Policies and institutions for global financial stability

The last two decades have seen a marked increase in financial instability. When we compare the period 1945-73 with the last 20 years, the number of banking and currency crises has increased both in frequency and GDP losses.(1) 

We have in fact seen all types of financial instability: banking crises resulting from crashes in real estate markets, asset price inflation followed by recessions and currency crises - associated with debt default problems or otherwise. These different types of crises illustrate the definition of financial instability proposed by Frederic Mishkin (2) “financial instability occurs when shocks to the financial system interfere with information flows so that the financial system can no longer do its job of channelling funds to those with productive investment opportunities”. This includes everything that impairs the main role of the financial system and then affects the real economy.

This recent period of higher instability coincides with the liberalization of capital movements and deregulation that are associated with globalisation. For some the problems have resulted mainly from wrong policies pursued by governments and not so much from the difficulties in controlling the effects of more significant international flows.

For a time, the efficient market hypothesis seemed to dominate and according to it, financial markets correctly price assets at all times because they process all the available information. Both theory (3) and experience nevertheless refute the validity of this hypothesis. And it is difficult to deny that instability reflects the risks associated with financial liberalization as well as technical developments like the appearance of financial products with capital protection that imply hedging, the automatic selling by institutional investors when markets fall or the inclusion of trigger clauses in bond contracts.

These developments towards more open and sophisticated markets undoubtedly improve the efficiency of the system but they also contribute to potential volatility. Other changes may also have played a role in the same direction. The increased importance of market financing in detriment of relationship banking that ensures a closer monitoring of risks has been a factor underlined by Alexander Lamfalussy (1999). Another element may be related to the exclusive concentration of monetary policy with inflation, implying less concern with financial stability.

Since the Asian crisis in particular, one can detect a change of attitude both among national authorities and in international financial institutions. The first consequence was the search for a new International Financial Architecture and the abandoning of the idea to change the IMF Articles of Agreement in order to introduce the principle of total freedom of capital movement as a condition for access to the Fund’s facilities. The new wisdom became the recommendation for an orderly sequencing of the opening of the capital account, making it conditional on the strengthening of the domestic financial system. In tandem, with the work for a new Basel Accord, the effort was initiated to reinforce prudential policy and the supervision of financial institutions. As the last element of the reaction to the volatility of markets, there are the ongoing discussions about the role of monetary policy in achieving financial stability.

In the time that I have, I will focus my comments on three subjects: the relationship between monetary policy and financial stability; the role of prudential policy and Basel II; and lastly the problem of new mechanisms to ensure private sector involvement in international crisis resolution.

Monetary policy and financial stability

In terms of the first problem, I would like to start by recalling that 15 years ago no one separated monetary stability from financial stability. The mission of Central Banks and their function as lenders of last resort included ipso facto the ultimate responsibility for the stability of the financial system and oversight of the payments system. In spite of possible tensions between those two functions, still present today, central banks always worked to achieve both objectives of monetary and financial stability, a task facilitated by the fact that in general central banks also had the responsibility for supervision of the banking system. The concern with systemic problems was naturally linked with macro prudential aspects of supervision. More recently, two developments have changed this environment somewhat. The theoretical drive to restrict the remit of central banks to price stability alone and the new fashion of giving supervision of banks to other institutions.

In relation to monetary policy, the two aspects of financial stability that are relevant are the excessive volatility of markets and the possible fragility of financial institutions. Avoiding both is important to guarantee the proper functioning of the financial system, as the excessive increase of asset prices creates the risk of post bubble recessions as we have seen in the case of Japan. Financial fragility may ensue and balance sheet weakness can lead to a credit crunch that aggravates the consequences for the real economy. This shows how important it is for monetary authorities to try and avoid such situations if they want to achieve stability of prices in the long run.

Of course the attainment of the primordial objective of monetary policy, low inflation in the market of goods and services, facilitates the maintenance of financial stability. But historically we have seen that it is sometimes not enough. According to some, like Alan Binder (1999) or Andrew Crockett (2000), the efforts to disinflate may even contribute to asset inflation. Subdued inflation and associated low interest rates can lead to high credit growth and unfounded expectations about returns in equity markets or the housing market. With monetary authorities exclusively focused on inflation, credit growth which fuels asset price increase may not be seen as a problem. This view is controversial and not dominant, but it has led to questioning the more conventional wisdom because of the recent bubble in the equity market and ongoing worries about the situation of the housing markets.

Nevertheless, the objections to the use of monetary policy for the purpose of fighting against asset markets overheating are well known. 

First, it is very difficult to pass judgment on market valuations and to be sure that there is a significant misalignment.

Secondly, as there is no reliable theory about the transmission from monetary policy to asset prices, there are doubts about the effectiveness of its use. 

Thirdly, there is the risk of a rise in interest rates triggering the sort of recessionary consequences that one wanted to avoid in the first place. 

A final difficulty stems from the possible contradiction of effects on different markets. For instance, an increase in interest rates to cool security markets may lead to currency appreciation which in turn can aggravate deflationary risks.

All those arguments show how difficult it could be to explain the use of monetary policy to counter asset price volatility. No adoption of a precise objective would ever be possible. In a well known analysis, Bernanke and Gertler (2000) have shown that in a general equilibrium model the attempt to achieve bubble-bursting leads in the long run to less stable inflation and output when compared with inflation targeting regimes or a simple Taylor rule. 

This illustrates that contrary to Cecchetti et al (2002) suggestions, a reaction to asset price misalignment is very difficult to be formally integrated in monetary policy objectives. On the other hand, monetary policy is not an exact science and should not ignore the indications given by asset markets. For instance, the wealth effects of asset valuations will always have to be considered. But besides this direct effect, the medium term view about the way financial stability can affect the possibilities of stable prices, in either direction, justifies some consideration being given to financial stability when monetary policy decisions are taken. After all, it is what everyone expects monetary authorities to do in the case of a severe downturn in asset markets. Providing an ample supply of liquidity to sustain the stability of financial markets and institutions is what central banks are supposed to do, even with temporary disregard to monetary stability. That was the case in the highly praised reaction of the FED to the 1987 crash. (and immediately after 9-11).

It is possible that monetary policy suffers from an endemic asymmetry, being efficient to counter possible financial meltdowns, but unsafe or ineffective to fight asset inflation. Nevertheless, the more traditional asymmetry attributed to monetary policy, the one that considers it to be more successful to control inflation than recessions, was never an obstacle to its use for both purposes. In the same vein, one could argue that even without defining precise targets, interest rate policy, in certain circumstances, should “lean against the wind” that blows in asset markets. In any case, the elusive search for anchors that could help to achieve both monetary and financial stability will continue, and monetary policy cannot ignore the problem. Nevertheless, in view of the difficulties described we must look for other policies.

To conclude this point, I want to stress its relevance for countries desinflating in transition to a new regime of pegged exchange rates or even monetary union. In fact, a credible transition to a low inflation regime with interest rates well below what they used to be can lead to an explosion of credit to economic agents that become solvent all of a sudden. Overheating can ensue with inflationary tensions in all markets as a result of the easier access to credit by both firms and households. As monetary policy is directed to reduce inflation, or simply not available for countries already members of a monetary union, are there any alternative policies that can be used to reduce the “boom and bust” risks existing in such a situation?

The question has been raised about some small peripheral countries of the euro area like Portugal. In our case, in spite of a credit growth rate above 20% from 1998 to 2000, we have had no significant asset inflation either in equities or the housing market. The consequences of high indebtness among economic agents have been manifest in the current account deficit. However, the high level of debt reached by the private sector explains why we are now going through a period of marked slowdown in domestic demand, which has aggravated the recessionary effect of the international deceleration of growth. This means that the risks of a backlash following the overheating phase can manifest themselves in different forms. Is there anything that can be done to offset these effects?
The more immediate response must come from a forcefully counter-cyclical fiscal policy. We should however be aware of the limits of such a policy to offset in a significant way the explosion of private expenditure after the change in the monetary regime. A recent OECD working paper (4) provides some simulations for small euro area economies and shows that the degree of restrictiveness of fiscal policy that would have been required to counterbalance the private demand explosion, would be very difficult to attain, if not outright impossible.

Regulation, supervision and financial stability

Another possibility to help ensure financial stability is of course prudential policy. At the macro prudential level this refers in particular to all measures that can reduce the usual pro-cyclicality of financial regulations, measures that strengthen defences in good times that can be used in the downturn. These include, in particular, rules on discretionary adjustments in capital standards, statistical or anti-cyclical provisions, tighter “loan to value” ratios and greater reliance on stress testing.

These measures may have some effect because in general the way the banking system works and is regulated is strongly pro-cyclical. In the strong phase of the cycle, risk assessment tends to be more relaxed, costs with provisions for non performing loans are lower and as a result credit growth accelerates. When the economy is weak the opposite occurs, reinforcing recessionary tendencies. To be fully aware of these developments, the authorities should improve the use of macro prudential analysis and draw the necessary consequences for monetary policy. A recent example of this behaviour, according to market analysts, has been the Bank of England decision not to cut rates because of the overheated housing market which, should it continue, could lead to future severe risks for the banking sector, as in the early nineties.

The problem of pro-cyclicality has been a point of concern in the preparation of the Basel II Accord. One positive and very important aspect of the new approach is the fine-tuning of capital requirements, adjusting them to a better assessment of actual risks. This is true either in the standard methods or in the internal-ratings-based approach. However, the better connection with risks as they develop could exacerbate the pro-cyclical nature of capital regulation. In the ensuing discussion about the best ways of dealing with the problem, three methods have been analysed: an accounting standard contemplating an allowance for a loan value impairment charge to the income statement, full fair value accounting and statistical provisions.

The first method is used in IAS 39 but the way it defines the “objective evidence” to determine the loan impairment makes the method too backward-looking. In the meantime, full fair value accounting has fortunately been dismissed as non-appropriate to banking. Statistical provisions seem altogether a preferable method: it consists of imposing general provisions based on an estimate of the long-term expected losses from defaults; it results from a statistical calculation of losses already present in the loan portfolio and it respects the principle of reserving capital for non-expected losses and builds up a buffer in times of strong growth that can be used when an economic slowdown occurs. In fact, these provisions accumulate in a “statistical fund” which can be used when bad debts materialize and determine specific provisions higher than the average annual charge for general provisions.

In this case, specific provisions over the “average statistical charge” can be made against the statistical fund instead of the current year’s profits. Costs with provisioning are therefore smoothed over the cycle and the backward-looking and pro-cyclical nature of the method now used is eliminated. By imposing higher costs when the economic situation is good, the statistical provisions method introduces a small brake in the process of euphoric credit growth. Unfortunately, Basel II will not directly address the question of pro-cyclicality in financial regulation. Under the second pillar though, supervisory authorities are given sufficient scope to adopt measures they consider necessary or useful. This is very important because in the present circumstances the banking sector needs a stronger capital base and not relaxation of regulatory requirements which for some could result from the new Accord.

International financial crises

One positive aspect of Basel II is the way it imposes a better consideration of the risks involved in inter-bank lending. The present Accord is too lax, especially in relation to emerging markets. 

This leads me to the third and last topic I want to address - the international dimension of financial stability. This is the most difficult aspect of our subject as crises have been recurring in the recent past and the progress made in dealing with them seems to be insufficient. The last decade was indeed the most unstable since the Second World War and also the one that registered the lowest growth rate of the world economy. The succession of financial crises contributed to this result and mostly affected the developing countries, with an average cost to them of 8% of GDP or 18% when a currency and banking crisis occurred together. It is not surprising then that since 1997 some initiatives have tried to devise a new international financial architecture to deal with that instability. Besides this objective, it is necessary to consider also the need to deal with increasing inequality and poverty in the world, a fact that has been described as “the ultimate systemic threat”. 

It is unavoidable to associate the increasing recurrence of financial crises with the successive opening of their capital accounts by more countries. There are indeed clear advantages that come from easier circulation of capital resources, as savings are made available where they are most needed and productive, thus increasing efficiency in the world economy. For the receiving countries however, the benefits depend on certain conditions that are not always met and those benefits are not as clear-cut as the ones deriving from free trade (5). As Stanley Fisher (1999), the former IMF vice-president, reminded us “there is as yet little convincing econometric evidence bearing on the benefits or costs of open capital markets”. International financial markets are potentially unstable as a result of asymmetric information between lenders and borrowers, the pervasiveness of incomplete markets to deal with risk and the difficulty of international enforcement of contracts. As is well known, the evolution of domestic financial systems has been dominated by attempts to reduce the effects of asymmetric and imperfect information. The role of government intervention in this field also has the same origin. Specific measures have therefore followed the most recent crises. After Mexico 94, the lessons drawn led to a stress on the necessity for full information and transparency. Initiatives about data dissemination standards ensued. After Asia and Russia 97/98, the important conclusions referred to surveillance and supervision of financial systems as well as the first attempts to bring about private sector involvement in crisis resolution. After Argentina the last aspect gained prominence.

There is still no consensus regarding the explanation of these international crises affecting emerging countries. One school of thought links them to excessive borrowing and lending stemming from moral hazard. According to an alternative view, they correspond to a jump to a bad equilibrium in an inherently fragile environment. This latter view is related to the so-called second generation models of external crises. There has been a move in the literature towards this type of self-fulfilling panics associated with two types of mismatches that afflict emerging countries: a maturity mismatch and a currency mismatch. Both have their source in what Eichengreen and Haussman (1999) have called the “original sin”, meaning by this the impossibility of those countries borrowing from non residents in their own currency and the difficulty in borrowing long term domestically, also in their own currency. Incomplete markets and weak institutions lead to financial fragility and to multiple equilibriums. The possibility of “sudden stops” of capital flows and of contagion can move a country to a poor equilibrium even when fundamentals could lead to a good one, provided market expectations were different. The usual comparison is between Brazil in 97, compared with Australia or New Zealand, countries with apparent weaker external positions (higher gross debt as a percentage of GDP) but that do not suffer from “original sin”. The second generation models that support this view get their inspiration from the classical Diamond-Dybvig model of bank runs (6) and tend to stress the liquidity aspect of financial crises. Too small financial flows seem to be the problem, just the opposite of the view based on moral hazard considerations. It is true that medium term international capital flows are less important now than in the heyday of the gold standard for the countries participating in the world economy. With low levels of capital flows the convergence of capital/labour ratios among nations is not feasible, something that is contrary to the theoretical conclusion concerning a more efficient allocation of resources worldwide. A dominant liquidity interpretation of financial crises leads to the defence of solutions based on the creation of a true international lender of last resort (Stanley Fisher, 1999) or an institution to provide insurance for emerging countries debt (Soros, 1998). However, these solutions have too many problems and are impractical in the present environment. 

In any case, the dominant view is the one that maintains that the problems come from too much borrowing by emerging markets because of implicit guarantees that create moral hazard. Pegged exchange rates and IMF bailouts are the two main forms of those guarantees. The obvious solutions leading to reduced moral hazard are therefore a more vigorous regulation and supervision of the financial system and the partial substitution of big IMF bailouts by workouts that require private sector burden-sharing. 

The initial steps in building a new international financial architecture relied mostly on the imposition of codes and standards to improve disclosure of information and to step up supervision of the financial system. Further work on sustainability analysis and early warning systems needs to be forthcoming. Financial stability reports should stress the need for financial institutions in advanced countries to implement proper risk management systems and build a stronger capital base. Other developments helpful for crisis prevention have been the change to a more careful consideration of capital account liberalization and the recognition of the advantages of a more flexible exchange rate regime. One of the lessons drawn from the Asian crisis was that pegged exchange rates provided a promise of stability that led to high unhedged foreign borrowing at lower interest rates. These large capital inflows created imbalances and led ultimately to currency depreciation. The double mismatch I mentioned before creates an almost impossible policy dilemma. Increasing interest rates to defend the currency overstretches balance sheets andcan lead to bankruptcies. But the same may happen if the exchange rate is allowed to fall and the currency mismatch produces its effects. 

The conclusion has been that the only possible exchange rate regimes are either a rigidly fixed rate, like in a currency board or complete dollarization, and the alternative of floating rates. The Argentinean case has called into question currency board regimes. On the other hand, floating rates can become very volatile and thus increase the cost of hedging. Structurally, these countries are subject to the currency mismatch and so a severe depreciation can create significant disruption. This explains why, in Calvo’s expression, these economies have “fear of floating”. To counter this, interest rates become more volatile in the attempt to stabilize the exchange rate and this will make it more difficult to develop a long-term bond market. In general, both exchange rate and interest rate volatility will lead to a higher level of interest rates. These realities imply that, as a result of structural weakness, all regimes have problems of their own and will not overcome the possible causes of financial crises. Countries need to be very prudent, and let deficits or debts increase too much. This will nevertheless have a high cost in terms of development opportunities lost. Besides, as the Asian crisis showed, in a liberalized economy the private sector may become the agent of indebtedness and not the Government. Also, when capital is flowing in, the exchange rate appreciates and as assets get more valuable this can trigger herd behaviour and the inflow may increase excessively. The implicit guarantee given by international bailouts can influence foreign investors to become less careful and inflows will continue only to augment the risk of a big outflow later on. Moral hazard also affects investors in a significant way and financial crises continue to be a possibility. 

This explains why the more recent efforts to improve the system have focused on private sector burden-sharing in crisis resolution. Private sector involvement (PSI) is an important way of reducing moral hazard and stimulating investors to maintain a more prudent attitude. Voluntary negotiations between creditors and debtors were already attempted with success in the cases of South Korea and Brazil in 97/98. Banks accepted to roll over short term lines of credit but there was an element of real financial burden involved. PSI is really relevant in the context of debt restructuring exercises. The central idea is that in many cases of real insolvency it is better to recognise it and organize an orderly process of debt-restructuring that may imply some financial sacrifice to creditors. After all, the high interest rates that are usually being charged before a crisis occurs already remunerated the existing degree of risk. 

The first approach to facilitate PSI supposes the introduction of “collective action clauses (CAC)” in all contracts and bonds. These clauses include representation and litigation rules with the purpose of overcoming the coordination and “free rider” problems that normally beleaguer negotiations between creditors and debtors. CACs will have to address several aspects of any negotiation. Imposing for instance a 75 percent vote on restructuring terms would bind all bondholders; requiring that minimum percentage of the bondholders would be necessary before legal action can be initiated; requiring the proceeds from any possible litigation to be shared with the entire group of creditors, renouncing the right to use the proceeds of any new financing obtained during the negotiations for repayments. 

This contractual approach would certainly ease the process of voluntary negotiations triggered by the debtor country. It nevertheless faces many problems:

• a) It requires great coordination to achieve the degree of generalized implementation that is required. In the end, a possible change in the IMF Articles of Agreement may be necessary to impose the inclusion of a uniform set of collective clauses. (7)
• b) The clauses apply to individual loans but the objective must be to involve the government debt as a whole. The question of aggregating different types of debt and groups of creditors seems to be difficult in such an approach.
• c) To apply the method to older debt implies a transition during which a voluntary exchange for new debt with CACs will have to be organized. Incentives with a certain financial cost will have to be introduced and the whole process will take time.

Collective action clauses will be a very useful instrument but the problems just referred to mean that in some cases they may not be enough to ensure a timely and sufficient debt restructuring. That is why I think that we need as a complement the introduction of the more ambitious mandatory scheme proposed by the IMF First Deputy Managing Director Anne Krueger (2001 and 2002). This proposal implies that the IMF could trigger a standstill of debt obligation during a short period (e.g. 90 days) and this could only be prolonged if the creditors agree. This would provide a shelter for the restructuring negotiations to take place, involving all the debt. The IMF or a new international body would act as a sort of mediator to adjudicate disputes among creditors or between the debtor and the creditors. The IMF would apply conditionality to the country and would lend into arrears as it is already entitled to do, and would help to finance essential imports to support economic activity. New private finance would be granted a privileged status over older debt. At the end of the whole process, a majority of creditors would have to approve the terms of the restructuring and this would then be binding for all creditors. 

The mandatory sheltering of debtors, though temporary, clearly facilitates restructuring and is designed to ease the pressure on the IMF to organize packages with sizable lending. There is misplaced resistance to this reforms. Banks clearly prefer a contractual approach and fear the limitation introduced to litigation rights. Debtor countries worry about possible (but doubtful) increased cost of the debt and the loss of control over the declaration of unsustainability. 

All those concerned, however should reflect that improved crisis management is in everyone’s interest. If indeed PSI is essential for a better functioning of the international financial system then generalized CACs and the SDRM (Statutory Debt Restructuring Mechanism) are necessary instruments. These reforms plus a clearer definition and a rigorous implementation of the rules about access to IMF facilities are the more important changes that need to be introduced. The promise of capital flows that avoid overshooting and excesses in both directions is important for global financial stability. There is of course the risk that financing flows may stabilize at levels below what would be required for a more efficient world economy where real convergence was better promoted. This is the risk that stems from focusing mostly on moral hazard and excessive borrowing as the source of financial crises. But as no other feasible solutions are available, we should concentrate on the hope that more stable global financial markets may in the end ensure the proper level of capital flows necessary for a more efficient international open system. 

(1) See Michael Bordo et al. (2001) 
(2) See Mishkin, Frederic (1999) 
(3) See Grossman, S. and Stiglitz, J. (1980) 
(4) Hoeller, Peter, Claude Giorno and Christine de la Maisonneuve (2002) 
(5) See Rodrik, Dani and A. Velasco (1999). 
(6) Diamond, Douglas and Philip Dybvig (1983)
(7) See Eichengreen, Barry (2002) References

Bernanke, Ben and Mark Gertler (2000) “Monetary policy and asset price volatility” NBER Working Paper nr. 7559.
Blinder, Alan (1999) “General discussion: monetary policy and asset price volatility”, Federal Reserve Bank of Kansas City Economic Review, nr. 4.
Bordo, Michael et al. (2001) “ Financial crises: lessons from the last 120 years” Economic Policy, Issue 32, April.
Cecchetti, S., Hans Genberg and S. Wadhwani (2002) “Asset prices in a flexible inflation targeting framework” NBER Working Paper nr. 8970.
Crockett, A. (2000) “In search of anchors for financial and monetary stability” SUERF Colloquim, Vienna (see BIS site).
Diamond, Douglas and Philip Dybvig (1983) “Bank runs, deposit insurance and liquidity” Journal of Political Economy vol 91.
Eichengreen, Barry (2002) “Crisis resolution: why we need a Krueger-like process to obtain a Taylor-like result” mimeo, April 29.
Eichengreen, B. and R. Haussman (1997) “Exchange rates and financial fragility” in “New Challenges for Monetary Policy” Federal Reserve of Kansas City, Jackson Hole Symposium.
Fisher, Stanley (1999) “On the need for an international lender of last resort” Journal of Economic Perspectives, Vol. 13, nr. 4.
Grossman, S and Stiglitz, J (1980) “On the impossibility of informationally efficient markets” American Economic Review, vol 70.
Hoeller, Peter, Claude Giorno and Christine de la Maisonneuve (2002) “Overheating in small euro area economies: should fiscal policy react?” OECD, Economics Department Working Paper nr. 323.
Krueger, Anne (2001) “International Financial Architecture for 2002: A New Approach to Sovereign Debt Restructuring”, IMF (www.imf.org).
Krueger, Anne (2002) “New Approaches to Sovereign Debt Restructuring: An Update on Our Thinking”, IMF (www.imf.org).
Lamfalussy, Alexandre (2000) “Financial crises in emerging markets”, Yale University Press
Mishkin, Frederic (1999) “Global financial instability: framework, events, issues” Journal of Economic Perspectives, Vol 13, nr. 4, Fall.
Rodrik, Dani and A. Velasco (1999).”Short-term capital flows” NBER Working Paper no 7364
Soros, Georges (1998) “The crisis of global capitalism”, Little Brown.