## Andrew Large: Basel II and systemic stability
Speech by Sir Andrew Large, Deputy Governor of the Bank of England, at the British Bankers' Association - Basel II/CAD 3 Conference, London, 13 March 2003.
* * *
## Introduction
Welcome to all present - particularly Bill Rutledge of the New York Fed, whose President, Bill McDonough, is rightly credited with the paternity of Basel II. Basel II is a significant step forward. It brings life to the concept of capital requirements as a function of the actual risks which banks undertake. And it extends many of the principles long developed in the area of market risk, into that of credit risk. The very essence of the banking business.
This was a courageous move. And it has required Herculean labour, intellectual agility, and sheer hard work. It seeks to address the public policy objectives of regulation and supervision on the one hand, and the real-world best practice of how banks manage their businesses on the other. It has sought to provide incentives to private firms to encourage them to manage themselves in ways which reinforce their own strengths, and hence to contribute to financial stability as a whole. But equally it is designed to enable supervisors to develop resources to sharpen their own judgement, to encourage them to take account of individual circumstances, to avoid box ticking, and to add real value to the process.
These are worthy ambitions. They rely on the interaction of the three pillars to achieve it. Having worked on both sides of this argument, I am more than aware of the anxieties - and indeed disagreements - which have arisen during the long debate that has got us to where we are. But the direction seems clear even though storm clouds may appear from time to time. But what I would like to do today is to look beyond today's situation a little, and to share with you a few reflections both on the Basel process itself, and, looking forward, on some related issues.
## Importance of all three pillars in the new Accord
Firstly a couple of observations on the Accord itself.
One is to remind ourselves that the Accord has three pillars - not just one. Pillar one has had most of the attention hitherto. The challenge of producing greater risk sensitivity has been considerable. Technical experts from all areas of the business, model builders, and others, have been involved. This has generated a tendency to seek greater and greater granularity and has resulted, perhaps inevitably, in considerable complexity. Striking a balance between risk sensitivity and detail has not been easy. But faced with the detail of pillar one, the challenge going forward is not to lose sight of the other two pillars, and the role they can play in mitigating complexity.
## Pillar two
Pillar two, it should be remembered, was designed to alleviate the need for excessive detail. Supervisory oversight was felt to be a better way to address the many complexities than detailed rules.
Some in the private sector complain that there might be a lack of regulatory level playing field. But my own experience leads me to believe that supervisors themselves are just as aware of this. So I would encourage those in the private sector to work with supervisors in different countries to ensure the delivery of fair supervision across frontiers. If we can achieve confidence in that, then the intention of pillar two acting as a mitigant against greater prescription will I believe be realised in practice.
That said, I recognise that in some jurisdictions the thinking behind pillar two represents a new departure in the philosophy, and perhaps legal basis, behind banking supervision. And in that way it represents a considerable challenge.
## Maintaining the link with market practice
Basel II has been a long process. But it is not an end in itself. Indeed if it were we would never achieve it. The markets, and the risks they give rise to, are not static. It is instructive, for instance, to reflect on just how much risk management theory and practice has advanced in the five years since negotiations on the Accord began.
- I have long felt that you need some mechanism to ensure that prudential standards remain appropriate as the state of the art moves on. That does not mean that the standards should be in a state of continuous flux. There is a good case for stability in the initial period while the Accord is being adopted. And any review mechanism must take careful account of the implementation costs incurred by banks when the rules change. But it is costly, too, to persist with regulatory standards where they are clearly out of line with market practice.
Real effort has been made to tie the Accord to best market practice and thus to changes that banks will need to introduce anyway. Going forward, the guiding principle, difficult I know to achieve in practice, should be to keep to a minimum the difference between expenditure that banks would be incurring in any event and that required by regulators. Part of the trick is to ensure sufficient continuity of core aspects of the standards, so that changes to basic IT systems etc are minimised or at any rate can take place over time. Time is, after all, a major help in this respect. Implementing systems changed as part of the software obsolescence cycle may involve only modest marginal cost. More immediate and discrete changes are always going to be more costly.
These issues will clearly need further discussion. In the EU context, however, it is particularly important that the arrangements for revising any Directive reflect, if not the letter, at least the spirit of Lamfalussy. It would be wholly inappropriate to try to hard code all the detail of the pillar one rules in an EU Directive, as if these can necessarily stand for a long period of time.
- I should say that the costs of implementing Basel II will not fall on banks alone - nor will the challenges. In most jurisdictions, regulators too will both face challenges and need to incur costs in expanding staff numbers, in training and in implementing systems. This will be important for them to handle the new approach. After all, it calls for a significantly greater level of expertise on the part of line supervisors than has the existing Basel Accord. And whilst the UK, and a few other jurisdictions, do have that expertise, we need to remember that we are looking at a global market place. And at a global level I detect a deficit of expertise which will require to be remedied.
## Pillar three and disclosure
I would also like to say a few words on pillar three.
Pillar three can, it seems to me, play two roles. First, of course, it can play a role within the Basel II context itself, by mitigating the need for complexity in pillar one. Disclosure about the way organisations actually manage their risks can indeed help a better understanding of their risk profile, compared to a study of complex numbers on their own. For example, understanding how a firm conducts its stress testing, or how valuations might be sensitive to different assumptions, could be of real value. But, second, such disclosure could be of value in the general accounting context itself. This gets to the complexity versus disclosure debate in the search for a better way forward in that arena; and should improve the ability of accounts to reflect changes in the underlying business.
All of this is really to say that pillar three needs to be seen in a broader context. In today's 'derivatised' world, the utilisation of derivatives and complex forms of contract, often with modelled valuation bases, is just about omnipresent. This is not a value judgment - it is a fact. I think we have to accept that any set of accounts, however drawn up, is likely to be considerably deficient, taken just as numbers, in terms of outlining the economic realities of risks within the balance sheet. And disclosure itself is arguably inadequate at present in relation to both off- as well as on-balance sheet activity. However much it may be a matter of regret, and despite best intentions, the transparency of today's accounts has become less and the opacity greater than was the case in former days, thereby giving a less than complete or reliable view of overall risks.
This is not just a question of SPVs; what is and what is not on the balance sheet; nor even whether historic or fair value accounting techniques are used. The fact that accounting standards setters are only too well aware of this, and grappling with the considerable issues involved, is very much to be applauded. For this is a fundamentally difficult area.
I particularly welcome the efforts of Sir David Tweedie and his colleagues at IASB in addressing so openly the hugely difficult task of getting the right balance and right principles into place. Many of us can see I think that the financial instruments standards IAS 32 and 39 have provoked significant and strong views. Wisdom, foresight and a certain forbearance will be needed if we are to achieve a good solution. So we should all welcome the present initiative of the IASB in holding a series of roundtables with respondents to their exposure drafts on revisions to IAS 32 and 39. Debate and efforts in relation to these particular standards seem to me to be vital. They get to the heart of the dilemmas and challenges presented by new financial techniques and often complex instruments. It has after all been shown, in the Enron case if no others, that detailed and prescriptive accounting standards, in the wrong hands, can engender a moral hazard in the search for ways round them which investment bankers and lawyers may well be expected to find.
A similar logic applies in the prolonged debate between historic and mark-to-market accounting principles. Once again, whichever approach is used, neither of them will tell the whole story in terms of the underlying economic realities of a bank's risk and return, nor how the risks are managed. But forms of disclosure may help.
## Looking ahead: the longer term financial stability policy agenda
Perhaps I can turn now to some observations for the longer term. These go beyond Basel II and what its architects had in mind for it. But they nonetheless require us to think in terms of the same combination of evolving best practice within the financial services industry on the one hand and public policy needs on the other.
## I The challenges of an increasingly integrated financial sector
Firstly, the new Accord is mainly designed for credit risks and hence essentially for banks. Yet we are all becoming clearer that the process of intermediation of risks is itself changing mightily.
The silos that were securities, lending, and insurance, are no longer silos. Instead we are seeing the increasing development of a single financial network embracing not only these areas, but also many organisations which fall outside today's definitions of what is a bank, what is a securities house, or what is an insurance company. This process has gathered pace in recent years, and has accelerated since the Basel II process itself began.
A couple of examples can illustrate what I mean. First is the growth of the credit risk transfer market. Alan Greenspan has recently noted that this may have benign effects, recently arguing that: 'If risk is properly dispersed, shocks to the overall economic system will be better absorbed and less likely to create cascading failures that could threaten financial stability'. But new forms of contract tend to bring uncertainty. The recent litigation in New York involving claims under surety bonds is perhaps a much reported incident of this sort. And uncertainties in a world dependent on expeditious performance of contracts has to be a real issue in financial stability terms.
Second, we have seen the emergence of large multifunctional financial groups.
So my observation is that whatever thinking might have gone into the Basel II Accord for banks, we would be wise to focus further, on the way in which capital adequacy and other prudential supervisory techniques fit together, for different parts of the financial world, including insurance. The wisdom in earlier days was that systemic risk, in a financial stability sense, was largely confined to banks. Not any more. It has been recognised for some time that the securities arena and the process of securitisation has changed all that. And more recently, growing interlinkages with the insurance sector are giving rise to the need for a further rethink.
## II Looking beyond capital adequacy: the importance of liquidity
Secondly a rather different angle. Viewed from the point of view of financial stability, capital adequacy is clearly a vital and valuable policy tool. That is why Basel II is so important. But systemic crises do not always emanate from capital inadequacy. The first indication of trouble ahead may well come from a completely different area. What I am talking about is the question of liquidity. And my observation here is that we would be advised to give more thought to the underlying issues, risks, and mitigants which could be ingredients in liquidity problems. What are the drivers? What may cause liquidity to be withdrawn? To what extent could uncertainties, referred to a few moments ago, impact liquidity in
unpredictable ways? What machinery/measures can best be devised to ensure that liquidity will continue to be available in times of stress? And, in the face of attempts to improve the resilience and efficiency of market infrastructure (for example, clearing and settlement), how will changing practices and perceptions affect the drivers of liquidity?
This is, I feel, an area where standards and principles could be further developed which engender both the best practice understandings of the private participants on the one hand, and the interface with public policy imperatives on the other.
## Conclusion
Basel II has been steered to this point of take-off for the hugely critical area of capital adequacy. I would identify four challenges for the next stage. First, to achieve the successful implementation of Basel II by banks and regulators across the globe. Second, to ensure that mechanisms exist to update it in the light of evolving best practice. Third, to give thought to whether analogous agreements would be appropriate in securities and insurance. And, fourth, to ensure that the scale of work on capital adequacy issues does not cause us to overlook the importance, either of liquidity management, or the development of a more robust market infrastructure in the maintenance of financial stability.