40,99 €
Mehr erfahren.
- Herausgeber: John Wiley & Sons
- Kategorie: Fachliteratur
- Sprache: Englisch
Effective risk management in today’s ever-changing world
Crisis Wasted? Leading Risk Managers on Risk Culture sheds light on today’s risk management landscape through a unique collection of interviews from risk leaders in both the banking and investment industries. These interviews zero in on the risk culture of organisations, effective risk management in practice, and the sometimes paradoxical effects of new regulations and how they affect decision-making in financial organisations They offer genuine insight into regulatory processes and priorities and their implications for the stability of the global financial system. As trending topics in the risk management field, each of these subject areas is relevant to the work of today’s risk management professionals. In addition to the forward-focused text, this reference provides access to a wealth of premium online content.
Risk management has become an area of focus for companies since the financial crises that shook the international community over the past decade, but, despite high levels of introspection and changes to key processes, many financial houses are still experiencing large losses. Understanding today’s risk environment can help you improve risk management tactics.
- Access essential information both in print and online
- Discover the most important topics in today’s risk management field
- Explore interviews with 1 risk management leaders
- Learn about ground-breaking recent innovations in risk management thinking
Crisis Wasted? Leading Risk Managers on Risk Culture is an integral resource for professionals responsible for minimising organisational risk, as well as those who want to better understand the risk culture of today’s world.
Sie lesen das E-Book in den Legimi-Apps auf:
Seitenzahl: 649
Veröffentlichungsjahr: 2015
Ähnliche
“I think we're too comfortable in our assumption that it has been mended. Rather, I think the conditions are setting us up for another very real crisis somewhere.”
Richard Meddings
“Arguments are even being made that banks should be allowed to regulate themselves. Imagine.”
Adrian Blundell-Wignall
“Look, being a … risk compliance guy means you are sitting there and you are watching all the kids in the pool and you know somebody is doing something they shouldn't be doing in the pool but you don't know who the hell it is.”
Michael Hintze
“For me it was more about who's making money, and why is he is making money, and can he explain to me in an intuitive way how he is making it?”
John Breit
“But whether we really need a regulatory rulebook with more than 10,000 pages (Dodd-Frank), I wonder.”
Hugo Bänziger
“I don't know how many pages of forms would give you the information that you get from meeting somebody face to face and asking some pertinent questions.”
Paul Bostok
CRISIS WASTED?
Leading Risk Managers on Risk Culture
Frances Cowell
Matthew Levins
This edition first published 2016 © 2016 John Wiley & Sons, Ltd
Registered officeJohn Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, United Kingdom
For details of our global editorial offices, for customer services and for information about how to apply for permission to reuse the copyright material in this book please see our website at www.wiley.com.
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher.
Wiley publishes in a variety of print and electronic formats and by print-on-demand. Some material included with standard print versions of this book may not be included in e-books or in print-on-demand. If this book refers to media such as a CD or DVD that is not included in the version you purchased, you may download this material at http://booksupport.wiley.com. For more information about Wiley products, visit www.wiley.com.
Designations used by companies to distinguish their products are often claimed as trademarks. All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners. The publisher is not associated with any product or vendor mentioned in this book.
Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. It is sold on the understanding that the publisher is not engaged in rendering professional services and neither the publisher nor the author shall be liable for damages arising herefrom. If professional advice or other expert assistance is required, the services of a competent professional should be sought.
Library of Congress Cataloging-in-Publication Data
Cowell, Frances, author. Crisis wasted? : leading risk managers on risk culture / Frances Cowell, Matthew Levins. pages cm Includes bibliographical references and index. ISBN 978-1-119-11585-4 (cloth) 1. Financial services industry–Risk management. 2. Financial institutions–Risk management. 3. Financial risk. 4. Investment advisors. I. Levins, Matthew, author. II. Title. HG173.C69 2016 332.1068′1dc23
2015025690
A catalogue record for this book is available from the British Library.
ISBN 978-1-119-11585-4 (hardback) ISBN 978-1-119-11587-8 (ebk) ISBN 978-1-119-11586-1 (ebk) ISBN 978-1-119-11588-5 (obk)
Cover design: Wiley Cover image: (c) andrey_l/Shutterstock
Frances dedicates the work to Josie.
Matthew dedicates the work to Sonia.
Contents
Preface
Themes and questions
Research
Engaging the faculty of interviewees
Execution
Acknowledgements
Notes
About the Authors
1 Setting the Scene
Notes
2 Background
Behaviour
Modelling and measurement
Regulation
Notes
3 Sir Michael Hintze
Notes
4 John Breit
Notes
5 Bill Muysken
Notes
6 Hugo Bänziger
Notes
7 Carol Alexander
Notes
8 Mark Lawrence
Notes
9 Paul Bostok
Notes
10 Todd Groome
Notes
11 Richard Meddings
Notes
12 Adrian Blundell-Wignall
Notes
13 Innovations
How to analyse risk appetite and risk tolerance
How to quantify risk appetite for an economy
Why models seem to fail
How to quantify illiquidity
Signals of market instability
Analysis of connectivity and contagion
Notes
14 Interpretation
Opportunity wasted?
Failure of regulation
What has been done so far and how useful is it?
Watering down
The burden of rules
Paradoxical and counterproductive
Could go either way
Time to reflect on theoretical underpinnings
Tower of Babel
Limits to regulation
Notes
A Risk Silos
Investment risk
Liquidity risk
Counterparty risk
Operations risk
Legal and compliance risk
Note
B The Mechanics of Selected Financial Products
Exchange-traded v over-the-counter transactions
Swaps
Securitisation and collateralisation
Asset-Backed Securities (ABS)
Structured Investment Vehicle (SIV)
Collateralised Debt Obligations (CDO) and Collateralised Loan Obligations (CLO)
“Do it yourself” structured products
Options and option implied volatility
Volatility and variance swaps
Option replication
Gaussian Copulas
Notes
C Basel I, II and III – Risk Weightings
Note
Glossary of Terms
Glossary of People
Further Reading
Bibliography
Index
EULA
List of Illustrations
Chapter 11
Figure 11.1
Asset Capacity of Banks
Figure 11.2
The Squeeze on Credit Capacity
Guide
Cover
Table of Contents
Preface
Pages
ix
x
xi
xii
xiii
xiv
xv
xvii
xviii
xix
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
295
296
297
298
299
300
301
302
303
304
305
306
307
309
310
311
312
313
314
315
316
317
318
319
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
347
348
349
350
351
352
353
354
355
356
357
358
359
361
362
363
364
365
367
368
369
370
371
Preface
In a Paris café in April 2013, the conversation turned to how the financial services industry was shaping up following the global financial crisis. We – Frances and Matthew – shared an uneasy intuition that some developments, especially in regulations – despite the best intentions to the contrary – could have the result of actually making the financial system more fragile. Risk management and governance failures, such as the 2012 London Whale, where JP Morgan incurred massive losses in CDS contracts, and similarly massive losses in 2008 at Société Générale caused by a rogue trader in stock index derivatives, continue to happen, suggesting that the professional risk management put in place by large, sophisticated organisations was not working as it should.
There was a sense that an opportunity was being wasted and the debate about what was needed to reinforce financial stability was somehow being hijacked by a combination of popular mis-information, political short-sightedness and special interests. But was this merely a perception stemming from our common perspective? What did others think?
From casual conversations, it was evident that many friends and colleagues shared our views, prompting questions about how widely they were shared outside our circles. In particular, what other factors were operating that we were perhaps unaware of? What extra insights could be had from the best informed market participants and observers?
Matthew proposed a formal exercise: to ask those who were most caught up in it as it all happened. The transcripts of the conversations would result in a readable and substantial book to make available to decision-makers, in the financial services industry and elsewhere, insights from the inside point of view, thereby leading to a more constructive and informed debate and perhaps even influencing decisions in favour of sensible regulation that would reinforce stability not undermine it.
Both having been risk takers turned risk managers, in banking and investment management respectively, since the early 1980s, we had therefore lived through a number of financial crises. This experience allowed us to select and engage an interesting and impressive faculty of interviewees, and to pose good, pointy questions that would fuel debate.
We reasoned that, like us, our readers would want to know why some developments since the crisis have been so disappointing. Like us, they would like better to understand the sub-agendas that operate on decision-making in financial organisations, their regulators and supervisors; to leverage the perspectives and insights of insiders and to understand what is really going on.
Constructive debate necessarily entails a forward-looking orientation. That said, given that the aim is to mitigate the effects of a future financial crisis, reflection on past events is inevitable and can be instructive. But the past must be kept in perspective, since the conditions that prevailed then will not be repeated exactly in the future. To paraphrase the ubiquitous risk warnings in prospectuses, the past is not necessarily a guide to what will happen in the future. Learn from the past, but don't extrapolate from it.
Our first steps were to gather initial feedback and ideas – for practical execution as well as to crystallise our thinking about issues we should address and that would emerge during our inquiry. In doing this, we were struck by the general enthusiasm for the project (only one person was not whole-heartedly supportive). This was of course due partly to its likely commercial appeal, but more important was that so many people saw it as a project that needed to happen. It seemed that even more people than we had imagined agreed that a new level of debate was called for.
Themes and questions
The next challenge was to decide how structured or otherwise the conversations should be. Certainly, readers would benefit from some comparability between the views of each interviewee, which suggests some structure; but interviews that are too constrained or structured would risk limiting the potential for unique insights. What people choose to talk about is as important as what they say.
This led to the idea of common themes, or threads, which would connect the conversations while allowing scope for each interviewee to address the issue closest to his or her heart. The themes were thus conceived as parameters rather than boundaries.
In developing the themes, we were keen to leap-frog arguments that are already well aired, such as conflicts inherent in the way many investment bankers are rewarded. But where we found that we could add a new perspective to an existing debate, we have done so, and so we include themes such as the complexity of models for valuation and risk, contradictions in regulatory regimes and possible unintended consequences.
Lots of “brainstorming” and plenty of war stories later, about 20 somewhat inter-connected themes emerged. These in turn boiled down to three macro themes: behaviour, models and regulation.
Within each theme, the challenge in framing the questions was to be specific enough to give the interviewees something to get their teeth into, but open enough to be relevant to more general issues.
Behaviour, the first theme, covers organisations and the people who work in them – what is increasingly referred to as organisations' risk culture: how it regards risk. A good indication of an organisation's attitude to risk is the status it accords risk management. This seemed to us a good place to start our questions too. Our experience of working in different organisations told us that this is about more than the organisational hierarchy, it is also about the relationship between the risk manager and risk takers – at all levels of the organisation, from the dealer and the risk analyst to the CEO and the CRO. For example, how equal are they in the event of a dispute? How equal can they be? What mix of personalities works best? What organisational structure works best? Which day-to-day practices work best?
The next bloc of questions is about how well equipped the risk manager is. How much has the organisation invested in the tools, and the skills needed to deploy them effectively, both to measure and report risk, and also to manage it? Aware that many readers are unfamiliar with how risk models work, we have done our best to present these questions in a way that does not demand any technical knowledge. (Technical explanations, for those interested, can be found in the appendices.) We have therefore aimed for questions that steer clear of the arcane details of risk measurement and focus instead on what their users think of them. How useful are the risk reports that inform supervisors, regulators and investors? The world is complex, should the models be too? Should everyone use the same one, or does this lead to herding that exacerbates systemic risk?
The last and most extensive theme concerns regulations. What is likely to work and what is not likely to work? Why? What might be the unintended consequences? Might there be a better way of doing it?
Another challenge was to avoid imposing our views on the interviews. This is easier said than done, since we draw heavily on our own experience to formulate questions. We therefore were at pains to couch questions in as even handed a way as possible, to avoid leading the witnesses. Nevertheless, we acknowledge that the very choice of questions can affect the course of conversations. Even so, as each conversation developed, our views are sometimes, unavoidably, drawn out.
Research
Reading around our themes and talking with friendly experts was an adventure in itself. It uncovered a number of rich veins of research, including valuable early and ongoing work on the theory of regulation, original approaches to understanding and estimating risk appetite and of calibrating market instability, as well as very thoughtful pieces on the pros and cons of analytical versus behavioural ways of managing risk: detailed VaR reports versus an hour in the pub with some traders.
Perhaps the most rewarding was that we learned about a number of innovations that promise to address some tricky problems that came up during our conversations. Because these innovations are concrete evidence that some good has indeed come from the crisis, they seem valuable enough to merit their own chapter. They are a stimulating challenge to the book's title.
For a selection of other books and articles that stood out, interested readers are directed to the Further Reading section.
Engaging the faculty of interviewees
Clearly the interviewees are selected to achieve the right complement of perspectives, without regard for the views that they might hold. They span banking, investment banking and investment management, which together can be thought of as capital formation, capital transfer and capital warehousing. Or, put another way, risk formation, risk transfer and risk warehousing. We regard the three functions as inseparable and co-dependent in a capitalist system in the same way that respiration and photosynthesis are mutually dependent in a world of carbon-based life forms. The views from academia, regulators and supervisors complement practitioners' perspectives.
Cliché it may be, but modern capital markets do operate within a single, global ecosystem. So we sought geographical diversification – though in reality the debate is dominated by North America and Europe, since that is where, for the most part, the crisis happened.
In an important sense, engaging faculty members was the easiest part of the project. Our invitations almost all met with enthusiastic acceptance, including two participants to whom we were unknown before the interviews and to whom we had no formal introduction – cold emails, in effect. This we received as an endorsement: people really believed that the project was needed; we were genuinely flattered that people of such calibre were so supportive! Nevertheless, two were obliged to decline to be interviewed. Unfortunately they were from the same geographical and professional group, which is therefore less represented than we would have liked it to be. Otherwise, the faculty of interviewees is a good illustrative sample of the financial industry.
Execution
Conducting the interviews was always going to be the fun bit of this project, but it was much more fun than either of us anticipated. We were genuinely delighted by how generous our faculty were with their time and the frankness with which they expressed themselves. Their initial enthusiasm did not wane throughout the interviews – if anything, the opposite is true, and we hope this is conveyed in the transcripts you are about to read.
A number of things stand out from the interviews themselves. The first is that nearly all our faculty attribute the success of their careers to close contact with very thoughtful (and often now famous) people. Many had the good fortune to be mentored or otherwise supported by thought leaders in economic and investment theory. But in pondering this luck we remind ourselves of the old adage that one makes one's own luck. If luck was involved, then it was, in each case, at least matched by talent, discipline and hard work. Working with our faculty was correspondingly stimulating and enjoyable.
Another thing that stands out is both the divergence and the commonality of views expressed in the interviews. We sought debate and we got it! It was tempting at times to raise an argument that had been put forward by a previous interviewee, but this impulse had to be contained in the interests of giving approximately equal air time to each point of view, and of not favouring subsequent interviewees over earlier ones. This is more important than it might seem: due to the practicalities of travel schedules, an unfortunate side-effect would have been that it would have favoured some continents over others, and any bias so introduced might therefore have been systematic.
Some observers of the book's gestation wondered aloud about the “tricky” task of deciding in what order interviews should appear: first, last and in the middle. Anyone who has organised, addressed or merely attended a conference or seminar knows how sensitive this can be. Exercising a mix of impartiality and judgement, we took the view that each interviewee, or at least most, would give more emphasis to some themes than to others, and that this would implicitly determine where in the book their interview should sit, once the themes' order in the book had been decided.
For the order of the themes themselves, it seemed logical to apply the order in which questions had been arranged, beginning with behavioural themes, such as the relationship of risk management to the rest of the organisation, followed by what we need to understand about risk tools and risk reports, and what it all means for regulation and policy and how we approach the possibility (or certainty, depending on your point of view) of another financial meltdown. Risk culture features both at the start, with the place of risk management within individual organisations; and at the end, with regulators' increasing interest in risk culture within organisations.
It turned out that each interviewee did tend to favour some themes over others, which is serendipitous, as it means that we were able to avoid that delicate decision and let participants place themselves.
To complete the book, we needed to decide how much and what kind of material we, as “authors”, should add. We believe that readers are perfectly capable of deriving their own inferences and conclusions from each interview, so we butted out, keeping our contribution to each interview chapter to a bare minimum. At the same time, each question and each answer derived from a narrative, or specific risk management issue, to which many, if not most, readers would not be privy. The Background chapter aims to provide a sort of primer on the themes, and the issues concerned, so that readers from any background can engage in and follow the conversations with a minimum of footnotes. Some of the background material supposes familiarity with some technical detail that, if included in the text, most readers would find an arcane disruption to the narrative, but cannot be left out entirely. We compromise by giving an intuitive explanation in the appendices.
Setting the Scene opens the book with a reminder of what set off the wave of regulation and reform of financial organisations, and why a new level of informed debate is now needed. Our other contributions are the Innovations chapter and Further Reading section already mentioned, and a final, Interpretation chapter, in which we bring the various themes together and suggest some conclusions.
In Crisis Wasted?, we believe we have succeeded in our aim of drawing out informed opinions. Readers will agree with some of the views expressed and disagree with others. We invite you to express your view on www.riskculture.today. We look forward to hearing from you.
Acknowledgements
The genesis of this book is due to a combination of the global financial crisis of 2007–08 and two collections of conversations with interesting people: The Super Analysts1 by Andrew Leeming, an erstwhile colleague, and Arjo Klamer's Conversations with Economists.2
As we developed our ideas and worked out the practicalities of preparing and writing the book, we were cheered on by nearly everyone we spoke to. We are particularly grateful to a number of friends and colleagues, past and present, for their encouragement and support.
Apart from giving us the idea for the book, Andrew also gave very generously of his time, encouraging us to benefit from his experience, helping us to avoid plenty of pitfalls and make the most of our resources. Our faculty of interviewees don't know it, but they have Andrew to thank too: he cut the questions that we asked them to think about before the interview down from 20 to five pages!
Andrew, together with John Beggs, Ron Bewley, Jillian Broadbent, Craig Davis, Mike Katz, Mark Lawrence, Rodney Maddock, Kevin Nixon and Peter Warne, all helped crystallise our thinking about the themes and questions, and directed us to useful background material. Ophelia Cowell came up with some helpful suggestions about how to prepare and present our questions to the interviewees. Mark Kritzman treated us to a very stimulating and absorbing hour and a half on topical innovations that have occurred as direct responses to the financial crisis. Gerald Ashley contributed an interesting and very informative conversation about the nature of systemic complexity and how to model it.
Andrew, Angela Campbell, Mike Katz, Mark Lawrence, Jason MacQueen, Rosemary Thorne and Peter Warne suggested and helped arrange contact with our faculty.
Eugenie White, Tom Wheelwright, Malcolm McIvor, Heather Loewenthal and Alberto Arabia all read the work in progress and gave generous, enthusiastic and constructive feedback.
Sonia and Brigitte Levins took on the critical task of transcribing the interviews from their audio format.
Sonia and Alberto held our respective hands.
Thomas Hyrkiel and Werner Coetzee of Wiley made us feel very special throughout the project with their enthusiasm, responsiveness and support. Jeremy Chia, also of Wiley, has been a constructive and very responsive editor.
Most of all, the people who gave their time and not a trivial amount of thought and effort, first to make themselves available to be interviewed and then to help with the hard work of editing the raw transcripts into a form that people will enjoy reading.
We owe a big thank you to each of them. Thank you!
Notes
1
About the Authors
Frances Cowell is a specialist investment risk consultant working with R-Squared Risk Management in Paris and London. She is a founding director of the London Quant Group, a not-for-profit that provides a forum for discussion of practical issues in quantitative investment techniques for the investment management industry. She previously worked as an investment manager at NatWest Investment Management in Australia, having helped pioneer risk-controlled trading in the nascent market for financial derivatives in Australia. Following her move to London, she was Head of Investment Risk at Morley Fund Management (now Aviva Investors) and CRO for CCLA Fund Management.
Matthew Levins is a risk consultant with a practice that spans banks in China, Asia and Australia. Previously, he directed trading, broking, capital market and risk practices in Australia, working for leading firms such as the Commonwealth Bank of Australia and Bankers Trust Australia. In the early 1980s, he took a leading role in the nascent market for financial derivatives in Australia. This entailed developing and refining robust decision rules to support underwriting and trading in fixed income swaps and options, as well as equity, soft commodity and foreign exchange options. He effectively helped pioneer much of the work that informs current hedging and risk control in modern banking.
1Setting the Scene
Between 12 and 21 December 2006 the data for asset-backed securities index futures (ABX 2006-1 AAA spread) showed a small but significant departure from its normal daily pattern of price fluctuations. This coincided with rumours that a trading division of a major US investment bank had expertly quit its collateralised debt obligation (CDO) portfolio. Over the ensuing months, analyst reports began to circulate that led some banks to delve into their loan portfolios to see if they had exposures to these securities. What sparked this activity? The US housing prices had begun to trend down, many “low start” loans were approaching anniversaries when their interest payments would “step up”. In February 2007, HSBC recognised the problem in its New York branch with a $10.5 billion charge. Market makers reacted by forcing a 350% increase in volatility.
Hitherto, the originating banks had bundled up these higher interest loans and securitised them. The resulting pools of securitised debt were re-packaged into CDOs, where equity and debt investors could participate in a preferred tranche and trade off a levered return against reduced exposure to losses on default relative to the overall pool of loans. New products were dreamt up, such as funds that borrowed to lever further the promised returns that would rid the originating investment banks of the inventories of higher risk equity tranches that were accumulating in their books.
Underpinning the market were assumptions about average rates of default that these loans would see. The US government had encouraged products like low-start loans as a means of providing home finance to borrowers who had until then been excluded from home ownership. With little credit history to draw on, analysts had to strike an educated guess as to the likely number of defaults. From the beginning of rapid growth in “no-document” loans in 2004, it took until early 2007, when low interest payment inducements were due to expire, for the divergent trend of high defaults to emerge.
Possibly dismissed as “too soon to call”, the “low-doc” CDO machine continued to revolve at a breakneck pace. Documentation processes for these instruments, and indeed many other credit derivative contracts, fell way behind. The NY Fed called several round table meetings to gauge the depth of the problem and set an agreed remediation path. Yet new tickets1 continued, seemingly unabated, to be written.
Re-engineering the higher expected default frequency into securitisation and CDO models showed that these structures were over-promising returns. Higher frequency of defaults, like peeling away the layers of an onion, meant for CDO structures not only that regular interest payments to the investors were jeopardised, but that the expected loss of all capital increased. Curiously to some, the valuation effect was often most damaging in the so-called “super senior” tranches, rated “AAA”.
By mid-2007, the approaching savage repricing had yet to occur but the market place seemed to have reached saturation, and the new issues market showed signs of imminent closure. Some structured credit funds managed by the US investment house Bear Stearns imploded. As if to preserve capital, banks began to enforce tighter underwriting standards across their businesses. There was little they could do for loans already “baked into their book”, so the emphasis turned to re-financings and holdings of traded securities.
This hardening of underwriting standards saw hitherto high-quality securitisations or revolving finance facilities supported by even low-risk assets, such as portfolios of shopping centres, falter (and borrowers forced to liquidation). Banks also withdrew their support for loan notes issued by securitised vehicles.
Little did they at first realise, but to get the securitised assets off their balance sheets in the first place, the banks were required to provide a standby line of liquidity that would purchase the loan notes should a market buyer not be found. These requirements were, for the most part, an unplanned call on their balance sheets. Liquidity started to become scarcer.
Organisations such as Northern Rock, whose liability management was reliant on an “open” securitisation market, fell to a run by its retail depositors. A Canadian money market fund broke the buck. A CDO fund had to close. Around this time, with sponsorship by the Fed, Bank of America purchased Countrywide, a major West Coast provider of “low doc” loans. Credit markets generally began to sell off in an orderly way, while most other markets adjusted, with a lag, to tightening credit spreads. It was not until the northern winter before equity markets began to reverse their positive trend.
The manufacturers of CDOs to varying degrees financed their warehouses and production lines using repurchase agreements2 (repos). Again, these contracts took assets and liabilities off balance sheet. But, as with the standby lines in securitisation structures, to maintain contracts off balance sheet the borrower was obliged to top up collateral as its value deteriorated. Liquidity tightened further. Parcels of CDOs were being sold off as repos struggled to be rolled over or renewed. The increased supply of paper for sale saw lower and lower prices register, while waning confidence in the value of CDOs aggravated the liquidity effect. Liquidity became unobtainable for some borrowers.
The pressure to liquidate a difficult-to-value balance sheet became too much for Bear Stearns. In February 2008, it was sold to JP Morgan after the NY Fed established a company to buy $30 billion worth of assets at just 7% of their price immediately before trading in Bear Stearns stock was suspended.
CDOs continued to lose value. Statutory profits and reporting fell hostage to the accounting classification for warehoused stock. With each reporting period, waves of securities became categorised “level 3”, automatically requiring greater capital to be assigned to them. As it reeled from these effects, it became increasingly obvious that the banking system in the US was systemically undercapitalised. The Fed and US Treasury stepped in to restore order, but it was not until the decision to allow Lehman Brothers to file for bankruptcy that it became clear that the regulators had overestimated the effectiveness of their regulations.
In 2008 Comptroller Dugan reflected that while “investors should never rely exclusively on credit ratings in making investment decisions, the plain fact is that triple A credit ratings are a powerful green light for conservative investors all over the world”.3 And so it had been that in the preceding period re-packaging of these securities had not only spread throughout the US banks and money market funds, but to US agencies and beyond its shores to the balance sheets of financial institutions in other countries.
Write-downs and illiquidity combined to take down banks that had aggressively made loans to property developers. Icelandic banks failed, with the assets of the three banks taken over by its supervisor equal to more than 11 times the country's GDP – an early indication of the effective limits of financial globalisation. German Landesbanks in time announced their losses and needed to be re-capitalised. In the UK, banks were crippled, partly due to write-downs to goodwill and loan impairment on the balance sheets of recently purchased Dutch banking assets. And so it played out across most of the northern hemisphere and parts of Asia.
From initial insouciance about the prospect of Lehman Brothers filing for Chapter 11 following its death spiral of the preceding months, the act itself led to something of an avalanche throughout the US – investors in money market funds that held now worthless Lehman Brothers debt were unable to redeem shares for their $1 face value. A run on money market funds ensued, clogging the repo market that broker dealers depended on, and obliging the US Treasury to step in to guarantee money market funds. The global insurer AIG also needed volumes of support, as it had been the dominant provider of credit derivatives through its London branch.
Money market funds sought to liquidate assets to meet redemptions and fund the additional collateral now needed to secure financing. At the same time, the margins required in the repo market continued to increase, and it became impossible to use some assets as collateral for loans. Organisations that could, called funds from their foreign affiliates in order to meet the obligations of their US operations. US affiliates, mostly branches, of foreign banks were under even greater liquidity pressure than were local banks, partly because their access to deposits from US citizens and residents was limited. This forced them to raise funds from outside the US, further contributing to international contagion of the liquidity squeeze, with the effect felt most acutely in London.
As valuation of financial assets became more an art than a science, asset markets such as equities were sold down around the world, as agents sought to raise liquidity wherever they could. Major exchanges generally behaved well, but the continued downtrend was further fuelled by investors withdrawing commitments to managed money, leading to the closure of many funds.
Part of the problem in funding liquidity was that some assets pledged as security had been re-hypothecated by holders of the collateral. When collateral was recalled, often the specific paper was unavailable, obliging investors to sell other assets to meet their own margin calls and redemptions.
One consequence of the turmoil was that collateral requirements could not be calculated in the usual way. This is because margins and collateral requirements are calculated using statistical processes that draw on recent asset price return history. The turbulence now prevailing meant that these processes broke down, particularly for non-traded and over-the-counter instruments. Futures exchanges and prime brokers were thus obliged, in effect, to guess the amount of collateral required. This meant that “Chinese walls”, intended to avoid conflicts of interests within organisations, were tested, as prime brokers set collateral levels at capitulation points for “clients” who were in some cases part of the same organisation.
Shortly after the failure of Lehman Brothers the system became unmanageable in private hands. Governments, advised by their regulators, stepped in.
All will remember the steps taken by governments to traverse the crisis with the Troubled Asset Relief Program, and ultimately quantitative easing by the Fed in the US, the recapitalisation of British banks with public money, the decision of the Irish government to guarantee all bank deposits (which later led to its call for help from the European Union) and the need for the Swiss authorities to save UBS.
After throwing a TARP-o-line over impaired assets and entities and providing extraordinary amounts of public support, regulators set about digesting the outcomes. The “live laboratory test” of failure presented them with many avenues to follow as they set about re-calibrating regulation and supervisory procedures.
Notes
1
Transactions.
2
An arrangement whereby a bond is sold with a contract to repurchase it at a given date and price in the future. In effect, short-term borrowing.
3
Comptroller Dugan Outlines Steps In Response to Losses by Banks and Investors Holding Tranches of Securities Considered Safe, 27 February 2008, Press Release 2008-22.
2Background
I think we're too comfortable in our assumption that it has been mended. Rather, I think the conditions are setting us up for another very real crisis somewhere.
Richard Meddings
Spurred on by bank failures and government rescues in the northern hemisphere, regulators at the global and sovereign state level have embarked upon a decade's work to overhaul the governance of financial institutions. The scope of this enterprise has been immense following the prodigal global financial crisis and its aftermath. Banking regulation has seen new capital measures proposed, new risks identified with derivatives and collateral markets, more conservative balance sheet treatments for “special investment vehicles”, efforts to align remuneration with outcomes, and renewed emphasis on the role and culture necessary to support risk activities. Efforts have also entailed forcing products onto exchanges in the name of transparency, but little appears to have been done to invite more cautious trading. Regulators and organisations are, however, focusing on behaviour, culture and risk appetite.
In periods of moderate market stress, complementary risk appetites of different organisations reinforce the relative resilience of the system because of the co-existence of long versus short risk horizons and risk appetites that range from very low to very high. For example, long-term investors are able to stand in as buyers of even the most distressed assets that would add to liquidity pressures for short-term investors. But resilience can, in extreme circumstances, give way to contagion, whereby failure in one part of the system can cause distress in a seemingly unrelated corner.
How and why this happens are questions that many people believe remain open – it may be that the product set of long-term investors is narrower than those in the domain of traders. How to prevent it happening is another; as is to what extent the regulations now coming into force help to improve the stability of the financial system – and protect the wider economy from the worst consequences of instability. These and other questions are the subject of dialogues that make up the next ten chapters.
Conversations with market participants sometimes turn to subjects that are hard for an outsider to follow without some important background. This chapter aims to fill in the background to issues, or themes, that are missing from the conversations themselves.
Each faculty member represents a different point of view from the others, so articulating a unique set of themes. As expected, themes overlap with each other and a number of them recur, which serves to highlight areas of common concern or controversy. To provide context and to juxtapose, as closely as possible, overlapping and related themes, we identify three macro-themes, which we call:
Behaviour
Risk modelling and measurement
Regulation.
Readers will also find in the Appendices more background information about some of the issues and mechanics of risk management within organisations and a basic guide to how some complex products work. Individual terms are given in the Glossary.
Behaviour
Themes that reflect the culture within financial organisations are:
The role of the risk manager
Asymmetries of the risk management role
Professional risk managers
Independence
Judgement versus analysis
Why become a risk manager?
Risk silos
Crisis management.
The role of the risk manager
Regulators and investors expect risk management to be pivotal to the operations of financial firms. This requirement is partly due to the leverage the firms are allowed to employ and the rapid and multi-faceted way financial market volatility impacts on the firm. One initiative has been to give more prominence and authority within organisations to the Chief Risk Officer (CRO) than they previously enjoyed. The aim is to enhance the influence of the risk management function, and therefore its presumed effectiveness, as a means of, among other things, containing the organisation's susceptibility to extreme events.
Yet an easy way for organisations to satisfy demands for more focus on risk management can sometimes be to hide behind accepted wisdom, such as “industry best practice”, or more generally, the “reasonable person test”. An example might be elevating the remuneration of the CRO while containing his real influence by selecting an individual lacking the necessary personal qualities – or simply of not the right calibre – to balance the influence of risk takers. Kaplan and Mikes note in their 2012 article: “For all the rhetoric about its importance and the money invested in it, risk management is too often treated as a compliance issue.”1
So despite the value now generally accorded risk management, manifest in typically well-resourced and influential risk management functions, we still hear reports of outspoken risk managers being shown the door or simply ignored, for example because they voice disagreement with the firm's strategy.
Asymmetries in the risk management role
One argument is that the occasional over-riding of risk managers is an inevitable consequence of the asymmetries inherent in the risk manager's role: asymmetries of status, information and incentives.
First, it is clear that the risk manager should have equal status with risk takers so that neither is in a position to override the decisions of the other. Some believe that the CRO therefore must be a board-level appointment. Others would go further and say that even this does not guarantee that the CRO will not be over-ridden by the Chief Executive Officer (CEO) or a coalition of other board members.
Second, information about risk-taking activities that is available to the risk manager vis-à-vis the risk taker is inevitably incomplete and delayed. This happens because the risk manager does not receive the same stream of market information and signals that the risk taker does, and as a result he is necessarily at a disadvantage when a position or concentration of risk needs to be challenged.
Third is that the risk manager must command superior or a broader spectrum of skills when compared with specialist divisional heads. For a start, he must be adept in the profit-generating, risk-taking activities he oversees in order to manage the risks they embed. Without a full understanding of the investment manager's or banking divisional head's techniques and strategies and their inter-play with market conditions, the risk manager cannot hope to spot potential areas of weakness and recommend pre-emptive or remedial actions.
Yet even with enhanced status within the organisation, it is not clear why individuals of this calibre would choose to become a risk manager rather than a – usually better paid – trader or portfolio manager. This is especially so, given that the risk manager's job is also more risky: while dealers' and portfolios managers' remuneration profiles combine potentially large gains that are often linked directly to successful risk taking, with losses mostly limited to the loss of the individual's job; the risk manager's job has limited and less visible upside potential (crises averted, losses contained), but plenty of downside risk (being held accountable for all failures of risk management, real and perceived).
Elevation of risk managers' remuneration and alignment of revenue generators' remuneration to longer-termed revenue outcomes can partly address this asymmetry, but by necessity cannot redress it completely.
Professional risk managers
Many very talented people still choose this career. Their career paths have, in the past, tended to start by gaining experience in the operations function of the organisation (where things like settlements, valuations, reconciliations and accounting take place) through performance measurement and analysis, then to risk management – often with the aim of a front office, risk-taking role such as portfolio analyst, manager or trader. The result is that some risk takers have a background in, and therefore sympathy with, risk management priorities. Risk managers, in their turn, understand issues that can cause problems in diverse parts of the organisation.
As more resources are directed to managing risk, we see the emergence of career risk professionals who benefit from specialist technical training, but have little experience of other areas of the organisation, and in some cases see their ultimate career goal in the senior ranks of risk management rather than in risk-taking roles. While some organisations recognise the value of “cross-fertilisation” and take active steps to ensure that risk takers and risk managers gain experience of both, for example, in the course of training of graduate hires, not all do. In the absence of active cross-fertilisation, trust and empathy between risk takers and risk managers can suffer, often to the detriment of the risk manager's influence and utility.
Risk manager independence
It is generally accepted that the risk manager should be independent of the risk-taking activities he oversees. This means that, at a minimum, the risk manager's remuneration must be unrelated to the profit generated by those risk-taking activities. But there are different degrees of independence.
Some organisations see risk management as a sort of collaboration between risk taker and risk manager, whereby the risk manager gives an informed second opinion of risk-taking decisions. Achieving this collabaration entails fairly close and frequent contact between the two, so that the risk manager is constantly aware of what risks are being put in place. In this scenario, the risk manager often relies on the risk taker volunteering information about his activities; while the risk taker has enough confidence in the risk manager to share information freely and to see value in the risk manager's opinions.
Many argue that such close collaboration dulls the risk manager's independence and so could lead to some risks not being scrutinised as closely as they deserve. They argue that risk management should be quite removed from risk-taking activities in order to maintain its objectivity.
The counter-argument is that if the risk manager is too independent he can miss signals, aggravating the asymmetry of information between risk taker and risk manager and eventually reducing the role to that of risk control, depending heavily on risk models and reports rather than a thorough knowledge of the business.
This in turn can foster a “policing” approach to risk management, which at its worst can deteriorate into a game of cat and mouse. Distance between risk managers and risk-taking teams can also deter some talented potential risk managers, who otherwise might have seen it as a step toward a career in investment management or trading.
Judgement v analysis
Regardless of how independent it is, risk oversight clearly needs to be objective. But as with independence, there are degrees of objectivity.
Intuition and objective analysis clearly are both valuable, and most people believe that the ideal is some mix of the two. Yet most risk managers lean toward one or the other, relying mostly on either quantitative risk analyses or judgement. Annette Mikes describes this in her 2011 paper,2 in which she identifies cultures of “quantitative enthusiasm”, dedicated to risk measurement and reporting, and “quantitative scepticism”. The latter is oriented primarily to judgement, where risk managers interpret alternative intuitive future economic scenarios and emerging risk issues. Each approach has its adherents and detractors.
Quantitative risk management is characterised by a disciplined process that demands risk managers with excellent technical skills who function within it. It is easy to report, since reports consist mainly of statistics that can be compared over time and across parts of the organisation. Quantitative reports are scalable in that risk measures and portfolios can be added at any time. They are also easily aggregated to the level of the whole organisation, so that it can be represented in terms of its risk exposures. On the other hand, risks that emerge ad hoc are easily missed if the manager is too reliant on pre-specified risk statistics. Being primarily process driven, quantitative risk processes accommodate smoothly the arrival of new risk managers and departure of existing ones. They also tend to encourage a very specialist risk manager career path with less cross-fertilisation with other parts of the organisation.
Judgement-based risk management relies more on key individuals and their experience working with risk takers, drawing from their experience and intuition in order to spot ad hoc risks that may not be captured by pre-defined risk analyses. A feel for the sorts of unanticipated risks that help the organisation to respond quickly and effectively to unanticipated risks is at least as important as advanced technical skills. The ability to communicate clearly both the nature of the risk and how it is managed is critical.
Risk silos
Most people believe that risk is best managed holistically – at the highest level possible so that it takes account of all relevant covariances, inter-dependencies and the causation that amplifies downside risk. But this can be impractical in a very large organisation where the natural response is to “compartmentalise” risk management, for example into silos specialising in things like investment risk (market and credit risk, asset allocation and stock selection risk), liquidity risk, counterparty risk and operations risk. This can introduce the danger that unanticipated sources of risk “fall between the cracks”, in that they fail to be captured by any designated risk function. Adding to this, the separation of accountability into silos for profit centres, budgets and bonus calculations discourages collaboration and risk sharing across departments.
Enterprise-level risk management that links investment, counterparty, operations, legal and compliance, reputational risks and more is sometimes proposed as a means of ensuring that all risk is “captured” even when it does not fit easily into a defined risk category. But this may be difficult to implement in practice, for example, because it demands exceptional skills and complex system support.
Crisis management
Each crisis is different and so by definition is hard to plan for. This is made even harder because people can behave out of character in a time of crisis. One of the main tools for estimating exposure to crises is scenario analyses that simulate the effects of a market event. These usually draw on data sampled from the past (and so implicitly the aggregate impacts of people's behaviour), but may not take account of how individuals within the organisation respond.
Human responses to crises can conform to patterns; for example, with the stress of a major market disruption, managers understandably can lose confidence in the risk information they have to hand: its completeness can be questioned and information can be regarded as stale. Extreme caution can take over within individual firms, which, when aggregated across many actors, exacerbates market illiquidity and even volatility. As staff move between organisations, often across continents, they bring with them skills and practices, which, together with the complexity of financial products and models, can contribute to industry-wide group-think, which can add to contagion.3 Andrew Haldane notes in a 2011 speech that “risk perceptions are as important as reality. Behavioural economics tells us that financial crashes can leave lasting psychological scars on risk-taking.”
Modelling and measurement
The faculty identifies five aspects of models and an important measurement issue.
Risk management:
Subjectivity
Complexity
Calibration
Standardisation
Mark-to-market valuation.
Risk models – subjectivity
Even the most objective and rigorous models embed some subjectivity. Indeed, the very choice of modelling methodology is necessarily a subjective one, relying on the judgement of a single individual or a small group of individuals. Selection of data sampling, choice of risk metrics, design of stress tests, too, are all products of human judgement. For example, the specification of parameters and data sample for simulation analyses and risk measurement is normally a human input. René Stultz4 identifies six subjective inputs to risk models that contributed to their apparent failure during the crisis, the second of which is selection of inappropriate risk measures. The result is that objective risk models can actually amplify subjectivity by concentrating judgement-based decisions on the part of a relatively small number of people.
Michael Power, in his 2004 article,5 writes that “[m]odels and measures would be part of broader organisational narratives of uncertainty” and “[risk management] would depend essentially on human capacities to imagine alternative futures to the present, rather than quantitative ambitions to predict the future”. But he qualifies this: “To the extent that process represents the codification of accumulated wisdom, it should be sustained subject to the possibility of constant challenge.”
Risk models – complexity
The very complexity of sophisticated models for asset valuation and for risk modelling can be a source of uncertainty and risk if the models are not well-enough understood by those who use them.
Advanced models add value by facilitating asset pricing and risk measurement for sophisticated products that allow organisations to target and manage risk and to improve returns. Analysts with good technical skills can encourage the use of increasingly complex asset and portfolio modelling techniques. The danger is that, as applications and variants of a complex model expand to accommodate “new” or “one-off” transactions, the model can become stretched beyond the scope of its original purpose, which in combination with often inadequate review of the assumptions embedded in it,6 can lead to mis-estimation of actual exposures, mis-allocation of risk and inefficient hedging. An example is the use of copula7 analysis in CDO modelling, chronicled by Mackenzie and Spears in their 2012 paper.8
Less complex models avoid this problem, but at the possible cost of over-simplifying the nature of risk exposures and thereby failing to capture material sources of risk, often understating it.
Risk models – calibration
It has become clear that many apparent risk model failures were in fact due not to shortcomings in the models themselves, but rather in how they were used. The most obvious misuse was unquestioning reliance on historical data aggravated by selection of inappropriate data intervals. In using historical data samples, risk models in effect extrapolate from the past and so assume that it is a good indication of what the future will be like. René Stultz identifies over-reliance on historical data as the first of his six items.
For example, bankers and investment bankers estimate short-term risks by sampling data from the most recent one or two years using daily observations. Meanwhile, investment managers, with longer investment horizons, usually sample data from the most recent two to five years, using weekly or monthly return observations. This means that, at the onset of the crisis in 2007, the maximum look-back period extended only to 2002, a period of exceptional calm in financial markets, while many look-back periods were much shorter. An obvious remedy to the problem of unrepresentative data samples is to sample data from earlier intervals or from periods that include significant volatility and market stress; but there are two reasons why this is not always satisfactory.
The first is that, even when suitable data going back far enough are available, they may embed (worse) problems of data quality and reliability, especially in less mature markets. For example, to capture, in 2007, the 1987 stock market crash, which is in some respects the last major event in global markets, would have entailed sampling at least 20 years of data.
The second reason is that long data sampling intervals can confound market conditions, such as different currency and interest rate regimes and stock market structures, introducing data that are irrelevant to the future in those markets. For example, a feature of several developed markets in the 1990s was the large-scale privatisation and listing of telecoms and other hitherto publicly owned utilities that changed market structures in important ways. Similarly, European currency and interest rate alignment has changed fundamentally since the late 1980s. In effect, data from 1987 were not necessarily relevant to market conditions prevailing at the onset of the global financial crisis.
Risk models – standardisation
There is an argument that comparable institutions should conform to agreed modelling principles and methodologies. This already happens to an extent: Merton's model for credit risk is considered to be standard by many, if not most, participants; Black-Scholes is generally accepted for option pricing, as are cross-sectional risk factors within variance-covariance models of equity risk. All have achieved the status of accepted wisdom (despite recognised flaws in some).
But organisations still tend to apply their own variations, which can amount to material differences of methodologies. And, just as different risk model methodologies can give different risk measures, even using the same input data, the answer a risk model delivers depends on the data fed into it. So variation of output from one organisation to another can be significant, and indeed, there are valid reasons why they sample from different sets of past data.
But different model variants can cause problems of comparability. Jon Danielsson of the LSE notes in the abstract to a 2014 paper9
