Information about Asia and the Pacific Asia y el Pacífico
Asian Financial crises

Chapter 43 An Analysis of Financial Crisis: Lessons for the International Financial System

International Monetary Fund
Published Date:
January 2001
  • ShareShare
Information about Asia and the Pacific Asia y el Pacífico
Show Summary Details


A financial crisis is a disturbance to financial markets that disrupts the market’s capacity to allocate capital—financial intermediation and hence investment come to a halt.

Investors should take risks, and risks mean that some proportion of investments will fail. This is why we have domestic bankruptcy laws that provide for orderly workouts when investments and the firms that made them do fail. Similarly, if the international capital markets are functioning well, mistakes will be made. Then one or another country (or its private sector borrowers) will fail: a crisis. History records many. We should not expect or even wish to prevent them all. That would be at the cost of insufficient, excessively risk-averse investment. So such crises will always recur—but capital markets forget

The term “financial crisis” is used too loosely, often to denote either a banking crisis, or a debt crisis, or a foreign exchange market crisis. It is perhaps preferable to invoke it only for the “big one”: a generalized, international financial crisis. This is a nexus of foreign exchange market disturbances, debt defaults (sovereign or private), and banking system failures: a triple crisis, in which the interactions are the key to causality, depth, and persistence (Eichengreen and Portes, 1987).

The widespread securitization of debt in recent years has not changed the picture—after all, one of the major historical examples is the 1930s crisis of defaults on sovereign bonds. Nor has it diminished the importance of banking sector fragility in provoking and exacerbating financial crises.

All crises are “crises of success” (Portes and Vines, 1997). The initial capital inflow that ultimately proves unsustainable (and perhaps unprofitable) is both a sign and—for a time—a cause of economic promise and success. But we have not yet learned how best to cope with the capital inflows, so success may lead to failure.

All crises raise the problem of distinguishing between insolvency and illiquidity. Today, some say that the “Asian miracle” economies are actually “hollowed out,” “zombie” economies; all that investment just went into creating excess capacity or driving up real estate prices. Others argue that the growth was real, that these economies should still come top in the World Competitiveness Report, and their problem is simply overvalued exchange rates and the classical “run”—a self-fulfilling crisis of liquidation of short-term loans. We will not know for some time which view is correct.

Because crises will always recur and because their causation is so complex, we shall never arrive at an international financial system—whatever its “architecture”—that can dispense with mechanisms for the resolution of financial crises. “Better information,” “early warning” and “preventive measures” will not remove the need for orderly workouts, whatever the source(s) of the crisis


The Asian macroeconomic fundamentals are good. And the Asian crises are primarily private-sector ones, in economies with high aggregate savings, sound fiscal positions, open and outward-looking policies, and relatively low sovereign debt. Rather than high inflation, the serious feature of the crises is debt deflation, or the bursting of a bubble. So this is very different from Mexico. In Russia, on the other hand, the key feature is fiscal imbalance—the inability of the tax system to provide an essential minimum level of government revenue.

The economies in crisis do have weak banking systems, and most started with somewhat overvalued exchange rates (some but not all pegged). Those are common features of financial crises. But the causes and current appearance of these crises are different from previous episodes. Common factors are de facto exchange rate pegs to the dollar, which resulted in overvaluation and large current account deficits, and excessive private sector foreign borrowing.

But the idiosyncratic factors are at least as important: Thailand’s central bank took forward positions that depleted net foreign reserves, and its finance companies accumulated massive nonperforming loans; Indonesian industrial structures favor a particular political force but are inefficient; Korean banks borrowed too much abroad on short maturities; and the Korean industrial sector is exceptionally highly leveraged.

Thus financial crises differ across a wide range of features:

  • The role of “fundamentals”
  • The relative importance of bank and securitized debt
  • The relative importance of private and sovereign debt
  • Exchange rate regimes and history
  • Underlying structure and dynamics


A taxonomy of financial crises is helpful (Radelet and Sachs, 1998). We can distinguish among the following features, which are perhaps best regarded as the proximate or initial causes of financial crises:

  • A speculative attack on the exchange rate—here we must distinguish between attacks based on deteriorating “fundamentals” and those that exhibit self-fulfilling expectations;
  • A “financial panic”—a bank run or its international analogue—in this case, no degree of “transparency” or better information will eliminate herd behavior, which is based on a collective action problem;
  • The collapse of an asset price bubble;
  • A crisis induced by moral hazard (implicit or explicit guarantees of bailout); and
  • The recognition of a “debt overhang,” followed by a disorderly workout.

Again, the key obstacle to foreseeing, preventing and indeed to resolving international financial crises is that they typically exhibit several of these features. There is no credible unicausal story, and in each crisis the nature of the interaction among the factors listed here is itself unique. In particular, we cannot simply regard the Asian financial crisis as a “financial panic” without recognizing the important roles of excessive nominal exchange rate rigidity, real estate price bubbles, and moral hazard that led to reckless behavior by lenders. Nor was the evident contagion that followed the initial crisis in Thailand just financial panic. There were common features in the economies that were hit, and there were significant links among them.


Contemporary policymakers are finally coming to understand what researchers (and the architects of Bretton Woods) recognized long ago: Fixed but adjustable pegs are simply unworkable with full capital mobility. Nevertheless, countries do introduce exchange-rate pegs for macroeconomic stabilization and subsequently cannot or will not “flexibilize” them soon enough to forestall speculative attacks. What policies are advisable when the attack materializes? Typically, financial fragility will make the interest rate defence ultimately infeasible, and intervention whether in spot or forward markets just feeds the speculators.

Thus if attacked, it is almost always best to save the reserves and let the exchange rate go. With appropriate monetary policies and adequate remaining reserves, it may be possible to bring the rate back to a reasonable and stable level fairly soon (e.g. France 1993, Czech Republic 1997). This is of course essential: a violent, uncontrolled depreciation will make it impossible to distinguish among “good” and “bad” domestic assets, in view of the high degree of foreign currency exposure that tends to arise in economies with pegged rates. It will only be possible to tell which assets are “good” when the exchange rate returns closer to “equilibrium”—and that is likely to take much longer if the defense of an indefensible peg has gone on too long.

The fashion for currency boards has already gone too far. A currency board is likely ultimately to fail—with potentially disastrous consequences—if the banking system or political discipline are weak and foreign exchange reserves are low.


The International Monetary Fund cannot be a true international lender of last resort. It cannot fulfill the essential ILLR role, because it cannot create money. Moreover, it has no recourse to a fiscal authority that can provide the resources to fill the gaps in balance sheets, when “liquidity crises” turn out to be “solvency crisesl.”

The sequence of events since summer 1997 demonstrates that in a world of full capital mobility, no “bailout package” can be sufficient if the markets are unconvinced—and even very large packages may not convince them. Moreover, the focus on foreign indebtedness, whether to banks or in securitized form, ignores the major role of domestic capital flight in most of the crises that we have observed. No feasible package will suffice to finance conversion of the entire domestic money supply (not just the monetary base) into foreign exchange.

Although the fund is conscious of contagion effects and systemic risk, its constitution and procedures require that it act country-by-country. In a systemic crisis, however, the domestic lender of last resort simply creates liquidity—”defends the money supply.” That function is not possible under the current international financial system. And the IMF cannot follow the Bagehot rules for additional important reasons: it cannot take collateral that would be immune to sovereign repudiation; and it may be impossible to value financial asset collateral in a case of global systemic crisis.


How to pick up the pieces once a crisis occurs will vary greatly from case to case. But there are some general principles (Eichengreen and Portes, 1995), some of which require action before any crisis hits.

Typically, an orderly workout of international debt will require both rescheduling and restructuring: partly the lengthening of maturities, partly conversion of debt into equity. It may also require writing off substantial amounts of debt—this has been well understood since the nineteenth century, however much creditors may resist it. The workout should not be a bailout of the creditors: implementation should put a high priority on minimizing moral hazard (the consequences of the Mexican bailout are still underestimated) and avoiding nationalization of private debt (an unfortunate feature of several of the Asian cases).

We have made several recommendations for contractual and institutional innovations to promote cooperation among holders of securitized debt (Eichengreen and Portes, 1995), and it is encouraging to see that these have been taken up seriously in the report of Working Group Three of the G-22.1 would stress the importance of collective action clauses and other ex ante commitments in debt contracts (e.g. to give seniority to new loans). I am confident that these would be worthwhile even if, as market participants claim, they would entail larger spreads (less favorable lending terms) and indeed lower levels of capital flows—although I believe this claim is much exaggerated. Still, if it were correct, that would suggest that the terms have been improperly and unsustainably favorable to borrowers.

Any new architecture should make explicit provision for temporary payments standstills (requiring exchange controls or other restrictions on capital outflows) and IMF lending into arrears on either bank lending or securitized debt. It is important, too, that there be a better codified institutional framework for workouts. The complicated interaction among the fund, the London Club and the Paris Club is time-consuming and in practice hinders the restructuring and writeoffs that are ultimately necessary.


There is a long list of obstacles to the construction of an agreed framework for orderly workouts. Most of them should be ignored, some of them can be avoided, and some must simply be regarded as inescapable costs of what will on balance be a much better system for dealing with financial crises.

  • i. It will always be desirable to have a “voluntary,” “consensual” suspension of payments and debt restructuring, but what if the creditors simply will not accept that? Then there must be some “mandatory” element. Here the fund or another agency may play a key role in achieving better behavior on the part of both debtors and creditors.
  • ii. Borrower countries may resist provisions for orderly workouts (e.g. qualified majority clauses in debt contracts), because they believe that having such provisions in place ex ante will worsen their market terms. An initiative by the G7 to include such clauses in their own sovereign debt contracts would set a good example.
  • iii. Lender institutions will doubtless resist any such provisions—market participants blocked any implementation of the recommendations of the G10 Deputies (Rey) Report of 1996. They should be ignored.
  • iv. Defining the scope of any suspension of payments may differ from case to case, and the extent and nature of the exchange controls or other restrictions on capital mobility will have to be tailored to each case too.
  • v. Any selective suspension raises problems, e.g.: would there be a run on debts not covered? What about cross-default clauses?
  • vi. It may be difficult to enforce the capital controls required for a standstill (but their duration should not be too long, if the management of the workout is expeditious).
  • vii. Offering seniority to new credits may encounter obstacles in existing pari passu clauses in loan contracts.
  • viii. It is claimed that no workout could be fully “orderly,” because trade credit would freeze up—but that has happened anyway in the Asian crisis countries.
  • ix. Market participants claim that any orderly workout provisions would pose a threat to the international interbank market. But the evidence for this threat is unclear.
  • x. Critics of any workout involving debt reduction claim that this would indefinitely postpone the debtor’s return to capital market access. But when Mexico went to the market in mid-1995, after its “successful” bailout that kept creditors whole, it got no better terms than did Poland at exactly the same time, although Poland had only six months previously agreed to a fifty percent writeoff with the London and Paris Clubs. And Poland did better than Hungary, too, although the latter had been punctually faithful on its exceptionally heavy debt service.


Finally, it is important to recognise that the IMF already fulfills an excessive set of roles in the international financial system—”excessive” because they involve potential conflicts of interest and counterproductive overlapping. That problem would be exacerbated by giving the fund additional responsibilities in debt workouts. Here we may just list the various functions—it is evident how they may conflict in practice:

  • Surveillance and analysis: macro and micro,
  • Policy recommendations,
  • Monitoring compliance with financial regulation and standards,
  • SDDS,
  • Providing information to the markets,
  • Possibly administering contingent credit facilities,
  • Other decisions on lending and conditionality,
  • Status as priority creditor,
  • Adjudicating on standstills,
  • Intermediating between creditor and debtor in debt restructuring, and
  • Certifying country adjustment programs for the Paris Club.

It is unfortunate that responses to the international financial crisis of 1997–98 have brought proposals to expand the range of the fund’s responsibilities and activities. The existing set is already conflictual and dysfunctional.


    Eichengreen, B., and R.Portes. 1987, “The Anatomy of Financial Crises,” in Threats to International Financial Stability, R.Portes and A.Swoboda(eds.), Cambridge University Cambridge Press, pp. 10–58.

    • Search Google Scholar
    • Export Citation

    Eichengreen, B., and R.Portes. 1995, Crisis? What Crisis? Orderly Workouts for Sovereign Debtors, London: CEPR.

    Portes, R., and D.Vines. 1997, “Coping with Capital Inflows,” Commonwealth Secretariat, London, economic paper No. 30.

    Radelet, S., and J.Sachs. 1998, “The East Asian financial crisis: diagnosis, remedies, prospects,” Brookings Papers on Economic Activity, 1, pp. 1–74.

    • Search Google Scholar
    • Export Citation

    Other Resources Citing This Publication