Lessons from the financial crisis
Speech by Callum McCarthy, Chairman, FSA
at the Manchester Business School
13 May 2008
It is a pleasure to be back in the city of my childhood. There is a strong link between the FSA and Manchester, in that both the chairmen of the FSA – the first chairman Howard Davies and now myself – went to Manchester Grammar School, Howard for all his secondary schooling, I for only two years (and I should add, by one of those strange accidents enjoyed by conspiracy theorists, both of us are graduates of Merton College, Oxford and the Graduate School of Business of Stanford University: those speculating as to who my successor will be in September should search for other MGS, Merton and Stanford GSB graduates – but I don’t think there are any).
Manchester in the 1950s was very different to today in many respects: Manchester United was said to have owed by modern standards its then limited success to its connection to the butchers’ trade, as likely apprentices were attracted by the prospect of their mothers receiving sausages over and above the meat ration; the Manchester Guardian was still Mancunian in both title and the siting of its editorial staff (there was, I think, a London letter to show that it didn’t entirely ignore metropolitan events); and – of course – Manchester Business School was still a decade in the future, awaiting the Franks Report.
In other respects, too, the world of Manchester as I knew it was greatly different. First, of course, there is the difference in the observer – then a pre-teen schoolboy, now a 60+ regulator. But the data to be observed were different. Then, Lancashire was still a county of cotton mills. The balance between manufacturing and services was dramatically different from today. And finance was still an industry dominated by a series of cartels: mortgages available only from building societies; competition between banks subject to all sorts of agreements which resulted in at best muted competition; the clear distinction between brokers and market makers in the exchanges; an almost complete absence of foreign firms in the British financial services sector. I can’t help but add that it wasn’t only finance which was ruled by restrictive practices. In 1955 the maximum legal weekly wage for a professional football player was £15 (equivalent to around £280 in today’s terms); Ronaldo earns £119,000 a week. There was, as I remember, only one non British Isles player on the Manchester teams – Bert Trautmann, the Manchester City goalkeeper, who had been a German POW in Britain. It is a world which appears immensely remote.
This evening I want to address the world of finance we now inhabit, where finance has grown hugely relative to the economy as a whole; where innovation is a constant feature; where the old cartels in finance have been eliminated; where competition is international (reflecting the fact that communication costs are independent of distance) – and where we are experiencing a period of turbulence and doubt which continues to cast deep shadows over economic prospects. In particular, I want to discuss how we have come to the present position, and then to address some of the wider issues which I believe will need to be addressed.
Before discussing how we come to our present position, let me establish what our present financial position is. I am tempted to describe it as being part way through a three act play, where Act I was liquidity, Act II is concerned with solvency; and Act III (which has not yet to occur, and may not occur – certainly not to the same extent in different economies) concerns the effects on the real economy of Acts I and II. I should add that in this play Acts can run simultaneously.
The liquidity issues are well known. In August 2007 markets in a range of financial instruments closed, and for effective purposes many of them have remained closed. The evidence of this is clear from the money markets: the three-month £ LIBOR:OIS spread (which a year ago was below 20 bp) rose to a peak of 113 bp in December 2007, and even today remains at 82 bp. The rise in € and $ three-month LIBOR spreads has been comparable, albeit slightly less. Funding markets based on securitisation have shown severe strains since August 2007, and in important respects remain closed. The UK markets are essentially closed for £ FRNs, for RMBS and for covered bonds. They remain open for CP and MTNs. Other products are best described as patchy. The scale of this change is very large: in 2007, the UK did 15 per cent of global RMBS issuance; this year there has been one such issue. The UK is reflecting changes which are global: RMBS issuance globally is 90 per cent down on 2007; high yield issuance is 70 per cent down; even high grade corporate debt issuance is down by approximately a third. The effect of this on UK banks, which have relied heavily on securitisation for funding, has been to increase their dependence on shorter‑term funding, with a resulting increase in rollover risk, greater pressures in the inter‑bank market, and the wide three‑month £ LIBOR spreads which I referred to.
The liquidity pressures I have described have led to heightened concerns about solvency, and these concerns have been exacerbated by the range of surprises which major institutions have given to their counterparties and to commentators. An element of the present circumstances which has to be recognised is that, as well as problems in quite specialised banks operating essentially within a single country (Northern Rock in the UK, IKB or Sachsen Bank in Germany, Countrywide in the US), the institutions which have surprised counterparties include very major, well regarded international institutions at the very centre of the financial system: in various ways, UBS, Merrill Lynch, Citigroup, Société Générale, AIG, RBS, Morgan Stanley are well founded and well respected institutions – but they have all surprised the market. The doubts about banks and brokers can be seen in the movement in CDS spreads – the price of insuring against failure of an institution – which have increased from about 10bp in June 2007 to between 50bp and 100bp today for British banks and from between 10 to 30 bps to between 70 to 170bps today for US banks.
The third act of the drama – the effects of the combined liquidity and solvency concerns on the real economy – is conjecture, not fact. The prognosis for economic growth in the US, the EU and Japan is unclear, with economic data providing conflicting signals. The position in the UK is comparable, as can be seen from the widely different analyses proferred by independent members of the Monetary Policy Committee. What is also unclear is whether, and if so to what extent, the liquidity and solvency pressures which I have described will affect the real economy via restricted credit growth.
I think it important to recognise the position from which we started, which was of continuing benign economic circumstances. In 2007, world growth in GDP was 4.9%; corporates have reported good profits, and have used them to reduce gearing on their balance sheets; banks have reported high returns on equity and went into the present difficulties well capitalised; provisions for reasons other than movements in the valuation of securitised products have remained low. This should be both reassuring in some respects, and sobering in others. Reassuring because the general economic circumstances were benign, and may even today on some but not all analyses remain benign. Sobering for all those concerned with financial services, whether as practitioners or supervisors, because the very real problems which now beset us have their origins not in the real economy, but in finance; and not in some marginal part of finance (they do not originate in hedge fund activities, as did the LTCM crisis of 1998, nor in emerging markets, as did the Russian debt and South East Asian crises of 1997), but in the mainstream activities of banks in the world’s most important economy and the world’s most sophisticated financial system. These problems have been incubated at the core of the financial system.
The problems of liquidity and doubts about solvency have their origins in substantial uncertainties which have led to the present fragile conditions. The first set of uncertainties derives from the difficulties in valuing particular financial instruments, especially those associated with recent vintages of US mortgage origination: loose underwriting has been accompanied by soft or hard fraud, and the results have been high delinquency rates. On top of this has been a failure to reflect this reality in the ratings attached to the securitised instruments created from these mortgages, so that ratings for complex instruments (which are not distinguished in sign from the ratings for corporate debt) have produced results very different from those many expected. There are many issues associated with this – the semiotics of ratings, due diligence by rating agencies, the responsibilities of purchasers of products to understand what they are buying. And third, in relation to the uncertainties affecting instruments, there has been the complexity of those instruments themselves: to anyone who doubts this problem, I would simply refer them to the pages of legal definition for a CDO, and remind them that there are CDOs of CDOs. Given these uncertainties about instruments, the lemons problems well known to economists has materialised – and left a very sour taste.
Linked to the uncertainty about instruments has been uncertainty about institutions, as counterparties have found it hard to estimate the exposure of particular financial institutions to various asset classes, and have often erred on the side of severe doubt. I think the position here is improving, for two reasons. First, disclosure has improved, with banks recognising that it is in their interest to describe the position as clearly and fully as possible. The FSA has encouraged them strongly to do this, with some success. For example, the recent RBS documentation of its exposure to, and valuation of, different asset classes was clear, and removed doubt. Giving more detail also makes it easier to explain and justify differences in valuations, relating them for example to different credit ratings or different vintages of mortgage. Better disclosure is lifting a fog of uncertainty. Second, there have been substantial recognition of losses and compensating capital raising which together have gone a long way to reassure counterparties that banks are recognising losses and can maintain the necessary capital ratios. The IMF report which estimated the total estimated loss associated with the sub-prime problems as US $945 billion thought that this would result in European bank losses of US $125 billion. In fact, provisions already declared by EU banks amount to more than US $130 billion. UK banks have been prominent in both recognising losses (HSBC and RBS) and in raising new capital (RBS and HBoS). These recapitalisations have substantially reduced doubts about solvency. They demonstrate that the market can – and will – provide fresh capital for banks with sound business models. They provide significant reassurance that corrections are taking place.
I have described the situation we find ourselves in, and the uncertainties which have caused us to be here. Let me turn to the lessons which we need to learn. I am not going to address the immediate, important but essentially tactical issues – central bank liquidity injections, the widening of eligible collateral accepted by central banks, reductions in interest rates, the provision of emergency liquidity support to systemically important financial institutions, or the encouragement with active financial support of the sort provided so successfully by the Federal Reserve to facilitate the Bear Stearns rescue, or other weapons in the armoury of central banks, supervisors and finance ministries. That is not because these tactical questions are unimportant – quite the opposite. Making the right tactical decisions will be essential if we are to deal with the immediate issues. Instead, however, I want to discuss some wider questions: not how do we escape from our present difficulties, but rather how do we make a reoccurrence of them less likely.
Let me start with lessons for the banks whose activities lie at the centre of our present problems. The first lesson is that in many cases risk management – both of credit risk and of liquidity risk – has been inadequate. This is a self-evident observation, and we (and the management and Boards of banks) need to understand better why this has occurred, and the steps to be taken to improve risk management. The FSA has been working with a group of regulators led by the Federal Reserve Bank of New York, which has examined risk management in some dozen institutions – more and less successful in managing their risks during the crisis – to establish good and bad practice. The group’s report was published in April and we expect the management of banks to act on its findings. The best firms succeeded in integrating risk management: credit, market, operations. They avoided dealing in silos. Risk management was addressed at the very top of the firm. Senior management in the best firms fostered a culture of challenge and flexibility as compared with a more mechanical and prescriptive approach in less successful firms. I welcome the various initiatives of the industry (the IIF Committee on Market Best Practices led by Rick Waugh, or the third round of work of the Counterparty Risk Management Group under the leadership of Jerry Corrigan). There is no shortage of understanding of what has to be done. We now need to ensure that the board and management of individual banks act.
Second, as both a specific part of risk management and as a contributing cause of poor risk management, the financial sector as a whole has to reconsider its incentive structure. The present bonus system places excessive emphasis on the short term performance of individual parts of an organisation; and it encourages the taking of trading positions whose immediate profitability (often unrealised but marked to market) is rewarded, but whose accompanying risk is neither properly recognised, nor disincentivised. This is a problem which has long been faced in traditional banking in respect of credit, where the immediate income on arranging a loan is offset by the provisions which inevitably attach to advancing loans. We need to have comparable incentives – and disincentives – on the market side. I think that there are also wider questions apart from the behavioural aspects of incentives. Shareholders of a bank who have seen the value of their investment fall by many tens of percent are likely to question the equity of a system which leaves the same bank’s bonus pool essentially unchanged, or even increased.
Answers to this problem will not be easy. It is worth observing that Bears Stearns’ employees owned some 30 per cent of the firm’s equity, but that did not prevent them making decisions which destroyed the firm: so long term equity shares are not in themselves the answer. Nor is there a place for regulation of individual rewards: that way madness lies.
Second, the lessons for the credit ratings agencies, which have a major job to do in bringing back confidence in the ratings system. The need for a ratings system in which users can have confidence is indisputable: ratings are integral to elements of Basel II and lie at the heart of a great range of transactions, from the mandates on investment given to asset managers by trustees to the risk controls within banks (they are, for example, a central part of the UK’s present Special Liquidity Scheme). The issues associated with the ratings agencies are many: conflicts of interest; methodologies; the melding of advice and actual rating; transparency. I am neither persuaded that all of these are real problems, nor that for all there are in fact solutions: I see, for example, no practical means of avoiding the rating agencies being paid by the issuers; this conflict of interest – like many in financial services – needs to be managed since it cannot be eliminated.
But two issues particularly concern me: the first is how to make sure the CRAs do not overtrade; the second is the need to move beyond the (correct) position that CRAs are not auditors to establish a better data underpinning for their work. On the first point, the growth in the scale of ratings of structured instruments issued by the CRAs over the last few years clearly outstripped the resources of the CRAs: politely, they became overextended; bluntly, they overtraded. They need to demonstrate the measures they will introduce to prevent that happening again. Second, a principal cause of the sub‑prime problems was that a serious degradation in the quality of data occurred as more and more sub‑prime mortgages were originated by essentially unregulated brokers, yet those data were treated by the CRAs as having equal validity with earlier and more reliable data. If they are not to be accused of applying sophisticated risk analysis to data of quality they should have been sceptical about – what bluntly might be regarded as an example of garbage in: garbage out – they have to make changes. The CRAs have to demonstrate a much greater concern with the data they use. It is clear that they are not the auditors of companies, but they need to establish better control over the data they use. The present position is unacceptable. The FSA has been involved in and supports ongoing international work, most notably in IOSCO, focused on defining appropriate standards for conduct of business by rating agencies in light of the lessons learnt since last summer. We will continue to monitor closely how the agencies respond in these areas in assessing the appropriate level of regulatory engagement with them.
Third, there is an essential but simple lesson for investors: do not invest in products whose risk you do not understand, on a naïve criterion of rating alone, however much the search for yield in a world with excess liquidity makes the apparent return attractive. Attention to what went wrong has focussed on the sell side. It is important that lessons are also learnt on the buy side, where many of the problems originated.
Fourth, there are some important lessons for standards-setters. Many of the present problems arise from doubts about valuations, where uncertainties have had widespread and damaging effects. The basis of valuation is undergoing great change with mark to market replacing historical cost valuations. The logic and need for these changes have long been set out, and I broadly accept them. But there are aspects of valuations of illiquid instruments, particularly the reliance on indices to establish value, which need urgent attention. There is good reason, admirably summarised in the Bank of England’s most recent Financial Stability Report (and cogently described by Donald Mackenzie in the most recent edition of a less likely journal for such a discussion, the London Review of Books) to believe that present valuation methodologies exaggerate the fall in the fundamental value of mortgage related complex securities. Operationally, we need a means of providing comfort to auditors that there are proper and defensible in litigation methods of valuation other than using indices.
Fifth, there are lessons for regulators and supervisors. First, we need to review the regulatory capital regime. Recent events have shown all too clearly failings in Basel I, particularly its treatment of off‑balance sheet vehicles – something which Basel II will improve. But we need to consider whether there are failings, and if so what they are, in Basel II, and correct them. There are both specific issues, like the capital charge for trading book credit loss, and the more general issue of how we use ratings, which are central to Basel II, and where the problems in ratings (particularly of structured instruments) should make us thoughtful. The Basel Committee, under the chairmanship of Nout Wellink of the DNB, is committed to urgent action to make these changes.
Second, we as supervisors need to pay more attention to, and overhaul, our liquidity regimes. The present troubles started with liquidity pressures, and evidently the present liquidity regimes are not up to the demands put upon them. That is why in the UK the FSA in December set out for consultation a series of measures designed to improve the UK's liquidity regime. I know that international banks would like to manage their liquidity globally, subject to one global, not several national, set of rules. But, unless there is a step change in the speed with which the Basel Committee progresses international agreement, regulators will be forced to deal with this on a national level. That is what we plan to do in the UK. I am of course very keen that the Basel Committee should actively consider whether it can develop a resilient approach to the cross-border management of liquidity for banking groups, which meets the most compelling requirements of regulators and industry alike. But we cannot delay.
There are other regulatory improvements which are needed: the desirability of finding ways of taking central bank action which does not stigmatise or destabilise the bank concerned is one; the desirability of establishing better means of co‑operation between different national authorities when dealing with major international banks is a second; changes to deposit insurance is a third; the prospect of more coherent approaches to defining eligible collateral by central banks is a fourth. The UK authorities in January published for consultation our proposals to deal with the first and third; I commend "Financial stability and depositor insurance: strengthening the framework" to all interested – as anyone banking in the UK should be. On the second, the FSA has been in the fore, with in particular the Swiss and US authorities, in developing practical solutions via the development of colleges of regulators.
What has happened over the last nine months has been both concerning and sobering. It is important we draw the correct lessons, and avoid false steps. I have set out some of the lessons which I would draw – and, at the least by implication, some of the steps I would avoid.