International standard setting in financial supervision
Lecture by Andrew Crockett, General Manager of the BIS and Chairman of the Financial Stability Forum, at the Cass Business School, City University, London, 5 February 2003.
It is a pleasure to be with you for this first lecture in the Institute of Economic Affairs/Cass Business School series. The IEA has long been a significant voice in making the intellectual case for open markets. I have not always found myself in agreement with the positions taken in their publications, but I doubt whether economic thinking and economic policy in the United Kingdom would be where it is today without the IEA's advocacy.
The City University Business School, now the Cass Business School, is of newer vintage, but promises to be equally influential. Under the leadership of David Currie and with its unique location and handsome new quarters, it too should be a powerful intellectual face of modern Britain, and its financial sector. It is, indeed, an honour to be the first speaker in this series.
This topic is somewhat wider than the one advertised, though it encompasses it. The reason for the extension is two-fold. First, as the first speaker in this series of seminars, it is perhaps appropriate for me to take a broad view of some of the issues to be dealt with in later lectures. Second, standard-setting in financial supervision needs to be seen in the context of the overall goal of improving the efficiency and resiliency of financial systems.
First, to definitions. I will take financial stability to be a situation in which the capacity of financial institutions and markets to efficiently mobilise savings, provide liquidity and allocate investment is maintained unimpaired.
Financial stability is distinguishable from monetary stability (though the two are often complementary). Monetary stability is usually taken to mean stability in the overall value of money - otherwise put, low and stable inflation. Financial stability means the absence of strains that curtail the intermediation function of the financial system.
Note that in my definition, financial stability can be consistent with the periodic failure of individual financial institutions, and with fluctuations of prices in markets for financial assets. The failure of individual institutions is of concern only if it leads (as it sometimes can) to an impairment of the basic intermediation role of the financial system at large. And asset price volatility is of concern only if it leads to a severe misallocation of capital.
Twenty-five years ago, the dominant monetary and financial issue facing the industrial world was the control of inflation. In 1980, consumer price increases in the OECD countries, though lower than their peak, still averaged over 10%, and seemed to be headed higher. Inflation was having serious social and political, as well as economic consequences. It had proved remarkably resistant to policies adopted to combat it.
By contrast, the issue of financial instability scarcely registered as a major concern of policy makers. To be sure, there had been isolated episodes of strain, such as the secondary banking crisis in the United Kingdom, and failures of individual institutions, such as Bankhaus Herstatt in Germany and Franklin National Bank in New York. None of these, however, had created wider financial or economic consequences.
Now, a quarter of a century later, there has been a remarkable evolution. Thanks to resolute action by central banks, the battle against inflation has been largely successful. Meanwhile the problem of financial instability has moved up the policy agenda. On the face of it, this is strange. Why has price stability not yielded a "peace dividend" of greater financial stability? And why has financial turmoil proved so troublesome to manage?
To anticipate the conclusions a bit, I will be making five interrelated points. First, following a wave of financial liberalisation, the financial system has come to play a much larger role in the allocation of resources than was the case twenty-five years ago. The capacity of financial system weaknesses to generate strains and even crisis has therefore grown. So have the real economic consequences when the system malfunctions.
Second, there are elements in the functioning of financial markets that naturally tend to overreaction. This "fear and greed" phenomenon is not simply driven by human nature (though this should not be underestimated) but also by elements specific to the dynamics of financial markets.
Third, while a stable monetary environment helps the efficient functioning of the financial system in many respects, it may not be sufficient to eliminate these tendencies to excess. In one respect, indeed, it may even encourage them. Confidence in the power of the authorities to manage the economy can encourage excessive risk taking, if economic agents come to discount downside risks.
Fourth, greater stability can be achieved through a supervisory approach that harnesses the prudential aspects of market disciplines. This means that the behavioural norms and prudential standards incorporated in systemic oversight need to focus on providing accurate financial information to market participants, and a framework of incentives that encourages a proper weighting of downside risks.
Fifth, and last, there is a need to strengthen the macroprudential (or system-wide) aspects of the supervisory framework, including the way in which standards and codes are implemented. We must be aware of the risk that even individually rational and prudent behaviour can at times become systemically destabilising.
As I have just noted one reason for the increased importance of financial stability is the fact that the financial system now plays a greater role in resource allocation than it did twenty-five years ago. There have been many contributory factors in this development, but it is helpful to think of the main driving forces being those of liberalisation and technological innovation. Liberalisation, in turn, has been driven both by ideological trends (the ascent of the free market philosophy) and by the sheer impracticability of maintaining restrictions and controls in the face of technological innovation.
Technology has affected the financial sector in two ways: first, by reducing the cost of processing and communicating information; and second, through the development of new instruments for the measurement and management of financial risk. A dramatic reduction in the cost of financial intermediation has not only drawn new users into the system (ultimate savers and borrowers) but, even more dramatically, encouraged a greater intensity of financial intermediation (i.e., more intermediate transactions between the ultimate lender and borrower). The development of new financial instruments has enabled a quantum leap in the scope of risk management - an advance that has facilitated risk-taking as well as risk-shedding.
The impact of technology and liberalisation on the role of the financial system has been compounded by two other trends in western societies. One is growing levels of personal wealth, and the other is the aging of populations. Together, these trends have resulted in greater volumes of financial savings seeking outlets in the capital markets and in financial intermediaries.
The growing role of the financial sector in the allocation of resources has significant potential advantages for the efficiency with which our economies function. If financial markets work well, they will direct resources to their most productive uses, as measured by relative rates of return adjusted for risk. Risks will be more accurately priced and will be borne by those most willing and able to do so. Real economic activity will proceed with fewer financial uncertainties. Investment should increase in quantity and improve in quality as a result.
There is another side to the coin, however. The fact that our economies have become more dependent on their financial systems means that, if the financial system malfunctions, the adverse consequences are likely to be more severe than they used to be. The past decade or so has provided ample evidence of the costs of financial instability. At the international level, think of the Mexican crisis of 1994-95; think of the East Asian crisis of 1997-98; think of the Argentine crisis that began in 2001 and is still far from reaching its end.
At the national level, there are also examples of costly financial instability in advanced countries. These include: banking crises of Spain and the Nordic countries in the 1980s; the S&L crisis in the United States; and the financial bubble in Japan, whose costs are still being felt today. Closer to home, if one defines financial instability broadly, is the ERM crisis of 1992. And the recovery of the United States from the recession of the early 1990s was delayed by financial "headwinds" resulting from strained balance sheets.
The growing role of financial markets does not by itself explain why financial instability has become more prevalent. So let me now turn to a closer analysis of why open financial markets have proved vulnerable to periodic episodes of stress.
As western countries embarked on the process of liberalising their financial markets, little thought was given to the possibility that this might result in an increase in financial instability. It was generally assumed that in financial markets, as in those for other goods and services, open competition would produce stable equilibrium prices. If, in addition, low inflation could be achieved, that would further support overall financial stability.
This sanguine view did not take into account some particular characteristics of financial markets that differentiate them from the conventional model of equilibrium price determination. Let me note four such characteristics.
First, fundamental value is extremely hard to assess. We can define a financial asset's value as the product of its expected flow of income, a discount rate and a risk premium. But this does not get us very far. The key element in judging the value of a financial asset is how much it can be sold for in the market. A function of financial claims is to telescope into the present intrinsically uncertain cash flow streams. In assessing these uncertainties, there are strong psychological incentives to extrapolate recent experience, and to fall victim to current fashions about how assets should be valued.
Such partial vision is true of individual agents taken in isolation. It is even stronger in the social behavioural patterns reflected in market prices. Price reactions to "news" can go through phases in which, whatever the intrinsic information content, the news is interpreted as reinforcing the prevailing paradigm. The "new economy" euphoria of the late 1990s is only the most recent example of such a phenomenon.
Second, in the financial sector, the process by which equilibrium is maintained does not work in quite the same way as in other industries. Normally, we think of supply and demand curves as being well-behaved. The increased supply of a product exerts downward pressure on its price, thus limiting the eventual increase in supply. In the case of credit, however, an expansion in supply can, for a time, strengthen economic activity and boost asset prices. By thereby improving the balance sheet position of both borrowers and lenders, it can sustain further increases in the supply of credit. Excess capacity and risk can build up partly unnoticed. The mutually reinforcing process between perceived wealth and access to external funding masks the extent of the underlying financial imbalance, until the process, when it goes too far, at some point unwinds. The amplitude of the financial cycle is thereby augmented.
Third, the leverage inherent in financial intermediation gives rise to fragile balance sheet structures. The sudden and sometimes indiscriminate retrenchment of suppliers of funds can cause institutions and markets to be starved of liquidity, exacerbating price declines and impairing the functioning of markets. Bank runs are the textbook example. But there are also cases of markets functioning in a similar way. In the wake of the LTCM crisis, for example, liquidity virtually dried up for a while, and the financial system was perilously close to a full-blown crisis. In my own institution, the BIS, the value-at-risk in our securities portfolio doubled within a few weeks, with essentially no changes in portfolio composition.
Finally, moral hazard plays a potentially important role: the forms of official protection put in place in response to past episodes of financial instability have weakened market discipline without providing offsetting prudential incentives.
While the above characteristics are inherent in financial activity, and in the institutional safeguards put in place in the first half of the 20th century, a number of changes in the financial regime over the past twenty-five years have arguably increased the potential for financial instability. All of them can ultimately be traced back to the financial liberalisation and technological innovation that has gathered pace during the period. This process has resulted in a broader range of services, at lower prices, and more accessible terms than ever before. But these great benefits have not come for free. Let me focus on four implications of financial liberalisation and technological innovation.
First, competitive pressures have vastly increased. This means that the rents that licensed financial institutions could previously extract from their protected franchises have diminished, if not disappeared altogether. The cushion available to absorb mistakes or misfortune has become much thinner. Previously sheltered financial institutions have had to learn to manage risk more actively, with a smaller margin for mistakes. Frequently, they have had to compete with new entrants, not saddled with burdensome cost structures inherited from the past. Net operating margins have thus come under pressure, making it harder to earn a given return for the same amount of risk. Consequently, the incentives to enhance returns by taking on added risk have grown.
Second, the new environment has structurally increased liquidity and the potential for leverage. The development of new financial instruments, and financial engineering more generally, has made it easier for firms, both financial and non-financial, to realise value from assets, whether tangible or intangible. This has, in a sense, "liquefied" a wider range of income streams, and by the same token, increased the potential for leverage. Moreover, the hugely increased emphasis on stock market value has encouraged the exploitation of leverage. In a rising market, leverage is the winning formula. If the period of rising asset prices is protracted, market participants can come to forget the warning that regulators now insist be included in the small point of stock offerings: "Prices can go down as well as up".
Third, the new environment has tended to raise the option value implicit in safety nets. The reason is simple. Ceteris paribus, guarantees become more valuable as the environment becomes riskier. So the hidden subsidy provided by the official sector has become greater, and the danger of resource misallocation through the mispricing of risk has grown.
Finally, financial globalisation has transformed geography, with significant implications for the character of instability. Globalisation has heightened the significance of common factors in originating and spreading financial distress. It has done so by extending and tightening financial linkages across institutions, markets and countries; by increasing the uniformity of the information set available to economic agents; and by encouraging greater similarity in the assessment of that information.
Freedom of capital movements has exposed emerging market countries to potential volatility of access to external funding. Portfolio adjustments that are comparatively minor for institutions in the countries originating capital flows can be of first order significance for the recipients. This greatly increases the recipients' vulnerability to changes in sentiment, whether these are due to revised perceptions of economic conditions in the periphery or to developments at the core.
Some of the environmental changes I have just described are particularly acute during the transition from a sheltered to a liberalised environment. Others, I suspect, have a more permanent character. The bottom line is that market discipline alone may be insufficient to ensure the necessary degree of stability. Hence the issue of whether additional policy action is needed to protect the system against instability. But before I discuss this question, let me say something about the link between financial instability and the monetary regime.
It is not uncommon for economists and financial practitioners to argue that monetary stability should yield, as a by-product, improved financial stability. There is, indeed, much validity in this contention. Inflation has always provided fertile ground for resource misallocation and facilitated the build-up of over-extended balance sheets. Inflation makes it harder to distinguish between real and nominal magnitudes. Moreover, since high inflation is almost invariably unstable inflation, the expectations on which financial judgements are based, are rendered even more difficult to form with confidence.
However, it would not be right to say that price stability is a sufficient criterion for financial stability. There are numerous counter-examples, of which the Japanese and East Asian cases are only the most prominent.
Two possible factors help explain why financial instability seems to persist, even in a world in which price stability has been credibly established. One lies in the paradox that success in taming inflation can make economies even more vulnerable to those waves of excessive optimism that breed unsustainable asset price dynamics. A danger sign is the increased prevalence in upswings of articles heralding the "end of the business cycle". In such circumstances, many may come to believe that low inflation removes a common cause of the termination of economic expansions, namely a sharp tightening of monetary policy. They may thus be tempted to accept balance sheet structures that are particularly vulnerable to changes in financial conditions.
The second possible factor is a more controversial conjecture. It is the following: in a monetary regime in which the central bank's operational objective is expressed exclusively in terms of short-term inflation, there may be insufficient protection against that build up of financial imbalances that lies at the root of much of the financial instability we observe. This could be so if the focus on short-term inflation control meant that the authorities did not tighten monetary policy sufficiently pre-emptively to lean against excessive credit expansion and asset price increases. In jargon, if the monetary policy reaction function does not incorporate financial imbalances, the monetary anchor may fail to deliver financial stability.
One response to this conundrum could be to modify, at least at the margin, the monetary strategies. I have addressed this controversial question on other occasions, and I do not want to return to it here today. Rather, I simply want to make the point that, if the monetary anchor is, along with conventional market forces, insufficient to produce financial stability, then additional policies may be needed to achieve these objectives. This brings me to the final part of my remarks, on the design of prudential standards and codes to enhance the stability of the financial system.
In principle, a competitive financial system should eventually eliminate poorly performing institutions or market platforms, and should encourage the development of prudent and efficient practices. In other words, competition should foster convergence towards best practice in risk management. For various reasons, however, this may not happen, or may not happen quickly enough. Hence the justification for official oversight of the financial sector, to strengthen prudential management.
Prudential supervision of financial institutions has a long history. In the past, financial sector regulation tended to focus on the authorisation of financial institutions, on the definition of their permitted spheres of activity, and on required balance sheet ratios. More recently, however, growing attention has been devoted to the prudential management of risk. Several trends in supervisory standard-setting are worthy of note.
First, there is growing recognition that the most effective regulatory framework is one that works with the grain of market forces, and allows the greatest play to market disciplines. Hence the search to relate regulatory requirements to risk management practices, and to find ways to increase transparency.
Second, it is increasingly accepted that globalisation of financial activity means that prudential norms have to be internationally consistent. Otherwise, the twin risks of regulatory arbitrage and competition in laxity are likely to present themselves. The international dimension of standards and codes raises the issue of how such standards should be developed, a subject to which I will return in a moment.
Third, it is now recognised that efficient financial intermediation requires a high-quality financial infrastructure. By infrastructure, I mean the network of conventions, practices, and information that underlie market activity. This includes the system of contract law and law enforcement, bankruptcy procedures, the accounting framework and auditing standards, corporate governance practices, and requirements for transparency and data dissemination.
Fourth, and last, it has been realised that prudential standards are interdependent. Minimum capital requirements for banks are of little use if the accounting conventions used to value a bank's assets and liabilities are flawed. And accounting conventions are only as good as the auditing standards used to apply them. More generally, market disciplines require accurate information, a legal environment that provides adequate security to market participants, and a payment system that can be relied on.
These underlying trends in the prudential oversight of the financial system have come together in the international effort to develop a framework of codes and standards for the financial sector. This effort crystallised in the wake of the Asian crisis, and owes much to the forceful advocacy of Gordon Brown in his capacity as Chairman of the IMF's International Monetary and Financial Committee.
The strategy is (i) to define those areas of financial activity in which it is desirable to develop internationally agreed standards of best practice; (ii) to assign the role of standard-setting to an appropriate international grouping; and (iii) to devise mechanisms that foster convergence on this best practice by as wide a range of countries as possible.
The first issue is the scope of standards. I have just argued that the financial sector is marked by a considerable degree of interdependence and complementarity. This suggests a broad scope for standards, which can be grouped in three main areas. First, guidelines for supervision of the main types of financial intermediary -- banks, securities issuers, and insurance companies. Second, standards for the transparent disclosure of the financial information needed to facilitate the efficient performance of markets -- macroeconomic information provided by governments, and microeconomic information related to the financial position of market participants and their counterparties. Third, codes for the robust underpinning of market infrastructure -- standards for contract law and law enforcement, corporate governance, accounting conventions, auditing practices, safety and soundness in the payment system, and so on.
The second issue is who should set the standards. Since standards are comprehensive in scope, interdependent in nature and global in their impact, it might seem logical to have a single authority responsible for their formulation. Some, including John Eatwell, have put the case on these grounds for a "World Financial Authority" with broad powers of regulatory design and supervisory oversight.
There is some logic in this idea, but as a matter of sheer politics, it does not seem to me to be practically feasible for quite some time to come. A more promising approach is to assign the responsibility for developing standards in individual areas to groups of national experts. In this way, the relevance of the resulting standards is enhanced and their acceptability in national jurisdictions is strengthened. However, means must be sought to ensure that the resulting standards represent a convergence to best practice and not a lowest common denominator.
This leads me to the third issue, that of how standards are to be implemented. The experience of the Basel Committee on Banking Supervision is instructive in this context. The Basel Committee is composed of senior supervisors from the most advanced countries, who have issued supervisory guidance on a wide range of topics. Basel Committee recommendations have no legal force. But since they have been adopted by consensus, they have been applied in all countries represented on the Committee. Interestingly, they have also been almost universally applied in non-member countries -- a telling example of the power of peer pressure and market forces to promote the adoption of best practice and to enforce what I have elsewhere called "soft law".
The success of the Basel Committee process is not just an academic matter of securing a common regulatory approach. It has produced real economic benefits. I would hazard the view that the absence of significant difficulties in the banking systems of Europe and America in the past couple of years, despite significant economic shocks, owes much to the strengthening of risk management that has taken place under the aegis of the Basel Committee's standards.
Realistically, of course, peer pressure will not be sufficient, by itself, to secure the prompt implementation of the wide range of standards that have now been drawn up. The international institutions, principally the IMF and World Bank have an important role to play here. They are using their consultation missions with member countries to carry out Financial Sector Assessment Programs (FSAPs) and Reports on Standards and Codes (ROSCs). FSAPs and ROSCs are a key means of enabling countries to "benchmark" their current standards on international best practice, to identify weaknesses, and to devise means for dealing with them. Giving publicity to the state of a country's financial sector may, of course, be the best way to ensure that the market is able to reward progress, through greater access to finance, on more favourable terms.
So far, I have been talking mainly about how to use codes and standards to strengthen the supervision of individual financial institutions. Improved risk management at the level of individual institutions can go a long way to strengthening the resilience of the system at large. Still, by itself, it may not be quite enough. My earlier remarks implied that the dynamics of financial markets can introduce inherent pro-cyclicality into market behaviour. How can this be dealt with?
The first step is clearly to understand the underlying causes of this procyclicality. In part, it lies in the short-term nature of many risk measures. Risk measurement is often based on assessments of the recent past and the immediate future. Risk comparisons are made at a point in time on the basis of how an institution compares with its peers. But risk has a time dimensional as well as a cross-sectional character. Existing techniques of risk assessment arguably pay insufficient attention to the movement of risk through the cycle, and the evolution of common risk exposures.
With the benefit of hindsight, we can see that risks tend to accumulate during the upswing of a cycle, then to materialise when the economy turns down. At the time, however, risks may appear to be diminishing the longer economic expansion continues. In other words, conventional risk management tools lack the capacity to identify the emerging over-extension of balance sheets at a system-wide level.
Another limitation of an institutional focus in risk management lies in the interdependence of the actions and assessments of market participants. Risk models typically treat the external environment as independent of the actions of the entity managing risk. In fact, however, not only are some players large enough to impact markets by themselves, many use similar models to guide their behaviour. What this means is that "one-way markets" can develop more easily than theory would suggest. Thus, what is sensible and rational for an individual market participant, acting in isolation, may produce a destabilising outcome for the market as a whole.
A couple of simple examples can help to make the point. When a lending institution faces a possible slowdown in economic activity, it may seek to cut back its lending activities to reduce its risk exposure. Of course, if all lenders act in the same way, they may well produce the result they are seeking to protect themselves against. A similar process can take place in markets for traded assets. If an external shock produces a downward price adjustment, the consequent increase in measured value-at-risk may produce further sales, additional downward movements in asset prices, further increases in VaR and so on.
It is not easy to find solutions for these problems. Nevertheless, some avenues offer useful prospects. First, it would be good for financial institutions to adopt risk-management practices that take better account of the evolution of risk over time. Techniques such as through-the-cycle credit assessment and pre-provisioning may have a role to play here.
Second, supervisors should encourage the use of stress-testing to assess the vulnerability of institutions to "multiple-sigma" events. They may also need to think about how the information from stress-tests at individual institutions can be aggregated for the system as a whole. In other words, how might endogenous reactions to exogenous shocks amplify disturbances in potentially troublesome ways?
Third, it may be desirable to develop oversight structures that enable the authorities to track emerging vulnerabilities in the financial system. Many central banks, like the Bank of England, now issue "Financial Stability Reviews" and I understand that the Treasury, the Bank of England and the FSA now have regular meetings to consider potential areas of weakness in the financial system.
Something similar has taken place at the international level. Following the Asian crisis and as part of their search for a "new financial architecture" the Ministers and Governors of the G7 countries established the Financial Stability Forum (FSF). This brings together, at very senior level, the key authorities responsible for international financial stability. They comprise: the finance ministries, central banks and regulatory authorities from the countries with the largest financial markets; the main international organisations (IMF, World Bank, BIS and OECD), and the principal standard setting bodies (the Basel Committee, International Accounting Standards Board, IOSCO and others).
The FSF has not, perhaps captured the public imagination through dramatic crisis intervention. Nevertheless, it has done much useful work in focussing attention on common sources of financial vulnerability, and in providing an impetus for tackling them. Perhaps even more important, the FSF is welding together institutions and groupings that need to work increasingly closely to promote financial market efficiency and stability. I believe it is a promising tool of international cooperation that can, and should, be further developed in the years ahead.
We have come a long way in the past twenty-five years in understanding the way in which a liberalised financial systems works and its vulnerability to episodes of stress. There have been important conceptual advances, such as new ways to measure risk and price assets, and new insights into financial behaviour, through game theory and behavioural finance.
Yet we still have much to do in applying these insights to achieving the goal of a safe and efficient financial system. The task will be to develop mechanisms of supervisory oversight that make markets work better, not by suppressing the symptoms of market failure, but by addressing their causes. Setting financial standards that attempt to harness prudential instincts and deal with the underlying sources of market failure should play in important role in this endeavour. I like to think it is an endeavour that is fully in accord with the goals and purposes of the IEA and the Cass Business School.