GEI Newsletter Issue No. 6

Andrew Crockett

Also in this issue:
Editorial
Reforming the World Bank


The Theory and Practice of Financial Stability

Andrew Crockett

This talk was presented at a GEI conference on the Origins and Management of Financial Crises on held at Gonville and Caius College Cambridge on 11/12 July, and at the Bank of England on 14 July 1997. It is loosely based on a paper of the same name which appeared in De Economist (vol 144, no 4, pp 531–8). A full report on the GEI conference will appear in Issue 7 of this Newsletter.

What can we say from first principles about the question of financial stability. To answer this question is the task which I have set myself in this talk. Much of what I have to say will, as a consequence, be familiar to people in this audience. Nevertheless much of the literature starts off from the closed economy case and doesn’t deal in detail - as we have been dealing this morning - with the international aspects of financial stability, and with the manifestation of international instability in the form of currency crises.

I am going to divide my remarks into three broad areas.

First of all, what is financial stability? I will take a broader definition than is sometimes used in central banks, when they confine their objective for financial stability to be that of ensuring the stability of the banking system. I’m going to define financial stability as stability in financial institutions and markets more generally. I will touch a little on the question, are banks special? I think that an emphasis on “specialness” can be overdone.

Secondly, I am going to ask the question, why should financial stability be a public policy concern? Obviously it is in the public interest to have stable financial markets. But this is true of all markets, so why are we concerned not just with competition, advertising, health and safety, and so on, as in other markets, but also with stability issues. I think that this is an important question to answer before going on to the issue of how to frame public policy to maintain stability.

My third question is to be: how can stability be achieved? I will touch on the role of safety nets, on the role of regulations, and on the role of market forces.

What is Financial Stability?

Let me begin by defining terms. I don’t think that there is a readily accepted definition of financial stability. By contrast, for example, there is a well-accepted definition of monetary stability, which we define in terms of the degree of inflation. That being so, for monetary stability the remaining questions centre around the relatively minor aspects of how to choose a number for the degree of inflation which is to be associated with monetary stability.

I’m going to define financial stability as an absence of instability, even although I was told, when I originally wrote the paper on which this talk is based, that the title ought to focus on the theory and practise of financial stability, and that it was wrong for a Central Banker to talk about instability! Basically, however, I think we do define stability with reference to its inverse. The definition of instability that I employ for the purpose of this paper is a situation in which economic performance is potentially impaired by fluctuations in the price of financial assets or by an inability of financial institutions to meet their contractual obligations. I would like to focus on four aspects of this definition.

Firstly, there should be real economic costs. I don’t think that we would want to define as financial instability something which did not have the capacity to cause measurable damage in the real economy. Secondly, it is the potential for damage rather than actual damage which matters: if, as a result of skilful management, significant economic damage does not result, such problems are ones - I would submit - that we still need to be concerned about. If one thinks, as an example, of the drop of the stock market in 1987, in fact there was very little measurable effect on the real economy but I would guess that many of us would think that this event revealed a troubling degree of instability. Thirdly, my definition refers to the price of financial assets as well as to the position of financial institutions. In other words it refers not just to banks but to non-banks, and to markets as well as to institutions.

Fourth, my definition allows me to address the question of whether banks are special. In one strand of the literature, financial crises are defined as crises that effect the money supply, and then effect economic activity through the money supply. This is a caricature, but one that reflects the monetarist strand of the literature. It is clear that bank’s special position has been attributed to two factors. The first of these is that banks have long term assets and shorter term liabilities, and this makes them especially subject to runs. But this, for reasons which I will come to in a moment, is a difference in degree rather than a difference in kind, in comparison with other financial institutions and indeed with financial markets generally. Secondly, banks are the mechanism by which other institutions and financial systems are linked together through payments and settlements. But that too is much less of a distinguishing characteristic than one might think: all institutions that have large exposures - all institutions that are largely interconnected whether or not they are themselves directly involved in the payments system - have the capacity, if they fail, to cause much widespread damage in the system.

The fact that I have said that I want to define financial instability as including markets as well as institutions will not come as a surprise or be controversial in this meeting, because this morning we have been talking about currency crises.

Despite all of this, in some - particularly in central banking and supervisory circles - there is a tendency to focus on a particular kind of instability, namely weaknesses in financial institutions, as being the major part of the systemic danger.

Why is Financial Stability a Public Policy Concern?

Let me now pass to the second question. Why is financial stability a public policy concern? In one sense this might appear obvious. A stable financial environment is one that permits growth in financial intermediation, and such growth in financial intermediation facilitates the mobilisation of savings, improves the efficiency of allocation of capital, makes the system work more effectively, promotes investment and growth, and all other Good Things. Of course, too, the fact the Group of Seven, and the Group of Ten have focused in recent years on the issue of how to improve arrangements for financial stability, makes in seem perfectly natural that public policy officials should devote attention to intervention of one kind or another within the financial system to promote stability. But the question can be put and is put, why not leave it to the market?

I think that there are two arguments or maybe three arguments, that on the whole I find persuasive, as to why there are differences between the financial system and other markets. One argument involves something that we have been discussing this morning. This is what one might call the “instability bias” in the financial system which is probably not present in most other markets, and certainly not present to the same degree. Self fulfilling runs on currencies, self fulfilling runs on banks, the presence of multiple equilibria, and the possibilities of swings from one equilibrium to the other, all of these argue for some kind of outside intervention to prevent instability from being generated by outside events.

Secondly, the externalities imposed by disturbances in financial markets are argued to be particularly significant.

Thirdly, the financial system ramifies throughout all other markets in a way that no other single market does. I will say a few words about the causes of this so called instability bias. We know more now about the dynamics of bank runs than we used to. These result from the fact that the institution of a bank is quintessentially vulnerable to this phenomena, having assets which are non-marketable and longer term and hard to value, but liabilities which are short term and redeemable on demand at a fixed price. The dynamics are that when the equilibrium between withdrawals and the returns on assets is disturbed by some outside event, it is in the interest of creditors to run as quickly as possible, to get their assets out before the value of the financial institution declines as a result of the forced sale of assets. The point that I want to make is that, whilst this problem is a clear one in the case of banks (as we have been discussing today ) it can also exist for markets. One example is when there is a run on a currency. In the situation in which the liabilities of the country do not always serve as good collateral, the creditor that first fails to roll over its credits will protect itself at the expense of other creditors. But we should be clear that this problem is a rather general one: it can also apply to any institution that has assets of a longer maturity than its liabilities

I would suspect that it is not such internally self fulfilling runs that have been at the source of most banking sector crises in developed countries in recent years. It has been plain simple bad lending. This has created problems that have come home to roost after a period of years, or when the macro economic situation deteriorates. The problems have been due, that is, to credit cycles. These are caused by a variety of factors which have become familiar through the literature on the asymmetric information: fading memories of previous crises, or of the difficult periods of previous recessions, encourage optimistic views on the part of lenders; disaster myopia; perverse incentives created by compensation structures; principle agent problems. All of these generate incentives for financial institutions to make loans in fashions that eventually result in the kind of problems and crises that we have seen. This appears to be why individual institutions get into trouble, why there is a potential for a bias to instability. Such instability can spill over in contagion, through the well known channels of the payments and settlements systems, and through the existence of imperfect or asymmetric information, whereby difficulties of one bank or financial firm are interpreted as being signals of difficulties to others.

We are quite familiar with the costs of such spillovers from recent studies. These have focused first on the fiscal and resolution costs of banking sector crises. Much of the evidence is reviewed in Maurice Goldstein’s recent booklet. This makes us realise the size of the costs that fall on the budget, or on whatever institution is charged with the resolution of the financial difficulties. These are, of course, transfer costs - ie they are not necessarily real economic costs. But both theory and the experience of many of the countries, not all of whom have got into banking sector difficulties, suggest that there are real economic costs as well. In a number of cases, particularly in developing countries but also in Scandinavian countries, there have been significant drops in GNP as a result of such financial sector crises. These spillover costs can be taken, together with the instability bias, as a justification for the public sector taking an interest in how public intervention can help to stabilise the financial system.

I want to say that some of these kinds of insights - many of which come from game theory - can also be used in the area of financial markets, where instability and imperfect information can create destabilising cycles. In addition, bubbles, when they subsequently burst, can generate financial difficulties, both because they imply misallocation of resources when prices are away from equilibrium and also because they can generate difficulties for financial institutions that have made contracts based on prices that may no longer be justified.

One particular financial market that we have been focusing on is the currency market. Problems can arise with fixed exchange-rate schemes. Both the defence of the currency, and the subsequent failure of that defence (when that happens), can impose significant costs on the real economy. But problems can also arise in flexible exchange rate schemes: in cases where there are large swings in currency values these must create questions as to whether misallocation of resources has not been generated. When one thinks that in a little more than the period of a year the Japanese yen has depreciated by something like 50% against the dollar, one has to question whether this is a situation that is optimal.

I haven’t spoken much yet about equity markets. Equity markets have traditionally [not?] been the focus of the commercial crisis literature [but?] I think it is true to say that they create concerns and worries among public policy makers. There are channels through which fluctuations in equity markets can - at least in theory - have significant wealth effects. These can create turbulence through effects on the strength of the financial intermediaries: think of the dependence of the Japanese banking system on the equity market. They can also generally impact on the state of confidence.

I will not comment on the role of other markets, for example the real estate market, apart from to noting that in a number of cases of financial instability one can look to the real estate market as a cause, because of the importance of real estate, either as collateral, or as a basis of which lending has been made. Fluctuations in real estate values add important effects both in generating excessive lending and, subsequently, in undermining the position of financial intermediaries involved in that market.

How Can Financial Stability be Achieved?

I now come to the third and final one of my three questions: how can financial stability be improved and strengthened?

I will not talk about proposals for narrow banking, or about free banking. These are interesting ideas, but I take it that they are not on the agenda as practical policies.

Safety nets are in most people’s minds as a clear aspect of how to improve the stability of financial systems. Alan Greenspan has said in a number of speeches recently that you cannot expect the financial system to carry enough capital to support all states of the world. As a result of this, I imagine that there will be occasions - even if they may be very rare ones - in which the support of the financial system will have to come from the Central Bank. This is not “bail-out”, it is simply a rational use of society’s capital.

The lender of last resort function, which was initially was intended for situations when liquidity was the problem and the self-generated run was the danger, is generally now also supplemented by deposit insurance, either explicitly or implicitly, in most countries. That has the well known consequence of moral hazard. I won’t go into this problem in detail except to note that it is a very familiar aspect of risk taking: risk taking may not be properly priced. There have been various approaches to trying to reduce such moral hazard (and these are characterised in some detail in the written paper). But all of these involve trying to make a distinction between illiquid and insolvent institutions, and this is much harder in practice than it is in principle. Furthermore, even if you could make such a distinction, doing so does not fully solve the problem. Sometimes the crisis that you want to prevent is caused by insolvent institutions, not just by illiquid ones. “Constructive ambiguity” about the availability of the lender of last resort function is a phrase beloved of central bankers, and it can help, But it alone does not go all the way to solving the moral hazard problem.

Therefore we come, inevitably, to regulation. Regulation, it is interesting to reflect, has passed through three different phases.

The original kind of regulation, which defined fields of activity for banks and protected their franchise within those fields of activity, was an effective method of preventing financial instability because in effect it gave banks monopoly rents within their area of franchise. [ ? What is the argument - that therefore they did not need to peruse so many risks? ] This is an inefficient method of regulation. But of course, not only is it inefficient but, with the development of markets and the growth of arbitrage amongst markets, it became an increasingly impractical form of regulation in the 1970s and the 1980s.

That is when we saw the development of risk-based, capital adequacy requirements. Fundamentally these have been an attempt to reflect the principles, and the regulatory rules, that ought to guide the holding of capital. These requirements became widespread in the 1980s. I have no doubt that they have improved the strength of the international financial system. But as time goes on, we are realising that risk-based capital adequacy requirements, with risk-weights assigned by supervisors, are an ineffective means of reflecting the actual risks in a bank.

I think what we are now seeing is a move towards trying to enlist market practices, trying to make supervisory rules more reflect the actual practice of banks. We have this in market risk lending [weightings?], whereby banks themselves are allowed to calculate volatility in the value of their portfolios and to hold the prescribed degree of capital against that. My suspicion is that eventually this will be an approach that will spread more widely as a measurement of risk by banks. I don’t think that this is going to happen in the immediate future, but I do think that is much more incentive-compatible approach to structuring regulation so that it avoids (or at least limits) moral hazard, and at the same strengthens financial systems. We see suggestions in other areas, for example, in this morning’s reports of proposals by the Group of Thirty to develop industry led standards for risk measurement and for risk monitoring and management. We are almost bound to move in this direction in the future. That is to say, we will inevitably move in the direction of exploiting the self-regulating characteristics of any market, and strengthening these, rather than trying to superimpose outside criteria.

Let me finally, in considering how better to achieve financial stability, come to the question of what to do with markets that display instability. We often find ourselves reduced to the true - but often not very helpful - statement that we need to improve macro economic policies. I think that this is important, but it is much easier to say that macro economic policies should be improved than it is, as outsiders, to do a great deal about this.

I would say in the area of currency crisis, which we have talked a lot about this morning, that one unifying element among several crises is the question of the choice of the exchange-rate regime. Many emerging market countries trying to bring down high inflation rates have chosen some kind of currency anchor. They have done this for reasons that are, I think, quite understandable. They have then got into difficulties because they have been unable to maintain the currency anchor in situations where the balance of payments changes, or where there has been some kind of domestic shock of an economic or political character. Some of the episodes which we are now observing in South East Asia are the result of attempting to hold on to the exchange rate peg for too long. I don’t have any ready answer for this sort of problem, except in general to say that choice of a fixed exchange rate is going to end up being a problem for countries which have chosen this route to reducing inflation, unless, that is, there exists a totally credible means of ensuring and maintaining market belief in this fixed exchange rate. Such a belief might exist in Hong Kong, and might possibly exist in Argentina, but I don’t think that it does exist in very many countries.

Conclusion In conclusion, I would say that the growth in arbitrage transactions, the growth in volume of financial transactions, and the increasing technical perfection of financial markets all make it harder to have any approach to financial stability that does not work with the grain, whether you are aiming to stabilise market prices or whether you are aiming to underpin the strength of individual institutions. Market forces are now so strong - and there are so many means by which participants in the markets can find the ways around restrictions that are incompatible with market forces - that I just don’t think that any other way can work.

[Andrew Crockett is General Manager of the Bank for International Settlements in Basle. From 1972 to 1988 he was at the International Monetary Fund, becoming Deputy Director of the Research Department, and from 1988 to 1994 he was an Executive Director of the Bank of England. Mr Crockett is Chairman of the Steering Committee of the Global Economic Institutions Research Programme]

The Newsletter of the GEI programme is published three times annually to inform policy-makers and the academic community of research, meetings, conferences, and Working Papers of the GEI programme.

To receive regular issues of the Newsletter of the Global Economic Institutions programme or to receive copies of back issues send an email containing your address and contact numbers to gei@cepr.org.