We recognize that structural banking reforms can facilitate resolvability and call on the FSB, in collaboration with the IMF and the OECD, to assess cross-border consistencies and global financial stability implications, taking into account country-specific circumstances, and report to our next Summit.” As argued above, structural reform is a necessary complement to the other legislative initiatives and intends to increase the effectiveness and credibility of the existing regulatory reform agenda. Basel 3 has already been front-loaded to a large extent through market pressure, and we have not yet witnessed a great deal of bank restructuring in the EU banking sector. Aggregate bank balance sheets have not declined significantly, and the EU banking sector has become even more concentrated. Expectations of contingent state support have been confirmed and may have become more entrenched than ever, rather than credibly removed, according to the perception of markets and credit rating agencies (IMF (2014)). From this perspective, we cannot afford to wait until 2019 to start the lengthy phase-in period that is required for

Bank Structural Reform and Too-big-to-fail

85

structural reform. Banks need to be structurally reformed to reduce the distortive implicit subsidies and the resulting moral hazard, further bank balance sheet growth and competition distortions. Only upon structural reform implementation, creditors are offered a clear and meaningful choice between deposit banks and trading entities. Some of the current regulatory reforms will only be phased in gradually over time. Nevertheless, important reforms have been pulled forward due to market pressure on banks (e. g. increased capital ratios for banks). Despite this frontloading of reform measures, EU banks remain fragile and unable to fully perform their important role in the EU economy. Some European banks are still on life support through the LTROs and other implicit aid. Structural reform is an avenue through which EU banks can regain stability and confidence of European citizens. Banks have (ab)used the public safety nets to artificially expand in trading activities that are not linked to core banking activities. Aggregate bank balance sheets have ballooned from a fraction of GDP to a multiple of GDP, without a documented positive impact on economic growth or liquidity (in normal times, let alone in in crisis times when liquidity really matters).³¹ Structural reform is not a panacea, but is an essential and complementary building block of the regulatory reform agenda. Structural reform refocuses the public safety nets to those core and relationship-oriented banking activities that they were always supposed to support, whilst shielding taxpayers from risks that belong in the private sector. This will ensure that bank risk-taking is subject to adequate market discipline. Given the devastating banking crisis, it is time to (re‐)introduce structural bank regulation as a necessary complement to the conduct-based and incremental regulatory approach of the last decades, to generate much needed transparency and simplicity. The current regulatory toolbox is not geared to reverse the trend away from a short term-oriented transactions oriented banking culture. Structural reform puts the basic framework in place for the regulation of the next couple of decades. This process allows for a gradual and orderly implementation, with a maximum of legal certainty. There is overwhelming support for an ambitious structural reform initiative. Both the general public and many top financial experts from the academic and regulatory community support an ambitious reform. Structural reform is in the public interest, but it is even in the bankers’ private interest in the medium  Strikingly, large European banks have become equally important as US commercial banks in channeling US savings towards US borrowers, by relying on US short-term wholesale market funding and investing the funds into US securitized assets. European banks have become greater sponsors of USD ABCP conduits than US banks.

86

Miguel de la Mano

term, as it will allow them to regain their reputation, and the respect and trust of the public at large. The concerns voiced against ambitious structural reform are self-serving and can be rebutted with economic analysis and facts. The end of “too big to fail” will reduce large banks’ funding advantage, forcing them to compete on the basis of products, price, and service rather than implicit government subsidies. Increased competition will reduce the margins on fee-driven businesses such as securitization, trading, and derivatives, putting pressure on large banks’ profits. A larger group of competitors will also make it harder for major banks to divert such a large proportion of their profits to employee compensation; bonuses for traders and investment bankers should fall from their inappropriately high levels. With more competition, it will be harder for a handful of firms to dominate the cultural landscape, and perhaps smart college graduates will find the city of London a little less compelling. Finance will never go back to being boring—globalization and computers have seen to that— but it should become a little less exciting and it should give less rise to public concern. Structural reform cannot be adopted overnight, but the arguments that support it are based on sound economic theory and solid empirical evidence, for which exhibit one is the near-meltdown of the financial system in 2008, after highly leveraged and trading oriented financial institutions had built balance sheets the size of large EU economies, on the back of implicit state solvency support. As Johnson and Kwak (2011) pointed out in their seminal book on the financial crisis “13 Bankers”: “in 1900, almost no reasonable person thought there was any basis for constraining the market power of large dominant firms or for restricting the degree of interconnection among rival firms. There was no well-developed economic theory supporting such a position, no common law tradition, and nothing in the US Constitution (as interpreted by the Supreme Court)”. And yet it was political bravery, wisdom and a commitment to bettering the common good that led to the development of one of the core pillars of modern capitalism and democratic economies, the notion that the competitive process needs to be protected and competition enforcement is required to constrain the activities of firms, including their conduct, their freedom to associate with rivals and the freedom to alter firm structure through mergers. Indeed, when Theodore Roosevelt sued Northern Securities in 1902, the conventional wisdom was that the industrial trusts were a fact of nature. The trusts had strong backers in Washington, including the key power brokers in Roosevelt’s own Republican Party. For Roosevelt, however, any economic benefits that might be provided by the trusts did not outweigh the costs they imposed on society, both by charging monopoly prices and by stifling competition. In

Bank Structural Reform and Too-big-to-fail

87

less than a decade, by 1910, the consensus view had shifted dramatically. The power of the industrial trusts and the details of their anticompetitive behavior were sufficiently obvious to provoke a political backlash. The antitrust movement had a lasting effect; few people think today that unrestrained monopolies are good either for our economy or for our political system. Similarly, the current crisis has prompted many (see annex) to challenge the conventional wisdom about TBTF banks. If the implicit subsidies unduly benefiting certain trading activities of large systemic banks are not removed, we concur with Johnson and Kwak (2011), that the result will be a financial sector that remains as concentrated as ever, retains as robust a guarantee of government assistance as ever, and engages in reckless risk taking all the way into the next crisis. In conclusion, the rationale for introducing structural measures constraining the largest, highly complex and interconnected banks is straightforward and intuitive. If there are no financial institutions that are too big to fail, there will be no implicit subsidies discriminating in favor of some banks and distorting competition; counterparties and supervisors will play their necessary role of ensuring that banks do not take on too much risk and cause the next financial crisis; and those responsible for the failure of banks will not have to be bailed out at taxpayer expense.

References Acharya, V., T. Cooley, M. Richardson, and I. Walter (2009), “Manufacturing tail risk: a perspective on the financial crisis of 2007 – 2009”, Foundations and Trends in Finance, 4, 4, pp. 247 – 325. Admati, A. and M. Hellwig (2013), The Bankers’ New Clothes: What’s Wrong with Banking and What to Do about It, Princeton University Press. Adrian, T. and H. Shin (2010a), “Liquidity and leverage”, Journal of Financial Intermediation, 19, 3, pp. 418 – 437. Adrian, T. and H. Shin (2010b), “The changing nature of financial intermediation and the financial crisis of 2007 – 2009”, Annual Review of Economics, 2, pp. 603 – 618. Alessandri, P. and A. Haldane (2009), “Banking on the State”, paper underlying a presentation delivered at the FRB Chicago International Banking Conference on 25/09/2009. Arcand, J-L., E. Berkes and U. Panizza (2012), “Too much finance?”, IMF Working Paper 12/161. Blundell-Wignall, A., P. Atkinson, and C. Roulet (2012), “The business models of large interconnected banks and the lessons of the financial crisis”, National Institute Economic Review, No. 221.

88

Miguel de la Mano

Blundell-Wignall, A., P. Atkinson, and C. Roulet (2013), “Bank business models and the Basel system: complexity and interconnectedness”, Financial Market trends, Issue 2. Blundell-Wignall, A., G. Wehinger and P. Slovik (2009), “The elephant in the room: The need to deal with what banks do”, Financial Market trends, Issue 2. Boot, A. and L. Ratnovski (2012), “Banking and trading”, IMF Working Paper 12/238. Bruno and Shin (2014), “Cross-border banking and global liquidity”, working paper, Princeton University. Buiter, W. (2008), “Lessons from the North Atlantic financial crisis”, Paper prepared for presentation at the conference “The Role of Money Markets” jointly organised by Columbia Business School and the Federal Reserve Bank of New York on May 29 – 30, 2008. Cecchetti, S. and E. Kharroubi (2012), “Reassessing the impact of finance on growth”, BIS Working Paper 381. Chow, J. and J. Surti (2011), “Making banks safer: Can Volcker and Vickers do it?”, IMF Working Paper 11/236. Dam, L. and M. Koetter (2012), “Bank bailouts and moral hazard: evidence from Germany”, Review of Financial Studies, 25, 8, pp. 2343 – 2380. Demirgüc-Kunt, A., Karacaovali, B., and L. Laeven (2005), “Deposit insurance around the world: A comprehensive database”, Working Paper, World Bank. Diamond, D. and P. Dybvig (1983), “Bank runs, Deposit Insurance, and Liquidity”, Journal of Political Economy, 91, pp. 401 – 419. ESRB (2014), “Is Europe overbanked?”, Reports of the Advisory Scientific Committee, No. 4, June. European Commission (2011), “The effects of temporary State aid rules adopted in the context of the financial and economic crisis”, Commission Staff Working Paper, 5 October, SEC(2011) 1126 final. European Commission (2012), “State aid scoreboard: Report on state aid granted by the EU Member States – Autumn Update”, 21 December, SEC(2012) 443 final. European Commission (2013), “Structural reform in the EU banking sector: motivation, scope and consequences”, Chapter 3 of the European Financial Stability and Integration Report 2012, April. European Commission (2014a),”Economic review of the financial regulation agenda”, Commission Staff Working Document No. 2014/158. European Commission (2014b), “Impact Assessment on Proposal on banking structural reform”, SWD(2014) 30. FSB (2013), “Progress and Next Steps Towards Ending “Too-Big-To-Fail” (TBTF)”, Progress report to the G20, September 2013. G20 (2013), G20 Leaders Declaration, St Petersburg, September 2013. Gambacorta, L. and A. van Rixtel (2013), “Structural bank regulation initiatives: approaches and implications”, BIS Working Paper No 412, April. Government Accountability Office (2014), “Large Bank Holding Companies: Expectations of Government Support”, GAO-14 – 809T: Published: Jul 31, 2014. Gropp, R., C. Gründl, and A. Güttler (2010), “The impact of public guarantees on bank risk taking: Evidence from a natural experiment”, ECB Working Paper No 1272. Haldane, A. (2010a), “The contribution of the financial sector – miracle or mirage?”, Bank of England. Haldane, A. (2010b), “The $100bn question”, Bank of England.

Bank Structural Reform and Too-big-to-fail

89

Haldane, A. (2012), “On being the right size”, Bank of England. Haldane, A. and V. Madouros (2012), “The dog and the frisbee”, speech at the Federal Reserve Bank of Kansas City’s 366th economic policy symposium, “The changing policy landscape”, Jackson Hole, Wyoming, 31 August 2012. ICB (2011), “Final report: Recommendations”, Independent Commission on Banking. IMF (2012), “Global Financial Stability Report”, October. IMF (2014), “Global Financial Stability Report: How big is the implicit subsidy for banks seen as too-important-to-fail?”, Chapter 3, April. Johnson, Simon, and James Kwak (2011), 13 bankers the Wall Street takeover and the next financial meltdown. Random House LLC. Kydland, Finn E., and Edward C. Prescott (1977): “Rules rather than discretion: The inconsistency of optimal plans.” The Journal of Political Economy 473 – 491. Laeven, L., L. Ratnovski, and H. Tong (2014), “Bank size and systemic risk”, IMF Staff Discussion Note 14/04. Liikanen (2012), “Final Report of the high-level expert group on reforming the structure of the EU banking sector”, 2 October, http://ec.europa.eu/internal_market/bank/docs/highlevel_expert_group/report_en.pdf. Lumpkin, S. (2011), “Risks in financial group structures”, OECD Journal: Financial Market Trends, Issue 2. Maes, S. (2014), “Shadow banking: a European perspective”, Proceedings of the Federal Reserve Bank of Chicago 16th Annual International Banking Conference, forthcoming. Mian and Sufi (2014), “House of Debt: How They (and You) Caused the Great Recession, and How We Can Prevent It from Happening Again”. University of Chicago Press Miles, D., J. Yang, G. Marcheggiano (2013), “Optimal bank capital”, Economic Journal, 123, 567, pp. 1 – 37. Miller, M., L. Zhang, and H. Li (2013), “When bigger isn’t better: bailouts and bank reform”, Oxford Economic Papers, 65, pp. 7 – 41. Richardson, M. (2013), “Why the Volcker Rule Is a Useful Tool for Managing Systemic Risk”, working paper, New York Stern School of Business. Schäfer , A., I. Schnabel and B. di Mauro (2013).”Financial Sector reform after the Crisis: Has Anything Happened?”, CEPR Discussion Paper Series, No. 9502. Shin, H. (2009), “Reflections on Northern Rock: The Bank Run that Heralded the Global Financial crisis”, Journal of Economic Perspectives, 23, 1, pp. 101 – 119. Shin, H. (2012), “Global banking glut and loan risk premium”, Mundell-Fleming Lecture, IMF Economic Review, 60, 2, pp. 155 – 192. Ueda, Kenichi, and Beatrice Weder di Mauro (2013). “Quantifying structural subsidy values for systemically important financial institutions.” Journal of Banking & Finance 37: 3830 – 3842. Zingales (2012), “Why I was won over by Glass-Steagall”, Wall Street Journal, June 10.

Annex The European Parliament adopted on 3 July 2013 the Own Initiative Report (McCarthy Report) on the structural reform of the EU banking sector. The report “Re-

90

Miguel de la Mano

forming the structure of the EU banking sector” was adopted by an overwhelming majority of members of the European Parliament. The European Parliament has welcomed the Commission’s intention to bring forward a proposal for structural reform to tackle problems arising from banks being TBTF in order to provide greater resilience against potential financial crises, restore trust and confidence in banks, remove risks to public finances and deliver a change in banking culture. As part of preparing its Impact Assessment (European Commission SWD (2014) 30 final), the Commission services held a public stakeholder consultation during the course of spring 2013 based on a consultation paper. The Commission services received 540 replies. These responses came from the expected type of respondents: banks and other financial institutions, corporate clients, investors, public authorities, consumer associations and individuals. The number of responses from individuals (439) and consumer associations (11) stood out in this particular consultation. The consultation responses highlight a distinct difference in view between the responses of banks on the one hand, and consumers and non-bank financials on the other hand. The former are to an overwhelming extent against structural separation (with the exception of some cooperative banks). The latter are largely in favour. The views of other categories are more balanced. Corporate customers, while acknowledging the need to address TBTF, express opposition based on the potential impact of such reforms on their cost of financing. The G20 is supportive of structural reform: “We recognize that structural banking reforms can facilitate resolvability and call on the FSB, in collaboration with the IMF and the OECD, to assess cross-border consistencies and global financial stability implications, taking into account country-specific circumstances, and report to our next Summit.” (G20 (2013)). The Bank for International Settlements has acknowledged the potential benefits of structural reform: “The line is generally drawn somewhere between “commercial” and “investment” banking businesses, restricting the universal banking model. Such a separation can, in principle, help in several ways. First, and most directly, it can shield the institutions carrying out the protected activities from losses incurred elsewhere. Second, it can prevent any subsidies that support the protected activities (e. g. central bank lending facilities and deposit guarantee schemes) from lowering the cost of risk-taking and encouraging moral hazard in other business lines. Third, it can reduce the complexity and possibly size of banking organisations, making them easier to manage, more transparent to outside stakeholders and easier to resolve; this in turn could improve risk management, contain moral hazard and strengthen market discipline. Fourth, it can prevent the aggressive risk culture of the riskier activities from infecting that of more traditional banking business, thus reducing the scope for

Bank Structural Reform and Too-big-to-fail

91

conflicts of interest. In addition, some observers have noted that smaller institutions would reduce the risk of regulatory capture. All these mechanisms would also help to limit taxpayers’ exposure to financial sector losses.” (Gambacorta and van Rixtel (2013)) A series of top economists and financial experts believe that the economy cannot recover unless TBTF banks are structurally reformed and/or broken up. – Nobel prize-winning economists Joseph Stiglitz, Ed Prescott, and Paul Krugman have all spoken in favour of bank structural reform. – Former chairmen of the Federal Reserve Alan Greenspan stated that “If they’re too big to fail, they’re too big.” – Paul Volcker “There is an expectation that very large and complicated financial institutions will not be allowed to fail. Unless that conviction is shaken, the natural result is that risk-taking will be encouraged and in fact subsidized beyond reasonable limits.” – Current Vice Chair and director of the Federal Deposit Insurance Corporation and former 20-year President of the Federal Reserve Bank of Kansas City Thomas Hoenig states that “Looking back, one sees that the crisis was inevitable, if for no other reason than that these TBTF firms would push the boundaries until there was a crisis.” “If we don’t make these changes, I think we’re destined to repeat the mistakes of the past. When you mix commercial banking and high-risk broker-dealer activities, you increase the risk overall and as a result you invite new problems.” – The former head of the FDIC, Sheila Bair states that “even with very good management these institutions are just too big to manage.” – Financial Stability Director at the Bank of England, Andrew Haldane, stated that “Have we solved ‘too big to fail’? No. That’s not my pessimistic verdict; it is the market’s. … Too-big-to-fail is far from gone. It is even more important it is not forgotten.” – In his testimony to the Financial Crisis Inquiry Commission, Former Superintendent Eric Dinallo argued that AIG wouldn’t have been so laden with risk if they hadn’t been able to function like a hedge fund. He says that the repeal of Glass-Steagall, “permitted AIG to operate an effectively unregulated hedge fund with grossly insufficient reserves to back up its promises. Had AIG Financial Products been a stand-alone company, it is unlikely that its counterparties would have been willing to do business with it because its commitments would never have carried a triple-A credit rating.” – Dean Baker, economist at the Center for Economic and Policy Research, states that “A break-up of the big banks will at least give the country some hope that things can change. As it stands now, the big banks are back on their feet, and in some cases more profitable than ever, feasting

92

–

–

–

–

–

–

Miguel de la Mano

on the now explicit government guarantee of support in the event of a crisis.” President of the Federal Reserve Bank of Dallas, Richard Fisher. “Given the danger these institutions pose to spreading debilitating viruses throughout the financial world, my preference is for a more prophylactic approach: an international accord to break up these institutions into ones of more manageable size—more manageable for both the executives of these institutions and their regulatory supervisors.” President of the Federal Reserve Bank of St. Louis, Thomas Bullard. “We do not need these companies to be as big as they are … We should say we want smaller institutions so that they can safely fail if they need to fail.” President of the Federal Reserve Bank of New York, William Dudley: “We are a long way from the desired situation in which large complex firms could be allowed to go bankrupt without major disruptions to the financial system and large costs to society.” The Director of Research at the Federal Reserve Bank of Dallas, Harvey Rosenblum. Russ Roberts, economist at Stanford’s Hoover Institute: “To me breaking up the banks would be better than the current situation, which is: fake capitalism where they [big banks] make money at taxpayers’ expense.” Luigi Zingales, Professor of entrepreneurship and finance at the Chicago Booth School of Business: “The too big to fail policy creates an unfair competitive advantage for large banks.” “breaking up the banks is not an unreasonable thing to do.” Willem Buiter (current Chief Economist, Citigroup) “Would there be significant efficiency losses as a result of breaking up banks and taxing bank size? What would be the social cost of taxing size in banking and other financial businesses and of breaking up banks? Why do banks and other financial enterprises become too big to fail? I believe there are four reasons. The exploitation of monopoly power (market power). The exploitation of ‘synergies of conflict of interest’. The exploitation of economies of scale and economies of scope. The joys of being too big to fail and the implicit subsidy provided by the tax payer guaranteeing the bank against default and insolvency. The pursuit of the benefits of subsidized liquidity and solvency support from the state clearly makes sense for the bank’s top management and board, for shareholders and for unsecured creditors: being too big, too interconnected, too complex and too international to fail is a major business asset. The universal banks, that dominate the European banking scene and much of cross-border banking and are now also increasingly dominant in the USA, exist for three of the four reasons outlined above – all but the third. Economies of scale and scope have long been exhausted and disecon-

Bank Structural Reform and Too-big-to-fail

–

–

–

93

omies of span of control compete with lack of focus as the main drivers of organisational inefficiency. But by bundling the systemically important activities with the not systemically important activities, the entire organisation falls under the government’s bail-out umbrella. It is time to see a lot more and a lot smaller banks.” Sandy Weill (CEO, Citigroup) “What we should probably do is go and split up investment banking from banking, have banks be deposit takers, have banks make commercial loans and real estate loans, have banks do something that’s not going to risk the taxpayer dollars, that’s not too big to fail.” David Komansky, former Merrill Lynch CEO: “Unfortunately, I was one of the people who led the charge to get Glass-Steagall repealed. As I sit here today, I regret those activities and wish we hadn’t done that.” Phil Purcell, former Chairman and CEO of Morgan Stanley: “There is one benefit of breakups that hasn’t gotten much publicity: Shareholders would get greater value from their investments.”

“Europe’s 500” is the European organisation and networking platform for growth companies and their entrepreneurs. The organization, founded in 1996, represents more than 2,300 growth companies from across Europe, with the common goal of contributing to growth and more employment in Europe and promoting entrepreneurship. Europe’s 500 issued a release on March 24 2013 in which it calls for Glass-Steagall like bank separation. As mentioned before, the Advisory scientific Committee of the ESRB, composed of top European academics, recently came out with a report that advocates structural reform to complement the current reform agenda (ESRB (2014)).

Debra Stone

The Volcker Rule

Thank you very much for inviting me today to speak about the Volcker Rule. Today, I will provide an overview of what the rule is meant to accomplish, and at a high level, the broad outlines of how it is intended to accomplish that. The views that I express today are my own and do not necessarily represent those of the Federal Reserve Bank of New York or the Federal Reserve System. First, a little background. In 2009, the Group of Thirty issued its report on Financial Reform – A Framework for Financial Stability.¹ Former Federal Reserve System Chairman Paul Volcker chaired the Steering Committee that produced the Report. One of the recommendations contained in that Report stated that “Large, systemically important banking institutions should be restricted in undertaking proprietary activities that present particularly high risks and serious conflicts of interest. Sponsorship and management of commingled private pools of capital (that is, hedge and private equity funds in which the banking institutions own capital is commingled with client funds) should ordinarily be prohibited …”.² This Report reflected Chairman Volcker’s view that would later be implemented as what is commonly known as the “Volcker Rule” – which was adopted as Section 619 of the Dodd-Frank Wall Street Reform and Consumer Protection Act in July of 2010 (“Dodd-Frank”).³ Section 619, which amends Section 13 of the Bank Holding Company Act, is titled “Prohibitions on Proprietary Trading and Certain Relationships with Hedge Funds and Private Equity Funds”. Before getting further into the specifics of the Volcker Rule, I will note that this conference is titled “Too Big to Fail III: Structural Reform Proposals – Should We Break up the Banks?” In connection with understanding the basics of the Volcker Rule and its implementing regulations, it is important to know two things. First, the Volcker Rule is largely a prohibition on activities and does not by its terms require reform to the structure of the banking institution. It does not, for example, require the ring-fencing of banking activities. Second, it is also important to understand that the prohibitions on activities contained in

 The Group of Thirty, Financial Reform: A Framework for Financial Stability (2009), available at http://www.group30.org/images/PDF/Financial_Reform-A_Framework_for_Financial_Stabil ity.pdf.  Id. at 28.  Dodd-Frank Wall Street Reform and Consumer Protection Act (“Dodd-Frank”), Pub. L. No. 111– 203, § 619, 124 Stat. 1620 (2010) (codified at 12 U.S.C. § 1851 (2010)).

96

Debra Stone

the rule apply not only to insured depository institutions, but to affiliates, parents and subsidiaries of such institutions. Simply put, the Volcker Rule prohibitions apply not only to banking institutions covered by it, but to their parents and affiliates on a worldwide basis. Before discussing the specifics of what the Volcker Rule prohibits and what it permits, I will describe my understanding of what the Volcker Rule is intended to accomplish. Congressional statements, as well as written comments provided by Chairman Volcker in connection with the rule proposal, are enlightening. U.S. Senators Jeff Merkley and Carl Levin were the principal authors of Section 619 of the Dodd-Frank Act and the Congressional Record reflects their views as to what the Volcker provisions of Dodd-Frank were intended to accomplish. According to Senator Merkley’s statement, “Section 619 is intended to limit proprietary trading by banking entities … Properly implemented, section 619’s limits will tamp down on the risk to the system arising from firms competing to obtain greater and greater returns by increasing the size, leverage and riskiness of their trades. This is a critical part of ending too big to fail financial firms.”⁴ Chairman Volcker stated the following in written comments to the proposed Volcker Rule: “Commercial bank proprietary trading is thus at odds with the basic objectives of financial reforms: to reduce excessive risk, to reinforce prudential supervision, and to assure the continuity of essential services”.⁵ In other statements, Chairman Volcker expressed the view that the cultural conflicts presented by proprietary trading – “impersonal trading dismissive of client relationships” – were difficult to reconcile with the provision of essential banking functions.⁶ Section 619 required three U.S. banking agencies, the Federal Reserve, the OCC and the FDIC, as well as the SEC and the CFTC, to adopt regulations to carry out Section 619. The Dodd-Frank Act also provided that the U.S. Treasury Secretary would be responsible for the coordination of the regulations to be issued by these Agencies. The regulations implementing the Volcker Rule were first proposed on October 11, 2011 and were open for public comment until February 13, 2012.⁷ The Agencies received over 18,000 comments, although the majority of

 156 Cong. Rec. S 5894 (daily ed. July 15, 2010) (statement of Sen. Merkley).  Paul A. Volcker, Commentary on the Restrictions on Proprietary Trading by Insured Depository Institutions 2 (2012).  Paul A. Volcker, The William Taylor Memorial Lecture: Three Years Later: Unfinished Business in Financial Reform 10 (Sept. 23, 2011).  Prohibitions and Restrictions on Proprietary Trading and Certain Interests in, and Relationships with, Hedge Funds and Private Equity Funds (“Proposing Release”), 76 Fed. Reg. 68846 (OCC et al. Nov. 7, 2011) (setting a comment deadline of January 13, 2012); Prohibitions and Re-

The Volcker Rule

97

the comments were a similar short-form comment letter expressing support for the proposed rule. However, hundreds of comment letters expressed a variety of more substantive comments on the proposed rule. After consideration of the comments and modifications to the provisions of the proposed rule, the final rule was released on December 10, 2013, with an effective date of April 1, 2014. The final rule is 71 pages, but the preamble, which provides supplemental information explaining the rule, runs close to 900 pages. There are two key dates from the perspective of both the institutions that are subject to the rule and the supervisors with responsibility for interpretation and enforcement of the rule. The first is July 21, 2015, which is the date by which institutions subject to the Volcker Rule must conform their activities to the requirements of the rule. The second date relates to metrics reporting, which is relevant only to a subset of firms with more substantial trading activities. I will mention those requirements later. With that background, I will now discuss how the rule accomplishes its objectives. Section 619 of Dodd-Frank contains two basic prohibitions that can be simply stated: The first is that “a banking entity shall not … engage in proprietary trading”.⁸ The second is that a “banking entity shall not … acquire or retain any equity, partnership, or other ownership interest in or sponsor a hedge fund or private equity fund.”⁹ Section 619 also provides for a variety of exemptions to these prohibitions. The challenges involved in developing the regulations to implement the statute related to defining and describing these various exemptions. For the purposes of today’s discussion, I will describe how the rule applies at a high-level, but please note that the provisions of the rule are complex and my comments only provide an overview. To understand how the Volcker Rule will apply, it is helpful to think of the requirements as part of a three-step analysis. This three step analysis applies both to proprietary trading and fund activities, but for purposes of today’s discussion, I will focus on proprietary trading, which has been the source of much of the public discussion regarding implementation of the rule. The three questions are the following: first, is the “banking entity” engaging as principal for the “trading account” of the “banking entity” in any purchase or sale of one or more “financial instruments”? Second, does an exemption apply? And

strictions on Proprietary Trading and Certain Interests in, and Relationships with, Hedge Funds and Private Equity Funds (“Extension of Comment Period”), 77 Fed. Reg. 23 (OCC et al. Jan. 3, 2013) (extending the comment period until February 13, 2012).  12 U.S.C. § 1851(a)(1)(A).  Id. § 1851(a)(1)(B).

98

Debra Stone

third, despite the application of an exemption, is the activity prohibited by a “backstop” provision of the Volcker Rule? I will now discuss these three questions in more detail. The first question focuses on the definition of “proprietary trading” that is contained in the rule. “Proprietary trading means engaging as principal for the trading account of the banking entity in the purchase or sale of one or more financial instruments”.¹⁰ This definition contains three terms that are key to understanding the scope of its application: trading account, banking entity and financial instruments. As I mentioned earlier, the prohibitions contained in the Volcker Rule apply not only to insured depository institutions, but to “banking entities”, which are defined in the Volcker Rule as insured depository institutions, entities that control insured depository institutions (for example, a bank holding company), any foreign bank treated as a bank holding company for purposes of Section 8(a) of the International Banking Act of 1978 (for example, foreign banks with a U.S. branch or agency) and parent companies of such foreign banks, as well as any affiliates or subsidiaries of any of the foregoing (for example, a broker-dealer subsidiary of a bank holding company).¹¹ The Volcker Rule prohibitions on proprietary trading apply only to activities that the banking entity is undertaking as principal for the “trading account”. The definition of “trading account” in the final rule implements the statutory provision that defined “trading account” as any account used to hold a financial instrument “principally for the purpose of selling in the near-term (or otherwise with an intent to resell in order to profit from short-term price movements)”.¹² The final rule implements this provision by including three types of accounts as trading accounts: accounts used to take positions for the purpose of benefitting from actual or expected short term price movements, accounts used by banking entities with a particular status as securities or swaps dealers in connection with such dealing activities, and accounts used to purchase or sell financial instruments of banking entities, or affiliates, of such banking entities that are covered positions or trading positions under the Market Risk Capital Rules.¹³ The third key definition is “financial instrument”. Instruments not included in the definition of “financial instrument” are not subject to the prohibitions on proprietary trading. Financial instruments include securities, derivatives, commodity for-

 Prohibitions and Restrictions on Proprietary Trading and Certain Interests in, and Relationships with, Hedge Funds and Private Equity Funds (“Final Release”), 79 Fed. Reg. 5535, 5781, Final Rule § _.3(a).  Final Release, 79 Fed. Reg. at 5779, Final Rule § _.2(c).  12 U.S.C. § 1851(h)(6).  Final Release, 79 Fed. Reg. at 5781, Final Rule § _.3(b).

The Volcker Rule

99

wards, and options on any of such instruments, but do not include loans, spot commodities or spot foreign exchange or currency transactions.¹⁴ While such instruments themselves are excluded, associated derivatives on such instruments, such as foreign exchange swaps and forwards, are not excluded from the definition. I have now described what is prohibited, the scope of which is seemingly broad. The second question is whether despite the prohibitions, is the activity permitted? The Volcker Rule specifically excludes certain purchases or sales of financial instruments that are not considered to fall under the definition proprietary trading.¹⁵ Examples are repurchase and reverse repurchase transactions, securities lending transactions, bona fide liquidity management, certain clearing transactions, purchases or sales in which the banking entity acts solely as agent, broker or custodian – generally the types of activities that would not involve an intent to profit from short-term price movements. The exclusion for liquidity management has been the subject of much discussion and I note that the final rule contains specific requirements intended to ensure that financial instruments purchased and sold as part of a liquidity management plan are highly liquid and not reasonably expected to give rise to appreciable profits or losses as a result of short-term price movements.¹⁶ The liquidity management exclusion does not broadly allow firms to engage in asset-liability management, earnings management or scenario hedging unless otherwise permitted by another exemption under the rule. I will now describe certain of the exemptions which have been the focus of public discussion and many of the comment letters. The rule contains a number of exemptions, which are for trading activity that falls within the definition of “proprietary trading” under the rule, but is nonetheless permitted if such trading satisfies the conditions of a specific exemption. The exemptions are for the following activities: underwriting and market making-related activities, risk-mitigating hedging activity, trading in U.S. government, agency, State and municipal obligations, trading in certain foreign government obligations, certain trading activities of foreign banking entities, trading on “behalf of customers” and trading by a regulated insurance company. I will provide more detail about a few of the exemptions that may be of particular interest to the conference attendees. The first is the exemption for market making-related activities.¹⁷ As Federal Reserve Board Governor Tarullo stated in connection with the adoption of the    

Id. Id. Id. Id.

§ _.3(c). § _.3(d). § _.3(d)(3). at 5783, § _.4(b).

100

Debra Stone

Volcker Rule, “ … the fundamental challenge is to distinguish between proprietary trading, on the one hand, and either market-making or hedging, on the other. The difficulty in doing so inheres in the fact that a specific trade may be either permissible or impermissible depending on the context and circumstances within which that trade is made.”¹⁸ After the Dodd-Frank Act was passed, a number of firms, in anticipation of the implementation of the Volcker Rule, exited their businesses that were solely focused on proprietary trading for the firm’s own account. The larger challenge related to the market-making businesses of the firms – how would the rule distinguish between the assumption of principal risk involved in permissible market-making activities that provide intermediation and liquidity from impermissible proprietary trading? The final rule focuses on providing a framework for assessing whether trading activities are consistent with market-marking, intended to accommodate the differences in market-making activities across different markets and asset classes. The focus of the final rule is on analyzing the overall market making-related activities of individual trading desks, and on meeting specific requirements that demonstrate that these activities (and related activities) are related to satisfying reasonably expected near term customer demands, rather than on requiring a transaction-by-transaction analysis of each trade. There are two exemptions that are of particular interest to foreign banks. The first is the exemption in the final rule that permits trading in certain foreign (non-U.S.) government obligations.¹⁹ This exemption was not included in the proposed rule and was the subject of a number of comment letters received by the Agencies in response to the proposed rule, both from foreign governments, market participants and others. This exemption is not specifically contemplated by the statute, but was adopted by the Agencies pursuant to their discretionary authority under the statute that permits the Agencies to adopt exemptions for any activity that the Agencies determine would promote and protect the safety and soundness of the banking entity and the financial stability of the United States. The exemption contains two parts, one applicable to non-U.S. controlled banking entities and one applicable to U.S. controlled banking entities. The exemption permits U.S. operations of foreign banking entities to engage in proprietary trading in the U.S. in the foreign sovereign debt of the foreign sovereign under whose laws the banking entity – or the banking entity that controls it – is organized (meaning home country obligations), and any multinational central  Daniel K. Tarullo, Governor, Federal Reserve System, Opening Statement at the Open Board Meeting on the Volcker Rule (Dec. , 2013), available at http://www.federalreserve.gov/newsevents/press/bcreg/bcreg20131210a-tarullo-statement.htm.  Final Release, 79 Fed. Reg. at 5785, Final Rule § _.6(b).

The Volcker Rule

101

bank of which the foreign sovereign is a member, so long as the purchase or sale is not made by an insured depository institution.²⁰ This exemption is intended to permit a foreign banking entity’s U.S. operations to engage in proprietary trading in the U.S. in the obligations of the foreign sovereign’s home country in order to permit these U.S. operations to support the smooth functioning of markets in such obligations in the same manner as U.S. banking entities are permitted to support the smooth functioning of markets in U.S. government and agency operations. The exemption also permits a foreign bank or foreign broker-dealer regulated as a securities dealer and controlled by a U.S. banking entity to engage in proprietary trading in the obligations of the foreign sovereign under whose laws the foreign entity is organized.²¹ This exemption allows U.S. banking organizations to continue to engage in proprietary trading activities in the sovereign debt of the country in which its foreign bank subsidiaries or broker-dealer subsidiaries are located. As is the case with the exemption for U.S. government and agency obligations, the exemption does not apply to derivatives of such obligations. It is made clear in the preamble to the rule that relying on this exemption is not the only means for a foreign bank to engage in trading activities relating to foreign sovereign obligations. The preamble to the final rule makes it clear that generally, banking entities may rely on either the market-making or underwriting exemptions for their primary dealer activities or the functional equivalent of such activities. However, reliance on these exemptions, as noted, requires the banking entity to comply with the specific requirements of the Volcker Rule that apply to such exemptions. In addition to those exemptions, the other exemption that is expected to be relied upon widely by foreign banking organizations is the exemption for permitted trading activities of foreign banking entities.²² This exemption is intended to implement the provision of the statute that permits foreign banking entities to engage in proprietary trading that occurs solely outside of the United States.²³ Banking entities controlled at the top tier by a U.S. entity cannot rely on this exemption and the exemption contains certain requirements designed to ensure that only organizations that conduct the majority of their activities outside the U.S. are permitted to rely on the exemption. This exemption contains a number of specific requirements, which are focused on ensuring that neither the trading activities, nor the decision-making related to the trading activities, undertaken in

   

Id. § _.6(b)(1). Id. § _.6(b)(2). Id. at 5786, § _.6(e). 12 U.S.C. § 1851(d)(1)(H).

102

Debra Stone

reliance on the exemption are conducted by entities or personnel located in or organized under the laws of the U.S. or any State. Earlier, I mentioned that there are three steps to analyzing any transaction under the Volcker Rule. The third question relates to what I will call the “backstop” provisions: whether or not an activity is permitted pursuant to one of the exemptions contained in the rule, no transactions or activities are permissible under the Volcker Rule if they result in the following: a “material conflict of interest”, material exposure to “high-risk assets” or “high-risk trading strategies”, pose a threat to safety and soundness of the banking entity or to the financial stability of the United States.²⁴ I will briefly mention the Volcker Rule’s second main prohibition – the prohibition by banking entities on the ownership and sponsoring of hedge funds and private equity funds, referred to in the rule as “covered funds”.²⁵ The definition of “covered fund” in the rule is quite detailed, but is intended to capture those types of funds that are used for the investment purposes that were the target of the restrictions contained in Section 619. As provided in the statute, the final rule permits a banking entity to invest in or sponsor a covered fund, subject to certain conditions and limitations, in connection with organizing and offering a covered fund. In addition, there are a number of exceptions to the prohibition, such as an exemption for activities that occur solely outside of the United States. There are two other significant areas of the rule that firms will be focused on in connection with implementation. The first is a requirement that any banking entity engaged in proprietary trading or covered fund activities develop and implement a compliance program reasonably designed to ensure and monitor compliance with the prohibitions and restrictions set forth in the provisions of the statute and the Volcker Rule.²⁶ The final rule applies enhanced standards for these compliance programs to banking entities with more significant assets and reduces the burden on smaller banking entities. For those banking entities subject to the enhanced compliance program requirements, the final rule includes a requirement that the CEO of the banking entity attest annually that the banking entity has in place a program reasonably designed to achieve compliance with the requirements of section 13 of the Bank Holding Company Act and the final rule. For a U.S. branch or agency of a foreign banking entity, this attestation can be provided for the entire U.S. operations by the senior management officer of the U.S. operations who is located in the U.S.

 Final Release, 79 Fed. Reg. at 5786, Final Rule § _.7(a).  Id. at 5787, § _.10(a).  Id. at 5796, § _.20.

The Volcker Rule

103

The second area that firms with significant trading activities will be focused on in the near term relates to the requirement for banking entities that meet certain thresholds with respect to their trading activities to report specific quantitative measurements for each of its trading desks engaged in trading activity covered by the rule.²⁷ The rule proposal contained 17 metrics; the final rule contains seven. The metrics reporting requirements are being phased-in, with the firms with the most significant activities required to begin reporting metrics in the summer of 2014. The metrics are designed to be a tool for triggering further scrutiny, both by the banking entity and by supervisors, in evaluating whether a banking entity is engaging in either prohibited proprietary trading or activities prohibited under the “backstop” requirements. The Agencies will evaluate the data collected and the preamble to the rule states that the Agencies will “revisit the metrics and determine, based on a review of the data collected by September 30, 2015, whether to modify, retain or replace the metrics.”²⁸ Approximately eight U.S. banking entities and 16 foreign banking entities will be required to report metrics when the metrics requirements are fully phased-in. I hope this overview has been helpful to you in order to gain a basic understanding of the Volcker Rule. As I mentioned at the beginning, the concepts are simple, but the challenges are in the details.

 Id. at 5797, Appendix A.  Final Release, 79 Fed. Reg. at 5772.

Adam S. Posen¹

Confronting the Reality of Structurally Unprofitable Safe Banking I served on the Bank of England’s Monetary Policy Committee during 2009 – 12. As such, I was not a financial regulator, and was not on the newer Financial Policy Committee. I have worked on a number of situations, including Japan in the 1990’s, where there have been financial crises, but previously I have tended to work on them as a macroeconomist looking at their impact on growth, long term systemic differences, as well as short term cyclical problems. As a result, I find the meagre state of the structural reform discussion, at large and as reflected in today’s conference, rather depressing. Consider our conference co-chair Patrick Kenadjian’s introduction to today’s proceedings in which he seemed to say, “Well, policymakers initially totally dismissed structural reform and now the agenda is full with radical initiatives like Liikanen, Vickers, and the Volker Rule.” To someone who like me who focuses on macroeconomic outcomes rather than legalities for banks, however, all I see are the same big banks doing largely the same things with the same or increased market share. Therefore, it does not look like structure has changed much at all, and therefore I think there has been no lasting reform. I think that it is misleading for anyone to pretend otherwise. It is particularly misleading in a world where millions of European and American citizens have had permanent hits to their lifetime income through asset market losses and through unemployment. We often talk about “structural reform in labor markets” (and in healthcare, and the services sector, and so on), and occasionally radical regulatory overhauls do occur, causing large displacements. Yet, the sweep of structural reform is never seriously applied to financial institutions in the same way as it is to labor markets. Perhaps this is because the institutions are deemed much too fragile and complex in their connections to be disrupted, as former US Treasury Secretary Timothy Geithner seems to argue. Sometimes it is because banks and private financial institutions are assumed to be sufficiently subject to market discipline, as former Federal Reserve Chair Alan Greenspan maintained. Others, like my Peterson Institute colleague

 President, PIIE. Contact: [email protected]. The author’s research in this area is supported by a major grant from the Alfred P. Sloan Foundation. The views expressed here are solely the author’s, and do not necessarily represent the views of the Peterson Institute, the Sloan Foundation, or any of their Boards.

106

Adam S. Posen

Simon Johnson, argue that the modesty of reforms is largely due to the ongoing political power of the banks in the EU and US. Whatever the reason, even if wellintentioned, I personally find where the FSB and the public state of debate now are to be something between rather disappointing and a betrayal. In the end, I think you do have to step back, even though more radical structural reform has clearly been stymied by many of the participants in this conference and your colleagues around the world. Let us step back and say, okay, if we were to try to do this right and treat financial markets the way we treat markets for energy, for healthcare, and especially for labor, what would we set for the ambition of reform? We should have a policy discussion about what we want the market for financial intermediation actually to be, and not be entirely bound by the legacy of the status quo. That gets us to a very different discussion about what was the underlying problem relating the recent financial crisis with most of the major ones in the last 40 years. What I want us to think about is ante-ex-ante. What has been the ongoing functioning and malfunctioning of the financial system including, but definitely not limited to, the repeated proclivity to crisis of banking – even given lenders of last resort and deposit insurance? We can all rattle off from our textbooks the idea that banks amass household savings and then allocate credit, and we need banks to alleviate information problems for small borrowers with collateral. Otherwise we want borrowers engaging in capital markets. If you take that fundamental perspective, it is very difficult to conclude anything but that our financial systems have failed us horribly over the last 40 years on both sides of the Atlantic (and parts of the Pacific as well). For those of us who are old enough to remember, or at least read enough old stuff to remember, we see generalizations taken for granted in the current discussion as weird about faces on the previous post-crisis received wisdom. Twentyfive years ago, it was widely proclaimed that the great thing about universal banking in Germany and elsewhere was that the universal banks still did staid, calm long-term lending. And this conservative behavior was attributed precisely to the fact that the universal banks did not have all these crazy regulations like Glass-Steagall and interstate banking limits in the US. In contrast, the American banks (and the British institutions who took advantage of them offshore in London) had to engage in financial innovation because they had all these crazy regulations in place. The virtues of financial innovation were debated, then and now, but many observers in the US in the 1970s and 1980s said that “financial innovation is a great side-effect of Glass-Steagall.” Now that discussion has swapped around, and Germany and the other universal banking systems are supposed to be beneficial because they can diversify across activities away from traditional banking, while the American and British

Confronting the Reality of Structurally Unprofitable Safe Banking

107

financial institutions are crying that financial innovation is impeded because they are over-regulated. All of which is to say, that whoever is managing the banks in either country at any given time wants to argue that there should be no change in bank structure, which is very nice for them, but not very compelling as a policy argument. Given the revealed emptiness of most reflexive status quo arguments, let me make some proposals for structural change of our financial systems. First, conglomerates do not work in general. This applies to almost every other business and has become a watchword of management and corporate governance. We had a brief period in the 1970s through the early 1980s where we had combinations of diversified businesses and these ultimately were seen as failures. They were failures not just because they ignored expertise and made it difficult to manage. They were failures also because the ultimate owners could not govern and monitor them even with large shareholdings. That complexity is an ongoing failure of corporate governance and impedes bank supervision as well as monitoring by ownership, and there is nothing particularly worthwhile about it. This holds true for large global banks. The markets have recognized that the combination of insurance and commercial banking was a mistake, and unwound most of those merged entities. In my view, however, the failure of conglomeration extends to universal banking narrowly defined when firms combine wealth and asset management, retail banking, commercial lending, and investment banking (not to mention prop trading). It is the specialists that have succeeded in investment banking (Morgan Stanley and Goldman Sachs are banks only in discount window name only), and the specialized asset managers that dominate that field of competition. None of the global banks claim a benefit from their retail networks, and most look at them as a cost (I will return to that later). More formally, finance researchers have spent decades trying to find either economies of scale or scope in banking, and – except for the fact that you can maybe economize on your backoffice once you get above a few million dollars in size of balance sheet – there is no robust evidence of any such economies. That is, except for too big to fail implicit subsidies, of course. Such empirical evidence should have been easy to find, and here, too, it is time to stop pretending otherwise. Second, consider the idea of robustness through variety in financial systems. There are a number of us who have written about this, and praise this virtue of variety both in terms of number and type of intermediaries. It is evident right now that the US and Germany benefitted in terms of macroeconomic resilience recovering from the crisis by virtue of having decentralized banking systems compared to the UK or Italy. In the US, while Citigroup, Bank of America, and a few other large banks got into roughly as much trouble as Royal Bank of Scotland and Lloyds HBOS did in the UK, for example, they represented a smaller

108

Adam S. Posen

share of the financial system. As a result, there were alternative means of getting US non-financial businesses credit when the major American money center banks were impaired. The broader economic impact of these institutions’ distress is less. You can make the similar comparison in crisis recovery for Germany vs Belgium or France. We can have either community banks or Sparkassen and Raiffeisenbanken. In the US, we also have specialized junk bonds and venture capitalists, among other entities, and that tiering and separation matters of financial intermediaries has benefits. It provides insulation and assures diversity the same as it would in any other industry. It does not solve everything in terms of crisis prevention, but it is evidently beneficial for resilience and diminishing impact. Moreover, we have direct evidence from the current crisis, in line with earlier smaller crises, that the more specialized and narrow financial firms were the ones who functioned better in many ways with higher survival rates. One fact that we have to keep reminding ourselves about is that most hedge funds either did not blow up during the crisis or if they did blow up, they did not take much of anyone else with them, which is exactly what they were supposed to do. They were one of the few parts of the system, that did pretty much what it was supposed to do: Risk only their own rich investors’ money and be clearly labeled as such. Similarly, insurance companies, both life and P&C, rode out the crisis much better than banks – partly due to superior regulation and supervision, in turn due to simpler product lines and governance. The insurers that got into trouble, like AIG, were the ones who undertook speculative businesses outside of standard insurance activities. Third, moving on to the supervision aspect, we need to enable regulators to see through excessive conglomeration and complexity of banks. For those of you who are here in Frankfurt or working in Europe you might not hear this as much, but in the US you will constantly be battered with assertions that “those public sector supervisors are not paid very much, so they cannot be very bright, so therefore they can’t possibly keep up with the whiz kids of the financial system. “ You hear this a lot in Washington or New York. Let us say that characterization is right in part. I don’t think it is justified, but let us say so for the sake of argument, as long as supervisors can be paid much more on the other side of the street. Therefore, we need to dumb supervision down to a level where any good civil servant can do it. That is what we do in medicine with the move to more nurse practitioners and hospitalers doing many initial diagnoses. That is what we do in the practice of law, as well, and increasingly that is what we do in various technical fields of chemistry and energy extraction. Though less familiar to most of us, the same shift holds true for more explicitly supervisory roles similar to banking in the regulated industries of transport, teaching, and government procurement. Your regulators and supervi-

Confronting the Reality of Structurally Unprofitable Safe Banking

109

sors do not have to be the trickiest, smartest, most expert people in the world, or even vis-à-vis the industry they are supervising. You simply have to be willing to constrain the industry enough and force sufficient disclosure such that the normal regulators and supervisors are able to do their jobs. That is not putting the cart before the horse – that is making sure that the horse is properly reined in and responsive if it is to pull the cart (and not be made into glue). I see no problem with moving to a structural change from that point of view to simpler organizations, products, and rules, and away from things like Value at Risk modelling. That leads us to the oft unspoken truth, which is the basis of my most fundamental structural critique: that traditional banking is fundamentally a money losing business except for certain suspect practices (some of which are actually illegal). Certainly a lot of commercial banking and most retail banking activities have been losing money for decades. Yet, there is very little exit of capacity from the industry. We have seen waves of consolidation through mergers, which leads to concentration in ownership (and profits and political power), but have not had true contraction of banking capacity the way we did in shipbuilding, coal mining, film developing, and other fundamentally unprofitable industries. That is where the savings and loan crisis came from in the US in the 1980s: too many fundamentally unsound institutions chasing profits when partial liberalization occurred. That really was the genesis of the banking crisis in Japan in the 1990s as well.² In both cases, you had partial deregulation of pieces of the banking system that were going to run out of money because they were completely uncompetitive. When they tried to compete, the American Savings and Loans and then the Japanese long-term credit banks gambled not on resurrection, but just on lengthening life to the duration of managers’ career windows.³ We have seen this very clearly in the UK high-street banks and mutuals, but we also see this with the strange experience of the Großbanken in Germany: traditional domestic retail and lending activities are a huge drag on profits versus their globally-oriented investment or trading activities. You cannot make a profit doing retail banking unless you are a very small local bank, and even then there are many challenges. Nothing has changed about that reality, but the post-crisis structure does not reflect this ongoing lack of profitability. If anything, post crisis

 See Ryoichi Mikitani and Adam S. Posen, eds., Japan’s Financial Crisis and its Parallels with US Experience, Peterson Institute for International Economics, 2000.  Some put the blame for the S&L crisis on the high interest rates from the Fed’s anti-inflation tightening. This misses the point. The rest of the US financial system, including institutions with outstanding long-term contracts at lower rates, survived fine, just the S&L’s did not. In Japan, the crisis took place in an environment of mildly rising rates, only a fraction as steep as under Volcker in the US.

110

Adam S. Posen

acceptance by regulators of greater concentration among banks, and pushing by supervisors for all kinds of institutions to have retail funding, exacerbates the problem in systemically risky institutions. So we have a world today in which there are ever increasing numbers of transactions that should be done directly via the capital markets (with reduced information problems), and anything that should be done in capital markets should basically be commoditized. If a product is commoditized, you cannot make any money on it. Or commercial banks have to do genuine credit appraisal and look at the actual collateral and business plan for small loans which might be profitable, and certainly serves the role society needs from some institutions. But doing so is far more expensive than other activities, and certainly does not have the upside profit potential or volume generated from securities issuance. As many of those commercial and personal lending processes get automated through credit scoring software utilizing law of large numbers, as frankly has already taken place, then that too becomes commoditized, unprofitable, and shuts out some potential good borrowers. So fundamentally, we have an underlying instability in our current financial system that we have not addressed because addressing it would require transforming the market structure of banks and other institutions. Financial conglomerates are inherently poorly monitored and governed. Disclosure and rules must be made commensurate with supervisory capabilities. More institutional variety and lower concentration are needed for systemic robustness. Most importantly, we must address the instability caused by the unprofitability of basic banking. Neither restrictions on activities as currently proposed nor capital requirements going up with size nor supposedly credible resolution threats will adequately address any of these structural problems. Think about the structure of supply and entry in energy markets, now that fracking and tight oil have transformed the cost structure. This has commoditized exploration and extraction, and thereby completely changed the nature of profitable activities in the industry. We have seen the oil majors take large losses on their spending on large oil fields and difficult extraction projects, despite billions invested in such exploration. Being a large incumbent actor with all kinds of specialized knowledge and technology used in searching for big fields does not help you compete in an industry where the money is now made by some individual person in a truck who gets a drilling bit from Schlumberger for a couple hundred thousand dollars and just goes and drills. Technology has taken energy exploration and supply back to the wild-catting days of a century ago, and removed most advantages to scale or scope. But there are no government guarantees for energy companies, so structural change and shifts in ca-

Confronting the Reality of Structurally Unprofitable Safe Banking

111

pacity are underway – and the conglomeration has no advantages for management using complexity to defeat monitoring by supervisors and owners. Underlying structure of an industry matters. Locking in the status quo in banking as much as we have done post-crisis, for all of the FSB’s efforts and Dodd-Frank, means that we are not addressing the underlying issue that banking is fundamentally an unprofitable pursuit, at least for most players. One empirical exercise which I do not think anyone has successfully done yet is to disentangle not only how much of the past financial sector profits were too-big-tofail subsidies, but then also take out how much of the remaining past profits were forms of rent-seeking or fraud. What is left? The answer is, I fear, very little for real profit. The exception would be speculative profits, which would be fine if done without depositors’ money (back to separating out the asset managers and hedge funds). So, to me, though many current regulators talk about preventing recurring crises, perhaps the source of recurring banking crises in the West are not being addressed by any of the current FSB agenda. To be fair, regulators and supervisors are perhaps making some progress on crisis mitigation. I would much rather have a 1980s savings and loans crisis than a 2008 – 10 global financial crisis and that is a worthwhile thing for policymakers to focus on. We should not kid ourselves, however, that they are getting at the fundamental source of financial instability by so doing. The fundamental source of financial instability is that it is not profitable to do normal banking. It is a commodity industry to the degree that it exists and you can’t make supernormal profits in a commodity industry except through extraordinary destabilizing means. Moreover, if one addresses this challenge structurally, more diverse, decentralized, and thus resilient financial systems will emerge, which gets you more profound mitigation of financial shocks as part of the package. In such a structural context, some financial activities are just plain public bads, especially when undertaken by banks. I do not mean just literally illegal activities. I mean the sneering references to “oh, there are people who think derivatives are terrible, but let’s not focus on such overly simple mindsets”. There are strong arguments to be made, not just intellectually empty calls by the Occupy movement, that there are a number of financial activities that are value-taking or -destroying or based on exploiting the lack of knowledge of the customer. In all of this conference’s discussions about separating activities, it misses the original intent of the Volcker Rule, and of the Vickers Commission as I understand them, i. e. prop trading by banks is genuinely bad, and should be prevented. It would not be terrible, for example, if a Mittelstand company comes to its long-time bank and says, “I need 10 million rubles for an export transaction,” and instead of the bank having an inventory of rubles on hand – because it

112

Adam S. Posen

has a department that has been speculating in currency markets – it has to say, “Okay, I’ll go buy some and charge you a fee.” I do not see the systemic threat in that. I do not even see the economic inefficiency in that. The business passes on the full cost of the transaction with a small mark-up and the other business client can decide whether or not to engage in the transaction. It is transparent and easily monitored. Sounds like capitalism! Sounds good! We really need in the banking system is a lot of shrinkage in capacity without further concentration. We have overcapacity in a lot of banking systems, and there is no profitability as a result. In the pre-crisis period, we made the mistake to believe that if you combine weaker banks, you might get economies to scale, but also you can watch them better via central supervisors and then its risk taking will be okay. Implicit was the idea that too big to fail properly supervised would mean profitability will go up. But it turns out the profitability goes up because of the political economy of regulatory capture and complexity and TBTF, not because of genuine efficiency gains. And this ends up being destabilizing rather than stabilizing. There are other industries that have undergone massive technological change. There are other industries, like tobacco or chemicals or energy or biomedical services or information technology or food for schools or automobiles, that have experienced rapid but lasting shifts in competitive structure due to changing technology. Regulation as well as businesses in these industries have been forced to adapt. Again, this notion that finance is somehow special and should not be subject to structural change without regulatory just boggles my mind. Especially since the lack of profitability is so obvious and long-standing. Sometimes we get the wishful excuse that financial instability is inherent. There are many scholars like George Akerlof and Robert Schiller coming up with complicated models of human irrationality, which emphasizes failures of logic on the consumer or saver side that will always induce asset booms and busts. There is the fatalism embodied in Reinhart and Rogoff’s This Time is Different. Let me emphasize something much more simple, which is an organizational behavior that I hope everyone in this room in familiar with. People in financial markets like to do what everyone else is doing. It is called herd behavior. Herd behavior becomes self-fulfilling. Again, one of the obvious advantages of separation of activities and structural reform is you put brakes on the extent of herd behavior. You do not allow large parts of the financial system to all go in the same direction at the same time. You do not allow people who are insurance brokers to suddenly think they are derivatives traders and vice versa. Let me turn now to the mistaken hope of the sincere supervisor. That is how I refer to improved resolution of distressed banks as a non-solution. The time ho-

Confronting the Reality of Structurally Unprofitable Safe Banking

113

rizons of bank management are really short. They often do not really care if the institution outlives their tenure as long as they get sufficient money out in the interim. And even if they did care and had a slightly longer time horizon, they tend to overestimate the worth and stability of what they’re doing. In other words, the idea that you are going to credibly threaten any banker to better behavior, especially any trader or investment banker, by the prospect of swift resolution is to me a false hope. And I believe this does nothing to reduce the incentives for regulatory forbearance, because the downside risk to supervisors comes from displeasing superiors and the public, and foreclosing future private-sector employment, not from messy failures. Probably resolution policy does have some uses in the sense of crisis mitigation for a Dexia or a Lehman Brothers, which is nice to have. But if we are stepping back further to ante exante, then resolution is a second order response at best. It will not fundamentally change the competitive structure, nor will it change the incentives of any of the people who are moving from bank to bank to bank in search of new bonuses. Lack of resolution has all too often been used as an excuse for inaction, though it is not a cause of crisis. As deep a respect as I have for leadership of the Federal Reserve during the crisis, I continue to be offended by the repeated claim that between the demise of Bear Stearns in April and the collapse of Lehman Brothers in September 2008, they were trying very hard to figure out something else to be done, but they just couldn’t because they did not have the legal authority for resolution. Those of us watching in real time, pointing out the imminent end for Lehman and Merrill Lynch immediately following Bear’s takeover, and referring to actions taken during the S&L crisis, know that was not the case. There is an old Hollywood saying, which is that when someone tells the studio head, “the lawyers tell us we couldn’t do that,” the boss replies “Well, then get me another lawyer.” As in all preceding US financial crises, use supervisory powers and even make closures as needed. Let the parties involved sue for restitution later- which in the US they have of course done anyway over both Bear and Lehman and a host of others. That said, the place where I am the most sympathetic to my friends in the banking industry, as far as their legitimate complaints with recent regulation, is that compliance has become an obscenely huge burden without public benefit.⁴ Even when sitting on the Monetary Policy Committee at the Bank of England, with no supervisory role, I saw the cost of enhanced regulatory compliance show

 Insurance companies and asset managers have a different set of legitimate complaints, namely being regulated and supervised as though they were banks, when they are not. But that is a topic for another day.

114

Adam S. Posen

up in the national GDP data and the productivity trends in various industries, not just banking. Of course, the financial services industry is a major industry in the UK. When employment in the industry is actually going up on net rather than down as a result of the crisis, something is wrong. This happened because 30 % of the banks’ workforce is being either reallocated to or replaced with compliance work. Compliance is, at best, an insurance payment but it looks like a deadweight loss in the absence of crisis. When the financial system is 12 % of GDP, and you reallocate 30 % of that industry’s work force to unproductive activities, you start adding up to several tenths of a percent of GDP growth lost annually. But that just brings us back to where I started by advocating reduction of financial conglomeration and complexity, making activities amenable to supervisors, addressing the unprofitability of traditional banking, and so on. Our real focus has to be about the will and capability to supervise. What we have seen repeatedly even from very well intentioned and trained supervisors is that when we try to align the incentives is it is never time-consistent to supervise strictly. It is very difficult to sustain any kind of strictness during or after a crisis. Credit booms and cognitive capture take place and even people who you would think would never succumb become loose and lax supervisors over the cycle. That brings us back to something I have called the quest for big dumb rules – and I mean every part of it. Big: the rules extend across all the banks restricting activities in a blunt way. Dumb: they are straightforward and visible, linked to structures, so people can follow them and interpret them. Rules: they give very little discretion to bank supervisors because then they cannot renege. I’m not an idealist or fantasist so I know that most of what I have talked about is completely off the table – for now. But when we get our next financial crisis, it will not be because we have insufficient resolution authority. And the crisis will not come because there is going to be another bubble due to consumers, and certainly not because monetary policy was too loose. We will have our next crisis because we will have done nothing to address the fact that much of banking is in its current form is fundamentally unprofitable. So it must become largely a rent-seeking or speculative activity, when regulation limits exit from the business. That is abetted by overcomplicating the governance of these institutions making then difficult to supervise and allowing concentration to proceed despite no economies of scale or scope beyond too-big-to-fail subsidies. I hope we can refocus our regulatory reform agenda in time.

Douglas J. Elliott

Ten Arguments Against Breaking Up the Big Banks I have been asked to present the key points that advocates of breaking up the big banks leave out or minimize. Let me be clear that I am, broadly speaking, a strong advocate of the substantial financial regulatory reforms that are included in the Dodd-Frank Act in the US, its equivalents in Europe, and the Basel 2.5 and Basel III agreements to increase the quality and quantity of capital and of liquidity at the banks on a global basis. I do have some disagreements with the details of these reforms; no sane and careful analyst would have designed the reforms exactly as they came out of the political and bureaucratic processes and there is plenty of room for analytical differences on these policy issues. But I am certainly not writing this piece as an opponent of reform. Indeed an underlying reason that I do not support the radical solution of breaking up the big banks is that the myriad of other reforms constitute a better way to deal with the problems that became clear in the global financial crisis. Further, I want to acknowledge that advocates of structural reforms in the banking industry do have some reasonable points. I believe that, on balance, the arguments against breaking up the banks far outweigh those in favor, but that does not mean I think there are no valid arguments in favor. The mandate of this paper is to present the arguments against, not to give a balanced view of the debate. For those who wish a more balanced view within a single paper, I recommend Andreas Dombret’s contribution to this book. Finally, I will be addressing both the proposals to break up the big banks based on size and those based on function, such as restoring Glass-Steagall to its full original scope, where it forbade US investment banks from being owned by groups that also owned commercial banks. In practice, advocates of break-up often meld the two proposals together, explicitly or implicitly, so it is difficult to address one without the other. With those caveats, here are 10 arguments against breaking up the big banks.

1. The experience of the recent crisis gives little support for break-up Proponents of breaking up the largest banks often explicitly or implicitly argue that the global financial crisis demonstrates the need to do this. To me, the US

116

Douglas J. Elliott

experience shows the opposite. The financial institutions that failed, or were rescued just short of failure, were predominantly fairly pure investment banks or fairly pure commercial banks, not the combined groups that break-up advocates warn against. Bear Stearns, Lehman, and Merrill Lynch were investment banks with relatively minor traditional commercial bank activities. For their part, Countrywide and WAMU were traditional banks with little or no investment banking activities and Wachovia was clearly much more focused on commercial banking, although it had investment banking activities of some scale. There were only two large banking groups that combined commercial and investment banking in a serious way and which required special support beyond that provided to the banking industry as a whole: Bank of America and Citigroup. However, Bank of America is not an example of such a group, pre-crisis. It was only their purchase of Merrill Lynch during the crisis that gave them a large investment banking arm to go with their traditional banking activities. Therefore, it is not fair to cite them as an example of a firm that got into trouble because of how they operated as a mixed entity. Instead, it is an example of a questionable gamble on a large acquisition in the middle of a financial crisis. The one real counter-example is therefore Citigroup and even there I would argue that its diversification helped preserve it through the crisis rather than being an underlying cause of its problems. There are indirect arguments that the existence of megabanks helped create the crisis, but they rely on difficult to prove assertions about the likely level of risk-taking in the counter-factual case where the banks had been broken up in advance of the crisis, arguments that I will address later. It is possible that a future financial crisis will have different characteristics, but it seems to me a powerful point that the global financial crisis shows little evidence that diversification across commercial and investment banking was a problem; in fact the evidence points in the other direction.

2. The key types of financial activities are intrinsically intertwined Advances in financial theory and practice, and vast improvements in computers and communications, have transformed finance from the simpler days to which many hearken back as an ideal. At this point, many underlying financial transactions can be structured to be a loan, a security, a derivative, an insurance contract, or another type of contract. Artificially dividing between these different structures is both difficult and often counter-productive. Most importantly, cor-

Ten Arguments Against Breaking Up the Big Banks

117

porations often have the choice of borrowing funds through a loan or through a debt offering. Further, they can transform the maturity or the fixed/floating rate nature of the loan, or the currency of payment, through derivatives, in order to lower their funding costs or reduce their risk. I have not observed any significant support in the corporate world for the proposition that banks should be forced to specialize between loans and securities activities. Quite the opposite, the corporate CFO’s and Treasurers that I know would actively oppose that change. Further, it is not remotely clear how this would be done without making very artificial distinctions that would practically beg for regulatory arbitrage activity. For example, large loans are now almost all structured so that they can be traded. If artificial divisions made it preferable for these to be securities instead, it would not be hard to do so. Or, if it were better for regulatory reasons to have a security be a loan, that would be easy as well, by tweaking a few of the terms. This is a very important example, since the traditional division in GlassSteagall is between lending and securities activities and it is really unclear to me how one would even divide those two types of activities.

3. The movement towards integration in finance is largely customer-driven I was an investment and commercial banker at JP Morgan for many years. During my time there, it was clear that there was strong pressure from clients for the bank to expand the range of its financial services, especially to be as strong in investment banking as it was in commercial banking. Many seem to imply that banks expanded their range of activities out of the greed of the bankers, but my own observation argues the opposite, for the most part. For example, when I first joined JP Morgan, in 1985, it had recently made the strategic decision that it could either follow its customer base into the securities markets by building investment banking services or it could continue to provide only its commercial banking services, but have to find a new customer base of smaller firms. They concluded that their customer relationships were more valuable than their historic position as a commercial bank. I have little doubt that the managers of the time would have preferred not to make the uncomfortable transition to investment banking if it had not been a strategic necessity. A survey by the Business Roundtable last year showed that CEO’s of large US corporations place a great value on the existing megabanks, which matches my own experience of dealing with customers and of talking to corporate Treasurers

118

Douglas J. Elliott

and CFO’s over the years. If we are truly trying to design a financial system that serves society well, it behooves us to seriously consider what the customers of the banks are telling us they want.

4. Financial markets are very important and banks are a key part Some of the arguments for separating traditional banking from securities and derivatives business implicitly assume that financial markets are less important than banks. The first problem with this argument is that securities markets, and the firms that specialize in them, can clearly be of crucial systemic importance, as we saw in the US with Bear Stearns and Lehman. So, it would be a mistake to focus solely on protecting the banks without regard to the effect on markets and the players in those markets. It sometimes seems to me that financial markets are bearing much of the brunt of financial regulatory reforms that are intended to protect the banks, such as the Volcker Rule or the original version of the counter-party credit limits proposed by the Federal Reserve. Another example is the adoption of a supplementary leverage ratio in the US that will make it much harder for banks to own and trade many low-risk securities. One of the main thrusts of structural reform proposals, including the Volcker Rule, the Vickers Commission proposal and the Liikanen proposal, is to distance securities activities from deposit-taking. However, deposits provide a great deal of stability to those firms that have access to them and their associated protections, such as deposit insurance and lender of last resort facilities. We could find ourselves in a situation where deposit-taking institutions become more stable in the first instance while financial markets become considerably more volatile and less stable. The result could be a net decrease in the stability of the banks due to their many continuing ties to these new, less stable markets. There are many ways that a crisis in financial markets could spread to traditional commercial banks, even when they are no longer in the same ownership group as an investment bank. Banks own many financial assets that are traded in markets or whose value goes up in down in tandem with traded securities. Further, many of their customers are strongly exposed to financial market conditions, over and above the important effects of markets on the economy as a whole, which clearly impacts banks. Perhaps even more importantly, financial markets provide a great deal of the funding used in our economies. In the US, banks only provide about a third of the credit provided to the economy and most of the rest is directly or indirectly

Ten Arguments Against Breaking Up the Big Banks

119

provided by the markets. Europe is more bank-based, but markets still matter and are likely to matter more over time. If we make financial markets more volatile and less liquid, this will have costs for the economy as a whole. Thus, I believe there are many potential costs to forcing the big banks to pull back from securities and derivatives activities.

5. Traditional banking is not necessarily safer One of the main arguments advanced for the Volcker Rule and other structural reforms is a cultural one. Proponents argue that forcing a separation of commercial and investment banking activities will allow traditional banking to shake off a culture of risk and greed that investment banking brought to the combined groups. Proprietary trading is often portrayed as an extreme case of a risky culture luring bankers away from their traditional prudence by the example of making seemingly easy money by taking excessive risk. The biggest flaw in this argument is that banks have not historically been all that prudent. History is replete with examples of banking crises that arose solely in the traditional lending activities. Continental Illinois was brought down by greedy loans made to the oil patch in the US when it appeared that no one could lose money on energy loans. They arranged most of these loans through a bank in Oklahoma called Penn Square that became a byword for crazy banking practices. The Latin American debt crisis of the 1970’s and 80’s involved straightforward bank lending, mostly to sovereigns. Subprime loans that helped lead to the global financial crisis were straightforward loans generally made by the traditional commercial banks. Some of them were created with securitization in mind, but banks made plenty of bad loans that they kept on their own books. The Cajas in Spain were traditional lenders who got carried away and they in turn sourced much of their funds from traditional German and French banks. Commercial bankers have the same human failings as investment bankers. I have worked extensively with both types of people and they, like virtually everyone, respond to the incentives they face. The key therefore is to ensure the right incentives, not to keep traditional bankers away from the baneful influence of traders and investment bankers.

120

Douglas J. Elliott

6. Existence of economies of scale and scope in banking Proponents of breaking up the big banks generally believe that there is little economic reason for a financial institution to be very large, other than to take advantage of unfair benefits of size, such as the potential to be viewed as Too Big to Fail and thereby benefit from an implicit government guarantee. There are two fundamental ways in which any firm may benefit from large size. First, they may find “economies of scale”, meaning that it is cheaper per unit of activity to do a lot of something than a little. For example, there may need to be a very considerable cost to create a computer system or to build an industrial plant of an efficient size, but ramping up production can spread the fixed costs over more units. There are areas in banking that clearly work in this manner, particularly transactional activities that require major investments in computer systems. But, the benefits can show up outside of information systems. It takes a certain scale of activity before it becomes efficient to hire an expert on a certain market or particular type of loan or investment, and loan officers are more efficient if they have more customers in a specific industry or geography. Second, there can be benefits from conducting a range of activities, known as “economies of scope.” Gas stations have become convenience stores because of the additional sales possibilities, just as airports have become malls. In finance, there are clear advantages to being able to offer a wide range of services, which are easiest to illustrate with regard to corporate customers. For instance, one reason that banks value mergers and acquisitions assignments so much is that they produce a wide range of related transactions which the bank is then in the best position to provide. There will be a need for financing, perhaps foreign exchange transactions, risk management products such as derivatives, and various operational services. It is most convenient for everyone to have the bank provide many of these services in combination. The original academic work on economies of scale in banking suggested that the benefits topped out at about $100 billion in assets, much smaller than the size of the largest US financial institutions. More recent academic work has been considerably more sophisticated and generally finds economies of scale going much higher. The Clearing House Association, an industry body, issued

Ten Arguments Against Breaking Up the Big Banks

121

a study that surveyed existing research,¹ concluding that large banks in the US provide a series of important economic benefits that would not be achievable to the same extent by smaller banks. The study found that the existence of big banks created $20 to $45 billion a year in benefits from economies of scale, figures they reached by looking in detail at various areas of banking where one would expect scale to matter. Similarly, benefits of a wide scope of product offerings at these banks and their affiliates produced another $15 to $35 billion a year of economic benefits. Finally, benefits that big banks provide through encouragement of financial innovation came to another $15 to $30 billion, bringing the total to $50 to $110 billion a year. The study argued that these findings were relatively conservative, pointing to a finding by Wheelock and Wilson of the Federal Reserve Bank of St. Louis² that even capping banks at $1 trillion in size, a relatively high limit compared to some proposals, would result in a loss of $79 billion in benefits. The Wheelock and Wilson study, and a Federal Reserve of Philadelphia working paper by Hughes and Mester,³ find economies of scale going up to much larger sizes than earlier studies. This is partly because the newer studies are substantially more sophisticated and partly because it appears that economies of scale in US banking have actually grown as technology and its application have improved and geographic and other legal barriers have been reduced. (The earlier studies are mostly a decade or so older.) Moreover, both these newer studies and the older ones understate the underlying economies of scale by treating compensation costs as mandatory, whereas there is an element of profit-sharing in them. Ronald Anderson attempted to estimate the economic “rents” captured by the managers of financial institutions and concluded that they have been substantial. Managers have presumably siphoned some of the benefits of economies of scale and scope off, but they are nonetheless economic benefits that need to be considered. The various economic pressures being placed on the financial industry in response to the financial crisis are likely to squeeze much of these rents back out, reallocating them to shareholders and customers.

 The Clearing House Association, “Understanding the Economics of Large Banks,” November 7, 2011.  Wheelock, David C. and Paul W. Wilson, “Are U.S. Banks too Large,” Federal Reserve Bank of St. Louis Working Paper, December 2009.  Hughes, Joseph P. and Loretta J. Mester, “Who Said Large Banks Don’t Experience Scale Economies? Evidence From a Risk-Return-Driven Cost Function,” Federal Reserve Bank of Philadelphia Working Paper, July 2011.

122

Douglas J. Elliott

7. Big crises tend to be created by causes of a sweeping nature One of the reasons that I reach different conclusions from advocates of breaking up the banks is that I believe the contribution of bank size and scope to the financial crisis is small compared to the underlying causes. To hear some advocates, it sounds as if the existence of megabanks accounts for half or more of the crisis, whereas I suspect the largest reasonable percentage one could ascribe would be 5 – 10 % and that would only be if you disregard all of my arguments as to the offsetting benefits of big banks and assume they only increase risk. I believe that the recent financial crisis was much more about system-wide problems than about issues resulting from the excessive size of financial institutions. A simple thought experiment illustrates this argument. If we had broken up the big banks a decade ago into 10 or 20 pieces each, the great bulk of the problems of the crisis would still have occurred. There would still have been excessive investments in both residential and commercial real estate, financial institutions would still have operated with far too little capital and liquidity, opaque and excessively complicated securitizations and derivatives would still have been in vogue, compensation structures would still have encouraged excessive risk-taking, markets would have been much too cavalier about the risks that were taken, US government policy would still have pushed homeownership too hard, monetary policy might still have been too loose, etc. In fact, it is difficult to find many significant ways in which the crisis would have been different. Indeed there is a chance that the US clean-up would have been more difficult without the ability to pull 17 key CEOs into a room and force them to accept the TARP arrangements.

8. Broken-up banks may not be much easier to manage Advocates of bank break-up often argue that the megabanks are too big to manage. It is clear that the largest banks are capable of making very large mistakes in absolute dollar terms, such as the $6 billion of losses from JP Morgan’s “London Whale” episode. It also seems intuitive that it must be harder to manage a very large bank than one that is smaller. Many then take the step of arguing that the largest banks are too big to manage and that breaking them up would improve

Ten Arguments Against Breaking Up the Big Banks

123

decision-making. A similar argument is made for supervision of these banks, that their sheer size and complexity makes it too difficult to follow their actions. However, this is actually a much more complicated set of questions than it might appear and there is little or no empirical evidence about the quality of management decisions. It is important to remember that we are not talking about the difference in complexity between a big bank and a small one. In considering breaking up the largest banks, we are comparing the aggregate management problems of, say, five successor banks, all still of quite large size, as compared to the existing megabank. If the five successors have the same range of activities, just each at one-fifth of the scale, then it may not be much easier to manage any one of them. In both cases, the top managers will be managing other senior managers who will be managing middle level managers who will be managing a next tier of managers, etc. The range of issues will still be too broad for the most senior managers to understand in depth each of the areas and they will remain reliant on the expertise of those under them. Even splitting the banks between commercial and investment banking activity would still leave each piece highly complex and large. It is also possible that the average level of management expertise of CEOs and other senior managers is positively correlated with the size of their institutions. One would expect this to at least some extent, since career advancement is almost always in the direction of moving to larger firms that pay better and allow an individual to make a greater impact as well as garner more prestige. If so, then the average level of management expertise of the CEOs of the five successor firms should be expected to be lower. It is at least theoretically possible that the impact of worse CEOs would have a significant effect on management’s effectiveness.

9. Implementing a break-up of the banks would be difficult Few advocates of breaking up the banks have come forward with concrete suggestions for how it ought to be done. There are a host of questions that would need to be addressed by any detailed plan, including: – What will determine whether a bank should be broken up? – What asset size is acceptable for a bank? – Are only domestic assets counted? – Who will decide how a banking group is broken up? – Will banks be allowed to grow above the size threshold?

124

– – – –

Douglas J. Elliott

Will the threshold be adjusted over time? Who decides where different customers end up? Who decides where financial commitments, including deposits and debt, belong? Over what period should a break-up be implemented?

These are all challenging questions and it will be difficult to become comfortable with break-up proposals without knowing the answers, since bad choices could compound the costs and decrease any safety benefits.

10. There would be many transition issues, whatever the ultimate outcome There are a series of transition costs or issues that would face the financial system and likely impact the wider economy, including: – A loss of customer focus. – Disruption of customer relationships when the break-up occurred. – Potential mistakes by successor firms as they gain their sea legs. – Litigation against firms and their regulators. – Funding difficulties, and higher cost of debt, until successor firms are clearly established. – Movement of activity to less regulated, and potentially riskier, firms. If there were a compelling case for breaking up the banks, then the right answer would be to find a way to minimize the costs and risks of these transition problems while proceeding forward. However, these are not minimal issues, and problems could last for a number of years, so I do believe they represent a significant negative in their own right.

Summary Although there are some arguments for breaking up the big banks, the arguments against are considerably more compelling to me and certainly merit more discussion and analysis than they typically receive.

Randall D. Guynn and Patrick S. Kenadjian

Structural Solutions: Blinded by Volcker, Vickers, Liikanen, Glass Steagall and Narrow Banking The phrase “structural solution” is a polite term for breaking up the banks. These “solutions” are presented as the key to solving the problem of “too big to fail” (TBTF).¹ They come in three main varieties, one in the United States (the Volcker Rule), one in the United Kingdom (the Vickers Report) and one in the European Union (the Liikanen Report). The problem they seek to address is real. It is the dilemma faced by public authorities when confronted with the potential failure of a financial institution that could destabilize the financial system in a country or region in the absence of appropriate tools to recapitalize or wind down the institution without such instability. Enormous amounts of public resources were devoted to supporting the financial system in 2008/2009. In some cases this support resulted in increasing public sector debt to levels that make a repeat of public support on a similar level hard to conceive of even if public opinion would tolerate it. Thus, solving TBTF is crucial both to the financial sector and to the common good. But the “structural solutions” proposed are illusory, at best are complementary to other more targeted measures being actively pursued, and at worst divert valuable attention and resources away from more targeted solutions.² They are perhaps best understood as threats of what could happen to financial institutions if these other measures are not implemented in a timely and credible manner, rather than as real solutions to the TBTF problem.

 For a good description of the TBTF problem and why it arises, see John F. Bovenzi, Randall D. Guynn & Thomas H. Jackson, Too Big to Fail: The Path to a Solution, Report of the Failure Resolution, Report of the Failure Resolution Task Force of the Financial Regulatory Reform Initiative of the Bipartisan Policy Center, May 2013, pp. 1– 2.  In its recent report to the G-20 Leaders on structural banking reforms, the Financial Stability Board was carefully neutral on the advisability of the proposed reforms, but does cite a number of concerns raised by jurisdictions which have not adopted such reforms relating to the spillover effects that the reforms may have. Prominent among these concerns is the potential for interfering with the efficient resolution of financial institutions if liquidity or capital are trapped in domestic silos and cannot be freely used for resolution. In particular, the strategy of resolution through the ’single point of entry’ approach could be impaired. Financial Stability Board, Structural Banking Reforms: Cross-border consistencies and global financial stability implications, Report to G20 Leaders for the November 2014 Summit, October 27, 2014.

126

Randall D. Guynn and Patrick S. Kenadjian

The more targeted measures include increases in the capital and liquidity requirements decided upon by the Basel Committee on Banking Supervision in the package of reforms known as Basel III and the development or expansion of resolution regimes and strategies such as the single-point-of-entry (SPOE) strategy in the US, the UK, France, Germany and at the Community level in the EU.³ These are the real solutions to TBTF because they aim to reduce both the risk of failure of financial institutions, based on an analysis of what caused the failures in 2008/2009 (through the Basel III reforms), and the loss given failure through the adoption of appropriate resolution regimes which would allocate losses upon failure to existing investors in the failed institutions without fostering contagious panic, thus preventing the failure of one institution from having systemic consequences. They are solutions endorsed by the Financial Stability Board acting to carry out decisions of the G-20 heads of state and governments on a coordinated worldwide basis. Their adoption will increase uniformity of regulation and outcomes worldwide, reducing the chances for regulatory arbitrage and contributing to financial stability. The structural solutions, in contrast, are ad hoc regional initiatives, in part contradictory, which will lead to fragmentation and to a more brittle financial system. They are based on theory, ideology, wishful thinking and nostalgia. They will not work and will in fact be counterproductive by diverting valuable time and resources away from real solutions. The term too big to fail predates the crisis of 2008/2009. It was coined in the 1980s in the United States as a result of the rescue of Continental Illinois Bank and Trust, but it is fair to say that it was not top of mind among bankers and regulators in the run up to the financial crises until March and September 2008. It was revived then to describe the problem created by the rescue of a non-bank financial institution, Bear Stearns, in March 2008 and the failure to rescue another non-bank financial institution, Lehman Brothers, in September 2008. The first action created an expectation that financial institutions above a certain size, regardless of whether they were deposit taking institutions as Continental Illinois had been, or pure investment banks as Bear and Lehman were, would not be allowed to fail. Bear was rescued for fear of what its failure would do to financial stability. The failure to rescue Lehman illustrated how severe those consequences could be. Lehman’s failure was followed in the US by the

 For a good description of the SPOE strategy, see Bovenzi, Guynn & Jackson, note 1 above, pp. 23 – 32. For a description of how the SPOE strategy was developed, and how quickly it has been accepted around the world as the most promising solution to the TBTF problem, see Randall D. Guynn, Framing the TBTF Problem: The Path to a Solution, in Across the Great Divide: New Perspectives on the Financial Crisis, Joint Publication by the Brookings Institution and the Hoover Institution, Edited by Martin N. Baily and John B. Taylor, 2014.

Structural Solutions

127

rescue of an insurance conglomerate, AIG, and of the money market mutual fund industry in general. None of these entities had insured deposits or access to the Federal Reserve System’s normal “lender of last resort” authority – the Discount Window. None of them ran a payments system or provided substantial amounts of credit to consumers or to small and medium sized enterprises outside the debt capital markets. Other than perhaps AIG, none of the entities involved would have been a candidate for being labeled a global systemically important financial institution, i. e. a G-SIFI. Based on this historical record, it is hard to understand how we arrived at the point where it is seriously argued that too big to fail is a problem caused by global systemically important banking groups, i. e., G-SIBs, and that the way to solve it is to separate deposit taking and lending from trading or other financial activities, especially since the degree of structural separation proposed in Europe by Vickers and Liikanen does not go beyond the structural separation in effect in the US in 2008 under Sections 16 and 21 of the Glass-Steagall Act, which survived the repeal of the rest of the Glass-Steagall Act in 1999.⁴

The Varieties of Structural Reform There are three main strains of structural solutions which have made it to international prominence, one for each of the US, the UK and the EU, showing that no one country or region has a monopoly on bad ideas. Since they are ably described and defended by other contributors to this volume, we will mention only a few of their salient features useful to convey our arguments in this chapter. The first in time is the Volcker Rule, technically section 619 of the Dodd-Frank Wall Street Reform and Consumer Protection Act of 2010, named after its progenitor, the highly respected former Chairman of the Federal Reserve System, Paul Volcker. It focuses on prohibiting deposit taking institutions and their separately incorporated non-deposit-taking affiliates from engaging in proprietary trading (but not long-term investing or market making) in certain financial instruments and from acquiring or retaining ownership interests in, sponsoring, or entering into certain lending and other covered transactions with related, hedge funds, private equity funds and many other vehicles (including most securitization vehicles) defined as “covered funds.” The Volcker Rule requires these activities to  As a purely technical matter, the Glass-Steagall Act consisted of Sections 16, 20, 21 and 32 of the US Banking Act of 1933. It has become common usage, however, to refer to these sections as sections of the Glass-Steagall Act, so we have adopted that convention in this chapter even though it is not technically accurate.

128

Randall D. Guynn and Patrick S. Kenadjian

be pushed out of banking groups entirely. As a result, we refer to it as a total separation solution, in contrast to Vickers or Liikanen, which are only partial separation solutions. Vickers and Liikanen do not require investment banking to be pushed out of banking groups entirely, but only out of deposit-taking legal entities into separately capitalized non-bank affiliates. The total separation required by the Volcker rule is not based on empirical evidence that either proprietary trading or investing in, sponsoring, or entering into covered transactions with related, hedge funds or private equity funds had any significant part in the failures of financial institutions in 2008/2009.⁵ Instead, it is based on a theoretical or cultural, not to say ideological view of what activities are appropriate for officially licensed deposit-taking institutions and their separately capitalized non-deposittaking affiliates. The second in time is the Vickers Report (technically the report of the UK Independent Banking Commission chaired by another respected former central banker, Sir John Vickers). It also seeks to separate trading activities from commercial banking, but allows the two to co-exist within the same banking group, so long as they are in separately capitalized entities which may only transact with each other on an arm’s length basis. This structural separation regime is similar to the partial separation regime that was in place in the United States after enactment of the Gramm-Leach-Bliley Act of 1999 but before the Volcker Rule. It adds a requirement that the entity engaging in commercial banking activities be subject to higher capital and lower leverage requirements than otherwise applicable under the Basel III standards and narrows substantially the activities in which this “ring-fenced” commercial bank may engage. Such a bank is to focus primarily on consumer, real estate and small and medium sized entity financing within the European Economic Area. While it may also engage in lending to larger corporate clients and deal in certain simple derivatives products, the “ring-fenced” bank cannot offer any more sophisticated products to its customers. The non-ring-fenced entities cannot take deposits from EEA individuals or SMEs or operate payment systems. The ring-fenced entity will thus have higher  Indeed, Representative Jeb Hensarling, the Chairman of the Financial Services Committee of the U.S. House of Representatives has repeatedly called the Volcker Rule “a solution in search of a problem.” See, e. g., The Impact of the Volcker Rule, Hearings Before the Committee on Financial Services, U.S. House of Representatives, January 15, 2014. The Volcker Rule was a last-minute addition to the Dodd-Frank Wall Street Reform and Consumer Protection Act of 2010 and has no clearly articulated purpose in the statutory text or legislative history. It is also impossible to infer any coherent purpose from the text since the Volcker Rule prohibits certain low-risk activities (e. g., short-term proprietary trading in highly liquid financial instruments) and permits certain high-risk activities (e. g., long-term investment in highly illiquid instruments such as commercial loans).

Structural Solutions

129

costs (due to the capital and leverage requirements), a focus on traditionally risky consumer, real estate and SME lending and be cut off from most sources of often more lucrative and sometimes less risky fee business. The fee business will be in the non-ring-fenced entity which can be part of the same corporate group but which will be deprived of access to an historically stable retail deposit base and thus will have to finance itself through the historically less stable capital and repo markets. What is striking about this “solution” is that the ostensibly safer ring-fenced entity will be required to focus on areas of lending, consumer, real estate and SMEs, which have traditionally been viewed as risky and carried correspondingly high risk weightings under the Basel system for risk capital. It was real estate lending which was the downfall of Northern Rock in the UK and Washington Mutual and Wachovia in the US⁶. And it was the lack of a stable deposit base, and reliance on funding from the capital and repo markets, which led to the downfall of Bear Stearns and Lehman Brothers in the US. Northern Rock’s failure was also triggered by an excessive reliance on capital market funding. Furthermore, the ring-fencing solution does little more than recreate the conditions which applied in the US in 2008/2009 under Sections 16 and 21 of the Glass-Steagall Act, which survived the repeal of the rest of the Glass-Steagall Act by the Gramm-Leach-Bliley Act of 1999 and which required separation of deposit-taking institutions and investment banks under a common bank holding company. These entities could only transact business on an arm’s length basis. However, this separation did not prevent a series of failures on both sides of the divide. As Thomas Huertas, former Alternate Chair of the European Banking Authority notes in his recently published book “Safe to Fail”: “This set-up has not brought financial stability to the United States. In the crisis, stand-alone investment banks failed, single purpose commercial banks failed and diversified financial holding companies failed. There is no conclusive evidence to suggest that the separation of investment and commercial banking limits risk, fosters safety or enhances stability.”⁷ Third in time is the report of the Liikanen EU High Level Expert Group in October 2012, which became the basis for a proposed directive by Commissioner Barnier in 2014. Liikanen, like Vickers, focuses on a separation of traditional commercial banking from trading within a single banking group, but includes a de minimis exception from mandatory separation where trading makes up a small percentage of the banking entity’s activities and focuses more on prevent Barth, J.R. and D. McCarthy, Trading Losses: A Little Perspective on a Large Problem, Milken Institute, November 2012.  Thomas F. Huertas, Safe to Fail, How Resolution Will Revolutionise Banking, Palgrave Macmillan, 2014, p. 42.

130

Randall D. Guynn and Patrick S. Kenadjian

ing insured deposits from being used to finance proprietary trading and on the ease of resolution of separated entities than on the dangers of trading per se. The Barnier proposal for an EU directive combines elements of Liikanen and Volcker, rather like the Chinese restaurants in the US of our youth where one could choose one dish from column A and one from column B on the menu. In the background of these three proposals can be found an older idea, that of “narrow banking” which has its origins in a proposal made in 1933 to President Franklin D. Roosevelt by a distinguished group of economists from the University of Chicago, including Frank Knight and Irving Fisher, and known as the “Chicago Plan”. It would have split banks’ two main functions, taking deposits and making loans. Banks would have had to keep 100 % of their deposits readily available for withdrawals. Lending would have been funded by equity or bank liabilities other than insured deposits or otherwise left to private investors. The US Congress rejected narrow banking in favor of deposit insurance as a better way to steer against bank collapses,⁸ but the idea of narrow banking has resurfaced since the financial crisis. The current incarnation of the idea has been most closely associated with John Kay in the UK and Lawrence Kotlikoff in the US.⁹ While their prescriptions differ in detail, the general idea is to create institutions which, in the words of John Kay, separate the “utility” functions of running payments systems and taking deposits from individuals and small and medium sized enterprises (“SMEs”), from the “casino” functions, basically most everything else they do, including anything to do with investment banking. Banks would be required to invest 100 % of their deposits primarily if not entirely in cash and government securities. There are differences among these proposals in exactly what else banks would be allowed to do (Kotlikoff sees them as providing mutual fund style investments to their customers) and the extent to which the activities cast out from within the banks would need to be regulated (Kay is not concerned with that at all; Kotlikoff is somewhat more concerned). However, all variations have in common pushing activities viewed as risky out of the bank and radically narrowing the scope of bank activities.¹⁰

 Narrow-minded, A radical proposal for making finance safer resurfaces, The Economist, June 7, 2014.  John Kay, Narrow Banking, The Reform of Banking Regulation, September 15, 2009; Lawrence Kotlikoff, Jimmy Stewart is Dead: Ending the World’s Ongoing Financial Plague with Limited Purpose Banking, John Wiley & Sons, 2011.  US Senators Elizabeth Warren, John McCain, Maria Cantwell and Angus King have introduced a narrow banking bill in the United States Senate. See http://www.warren.senate.gov/ files/documents/21stCenturyGlassSteagall.pdf. Although they have called it the 21st Century Glass-Steagall Act of 2013, it is substantially more restrictive than the Glass-Steagall Act of

Structural Solutions

131

Since none of these proposals has been adopted as legislation or regulation we will not linger further on them, except to make three short remarks. First, tying the credit of banks to the credit of their sovereigns turned out to be dangerous in the Euro debt crisis of 2010 – 2012. Second, these proposals present many of the same definitional problems we will discuss below concerning the three main “solutions”. For example, why is consumer and SME lending, traditionally viewed as risky, suddenly a utility function, less risky than underwriting highly rated corporate bonds? Why is that activity not as much a utility? Third, we would like to cite Charles Goodhart’s acute observations on narrow banks: “A problem with proposals of this kind is that they run counter to the revealed preferences of savers for financial products that are both liquid and safe, and of borrowers for loans that do not have to be repaid until some future date. It is one of the main functions of financial institutions to intermediate between the desires of savers and borrowers, i. e. to create financial mismatch. To make such a function illegal seems draconian.”¹¹

Martin Wolf who, over the years, has shown a certain sympathy for the arguments of John Kay, has noted that any fragility such proposals would banish from the banking system would be recreated outside of them,¹² and concluded in his recent book that however fascinating it might be to see one of these proposals enacted, “the difficulties involved in making such a transition would be huge. So let’s first consider less radical ways of buttressing a system that would be much more like our own.”¹³ To return to the three less radical structural solutions that have made it out of the starting blocks, they all target activities (trading, hedge funds and private equity funds) without explaining how these contributed to the last crisis. Vickers and Liikanen would recreate the same situation which existed in the US under Sections 16 and 21 of Glass-Steagall (as left in place by the Gramm-Leach-Bliley Act) on the eve of the 2008 financial crisis when Washington Mutual and Wachovia collapsed and Citigroup and Bank of America had to be rescued, without ex1933, both in terms of more extensive restrictions on the direct activities of banks and on their ability to affiliate with companies engaged in securities and other financial activities. The original Glass-Steagall Act only prohibited banks from being affiliated with securities affiliates that are “engaged principally” in underwriting and dealing in corporate securities. See Section 20 of the US Banking Act of 1933.  Charles A.E. Goodhart, The Optimal Financial Structure, LSE Financial Markets Group Paper Series, Special Paper 222, March 2013, p.5.  Martin Wolf, Why narrow banking alone is not the finance solution, Financial Times, September 29, 2009.  Martin Wolf, The Shifts and the Shocks, What we’ve learned – and still have to learn – from the financial crisis, Penguin Books, 2014, p.237.

132

Randall D. Guynn and Patrick S. Kenadjian

plaining why, if that degree of separation did not avoid collapse or the need to rescue those institutions in the US it will fare better in Europe. They also do not bring any evidence of internal contagion within a banking group from the investment banking and trading side to the commercial banking side. Vickers will clearly lead to creating palpably weaker institutions carved out of existing groups, with the ring-fenced banks making what have traditionally been considered the riskiest loans and the investment banks deprived of the support of a stable deposit base. Moreover, none of these three structural changes do anything to prevent trading and hedge fund activities from migrating to the shadow banking system and for the shadow banking system to compete for the same retail funds that might otherwise be stored in deposit accounts and used for payments through the official banking system by offering money-like substitutes such as mobile payments, BitCoins, google money or other demand or short-term credit instruments to the public. Thus, the three structural changes do little more than create a financial Maginot Line that existing and future shadow bankers will just drive around and which is almost sure to be the cause of some future financial crisis. The Volcker Rule already has led to the flight of proprietary trading, hedge fund, and private equity activity to the shadow banking system. And the Volcker Rule has done so without any explanation of how such flight would have solved the problems of the last crisis. Volcker stands out from the rest of the Dodd-Frank Act in that it violates a core principle of the Act, which was also a key lesson of the financial crisis. This is that activities should be regulated according to their nature and not according to the type of charter or license that the institution has in which they are conducted. If fund sponsorship and proprietary trading are dangerous, why are they dangerous only for commercial banks and their affiliates? The answer from the proponents of structural reform appears to be that this is because commercial banks and their nonbank affiliates benefit from a safety net (deposit insurance and access to liquidity from a central bank lender of last resort) and because payment system products and credit to consumers and SMEs are socially useful activities, whereas proprietary trading or providing hedge funds or private equity funds are not socially useful. They see the combination of deposit taking and lending activities within the same legal entity or the same group of affiliated legal entities as constituting a public subsidy through lower borrowing costs and as encouraging moral hazard in the form of riskier investments on the assumption that tax payers will end up bearing the cost of failure. This is often presented as if money were flowing from the government to the banking sector rather than from the willingness of the debt markets to accept a lower in-

Structural Solutions

133

terest rate.¹⁴ It is argued that these advantages were only meant to accrue to deposit taking institutions, which engage in traditional banking activities, and have been internally misappropriated by banks to support other extraneous and riskier activities. Thus, if only we could separate the subsidized socially useful parts of financial institutions from the riskier and socially useless parts, we would reduce the instances in which bail-outs would be needed because risky and socially useless shadow banks would not need to be rescued. While this line of reasoning may have a superficial plausibility, it is deeply flawed on a number of levels: historically, empirically and methodologically.

What is Wrong with this Picture? From an historical point of view, let us start with who was “bailed out” in the US in the last crisis. Putting aside the point that it was their creditors and not the shareholders of the institutions themselves who were protected from loss, in the US it was a broker dealer with no commercial banking operations, Bear Stearns, an insurance group, AIG, and the entire money market mutual fund industry. None of these were deposit-taking institutions and yet the threat of their collapse resulted in public authority intervention to prevent their disorderly collapse from having serious adverse effects on the U.S. financial system. Equally important, historically there is one institution that was not bailed out and whose collapse did create precisely the kind of contagion public authorities fear: Lehman Brothers. Lehman was not a deposit-taking institution and yet its collapse caused damage to the financial system and ultimately to the “real economy” entirely out of proportion to its size. It had a balance sheet on the order of $600 billion and its collapse was at least partly responsible for stock market losses of nearly $1 trillion in a single day. Thus, it is clear that not only do government authorities face pressure to bail out deposit-taking institutions to avoid the adverse consequences of their collapse, but that these pressures also apply in the case of certain non-deposit taking financial institutions. From an empirical point of view, proponents advance the subsidy-to-borrowing-costs thesis, based largely on contradictory analyses of ambiguous empirical data and on approaches to rating financial institutions by credit rating agencies. This battle of experts on subsidies resembles the similar battle of experts on  Because it is really the latter rather than the former, some commentators have started referring to it as a funding advantage rather than an implicit subsidy. See Martin N. Baily, Douglas J. Elliott & Phillip L. Swagel, Big Bank Theory: Breaking Down the Breakup Arguments, October 2014, pp. 14– 18.

134

Randall D. Guynn and Patrick S. Kenadjian

whether banks experience economies or diseconomies of scale as they grow larger.¹⁵ As lawyers, we must leave the analysis of empirical data to others and only say there appears to be no serious consensus on economies of scale, but that we find that the evidence adduced for a specific subsidy attributable to TBTF at the present point in time is insufficient to reach any actionable conclusion.¹⁶ With respect to ratings, it is true that for a period of time after the crisis the credit rating agencies (the same ones whose misjudgment of the creditworthiness of mortgage-backed securities cost them a good deal of public trust during the crisis) added an uplift to their ratings of financial institutions to reflect the potential for government support. These uplifts are in the process of disappearing, not because of the prospect of these institutions being divided up, but country-bycountry, according to the degree to which bank resolution laws have been adopted there, which will allow or even require the costs of failure to be internalized by the financial institutions and their creditors and not be shared with taxpayers.¹⁷ A recent study by the US Government Accountability Office confirms that whatever funding advantage there was to the largest US banks has disappeared or even turned negative in 2012 and 2013.¹⁸ An interesting observation in the IMF’s examination of the issue in Chapter 3 of its annual Global Financial Stability Report in April 2014 is that estimates of the size of the funding advantage in the EU are six times the size of the subsidy in the US.¹⁹ This would be consistent with the general analysis of the rating agencies that the single-point-of-entry approach to implementing the Orderly Liquidation Authority mandate in DoddFrank by the US Federal Deposit Insurance Corporation (FDIC) “is emerging as an accepted and credible alternative to tax payer-funded capital infusions from the government that might otherwise be needed to avoid the contagion and systemic risks from a disorderly failure of one or these institutions.”²⁰

 Compare Wheelock, D.C. and P.W. Wilson, Do Large Banks Have Lower Costs? New Estimates of Returns to Scale for U.S. Banks, Journal of Money, Credit and Banking, 44(1), 2012 and Hughes, J.P. and L.J. Mester, Who Said Large Banks Don’t Experience Scale Economies? Working Papers No. 11– 27, 2011 Federal Reserve Bank of Philadelphia.  Oliver Wyman, Do Bond Spreads Show Evidence of Too Big To Fail Effects, April 2014; Baily, Elliott & Swagel, note 14 above.  Moody’s Investor Service Special Comment: Moody’s Concludes Review of Systemically Important US Banks – Frequently Asked Questions. November 14, 2013. Fitch Ratings, Peer Review: Global Trading and Universal Banks, Special Report, March 27, 2014 and Sovereign Support for Banks, Rating Path Expectations, Special Report, March 27, 2014.  Government Accountability Office, Large Bank Holding Companies: Expectations of Government Support. July 31, 2014.  International Monetary Fund, Global Financial Stability Report, April 2014, p. 14.  Moody’s, note 17 above, p. 4.

Structural Solutions

135

This is in contrast to the continuing questions the agencies had at the time concerning how the EU’s draft Bank Recovery and Resolution Directive would function.²¹

Reducing the ex ante Risk of Failure This brings us to the methodological point. As Andy Haldane of the Bank of England reminded us at the conference out of which this chapter grew, there are basically two ways to solve a problem relating to the failure of an institution: ex ante to reduce the likelihood of failure and ex post to reduce the loss given failure. On the ex ante side, when we examine the sources of fragility of both individual financial institutions and the overall financial system in the run-up to the crisis, we see the following factors: some financial institutions were levered up to 30 and 40 times and had too little capital that was truly loss-absorbing on a going concern basis. Thus, even a small loss could easily reduce their capital drastically and lead to their becoming insolvent. These and other institutions (mainly investment and other shadow banks in the US, but also universal banks in Europe) suffered from the absence of a sufficiently large stable deposit base to support their assets, with the consequent need to rely increasingly on short-term capital market debt instruments. They also lacked liquidity reserves and concentrated on lending to or investing in financial instruments backed by various forms of real estate. It was bad lending decisions and holding on to or investing in the products of that bad lending, which caused the losses, not proprietary trading. These losses led to the potential for failure in highly leveraged institutions whose loss absorbing capital was quickly reduced and whose liquidity reserves were insufficient to allow them to weather the decisions of capital market investors not to roll over short-term debt, leaving them with no alternative other than having to engage in fire sales of assets, which led to further losses and to contagion to other financial institutions holding the same or similar assets. The reforms embodied in the Basel III rules and in related supervisory initiatives attack all of these weaknesses and thus reduce directly the risk of failure of financial institutions. How does structural separation reduce the risk of failure? The argument is by reducing the chance that losses on the trading side of the bank will bring

 Moody’s Investors Services, Special Comment, EU Bank Resolution: Draft Directive Offers Clarity on Future Support Framework, But Important Questions Remain Unanswered, September 30, 2013.

136

Randall D. Guynn and Patrick S. Kenadjian

down the traditional banking side. But that is not what we observed in the last crisis. The losses originated on the banking side, i. e. bad lending decisions, or in assets held in the banking book (bad loans) or as assets such as collateral debt obligations available for sale, not in the trading book. Trading, if one can call it that, predominantly played a role through fire sales once losses had begun to spread and financial institutions had to de-lever. What do the structuralists propose in addition to this? They underline how much easier it will be to resolve a financial institution if it is divided up into a trading and commercial lending operation and imply that there will be no need to bail out the trading entity if the exposure of the commercial lending entity to it is at arm’s length and capped. As noted above, this argument is contradicted by the historical record. Under the sort of structural solution mandated by Volcker, Vickers and Liikanen, the separation takes place along pre-ordained ideologically determined lines, not the relative riskiness of activities, which constantly changes over time. Another ex ante argument made by the structuralists is – overtly – the argument from complexity and – usually rather more indirectly – the issue of size. While a respectable argument could be made that a bank whose balance sheet is a multiple of the GDP in the country in which it is incorporated (which is the case in almost all European countries, but not the US) must present a greater danger to financial stability than a smaller institution, since rescuing it may be beyond the fiscal resources of the public purse, especially after the strains put on that purse by supporting the financial system in the last crisis, none of the three main flavors of structuralism address size directly. Only the US, in Dodd-Frank, has a limitation on size (10 % of total nationwide bank deposits or 10 % of the aggregate consolidated liabilities of all financial companies), but only if that size results from an acquisition and this restriction is merely an updated version of pre‐existing size limitations.²² This reluctance to attack size alone reflects the fact that TBTF rarely has much to do with the absolute size of the institutions involved. Lehman Brothers was only the fourth largest investment bank in the US and monoline US investment banks were considerably smaller than US commercial or universal banking groups. None of Northern Rock in the UK, Sachsen LB, IKB or Hypo Real Estate in Germany was a very large bank. This has led to variations on the label TBTF such as “too interconnected”, “too complex” or “too important” to fail. Part of the problem of finding an appropriate label derives from the fact that, in addition to elements endogenous to the institution, a  For the moment this prohibition would only prevent the two largest banking groups in the US from growing by acquisition, JP Morgan Chase and Bank of America. Barth, J.R. and Apanard Prabha, Breaking (Banks) Up is Hard to Do: New Perspectives on Too Big to Fail, December 2, 2012.

Structural Solutions

137

key factor in a TBTF scenario seems to be the fragility of other financial institutions and the entire financial system at the time one institution faces a crisis. If there are a sufficient number of other institutions which might be facing the same problem, so that if the first one collapses it could be taken as a sign that the others might also collapse, that first institution, regardless of its absolute size or level of interconnections, will present its government with the TBTF dilemma. Under conditions of uncertainty, if financial institutions engage in maturity transformation – i. e., they have demand or other short-term liabilities and long-term or other illiquid loans or other assets that do not have stable and transparent market prices – and some common shock causes investors to question the value of assets throughout the banking system, contagion can quickly grow into a system wide crisis. Faced with this uncertainty, a government may well opt for a “bail-out” regardless of the size of the institution involved unless it believes it has the tools needed to contain contagion by resolving the affected entity in an orderly fashion. The argument from complexity (too complex to manage) is largely based on anecdotal incidents involving operational risk and on the number of subsidiaries to be found in many financial groups (Lehman’s 3000 subsidiaries is an oft-cited figure). But operational risk in the form of traders who double down on losing bets and seek to circumvent internal exposure limits in doing so when the markets turn against them is an old story and where banks have been brought down by such activities, such as Baring Brothers in the UK in 1995, it has been smaller, simpler institutions rather than large, complex ones who seem to have failed.²³ Finally, much of the complexity in corporate structure has come from regulatory requirements to separate various activities into separate entities and both Vickers and Liikanen will result in additional complexity in this regard.

Reducing the ex post Losses Given Failure Turning now to the ex post side, reducing losses given failure, the key initiatives are the ones establishing or expanding resolution regimes tailored to financial institutions in the US, the UK and elsewhere in the EU both at the Member State and at the Community level (via the Bank Recovery and Resolution Directive). The fundamental problem faced by public authorities during the crisis  Barth & McCarthy, note 6 above, analyse 15 instances of trading losses of at least $1 billion between 1990 and 2012 (at hedge funds, manufacturers and oil refiners, investment banks, governments and commercial banks) and conclude that losses relative to equity were lowest at banks (5.3 %). At investment banks they amounted to 34.2 % of capital.

138

Randall D. Guynn and Patrick S. Kenadjian

when confronted with a potentially failing institution was the absence of appropriate tools to handle its failure in an orderly fashion. In the absence of such tools, the choices are a disorderly sale through the liquidation of financial assets at the bottom of the market during a financial panic or a value-destroying reorganization that takes so long to complete that the institution involved loses most of its value by the time the reorganization is approved. The experience with Lehman Brothers on both sides of the Atlantic made clear that ordinary bankruptcy or insolvency laws, at least as traditionally conceived and without advance resolution planning or certain amendments, were inappropriate tools to reduce the loss upon failure of a financial institution, for a number of reasons. First, unless traditional bankruptcy laws can be used in creative ways to separate the good and bad parts of a banking group quickly, such as through the quick transfer of the good parts of a failed banking group to a bridge entity (the good bank) under Section 363 of the U.S. Bankruptcy Code and the ability to leave the bad parts of the failed group (the bad bank) behind in a bankruptcy proceeding,²⁴ the bankruptcy process is too slow and financial institutions’ value evaporates too quickly. Second, because central to a bankruptcy proceeding is a general stay of claims against the bankrupt, and this stay interferes with what Tom Huertas has identified in his paper in a prior volume in this series as “the very essence of banking,” “the ability to make commitments to pay.”²⁵ Another problem is that bankruptcy courts have not traditionally viewed their roles as including a consideration of the need to maintain depositor confidence and financial stability. In the US, the FDIC has long had powers tailored to resolving commercial banks without producing financial instability, but these powers did not extend either to broker-dealers like Lehman or to bank holding companies like Citigroup or Bank of America Corporation. There were no comparable legal frameworks in Europe. The application of ordinary bankruptcy laws to Lehman Brothers, especially in the UK, in the absence of resolution planning and some of the creative thinking that has taken place since 2008 about how the U.S. Bankruptcy Code can be used to execute a good bank / bad bank resolution structure,²⁶ resulted in a very messy collapse with repercussions throughout the financial system in the US and the EU, with multiple uncoordinated national bankruptcy and insolvency proceedings seeking to liquidate assets for the benefit of local creditors and to the detriment of just about everyone else, including

 Bovenzi, Guynn & Jackson, note 1 above, pp.33 – 34, 38 – 39, 75 – 76.  Thomas F. Huertas, Resolution Requires Reform, in Patrick S. Kenadjian, Too Big To Fail – Brauchen wir ein Sonderinsolvenzrecht für Banken, Institute for Law and Finance Series, De Gruyter 2012.  Bovenzi, Guynn & Jackson, note 24 above.

Structural Solutions

139

the financial system. None of the judicial authorities involved had a mandate to take into consideration the health of the financial system. In addition, there was no advance resolution planning that might have resulted in recapitalization strategies being carried out in bankruptcy proceedings that would have taken those considerations into account. The consensus solution to this problem developed by the FSB and approved by the G-20 has been to design a system tailored to deal with the failure of financial institutions, through what has become known as “resolution”. The key idea is to recognize that financial institutions are different from other kinds of enterprises and that their failure needs to be dealt with in a way which preserves the stability of the financial system by allowing them to be recapitalized and reorganized outside of the regular bankruptcy or insolvency process, at least in the US if it is not possible to do so within the regular bankruptcy process. In Europe the focus has been on a procedure that internalizes the cost of failure as much as possible through the concept of “bail-in”. This can happen either outside of a resolution proceeding at the “point of non-viability” of the enterprise or as part of a resolution proceeding overseen by a resolution authority. Under “bail-in”, the long-term senior and subordinated creditors of the failing enterprise can be required to absorb its losses to the extent required to recapitalize it by having their debt written off or converted into an amount of equity sufficient to absorb the losses incurred and bring its capital ratio back up to that required by its license. They do this in reverse order of seniority and subject to certain exceptions, such as for insured deposits and secured debt which are either contractually, structurally or legally made senior to the claims of long-term creditors. Assuming the institution has sufficient total loss-absorbing capacity (TLAC) in the form of combined regulatory capital and long-term unsecured debt, bail-in prevents the institution from having to file for bankruptcy or insolvency protection from its creditors and assures depositors and other super senior short-term creditors that they need not run on the institution. Because the losses are borne by equity and long-term debt holders, the need for public capital is eliminated, the need for the institution to deleverage by reducing lending is also lessened and the effect on the “real economy” is correspondingly reduced. Under the rules of the European Union, “state aid” (i. e. public funds) is permitted to be provided to a financial institution only after equity and unsecured long-term debt holders have absorbed losses up to an amount equal to 8 % of the institution’s liabilities.²⁷ It is calculated that this amount of loss-absorbing capacity

 The Bank of England’s approach to resolution, October 24, 2014, p. 9, note 1.

140

Randall D. Guynn and Patrick S. Kenadjian

would have sufficed to repair the capital of the EU institutions that received bailouts during the crisis of 2008/2009. What do the structuralists propose that supplements or improves this? Little more than the assertion that we only provide deposit insurance and central bank lender-of-last-resort liquidity to deposit taking institutions because they provide socially useful products and services. In other words, they argue that if trading and other investment banking activities can be separated from deposit taking, the deposit taking institutions will not fail or any loss to the public purse will be smaller. As noted above, this flies in the face of history. Financial institutions were rescued regardless of whether they had insured deposits because public authorities feared contagion throughout a fragile financial system and did not have an instrument other than bail-out to avoid a messy bankruptcy with fire sale liquidations of illiquid assets. New resolution regimes, and creative new resolution strategies designed under those regimes or even under normal bankruptcy regimes, provide a solution out of the dilemma between bailouts and fire sale liquidations, regardless of the activities within the institution. Those tools can also be tailored to achieve in practice what structural solutions aim for in theory: once a financial institution is in resolution, the resolution authority can decide to divide it up, transferring parts of it to a bridge bank which is recapitalized by bailing-in existing debt and continue operations uninterrupted, while leaving other parts behind to be liquidated over time. To quote Bank of England Deputy Governor John Cunliffe: “The aim is to enable the critical parts of the group – the parts vital to the real economy and the parts that financial stability depends on – to keep operating so the group can be safely resolved over time. The ultimate solution could involve a mixture of sales, administration and run-off”.²⁸ And that is the true solution to TBTF.²⁹

But What Harm do Structural Solutions Cause? The fundamental problem with structural solutions is that they divert attention and resources away from real solutions to the TBTF problem by focusing attention on traditional regulated commercial banks and they create incentives for

 Sir John Cunliffe, Ending Too Big To Fail – progress to date and remaining issues. 13 May 2014. See also The Bank of England’s approach to resolution, October 24, 2014.  Interestingly enough, in an FAQ issued in connection with this pamphlet cited in footnote 9 above, John Kay acknowledges that “the plans for ‘living wills’, combined with a proper resolution regime” probably would “do the trick”. John Kay, Narrow Banking FAQs, available at www. johnkay.com/wp-content/uploads/2009/09/Narrow-Banking, p. 7.

Structural Solutions

141

that portion of the overall financial system to shrink relative to the part of the financial sector typically referred to as shadow banking. That is, of course, a term which covers a very broad array of institutions, and which has changed over the years;³⁰ today it includes investment banks, money market mutual funds, hedge funds, on-line peer-to-peer lending clubs, various kinds of crowd funding arrangements, alternative payment systems on line or via mobile networks and various issuers of virtual currencies such as BitCoin. Tomorrow it could include Walmart, Amazon, Facebook and Google. We know from a quick review of the 2008/2009 crisis that many of the entities that needed to be bailed out were part of what was then considered the shadow banking sector. We have noted Bear Stearns, Lehman and AIG. We can add the two government sponsored entities active in mortgage financing, Fannie Mae and Freddie Mac, commercial paper conduits and other securitization vehicles and note that many of the commercial banks that failed did so because they relied excessively on the shadow banking sector for financing or were excessively exposed to loans made to the shadow banking sector in the form of investments in collateralized debt obligations. Many of their problems resulted from their failure to fully understand their interconnections, or the risks of those interconnections, to the shadow banking sector. The Financial Stability Board has recognized this issue and created five working groups to focus on shadow banking, but only two of them, the one focusing on money market mutual funds and the one focusing on secured lending seem to have made much progress. The US and the EU have adopted inconsistent approaches to money market mutual fund regulation and there has been no coordinated action on secured lending, despite significant concerns about repos and the existence of FSB recommendations on their reform.³¹ Focusing on illusory structural solutions for the official

 Shadow banks of the past have included (1) unofficial banks or colonial governments (e. g., Rhode Island) that issued money or money-like instruments that were not legal tender under the then-existing laws; (2) commodities merchants that provided commodities money such as tobacco, grain, gold or silver that functioned as official or unofficial money; (3) merchants and farmers that issued notes and other IOUs that functioned as money in local communities; (4) private banks and other unchartered banks (i. e., nonbank banks) that issued bank notes or took deposits; (5) wildcat banks that issued paper money on the American frontier with insufficient gold and silver reserves for purposes of converting the paper money to gold or silver; and (6) nonbank banks that escaped regulation in the 1980s under the US Bank Holding Company Act of 1956 by issuing deposits that were withdrawable upon demand as a practical matter but were legally subject to a 7-day delay. See Bray Hammond, Banks and Politics in America from the Revolution to the Civil War, 1957, for examples 1– 5.  Financial Stability Board, Strengthening Oversight and Regulation of Shadow Banking: Policy Framework for Addressing Shadow Banking Risks in Securities Lending and Repos, 2013.

142

Randall D. Guynn and Patrick S. Kenadjian

banking sector detracts valuable attention and resources from dealing with the very real issues arising from the shift of maturity transformation from the official banking sector to the shadow banking sector that has taken place and will continue to take place despite and in part because of the so-called structural solutions. It is a bit like the British guns in Singapore facing the wrong way and equipped with the wrong kind of shells to be effective against infantry rather than battleships in 1942. It gives a false sense of security against yesterday’s perceived problems while being useless against tomorrow’s ever-evolving real dangers. A second problem is that the solutions that the structuralists propose are both brittle and rigid. Structural solutions will clearly reduce the flexibility of the commercial banking sector to respond to market changes, much in the way the Glass-Steagall Act of 1933 did. The original judgment which underlay Glass-Steagall that underwriting and dealing in corporate securities was riskier than lending was overtaken in time by changes in the breadth and depth of the US capital markets which made it cheaper for the best corporate credits to raise debt from the capital markets than to borrow from commercial banks. As a result, by the 1980s, the market for commercial bank lending and the profits from them had shrunk, forcing the commercial banks to concentrate on lending to riskier and riskier credits, which resulted in their risk exposures to soar instead of staying the same or decreasing. At the same time, the market for investment banks had grown by the 1980s, as did their profits, and their risks had declined. The investment banks had also found ways around the restrictions on deposit-taking and money creation through the development of overnight repos, money market funds and securitization of bank loans. Talent followed profits from the commercial banks to the investment banks. It is not too hard to conjure up a similar scenario as a result of today’s proposed structural solutions. One can see the size and the profitability of the commercial banking sector continuing to shrink under ever-heavier regulation and the growing competitive advantage of the shadow banking sector. Taking, for example, the Volcker Rule and comparing it to Glass-Steagall, Volcker reflects a similar judgment that proprietary trading and investing in certain funds is riskier than lending and other commercial banking activities. Putting aside the fact that this judgment is highly questionable based on the historical record, it is also unlikely to stand the test of time for the same reason the judgment underlying Glass-Steagall did not. The US financial markets will evolve in ways not anticipated by the Volcker Rule. While this is happening, the official banking sector will shrink, become less profitable and more risky relative to the shadow banking sector. The shadow banking sector in turn will grow, become more profitable and less risky relative

Structural Solutions

143

to the official banking system. Those shadow banks will continue to develop alternatives to whatever restrictions on deposit banking or money creation they are under and continue to attract talent away from the official banking sector. As the shadow banks increase their share of the financial system, so will the danger of contagion and market meltdown increase, in the event one of them fails, thus increasing the chances that public authorities will feel they have no choice but to bail them out, unless they too are subject to a resolution system comparable to that being applied to official banking groups. So, having divided up the banks, ring-fenced their traditional commercial banking activities and kept them out of the profitable fee generating businesses, will not end TBTF. It will only result in the migration of some prohibited activities from the official banking sector to the shadow banking system and encourage the official banking system to engage in riskier and riskier versions of their permitted activities in order to remain competitive. As the shadow banks account for a larger and larger percentage of the US financial system, the likelihood of their being bailed out can only grow. It would be a fool’s errand to try to prohibit shadow banks from providing money-like instruments to compete with the official banking system. History is littered with attempts by governments to grant monopolies over the money creation process to certain specially licensed institutions or groups of institutions. The best example of this is the ill-fated attempt by the U.S. government to give a monopoly over the paper money creation process to a new national banking system in 1864 and drive the state-chartered banks out of existence.³² The

 Other examples include (1) the British Tunnage Act of 1694 and the Bubble Act of 1720, which attempted to grant money-making monopolies to the Bank of England and existing corporations such as the South Seas Company; the shadow banks of the day in the American colonies (e. g., pools of merchants or US colonial governments, e. g., Rhode Island) responded by issuing bills of credit which were not legal tender but nevertheless functioned as money; (2) the British Acts of 1741, 1751 and 1764, which severely restricted the power of the colonial shadow banking system to create money; these acts caused a public uproar in the colonies, were cited by Benjamin Franklin in a 1767 speech as one of the reasons for growing colonial hostility to the British Parliament, and were ineffective as the shadow banks in the colonies continued to circulate unofficial paper money that was used in local commercial transactions; (3) American Constitution of 1789, which gave the federal government exclusive control over the money creation process and expressly prohibited the states from coining money, emitting bills of credit or making any thing but gold or silver legal tender; the states responded by chartering banks which issued bills of credit and other forms of paper money and took deposits that could be debited or credited to make payments; (4) the Glass-Steagall Act of 1933, which prohibited investment banks and other shadow banks (e. g., money market mutual funds) from engaging in the business of taking deposits; the investment banks responded by taking overnight repos and other forms of shortterm credit that could be transferred by book-entry to pay for transactions and money market

144

Randall D. Guynn and Patrick S. Kenadjian

state-chartered banks responded by offering checking accounts to their customers instead of bank notes (paper money). Checking accounts proved to be as efficient as, and in some cases more efficient than, bank notes as a medium of exchange and store of value – i. e., money. If the government had prohibited the state-chartered banks from offering checking accounts, they almost certainly would have invented some other form of money substitute that was not within the ambit of the prohibition.

Conclusion In conclusion, there is nothing that the structural solutions propose to do to solve the TBTF problem that cannot be done better by more targeted measures such as increased capital and liquidity requirements under Basel III and various new resolution regimes and strategies such as the SPOE bail-in or recapitalization strategy. Basel III should reduce the risk that financial institutions will fail in the first place and the new TLAC requirements, resolution regimes and resolution strategies should allow the cost of failure to be imposed on the private sector rather than taxpayers, without resulting in a destabilization or collapse of the financial system and related collateral consequences to the “real economy”. Structural solutions mostly serve to distract attention and resources away from these real solutions.

mutual funds responded by issuing debt securities that were redeemable upon demand or within a matter of days. Bray Hammond, note 30, for examples 1– 3.

Simon Gleeson

Structural Separation and Bank Resolution 1. Why Structural Separation? The recollection in tranquillity of events of great stress usually discloses ideas which, passionately believed at the time, now appear fantastical. The first world war urban legend of Russian soldiers marching through the English countryside “with snow on their boots” was passionately believed despite its utter implausibility. A similar process of thought during the financial crisis seems to have led to an outbreak of faith in “narrow banking”; a manifestation of the belief that if only modern banks could be wished away, we could return to the comforting certainties of the fictional bankers of “Dad’s Army” and “Its a Wonderful Life” This approach – of wishing away the world as it is in favour of a mythological past – is as old as recorded civilisation. However it had a curious offshoot in the form of regulatory proposals to separate out “Captain Mainwaring-style” banking from the frightening business of derivatives, securitisations et. al. which were confidently believed by the uninformed to have “caused the crisis”. Complicating matters further is the fact that such structural separation proposals came in a variety of different flavours. In addition to “Volcker” type proposals, which prohibit bank groups from engaging in certain activities, Europe has managed to generate two different sets of proposals for separating activities within bank groups. One of these is the UK “Vickers” proposal, which is aimed at ring-fencing retail deposit taking into a separate legal entity; the other is the EU “Liikanen” proposal, which is aimed at separating market trading activity into a separate legal entity. These differing policies can best be explained graphically as follows:

The point which this illustrates is that to impose all of these simultaneously on a single entity would, in effect, involve dividing it into four, with the Volcker part being squeezed into a separate corporate group and the remaining three compo-

146

Simon Gleeson

nents coexisting under a common holding company. No policymaker currently considers that imposing all of these restrictions on a single entity would be a good idea (although this may be the ultimate fate of UK deposit-taking banks). However, it does illustrate the fact that to talk about structural separation as if it were a single concept is to misunderstand the nature of the problem. This proliferation of structural separation proposals needs to be systematised before the impact of structural separation on resolution can be considered. Ordinarily, this could be done by reference to the policy aims which the separation is intended to secure. However, identifying these aims is curiously difficult – the documents which propose ring‐fencing tend to proceed by assuming its benefits rather than setting out a case for them.

Vickers-style separation In this regard, the observations of Sir John Vickers, the chair of the UK Independent Banking Commission report (the “Vickers Comission”) are particularly enlightening “… the Commission sees merit in a UK retail ‘ring-fence’. This would require universal banks to maintain the UK retail capital ratio – that is, not to run down the capital supporting UK retail activities below the required level in order to shift it, say, to global wholesale and investment banking. Our current view is that such a limit on banks’ freedom to deplete capital would be proportionate and in the public interest, and would preserve benefits of universal banking while reducing risks. Without it, capital requirements higher than 10 % across the board might well be called for”.¹

Leaving aside the fact that capital requirements of 10 % are now a distant memory – requirements for global banks are much higher – the underlying idea held by the commission of how a bank works, with capital being “shifted” within a single legal entity from one part to another, is particularly disturbing, in terms of highlighting the grasp of those writing the report of the topic concerned. A slightly better effort was made by HM Treasury, whose paper on the implementation of the Vickers ring-fence summed up its objectives as being:– to insulate critical banking services from shocks elsewhere in the financial system; and

 http://web.archive.org/web/20110914044642/http:/s3-eu-west-1.amazonaws.com/htcdn/Interim+Report+publication+JV+opening+remarks+-+check+against+delivery.pdf.

Structural Separation and Bank Resolution

–

147

to make it easier to preserve the continuity of those services, while resolving financial institutions in an orderly manner and without injecting taxpayer funds.²

It should be reasonably clear that ring-fencing certain activities does nothing to insulate them from shocks in the financial system. If the financial system suffers a shock, that shock will be manifested in the value of the assets held by financial institutions, the extent of the impact will depend on the composition of the asset book of the institution concerned, and the Vickers report has almost nothing to say about the riskiness of the asset portfolio of the ring-fenced institutions.³ Thus we seem to conclude that the primary justification for the creation of a Vickersstyle ring-fence is to facilitate resolution, enabling the retail banking part of the enterprise to be separated out and rescued.

Liikanen-style separation The section of the Liikanen report which summarises the objectives of its structural separation proposals reads as follows. The central objectives of the separation are to make banking groups, especially their socially most vital parts (mainly deposit-taking and providing financial services to the non-financial sectors in the economy) safer and less connected to trading activities; and to limit the implicit or explicit stake taxpayer has in the trading parts of banking groups. The Group’s recommendations regarding separation concern businesses which are considered to represent the riskiest parts of investment banking activities and where risk positions can change most rapidly. Separation of these activities into separate legal entities is the most direct way of tackling banks’ complexity and interconnectedness. As the separation would make banking groups simpler and more transparent, it would also facilitate market discipline and supervision and, ultimately, recovery and resolution.

This piece of text proceeds on the unexamined assumption that retail deposit taking is somehow “less risky” than market-based activities, and we will return to this topic in section 4. However, even assuming it were possible to separate a

 Banking reform: delivering stability and supporting a sustainable economy; June 2012, https:// www.gov.uk/government/uploads/system/uploads/attachment_data/file/32556/whitepaper_banking_reform_140512.pdf.  Ring-fenced deposit-funded institutions have traditionally been forced to improve their returns by taking on high-risk loans (typically in the commercial real estate sector) which are highly sensitive to downturns in the financial system as a whole.

148

Simon Gleeson

bank into “more risky” and “less risky” activities, it should be relatively obvious that separation of activities conducted within an institution into separate booking vehicles will not improve transparency or market discipline, since these ends are achieved by disclosure requirements and not by legal restructuring. However, if we approach the report in a sympathetic spirit and try and identify what the thought was which animated the conclusion, what appears to come through is the idea that in the public mind the customer should perceive the group as two entities rather than one, be aware of which entity he has contracted with, and be comfortable that it is possible for one of these entities to fail without his claim on the other being affected.

2. The US Experience The enthusiastic proponents of structural separation in Europe seem to have been unaware that what they were propounding was, and had been for many decades, the state of affairs in the United States. The US regime is based on two ring-fences. The broader ring-fence is set out in the Bank Holding Company Act 1956, which effectively sets a ring-fence around the banking group and determines what any member of that group may and may not do.⁴ The narrower ringfence is the bank charter itself. This is the charter granted to the bank subsidiary of a bank holding company, and this limits what the bank itself and its subsidiaries may do.⁵ The effect of these limitations is to confine the bank itself fairly strictly to deposit-taking, lending and money-market activities, with the consequences that other activities – notably securities trading – must, to the extent that they are permitted within the group at all, be carried out within the nonbank part of the group. In order to give effect to these sections 23A and 23B of the Federal Reserve Act place strict limits on the transactions which may be entered into between the bank and its subsidiaries (generally referred to as the “bank chain”) and the other members of the group (the “non-bank chain”)⁶. The purpose of the 23A/23B regime was (and is) to prevent the activities of the nonbank chain receiving any cross-subsidy from the bank, on the basis that

 The details of these restrictions are set out in Regulation Y http://www.ecfr.gov/cgi-bin/textidx?SID=6f7456d810ed1f37c2ae4c2b56a0f254&node=12:3.0.1.1.6.1.6.1&rgn=div8.  For a comprehensive list of these see the OCCs statement of Activities Permissible for a National Bank, Cumulative 2011 Annual Edition April 2012 http://www.occ.gov/publications/publications-by-type/other-publications-reports/bankact.pdf.  The details of these are set out in regulation W http://www.gpo.gov/fdsys/granule/CFR-2012title12-vol3/CFR-2012-title12-vol3-part223/content-detail.html.

Structural Separation and Bank Resolution

149

since the bank’s cost of funding is effectively subsidised through the existence of FDIC and deposit insurance, the benefit of that subsidy should be retained within the bank for the purpose of minimising losses to depositors (preservation of competitive equality between free-standing securities firms and bank securities trading subsidiaries was also a concern). It is fair to say that the fact this fact remained as unknown in Europe after the crisis as it was before the crisis suggests that the separation itself may not have been a major factor in addressing the crisis. The idea that the problems at (say) Citibank could have been usefully addressed by separating the entity into its bank and its non-bank components and resolving the two separately does not appear to have been seriously considered by any relevant authority at any time, and it is not too hard to see why.

3. Separating Market Risk and Credit Risk It was noted above that a canonical belief of the Liikanen group was that “the trading parts of banking groups” are somehow more risky than the other parts. At first sight this proposition seems counter-intuitive. Even leaving aside the fact that the riskiness of a bank is a function of the composition of its asset book, risk in investment banking is generally monitored and controlled far more carefully than risk elsewhere in a bank. It is notable in this context that the investment banks in the crisis survived fluctuations in asset values which would have flattened a conventional bank, and the reason for this is that investment banks, unlike conventional banks, hedge almost every aspect of their portfolios. The problem which was encountered during the crisis is that market-based activities are exceptionally vulnerable to market failures. The primary characteristic of investment banking which distinguishes it from commercial banking is – in theory – that the assets and liabilities of an investment bank are marked to market – including, importantly, for regulatory capital purposes. It is now generally accepted that, whatever the cause of the crisis is believed to be, one of the most important factors leading up to it was the idea that markets had moved away from irrational boom and bust cycles and towards a modern, scientific world in which participants would always be prepared to purchase assets at the mathematically modelled “correct” valuation, and that liquidity would therefore always be available at that valuation. The construction of a financial system based on that model meant that when the market did return to boom and bust, the resulting swings in market prices were literally inconceivable to those who had built the models. This problem was compounded by the fact that the liquid-

150

Simon Gleeson

ity whose perpetual continuation had been so confidently predicted stubbornly refused to appear. This created a vicious circle which drove prices even lower, since buyers will not buy, no matter what the valuation, if they have no money to buy with. The impact of a market crash has a real-time impact on the capital position of a bank which holds mark to market assets. In particular, the fact that the market has disappeared, or can be accessed only at rock-bottom prices, results in a wholesale destruction of value on the accounting and regulatory balance sheet of the institution concerned. Viewed from this perspective a business which holds mark to market assets can do enormous damage to a bank balance sheet simply by existing, regardless of how well or badly it is run. Given this, it is worth examining the nexus between trading losses and credit losses to see whether the combination of the two activities in the same entity provides a transmission mechanism which could be broken by structural separation. It is true that losses experienced in the trading arm reduce the capital of the institution as a whole and therefore its ability to lend. However this would be true on a consolidated basis for any institution, whether the activities were separated or not. Possibly more importantly, if institutions with separated trading and credit businesses reacted to market crises by allowing their trading businesses to fail, it is more or less certain that one consequence of this would be that the market crisis would become worse than it would have been if those losses had been absorbed by capital elsewhere in the group. The question is therefore whether the vector of contagion between market crisis and credit crisis is the internal linkage within individual groups or the external linkage provided by the market as a whole. A fairly cursory examination of the bank failures of the credit part of the crisis would seem to give a clear answer to this question. If the vector of contagion were internal connections within banks, then we would expect to see the largest bank failures as those entities having the largest market-facing or trading businesses. This was not the case. Some of the worst bank failures – such as HBOS in the UK – had almost no investment banking or trading activities, and it is generally accepted that there was no one business model which was either particularly robust or particularly vulnerable during the credit phase of the crisis⁷. It therefore appears – in accordance with intuition – that the connecting factor between market crisis and credit crisis is the market itself.  For a detailed consideration of the potential linkage between business models and crisis vulnerability, see ECB Working Paper 1394/November 2011 Bank Risk During The Financial Crisis: Do Business Models Matter? by Yener Altunbas, Simone Manganelli and David Marques-Ibanez http://www.ecb.europa.eu/pub/pdf/scpwps/ecbwp1394.pdf.

Structural Separation and Bank Resolution

151

It is however true that if the 2008 crisis been nothing more than a market crisis, it would have had little more impact on the global economy that the dot-com crisis of the early 2000’s. The crisis was in fact a very old-fashioned crisis, in that it was constituted by a credit-fuelled expansion in the value of a particular asset, and in that regard it differed not at all from previous crises involving tulip‐bulbs, South American railway shares or commercial property development. The difference this time around was that the asset class which boomed during the 2000s was credit itself. Thus when the market crashed, the availability of credit within the global financial system collapsed, and significant pressures were brought to bear on the credit position of all those – notably sovereigns and banks – who required recourse to that market. It was this collapse of credit availability which brought on the real crisis. Mark-to-market assets are a canary in a coal-mine – they provide an early warning of problems. Loan asset books, however, have no such early warning feature – they fester in silence until they collapse. The large European banks which failed during the crisis turned out to have been taking excess risks in their loan books for years, and it was only when they were placed under pressure by the drying up of the credit markets that these failures came to the fore. Conversely, the losses actually suffered in the sub-prime crisis were relatively small – the US Financial Crisis Inquiry Commission (FCIC) found that “Overall, for 2005 to 2007 vintage tranches of mortgage-backed securities originally rated triple-A, despite the mass downgrades, only about 10 % of Alt-A and 4 % of subprime securities had been ’materially impaired’-meaning that losses were imminent or had already been suffered-by the end of 2009”⁸, and it has been calculated that the realized principal losses on the $1.9 trillion of AAA/Aaa-rated subprime bonds issued between 2004 and 2007 were 17 basis points as of February 2011.⁹ The problem in the market was not that large losses were suffered, but that the small losses that were suffered on a few products led to a major writing-down of the value of prices in the market as a whole. The other risk area in investment banking is of course funding. Investment banks fund and hedge portfolios of securities on the basis that they can be sold in relatively short order. When Lehman Brothers collapsed over the weekend of 12/13 September 2008, market liquidity disappeared, and it became impossible to liquidate portfolios. This meant that what had been perceived as shortterm holdings became, in effect, long term assets. Where these portfolios were  Report (2011) p. 228 – 29.  Park, Sunyoung. 2011. “The Size of the Subprime Shock.” Working Paper, Korea Advanced Institute of Science and Technology.”, and see generally “Collateral Crises” Gary Gorton and Guillermo Ordonez, March 2012.

152

Simon Gleeson

funded through the short-term money market, through ABCP or through securitisation, the maturity of the portfolio lengthened just as the supply of liquidity dried up. Whilst the disappearance of liquidity in the funding markets was the most visible effect of the collapse of Lehman, liquidity in other capital market instruments disappeared as well. Spreads between cash and index markets, corporate bond interest rates and their respective credit default swaps (CDSs) widened sharply, and in the equity option markets, liquidity for long-term options dried up whilst short-term options remained available. This removed the suite of products that investment banks and dealers had traditionally used to hedge their portfolios of market products, leaving them vulnerable to market movements. In addition, collateral calls were made on banks at the very moment when their access to liquid collateral was most restricted. In principle this is also true of retail banks, in that the vast majority of their funding is in the form of deposits which can in theory be withdrawn at any time. However, historic experience during the crisis is that in practice depositors do not run to any significant extent, and deposit funding constituted reliable long-term funding throughout the crisis.

4. Could Structural Separation be Sufficiently Refined to be made Workable? What this analysis makes clear is that the crisis was divided into a market phase and a credit phase. In broad terms, market-facing businesses bore the brunt of the first phase of the crisis, whilst credit-based lending businesses bore the brunt of the second phase. What this means in practice is that for a crisis of this kind, market-facing businesses will be disproportionately and rapidly affected by a market crisis. This point, however, illustrates the importance of clear thinking with regard to nomenclature. What is affected by a market crash per se is not necessarily market trading, or even investment banking more generally. A market crash primarily impacts entities with large unhedged portfolios of mark-to-market assets. At the commencement of a market crisis this is unlikely to include investment banks or dealers. If the crisis continues for an extended period of time then these businesses may be affected, as a lack of liquidity may make it harder to maintain existing hedges or to put in place new hedges as old ones expire. However in general an investment bank is probably less likely to own a directional unhedged portfolio of market assets than a commercial bank – significant prin-

Structural Separation and Bank Resolution

153

cipal position taking in credit assets is what commercial banks are for, although it is a relative rarity amongst investment banks and securities dealers. The difference is supposed to be that commercial banks do not take significant positions in market traded assets, and can therefore use their freedom not to mark assets to market to preserve their credit until the market crisis is over. However, the first of the problems uncovered by the 2008 market crash was the extent to which commercial banks had become holders of mark-to-market assets. Because of the incentives within the Basel II system to move assets from banking to trading book, banks had been heavily incentivised to take banking book assets (such as mortgage loans) and repackage them into the form of trading book assets.¹⁰ The paradigm of this position was Citibank, which shortly before the crisis had acquired a very substantial portfolio of mortgages in the form of mortgage-backed securities. The result of this was that for the 2008 financial year it reported a loss of $27.7 billion because of the $32 billion of revenue markto-market losses on assets.¹¹ In 2008 the most recent market crisis referent was the dot-com crash of 1999 – 2001. This had revealed that many of the dot‐com equities which had traded at substantial premia during the 1990s were nearly (or in some cases completely) worthless. Consequently it was held to be entirely possible that the same might turn out to be true of mortgage-backed securities, and lurid rumours circulated of entire portfolios of mortgages originated through loans to insolvent borrowers for the sole purpose of being securitised. The practical effect of this was that the economic value of these parts of the bank’s balance sheet declined to almost zero. The reason for this is that once an asset is transformed into a traded asset, there is almost no way of moving it back again.¹² Accounting convention – quite properly – does not permit firms to say “the open market value of this investment is 2, but I will value it at 20 because I think that the market is wrong and 20 is its true value”. There is an exception to this where the market is either notional or non-existent – the “level 2” and “level 3” classifications  The logic of this incentive was entirely sound in the context of the time. Since an asset which can be sold in a liquid market is ex hypothesi less risky than one which must be held to maturity (since the riskiness of a transaction increases with the length of time you are locked into it), once it has been marked to market, that value should attract a lower risk charge. The problem with this logic is the same problem which ran through the whole of the Basel framework – that it assumed the existence at all times of a liquid market.  Citi Annual Report 2008, http://www.citigroup.com/citi/fin/data/ar08c_en.pdf.  It is theoretically possible to acquire 100 % of the bonds issued by a securitisation SPY and take back the underlying assets, and this would solve the mark-to-market problem. However viewed from the perspective of an institution with a day-to-day valuation problem this would go in the “takes too long and costs too much” box.

154

Simon Gleeson

of FASB 157 – and crisis veterans will recall the extraordinary accounting battles which broke out over when markets had disappeared to such an extent that the assets concerned were no longer level 1 and could be valued using models (which invariably gave a significantly higher valuation). However, the brutal truth of marking to market – that as long as there is one desperate seller in the market there is an independently determined market price, and every relevant asset must be written down to the price of that one desperate trade – eviscerated balance sheets. It is the fact that the market crisis preceded the true crisis which explains why the US banks were disproportionately affected in the first part of it. US banks, for a variety of reasons including regulatory reasons, had been more assiduous in moving assets from the bank to the securities side of their firms, and even within the banks the holding of credit risk assets in the form of securitised mark-to-market assets provided significant advantages.¹³ Thus, when the markets crashed, it was those with the largest portfolios of mark-to-market assets which were hit hardest. It is the depth and suddenness of this impact which gave rise to the perception that investment banking was “risky”. However, it was the impact of the market crisis on the business of credit creation as a whole which caused the real impact. The question which falls to be answered here is does the combination of these two problems in the same legal entity make the issues better or worse. It should be noted that the focus on “trading” is both misguided and unhelpful, and we can take it as axiomatic that a separation between trading and non-trading activities is irrelevant to the problem at hand. The question is whether we should divide the bank into an entity which is only permitted to hold mark-tomarket assets (say securities) and an entity which is only permitted to hold non-mark-to-market assets (say loans). This arrangement would have a number of appealing features; not least that in theory a devastating market crash would leave the non-mark-to-market entity untouched. However, on closer examination its defects become apparent. Markto-market and non-mark-to-market are, at bottom, simply valuation conventions. If the value of a £100 bond issued by a corporation has sunk to 50p in the £, noone believes that the value of a £100 loan to that corporation remains at £100.

 It is interesting to note that the reason that the US banks were so strongly incentivised to do this was that they were not permitted by their regulators to benefit in full from the “risk modelling” approach to capital requirements permitted by Basel II and widely employed by EU banks. This approach enabled EU banks to significantly reduce capital requirements against loans to high-quality borrowers without necessitating their transfer to the trading book.

Structural Separation and Bank Resolution

155

Accounting convention may permit it to be accounted for as a £100 asset, but the reporting entity which does so risks undermining all faith in its accounts. It is all very well to say that market valuations vary faster than annually revalued assets, but that is a statement of accounting practice, not credit reality. If the credit markets drop substantially across the board, a business whose balance sheet is wholly composed of credit assets which claims to be untouched is likely to be abandoned by its credit suppliers even faster than it would be if it were to provide an honest estimate of the revised value of those assets. Ultimately what this boils down to is that credit is credit, and that the collapse of the credit markets will impose a pressure on all credit businesses which is likely to end in the failure of some of them. Market asset based businesses may fail sooner, whilst more opaque businesses may take longer to fail – and, as a result, fail in a larger way with greater losses. However, ultimately it is a category mistake to confuse valuation methodology with solvency, and this turns out, on examination, to be the fundamental basis of ideas of institutional separation within credit businesses.

5. Crisis and Resolution – What is the Problem to be Solved? Where the challenge for a resolution authority is to resolve a business which holds largely mark-to-market assets, the problem is generally one of timing. “Markets can remain irrational longer than you can remain solvent” is an aphorism sometimes attributed to Keynes, but in the context of a government-backed resolution exactly the opposite is true – government can provide stability, in the form of liquidity, longer than the markets can remain wrong. Such liquidity is, in a perfect world, provided by central banks, but various resolution mechanisms have been designed to provide temporary liquidity when the central bank eligible collateral runs out. Government is the only entity which can reliably beat the market in a stand-off, and for this reason a government backed rescue of a market-based institution, even if it is not accompanied by any new capital, is likely to be effective in the medium term. The position is different where the problem is a bad asset problem. Bad asset resolutions are always a long game, mostly because it takes so long to find out how bad the assets are. Anyone who has been involved in a bad asset resolution will be familiar with the process whereby a final, worst case balance sheet is drawn up on a Friday, only for further losses to be uncovered the following Mon-

156

Simon Gleeson

day. Bad asset resolutions require both (extensive) interim liquidity support and capital reconstruction. These differences mean that the approach taken by a resolution authority resolving an institution which has been hit by a market crisis is likely to be very different from the approach taken by an authority resolving an institution which has been hit by a credit crisis. The first involves large amounts of shortterm liquidity provision, some credit support and the likelihood of a positive return on any capital provided. The second involves the provision of long-term credit support, the certainty of large losses being incurred, and a difficult policy issue as to how those losses should be borne and by whom. However, these issues exist alongside the structural separation issue, and are not helped or hindered by it. Thus the primary issue arising out of bank resolution is the interaction of structural separation within the institution to be resolved and the group resolution technique adopted as regards the group concerned.

6. Resolution Techniques and Bank Group Structures In order to think about the impact of structural separation within bank groups on resolution, it is necessary to begin with the likely course of that resolution. In general, bank resolution can be conducted in one of two ways; single point of entry (SPE) or multiple point of entry (MPE). The Financial Stability Board¹⁴ defines SPE and MPE as follows Single point of entry (SPE) involves the application of resolution powers, for example, bail-in and/or transfer tools, at the top parent or holding company level by a single resolution authority – probably in the jurisdiction responsible for the global consolidated supervision of a group. An SPE strategy operates through the absorption of losses incurred within the group by the top parent or holding company through, for example, the writedown and/or mandatory conversion of unsecured debt issued by that top company into equity (“bail-in”). Provided that sufficient LAC is available at the top parent or holding level, operating subsidiaries should be able to continue as a going concern without entering resolution. 9 However, host authorities may need to exercise powers to support the resolution led by the home authorities. Multiple point of entry (MPE) involves the application of resolution powers by two or more resolution authorities to different parts of the group, and is likely to result in a break-up of the group into two or more separate parts. The group could be split on a na-

 FSB July 2013 “Recovery and Resolution Planning for Systemically Important Financial Institutions: Guidance on Developing Effective Resolution Strategies”.

Structural Separation and Bank Resolution

157

tional or regional basis, along business lines, or some combination of each. The resolution powers applied to the separate parts need not be the same and could include resolution options, such as bail-in within resolution, use of a bridge entity, transfer of business or wind-down. MPE strategies nevertheless require actions to be coordinated across jurisdictions so as to avoid conflicts or inconsistencies that undermine the effectiveness of the separate resolution actions, a disorderly run on assets and contagion across the firm.

An important point here is that SPE and MPE are names of resolution approaches, not types of corporate structure. Some bank groups are structured in such a fashion as to facilitate SPE resolution (for example, by having a large amount of bail-in-able debt issued out of an otherwise “empty” holding company). Others are structured in such a fashion as to facilitate multiple point of entry (for example, by operating through separate national subsidiaries and limiting intra-group exposures and dependences). As a result, we tend to speak of banks as “MPE” or “SPE” structures. However, there can be no guarantee than an SPE bank will in fact be resolved by the home supervisor of the holding company, and equally there can be no guarantee that an MPE bank will in fact be resolved by the closure of a ringfenced entity. In practice the resolution authority will decide the resolution approach to be adopted in the circumstances of the crisis concerned, and there is no way to know in advance what that decision will be. However, that having been said, it remains true that a group which is structured to be most easily resolvable on an SPE basis is likely to be resolved in that way, and it is reasonable for all those involved in dealing with such a group to presume that this will be the case.

Single Point of Entry An SPE approach involves recapitalising a bank group at the holding company level, either through writing down debt issued directly by the holding company or by writing down debt elsewhere in the group and upstreaming the proceeds. SPE is generally underpinned by a belief that the group is worth more than the sum of its parts, and by dealing with problems at a group level the minimum value destruction can be achieved. MPE, by contrast, involves hiving off a troubled member of a group and allowing it to fail, allowing the remainder to continue operating. MPE is generally underpinned by a group structure which minimises interdependencies (financial, managerial or technological) within the group in order to ensure that one component of the group cannot drag down others. The essence of SPE resolution architecture is that intervention occurs under the supervision of a co-ordinating resolution authority at the ultimate holding

158

Simon Gleeson

company level. In theory, debt issued at that holding company level is written down, the proceeds used to recapitalise other members of the group, and those entities are restored to solvency without ever entering any formal process. The whole aim and object of an SPE resolution is to avoid any sort of formal process being commenced in relation to any of the operating subsidiaries of the holding company concerned. It is therefore a matter of complete indifference to the supervising resolution authority how the business which is undertaken by those entities is divided up between them. If the purpose if the exercise is not to touch the legal entities, the distribution of business between those entities cannot, by definition, be relevant to the conduct of the resolution. It is therefore clear that in order to argue that functional separation within a group improves resolvability, it is a necessary assumption that the group concerned will be resolved using an MPE rather than an SPE approach. Since the vast majority of global systemically important financial institutions have resolution plans based on an SPE approach (the major exceptions being HSBC, Santander and BBVA), this alone suggests that such separation does not, except in a few cases, have any utility as regards enhancing the resolvability of banks. However, it is now necessary to consider the issue as it applies to MPE banks.

Multiple Point of Entry The issues with MPE in this context are more complex. It is clear that an MPE structure does not mean that every single legal entity in the group will be separately resolved – MPE groups tend to operate in clumps of legal entities rather than as completely coenobitical arrangements. Thus when we speak of MPE resolution we are in reality speaking of a number (in some cases, a large number) of SPE resolutions. The first question for an MPE group is therefore whether it matters whether each of those separate sub-groups is divided by activity? The simplest approach to this is to say that the SPE argument as a whole will apply to any sub-group which is approached in the same way. If you hypothesise a sub-group of a global bank in country A, it is unlikely to matter very much whether its activities are undertaken as two separate entities or a single entity, provided that those two activities are conducted under a common country A holding company. If the MPE strategy results in the closure of some of the activities in country A, it is unlikely to be possible to preserve other activities in country A, whether they are under the same holding company or not. Thus, in order to examine this further, it is necessary to ask why the MPE bank is structured in this way in the first place.

Structural Separation and Bank Resolution

159

As an indefensible generalisation, the banks which have adopted MPE approaches have tended to be those with a significant amount of their activities established in emerging markets. These operations tend to be conducted through separately established and capitalised subsidiaries for two reasons. One is that national governments in such countries require domestic operations of any size to be conducted through a locally-established subsidiary – not least because this arrangement gives the government concerned more control over the domestic activities of the bank in question and provides it with a mechanism to control inflows and, more importantly, outflows of its currency¹⁵. The other is that local activities are very vulnerable to a collapse in the credit of the government of the country concerned, and a local activity can be bankrupted either by a local bank failure or by other measures (for example “pesoification” in Argentina). In such a case, it is useful to be able to ringfence local creditors within the bankrupt local subsidiary in order to avoid the bank becoming a de facto guarantor of the credit of the sovereign. Although these are the historic drivers of MPE group structures, they have recently been joined by a number of others. The most important of these is the collapse of trust and confidence between governments and national bank resolution authorities which resulted from the crisis. Put simply, for a bank incorporated in country A, the less confident any other authority is that the authorities in country A will manage a resolution in an efficient and non‐discriminatory way, the less likely they are to allow that bank to operate in their territory through a branch, and the more likely they are to require a subsidiary. This is not always in the best interests of the creditors concerned, since such ringfencing results in those creditors being limited to local assets in respect of their claims, as opposed to being able to pursue the worldwide assets of the entity. However, for national resolution authorities, this structure ensures that they are able to manage local resolution locally, without becoming dependent on foreign authorities, and – more importantly – ensures that any national taxpayer funding which is provided is clearly applied to the payment of local creditors, and is not sent abroad for the potential benefit of foreigners. The result of this is that global bank groups are currently under stronger pressure than ever before to divide themselves nationally into subsidiaries, and this trend (known as “balkanisation”) is clearly visible in the industry. We are some way away from seeing the end of cross-border branching, but it is under pressure.

 It should be noted at this point that this motivation is by no means confined to emerging markets – this is also the Federal Reserve’s justification for its Section 165 FBO rules.

160

Simon Gleeson

Thus a typical MPE bank will be divided along national lines, and the question which we have to address here is as to how a functional separation might interact with such a division. If we simply assume that each national component of an MPE bank would, all other things being equal, be a full-service bank, then the conceptual model that we end up with is that the two sets of separations can happily coexist with each other – thus, each national operation can be divided into a banking and a non-banking business. The problem here is that by separating the businesses economies of scale will be destroyed. Thus even if we take a group whose investment banking business, taken as a whole, is economical, if that business is separated into both national and functional units, it is likely that many of the national trading businesses may be unsustainable and will have to be closed. This, however, takes us to the second driver that investment banking businesses, as a result, tend to operate on a cross-border basis. A comparison of a typical global banking business with a typical global investment bank will reveal that the investment banking business operates out of a very small number of legal entities relative to the number of legal entities which exist within its group structure – for example Goldman Sachs, in its public resolution filing of 2013¹⁶ notes that it has offices in over 30 countries worldwide, and that 49 % of its staff were permanently located outside the US, but has material legal entities in only four countries.¹⁷ Thus one outcome for the MPE bank is to consolidate all of its national investment banking activities into a single group-wide investment banking entity which will then operate as a single firm. It is a matter of observable fact that in the largest MPE groups, individual banks tend to operate predominantly within a single country. In effect, the adoption of this strategy embeds an economic bargain between the group and the supervisor in the country concerned – we, the bank, will hold sufficient assets in your jurisdiction to ensure a surplus of assets over liabilities (equals capital), thereby providing you with reassurance that depositors in your country will not be left high and dry. Conversely, we the bank reserve the right to say to you the regulator that we will not provide new funding past a certain point to the subsidiary concerned, even if the group as a whole is solvent and could afford it. The supervisors and regulators of a subsidiary of an MPE group are likely to be very sensitive to the interconnectivity of the subsidiary with the remainder of

 http://www.federalreserve.gov/bankinforeg/resolution-plans/goldman-sachs-1g-20131001. pdf.  The US, the UK, Japan and – perhaps surprisingly – Mauritius.

Structural Separation and Bank Resolution

161

the group. Their aim will be to ensure that the individual national subsidiaries of the group are capable of continuing on unaffected if problems arise elsewhere within the group. This is why the FSB, in its guidance on MPE strategies, says “… exposures between group entities may need to be reduced, and any intragroup funding should be provided at arm’s length and subject to appropriate large exposure limits”.¹⁸ In general, the larger bank subsidiaries of the largest MPE groups will tend to have investment banking activities both at a group level and at national level. This is because transactions with the largest clients are likely to be booked in a single entity in order to maximise netting benefits, whereas transactions with smaller national clients are likely to be dealt with at local level for relationship reasons. The issue, therefore, is as to how structural separations within MPE banks interact with a predominantly national subsidiary structure. This to some extent takes us back to the nature of the separation concerned. Volcker-style separations are the easiest to comprehend on this basis – a Volckerstyle separation effectively operates at the group-wide level (subject to exemptions for overseas operations which may or may not be present). Vickers-style separations are resolutely national – they are aimed at protecting depositors in one country, and are likely to be rolled out on a one-by-one basis. It seems very unlikely that any other country will follow the UK down the Vickers route, so this issue can be put to one side for the time being. That leaves Liikanan-style separations which, by contrast, are aimed at excluding an activity which appears to be risky from an activity which appears to be less so, and as a result tend to operate at a group-wide level. The effect of a Liikanen-type separation is to “dice” the group concerned – that is, a group which is already “horizontally” separated into a number of national subsidiaries must now have each one of those subsidiaries “vertically” sliced into a market and a non-market business. The first thing to say about this outcome is that many of the national markets businesses thus separated are unlikely to be viable as free-standing business units on grounds of size. This presents the group concerned with a challenge – it must either close down an element of its customer-facing activities, thereby reducing its ability to service its clients and losing ground against local bank competitors not subject to an equivalent requirement; or it must aggregate its national activities into a larger multi-national markets business which has sufficient scale to operate profitably. There is nothing logically wrong with this idea – as noted above, the service of investment banking is frequently provided

 FSB, Recovery and Resolution Plans for Systemically Important Financial Institutions: Guidance on Developing Effective Resolution Strategies, July 2013.

162

Simon Gleeson

to many jurisdictions from establishments – and balance sheets – located in only one or two. The difficulty which is created, however, is that requiring a development of this kind would potentially seriously undermine the basis of the MPE structure itself. A host supervisor for a subsidiary of an MPE bank will place a great deal of reliance on the fact that all of the business of the group concerned in his jurisdiction is done out of a legal entity which is subject to local resolution and insolvency, and whose assets and liabilities will, in a crisis, be under his sole control. If the bank concerned then starts doing part of its business in that jurisdiction on a cross border basis, the host regulator may conclude that the whole basis of the MPE strategy is being undermined, since the cross-border investment firm appears to provide precisely the mechanism for rapid withdrawal of assets and liquidity from the jurisdiction which the host is likely to be eager to prevent.

7. Conclusion The arguments that structural separation has any benefit on the management of risk exposures within a bank are unsustainable, and the only argument which appears to have any real plausibility is the argument that such separation will make resolution easier. However, this is by definition irrelevant in a situation where – as with most global SIFIs – the group concerned is to be resolved on an SPE basis. As regards groups which are to be resolved on an MPE basis, the typical MPE approach is based on separation by jurisdiction, not by function. Overlaying functional separation on top of country-by-country separation may well force existing MPE groups away from their current model towards a hybrid model where some activities are effectively forced to be conducted on a crossborder basis. This development would interfere with the clean break concepts which are at the heart of the approach to MPE resolution, and arguably make the application of an MPE approach harder. Consequently we have been unable to identify a resolution strategy which would benefit from structural separation within the group to be resolved.