Central banking, financial stability and Basel II
Speech by Andrew Crockett, General Manager of the BIS and Chairman of the Financial Stability Forum, at the 38th SEACEN Governors' Conference, Manila, 13 February 2003.
It is a great pleasure to speak once again at the SEACEN Governors' conference. Five years ago, I came to this conference for the first time when it was held in Bali. Many things have changed since then - largely in the right direction. Where the Bali Conference was marked by reactions to the traumatic crisis that had just broken out in Asia, I am happy to note that today most economies in this region have recovered from the crisis, and several have resumed robust growth.
Restoring financial stability has been at the core of the efforts to revive economic performance. Promoting financial health, however, is not just a temporary challenge following a crisis. It is a task that has to be pursued on a continuous basis. Financial authorities, and central banks in particular, have to be vigilant all the time for potential threats to the health of their financial system. I have therefore chosen the theme of financial stability for my remarks today.
If we look back as central bankers at the past two or three decades, it is hard to escape the conclusion that this has been a particularly eventful period. In the industrial countries, the 1970s saw a surge in inflation that posed a threat, not just to economic performance, but to social and political stability as well. Inflation was a problem in virtually all emerging markets as well, though East Asia did better than many other regions in avoiding its worst effect. Fighting inflation was the key challenge facing central banks throughout the 1970s, 1980s, and much of the 1990s. It is only comparatively recently that we consider the battle against inflation to have been won. (Actually, the battle is never "won", but inflation is now sufficiently under control of policymakers to no longer be at the head of their policy concerns. It may even be the case that the problem of the future is that of deflation, not inflation.)
A second challenge faced by central banks over the past quarter century has been in some ways just as daunting. The period since the mid 1970s has been one of periodic bouts of acute financial turmoil. These episodes of volatile financial conditions have caused the issue of financial stability to move to the top of central banks' policy agenda. This increased focus on issues of systemic stability can be seen as a reversion to a traditional pre-occupation of central banks. Indeed, if we go back to Bagehot and the beginnings of central banking, the original function of central banks was seen as preserving the health of the banking system. By virtue of their liquidity-creating powers, central banks could act as others could not to stem a financial panic. And their authority and power of "moral suasion" could act as a supervisory and regulatory force, even when no explicit oversight of the financial system had yet been put in place.
Central banks have by now been largely successful in finding the tools to sustain monetary stability. The same cannot be said with confidence for financial stability. This morning, therefore, I will try and address the issue of how to use supervisory oversight to build a more robust banking system. Before doing so, I would like to sketch briefly the ways in which the financial landscape has changed over the past 25 years, and how this has affected the task of preserving financial stability.
The financial landscape has been transformed over the past two or three decades by the twin forces of liberalisation and innovation. Liberalisation brought the dismantling of the regulations and controls over financial activity that had been in place for much of the post war period. The key organising principle became that of the market mechanism. The process of giving more rein to market forces was further reinforced by technological innovation; in particular, the exponential growth in computing power and information technology. New financial instruments made restrictions less relevant and easier to circumvent. And they helped to make markets more complete, offering greater scope for hedging or managing financial exposures.
The transformation of the financial system from one in which administrative controls were widespread to one driven by market forces resulted in a substantial increase in the importance of the financial sector. A greater proportion of resources flowed through the financial system, and new techniques were developed to mobilise savings and allocate investment. Financial institutions came to view their activities, not simply as borrowing and lending, but as managing and trading portfolios of risk. If done well, this process could unlock a greater volume of investable resources and allocate them to higher-productivity outlets, thus promoting faster routes of growth and sounder economic development. If done badly, however, resources would be misallocated and, in a worst-case scenario, economies could become more vulnerable to crisis.
The process of financial deepening also affected market structures, complicating the task of regulatory oversight in several ways. First, financial instruments become more complex, with greater power to disaggregate and transfer risk but more opaque structures. Second, the boundaries between different types of financial market activity (lending, trading, insuring) become blurred, as financial instruments began to assume multiple characteristics. Third, more claims became securitised and were exchanged in markets for traded instruments. Fourth, much financial activity became international in nature, blurring boundaries of responsibility among supervisors.
There is little doubt in my mind that a system of financial intermediation based on deregulated and competitive markets is generally superior to a rule-based or government-controlled system. The experience of South-East Asia during the period of its "economic miracle" generally attests to the beneficial effects of market-determined resource allocation.
However, the experience of this region also demonstrates the downside of unfettered markets. The move from a tightly controlled financial system to a much more competitive environment has been associated with frequent and sometimes prolonged bouts of financial turmoil, with associated economic losses and social hardship.
It is not surprising that the risk level in a liberalised system is, almost by definition, higher than in an environment of financial repression. So if countries are to reap the benefits of financial liberalisation, without incurring the costs of instability, it is essential that their financial institutions move to a higher standard of risk awareness and management. The transition to a more open financial environment should therefore take place only when institutions have adequate systems of control in place, and the supervisory and regulatory institutions have the necessary capacity to monitor risk-taking.
However, the risk of financial instability is not only a consequence of inexperience during the transition phase It also needs to be recognised that financial markets contain disequilibrium tendencies and are vulnerable to certain weaknesses even in mature systems. It would take more time than I have today to go into these sources of market failure in detail, but a couple of examples will help make the point.
A first key consideration relates to the crucial role information plays in financial intermediation. Financial markets and institutions can be thought of as devices to mobilise and disseminate information, which is then reflected in prices for financial claims. But, as we all know, information may be incomplete or misleading, and may not be available on a comparable or timely basis. Moreover, it may not be available to all market participants on the same basis. Asymmetric information, we now understand, can distort market functioning and impede the development of mutually beneficial exchanges. Lack of proper information can allow financial imbalances to build up and then precipitate their disruptive unwinding. So a key focus of prudential oversight of a liberalised financial system is to ensure that the information is available to allow financial prices and flows to be efficiently determined and just as important, to respond in a smooth and stabilising manner to changes in circumstances.
A second area of inherent vulnerability in financial activity is that of the whole infrastructure which financial markets rely on to function properly. This financial market infrastructure includes contract law; law enforcement procedures; accounting practices and valuation standards; prudential regulations; effective supervision; appropriate disclosure requirements and the creation of well-functioning payment and settlement systems.
In economic jargon, these are "public goods". They are costly to create, but once created they benefit everyone. Since the costs of their provision exceed the benefits of an individual user, they are likely to be undersupplied in the absence of public intervention.
An important element in strengthening financial systems, therefore, is to ensure that these public goods are adequately provided. Indeed, recent research suggests that these "environmental" factors, defining and protecting property rights and the integrity of markets, are the single most important factor determining long-term differences in economic performance.
Building a strong infrastructure is not easy. It also needs to be recognised that the elements are interdependent. Minimum capital requirements for banks are of little use if the accounting conventions used to value banks' assets are flawed. Similarly, accounting is only as strong as the audit standards used to implement it. And asset valuation is fragile if uncertainties persist about the legal recoverability of claims.
Towards greater financial stability
Everybody will agree that financial instability, with its high social and economic costs, is unacceptable. Does this mean, as some argue, that we need a new financial architecture from the bottom up? In my view, this would be to throw the baby out with the bathwater. A market-based system remains the only satisfactory framework for promoting efficient resource allocation. What needs to be done, therefore, is to identify the ways in which free markets fail to work properly, and then to set about measures that correct or offset these market failures, without having undesirable side-effects.
How might this objective be pursued? There is now a growing consensus that the development of standards and codes of best practice can provide a useful basis for improved market functioning. Standards and codes, backed up where necessary by national legislation and supervisory oversight, can provide a "benchmark" against which individual institutions and national authorities can measure their actual situation. If drawn up by an international body and widely adopted, they can prevent divergent national practice, and reduce the scope this offers for unlevel playing fields and regulatory arbitrage.
Standards need not be juridically enforceable through international treaty. If they are accepted by markets as representing best practice, market forces and peer pressure will shoulder much of the burden of implementation and enforcement. In this sense, although the standards could be considered "soft law", they may be just as, or even more, effective, than "hard law" enshrined in treaty obligations. If we have learned one thing in the past decades of financial liberalisation, it is that financial engineers can find their way around legal restrictions that do not reflect market realities.
The strategy for strengthening financial systems is thus now firmly based on developing a network of standards, designed to make markets work better, and adopted voluntarily at the level of individual national jurisdictions. But this still leaves a number of questions to be answered. What should be the scope of standards? Who should develop them? Should they be uniform for all countries? And are we looking for minimum standards or best practice?
As far as the first issue is concerned, that of the scope of standards, it is apparent from what I said earlier that the various elements of a well-functioning financial system are interlinked. A financial sector can be thought of as an eco-system in which each element is dependent on all other elements. So it is not really possible to, say, build a strong banking system, without considering the health of the markets in which banks deal, the soundness of financial reporting, the robustness of payment and settlement systems, the reliability of the legal system and the effectiveness of corporate governance.
All this argues for a wide scope for standards. The website of the Financial Stability Forum (FSF) contains around 70 standards, of which 12 have been identified as core. I am well aware that pursuing a large number of standards simultaneously can stretch the administrative capacity of most countries. There is no easy answer to this dilemma; and we should recognise therefore that the task of upgrading financial systems will require sustained efforts over a long period. What is not a satisfactory option, however, is to deal only with one element of the picture (say, prudential supervision) without giving attention to those elements which are a precondition for its effective development.
A second sensitive issue is who should develop standards. As standards tend to be applied on a global scale, a natural answer would be to involve in the design process all those who have a stake. A number of standards have indeed been developed in universal bodies, such as the IMF, which has formulated guidelines for the transparency of macroeconomic policies and dissemination of data.
In many, often technical, areas, however, this procedure has proven impracticable and the design of standards and codes has been left to smaller groupings of national experts who, having worked or operated in the financial markets in which the issues concerned have progressed furthest, have accumulated extensive knowledge and experience. Examples are the work of the Basel-based committees with standard-setting functions - the Basel Committee on Banking Supervision and the Committee on Payment and Settlement Systems. Other examples are the International Accounting Standards Board, the International Association of Insurance Supervisors and the International Organization of Securities Commissions.
While the small size and expertise of these groups promote efficiency in the development of standards and rigour in their content, the legitimacy of the resulting norms (and thus the sense of ownership which is key to effective implementation), require that other stakeholders, including the private sector, are able to inject their views during the standard development stages. Like several other standard setters, the Basel-based committees have accordingly put in place elaborate mechanisms to consult broadly and regularly on the key issues under consideration.
I am aware, of course, that those not directly involved in standard setting do not always find the existing consultation mechanisms fully satisfactory. This is an issue that will require continuous attention. The standard-setting groupings may need to be willing to adapt their membership to reflect changing realities in the international financial system, taking care, of course, to keep their membership small enough to be effective, and to remain focussed on high-quality (rather than simply widely-acceptable) standards. Moreover, we can probably make consultation mechanisms even more effective. One of the reasons for the BIS to establish offices in Asia and Latin America, and to work closely with regional organisations of central banks and other regulators, is to establish an even closer dialogue on standard-setting issues.
A third issue is whether standards should be global and applicable to all. Or should they vary across countries, depending on circumstances? As often happens, there are arguments on both sides of this issue. Globality and consistency in application would seem necessary to create the level playing field in which competition can thrive. Moreover, the potential contagion from countries with lax standards or applications is an argument in favour of comprehensive and equal application of standards in all financial systems.
At the same time, however, countries do differ in their historical and cultural characteristics. This may call for some degree of differentiation in the application of the standards. An example is the management in an Islamic banking system of a range of risks which in a conventional banking system would be subsumed under interest rate risks. The level of sophistication of the domestic financial system may also condition the extent to which particular standards or best practices are adopted. Some standards contain a high degree of complexity. Requiring countries to apply standards in their full complexity may unnecessarily delay the implementation process and in some cases even deter countries from starting it. Sometimes a solution to this dilemma is sought by trying to set out best practice at the level of broad principles and to augment these principles with methodologies showing how principles should be applied in particular cases. This has been, for instance, the spirit of the Basel Committee's Core Principles of Effective Banking Supervision.
This brings me to a fourth related issue. Should standards be minimum requirements or best practices? On the one hand there is the practical consideration that reaching consensus on best practices across a wide range of countries is very difficult. Accordingly, there is the temptation to stipulate minimum requirements which can be quickly put in place. Yet we are also aware of the numerous examples of how costly it can be to have standards that are not up to scratch. Moreover, I doubt that many national regulators would want to be seen as falling below standards considered necessary in other parts of the world. Resource limitations in emerging market economies may sometimes require more gradual movement towards best practice. But the fact that best practice cannot always be achieved in the short term should not imply in my view that standard setters should look for the lowest common denominator.
In all this, what is the role of central banks? Given their responsibility for overseeing systemic stability, and given their reliance on well-functioning markets and well-managed counterparties for pursuing their other objective of price stability, central banks have a vital stake in the successful development and implementation of quality standards and codes of best practice. Central banks' crucial role as lender of last resort and core guardian of the payment and settlement system of most countries gives further weight to the financial stability objective. Even though in several countries supervisory functions have been assigned to non-central bank agencies, it would be unfortunate if this process were interpreted as a licence for marginalising central banks in the debate on financial stability. But it should also force the central banks to reflect carefully, perhaps more so than in the past, on how to give substance to the specific role they should play in the design of sound financial systems.
One important aspect of these reflections is that the role of central banks has to be seen not in isolation, but in conjunction with that of the other guardians of financial stability. Complexity, globality and institutional interdependence forcefully argue for close cooperation between central banks, regulatory authorities and other standard setters, and for open discussion of the nature and causes of the vulnerabilities of financial systems and the potential remedies for them. In a number of advanced countries, formal mechanisms of collaboration between regulators, finance ministries and central banks have been established. Such structures can ensure that the needed multidisciplinary perspective is brought to bear in preventing, and if necessary dealing with, financial vulnerabilities.
The conviction that increased cooperation amongst financial authorities was essential was at the origin of the creation of the Financial Stability Forum (FSF) some four years ago. The FSF brings together top officials from the central banks, regulatory authorities and finance ministries of countries with major financial markets, along with senior representatives of the main international financial organisations and standard-setting groups. It is thus uniquely placed to monitor the consistency and comprehensiveness of standards and codes, to share experience with issues of implementation, and to identify potential vulnerabilities in the international financial system.
It is important to be clear about what the Forum can and cannot do. In the first place, it is, as its name states, a forum for discussion and not an executive body. It can sensitise its members to the concerns of others, it can pool information and perspectives, and it can provide political and technical impetus to deal with prospective problems. But it cannot substitute for the responsibility of central banks and other national and international organisations, to take the concrete steps that are necessary to address potential vulnerabilities in the domestic and international financial system.
Second, the Forum has a limited membership. Some have therefore questioned its legitimacy to prescribe solutions for general application. As Chairman, I have tried to address this limitation by establishing regular regional meetings with financial stability authorities in Asia, Latin America and Central and Eastern Europe respectively. Last October, for example, we held a very successful regional meeting in Beijing, with top representatives from the principal Asian countries, as well as key members of the Forum. Beyond this, we have included representatives of non-member countries in our working groups. And finally, I have tried to consult the private sector and non-governmental organisations. In this way, I believe the FSF can develop into a broad-based process in which all those with a stake in financial stability can play a role.
Let me close with a few words about Basel 2. As you know, we are now approaching the final phase of agreement on the new Capital Accord. I expect this to be complete towards the end of this year, and to be implemented in 2006. It has been a long process, but I am confident that the result will be a more robust and risk-sensitive accord than the one it will replace.
There is not the time here today to go into the detail of the new proposals, so I would like to restrict myself to one general comment and to address five criticisms that are sometimes made.
The comment is that capital adequacy guidelines for banking systems are enormously important, and have proved their value. A well-capitalised banking system is essential to weather downturns in the economic climate. I doubt whether the economies of North America and Europe would have survived the recent series of economic shocks without significant financial turmoil, had not their banks built up their capital during the previous period of expansion. In emerging markets, which often face an even more volatile environment, robust capital standards are, if anything, even more important.
Of course, the importance of an adequate capital cushion does not tell us anything about the appropriate way to measure how much capital is needed. We know that the old rules under Basel I had become outdated. Banks had founds ways to manage their risks, and to "game" regulatory capital requirements. The critical link between risk and capital holding had therefore been weakened. Hence the attempt to develop new rules, with a closer relation between underlying risk and required capital.
Let me now focus on five of the criticisms that have been levelled at the new accord:
II. It reinforces the procyclicality of the financial system
III. It gives too much weight to the judgements of rating agencies
IV. It penalises small and medium sized enterprises (and some developing countries)
V. More generally, it does not take adequate account of the special situation and concerns of emerging countries.
To the criticism of complexity, there are two answers. First, banking itself has become more complex. Risk management is highly sophisticated, and the underlying approach of the old capital accord, that all loans could be assigned equal risk weights, (which, by the way, was never intended to imply that banks should treat them as equally risky) had become unrealistic. Pillar I of the new capital accord sets out to prescribe minimum capital holding on the basis of more differentiated assessments of banks' risk exposures. Simplicity and greater risk-sensitivity are not mutually compatible objectives.
The second answer is that, for banks with straightforward business models and non-complex loan portfolios, the new accord really adds very little in the way of complexity. Under Basel 2, there will be a range of options for calculating capital ratios. It will remain possible to use a standardised approach, although it will also be possible for banks to adopt more sophisticated approaches, where they have the desire and the capacity to do so.
A second criticism is that the new accord might reinforce pro-cyclicality in financial behaviour. If capital requirements are more sensitive to risk measurement, and if measured risk tends to go up in economic downturns, banks may be encouraged to step lending just at the time when the real economy is most vulnerable. A first point to make in response to this criticism is that procyclical behaviour of this type is endemic to financial systems, and not simply the result of regulatory requirements. Of course, it is important that regulation does not inadvertently amplify the economic cycle. Several approaches can help here, and have been incorporated by the Basel Committee. Firstly, it can be made clear that minimum capital requirements are just that - minimum requirements. Banks should ensure that their capital is adequate in all circumstances. They may therefore be expected to build up a cushion of excess capital in good times, so as to have a margin of protection in downturns. Second, capital requirements under Pillar I of the accord should not be applied blindly or mechanically. Pillar II, supervisory oversight, and Pillar III, market discipline, are also important in reaching judgements about a bank's capital position, reinforcing the incentive to maintain a cushion of capital above the minimum. Third, it is desirable to use measures of credit risk that are not excessively vulnerable to short-term revisions.
This brings me to the third criticism that has been directed to the new accord, namely the role that it assigns to rating agencies. Many emerging countries have pointed to the limited penetration of ratings in their economies, and have expressed doubts about the quality of rating agencies' judgements. Nobody can dispute that lending decisions should be based on rigorous credit appraisal, and that supervisors should have the means to review the way in which banks manage their credit risk. To the extent that the new accord spurs the development of high quality ratings, and/or encourages more robust internal credit appraisals, that is all to the good. I would not give too much weight to the fear that borrowers with low, or no, ratings will be disproportionately penalised. Banks are pretty well able to price credit risk, and a change in the regulatory capital charge is not likely to add more than a few basis points, if that, to borrowers' costs.
It follows from what I have just said that I feel the fourth criticism, that the accord penalises small and medium-sized enterprises (SMEs), is also exaggerated. In any event, let me recall that the purpose of the accord is to align capital holding with risk, not to promote any particular sector of the economy. In response to earlier criticisms, the Basel Committee has now taken another look at risks of lending to SMEs, and has significantly modified its original proposal. I am not sure that SMEs will be able legitimately to blame the capital accord for their cost of borrowing (though lenders, of course, may find it a convenient excuse to justify what they would have done anyway).
Finally, let me say something about the fifth criticism - namely, that Basel 2 is focussed on the concerns of the industrialised countries, and pays inadequate attention to the circumstances of emerging market economies. In a formal sense, of course, this has some truth to it. Basel Accord is intended to apply first to internationally active banks from the G10 countries. It is the Core Principles for Banking Supervision developed in 1997 by a representative group of industrialised and emerging market economies that provide the basis of banking system oversight more generally.
Nonetheless, it is clear that Basel Committee guidelines have considerable force, and for that reason it is appropriate that the concerns of all countries be taken properly into account. In that connection, I want to emphasise that the process of consultation over Basel 2 has been both broad and deep. In response to the three consultative documents, the Committee has received and carefully reviewed a total of nearly one thousand comments. In addition, the Committee has had continuous and fruitful discussions with regional supervisory groupings, and none more intense than with those in Southeast Asia. Many substantive aspects of the present proposals reflect the suggestions made in this process. Naturally, not all are satisfied with every aspect of the end-product. But this is the nature of any collective process.