A quarterly magazine of the IMF

September 1999, Volume 36, Number 3

Now that the Asian crisis is behind us, what lessons can we draw from the experience, and how can we use this knowledge to forestall future crises and minimize damage from those that occur?

The Asian financial crisis, which spread from Thailand to other countries in the region during the second half of 1997, plunged the countries affected into deep recessions that brought rising unemployment, poverty, and social dislocation. The outbreak, spread, and persistence of the crisis also challenged some basic assumptions: the countries most strongly affected were "tiger economies" that had few of the weaknesses usually associated with countries that turn to the IMF for help. They had fiscal surpluses, high private saving rates, and low inflation; and in most cases their exchange rates did not seem out of line.

The crisis now appears to be over: market conditions have stabilized and strong recoveries are under way, although the countries still face a long agenda of needed structural reforms. But the events of the past two years have raised many important questions. Why did crises occur in these countries? Why did the IMF-supported policy programs introduced in Indonesia, Korea, and Thailand initially fail to quickly stop the market panic and prevent a sharp recession?

Financial fragility

Many factors may have contributed to the onset and spread of the Asian crisis, but there is a growing consensus that the main ingredient was financial fragility. This involved four related aspects. First, many financial institutions and corporations in the countries affected had borrowed in foreign currencies without adequate hedging, making them vulnerable to currency depreciation. Second, much of the debt was short-term while assets were longer-term, creating the possibility of a liquidity attack, the effect of which would be similar to that of a bank run. Third, prices in these countries' equity and real estate markets had risen substantially before the crisis, increasing the likelihood of a sharp deflation in asset prices. Fourth, credit was often poorly allocated, contributing to increasingly visible problems at banks and other financial institutions before the crisis hit.

How did these countries' financial systems become so fragile? In part, this reflected ineffective financial supervision and regulation in the context of countries' financial sector liberalizations. Capital account liberalization was poorly sequenced, encouraging short-term borrowing, while limited exchange rate flexibility led borrowers to underestimate exchange risk. Monetary policies allowed domestic credit to expand at a breakneck pace. But if banks and corporations in these countries borrowed imprudently, foreign lenders also lent imprudently, possibly reflecting sloppy risk management, perceptions of implicit government guarantees, and the incomplete information available.

Given these vulnerabilities, once a crisis started, it became difficult to stop. As foreign and domestic investors rushed for the exits, a vicious circle was created: currencies depreciated, plunging more institutions into insolvency, further undermining creditors' prospects of repayment, and accelerating the exit of capital. Many unfavorable events aggravated these adverse dynamics after the IMF-supported programs were introduced: political events and initial hesitation in implementing agreed policies cast doubt on the authorities' commitment to reform programs; disturbing information (especially on weak international reserves) that had previously been withheld was revealed at the height of the crisis; and, in Indonesia, the much-needed closings of insolvent banks were accompanied by only limited and ill-publicized deposit guarantees; in addition, there were doubts about how much of the announced financing packages were actually available.

Crisis management

Given the nature of the financial crisis, the policy response, in the context of the IMF-supported programs, had three main elements: large official financing packages, together with some action to keep private money in place; an unprecedented body of structural reforms; and macroeconomic policies intended to counter the crisis itself. The policy content of the programs is addressed in greater depth elsewhere (see, for example, Lane and others, 1999), but some of the key points are discussed in this article.

Financing. The financing packages assembled to support the Asian crisis countries were impressive: $36 billion for Indonesia, $58 billion for Korea, and $17 billion for Thailand. But they were not so large in relation to potential private capital flows. Moreover, not all of the money was available—especially at the outset—to counter market pressures. First, the IMF support was phased—made available in tranches— over the life of the programs; this is standard procedure, intended to ensure that the authorities have a continuing incentive to adhere to the adjustment policies agreed under the program, but it did reduce the authorities' ability to counter immediate market pressures. Moreover, some of the bilateral money—the "second lines of defense"—promised to Indonesia and Korea was never actually disbursed.

In addition, the programs could unfold as planned only if a substantial amount of private financing remained in place. The initial emphasis was on impressing the markets sufficiently to ensure that this would happen voluntarily. Only Thailand tried to secure assurances from creditors at the outset. In Korea, which experienced massive private capital outflows in the first three weeks of its reform program, default was averted only by an eleventh-hour deal with key bank creditors to roll over credit lines, followed by an agreement to reschedule short-term debt. In Indonesia, a de facto payments standstill was imposed, and talks with creditors began in February 1998; restructuring there was complicated by the fact that almost half of total external debt was owed directly by private corporations rather than banks.

An obvious question is whether earlier and more aggressive steps should have been taken to "bail in" private creditors—that is, to require them to maintain their net lending to a debtor country—through debt restructuring and controls on capital outflows. This idea has apparent attractions: it offers a way to break the vicious circle of capital outflows, depreciation, and insolvency; and it could send a signal to private creditors that they would not necessarily be made whole by official financing packages. But bailing in the private sector in the midst of a crisis is by no means straightforward. A heavy-handed approach would almost certainly have exacerbated the contagion that did occur, and the net effect could well have been less, rather than more, private financing for the countries affected. The main emphasis in subsequent policy discussions has thus been on setting up arrangements beforehand to maintain private exposure rather than on taking aggressive action when a crisis occurs.

Structural reforms. Right from the start, structural reforms were a key feature of the countries' programs to address the root causes of the crisis and its consequences, as well as to set the stage for medium-term growth. Because many of the needed structural reforms fell outside the IMF's area of responsibility, details of the structural reform programs were, in many instances, worked out in cooperation with the World Bank and the Asian Development Bank.

The core of the structural reforms was in the financial and corporate sectors, where the strategy had two main strands. The first was to clear up the fallout from the crisis—insolvent institutions needed intervention, with unviable ones being closed and potentially viable ones strengthened—and actions had to be taken to limit the risks of bank runs and uncontrolled liquidity expansion. The second was to put the system on a sound footing—most important, by improving financial supervision and regulation—to minimize the likelihood that problems would recur. Each strand was essential for the success of the other: it would have made sense neither to rescue weak institutions so they could do business as usual in a poorly regulated system nor to set up a state-of-the-art regulatory system for institutions mired in (or hovering close to) insolvency.

Structural reforms covered many other areas. To enhance governance and competition, steps were taken to reform state-sponsored monopolies and cartels, strengthen competition laws, and increase the transparency of economic and financial data. International trade reforms were aimed mainly at continuing existing liberalization plans to prevent a lapse into beggar-my-neighbor restrictions.

Social sector reforms—intended to strengthen and broaden existing social safety nets—played a prominent role, in light of concerns over the effects of the crisis on the poor and vulnerable. These reforms included measures to raise income transfers; limit unemployment through various employment and training schemes; limit the impact of price increases on poor households by introducing or continuing subsidies for food, energy, and transportation; and maintain access by the poor to health care and education.

Given the extensive structural agenda, an important question is whether the programs may have suffered from an overload of structural reforms. This criticism cannot be dismissed lightly: while it may be broadly accepted that action was needed in the financial and corporate sectors, the immediate need for many other reforms included in the programs was less clear. Important questions remain regarding the appropriate pace and sequencing of reforms and the emphasis on different areas of reform—as, over time, the programs became more sharply focused on financial and corporate reform, as well as social issues.

Macroeconomic policies. When the initial IMF-supported programs were being formulated, massive market pressures had already forced the authorities in all three crisis countries to float their currencies. The programs took this as a fact and designed other policies around it. The main reasons for not trying to reestablish currency pegs were that the authorities had neither the reserves nor the commitment to use monetary policy unstintingly to defend them.

Given the decision to float and the turbulent market conditions at the time, monetary policy faced a difficult trade-off between the desire to resist market pressure and avert a spiral of depreciation and inflation, and the desire to limit the adverse effects of monetary tightening on the real economy. The results were quite different in the three countries. In Korea and Thailand, the pattern was similar to that of a classic case of successful stabilization: nominal interest rates were raised significantly, albeit after some initial hesitations, and real rates reached peaks of more than 20 percent in Korea and about 15 percent in Thailand before declining gradually. From mid-1998 on, nominal and real interest rates in both countries were at or below pre-crisis levels, as market pressures abated and currencies strengthened. In Indonesia, the situation was quite different: monetary policy veered far off course, driven by the massive provision of liquidity to a collapsing banking system in a setting of intense political and social turmoil. At the outset, monetary expansion accommodated the precipitous depreciation of the rupiah as inflation surged. Nominal interest rates were driven to high levels, but the increase corresponded to depreciation and increasing risk premiums; real interest rates remained negative from the beginning of the IMF-supported program through the summer of 1998. Real broad money and credit generally continued expanding, albeit with sharp month-to-month fluctuations associated with bursts of inflation. Thus, Indonesia presents a case in which basic monetary control was lost and market turmoil continued, leading into a vicious circle. Only in the latter months of 1998 did market conditions stabilize.

The experiences of these countries suggest that monetary tightening worked in stabilizing currencies—indeed, earlier, more resolute tightening might have been less costly. The bout of high interest rates was comparatively short-lived and did not result in an implosion of money and credit; although there was a negative impact on banks' portfolios, accommodating an ever-larger depreciation could have put banks in an even worse position. At the same time, questions remain about the appropriate degree of tightening and the effectiveness of tight monetary policies in a setting of widespread insolvency. There were also widespread concerns about a credit crunch that adversely affected many potential borrowers, particularly smaller firms. Such concerns point to the urgency of making progress in financial sector restructuring to clear away impediments to efficient credit allocation.

Before the crisis, fiscal policies had been relatively strong in these countries, which had run surpluses for some years. Only in Thailand did fiscal expansion contribute to the onset of the crisis. When fiscal policies were initially formulated as part of the programs, it was expected that growth would continue, albeit at a slower rate, and that the current account needed to be strengthened to cope with a reversal of external capital flows. Some fiscal adjustment was intended to alleviate the burden of adjustment on the private sector and to make room for the anticipated carrying costs of financial restructuring as well as increased social spending. The planned fiscal adjustment was stronger in Thailand, where the deficit had increased in the previous year, whereas in Indonesia and Korea, even the initial programs contained little overall adjustment, confining themselves mainly to finding room elsewhere in the budget to cover the prospective costs of recapitalizing banks. In hindsight, given the sharp decline in private sector demand that was under way, fiscal policy should have been more expansionary, and there was a major change in course as the situation became clear. The deteriorating economic environment led directly to substantial increases in fiscal deficits, which, from the beginning of 1998 on, were accommodated by easing the programs' fiscal objectives.

Lessons

The crisis raises a number of important issues for the international financial system, many of which are related to the development of a new international financial architecture. The unfolding of the crisis underscored the inherent difficulty of stopping a crisis once it has started, given the speed with which short-term capital can move in response to changing market sentiment: prevention is the key.

But how can countries prevent the buildup of vulnerabilities of the kind that led to the Asian crisis? Clearly, part of the answer is the maintenance of sound macroeconomic policies. The exchange rate regime is a particularly controversial aspect, because many observers have focused on the role of limited exchange rate flexibility in fostering capital inflows prior to the crisis. Some have concluded that the only viable options are full exchange rate flexibility or the opposite extreme of institutionalized fixity—a currency board or full dollarization—while others have expressed reservations over this "law of the excluded middle." This remains an active area of policy discussion and research.

Another key element is improved financial supervision and regulation in both debtor and creditor countries. The crisis also exposed possible flaws in risk management by international financial intermediaries, including inadequate stress-testing of exposures with regard to price movements that were possible but appeared unlikely.

Transparency is also important to crisis prevention. At the height of the Asian crisis, some unpleasant information was revealed—in particular, on the weaknesses of central banks' international reserve positions—that exacerbated market panic; it would have been much better if such information had instead been revealed earlier on, when it might have restrained the heady inflows of capital. In normal times, improvements in standards for data dissemination and steps to increase the transparency of policies could help markets to improve their pricing of risk, inhibiting the buildup of imbalances, and also spur policymakers to take timely action to address vulnerabilities. Greater transparency of the IMF itself is an integral part of this agenda.

Strengthened international surveillance with closer monitoring of the financial sector and a focus on international standards may also help alert policymakers to upcoming problems. Such surveillance should also incorporate a regional perspective to provide warnings of impending regional contagion of the kind that spread the Asian crisis.

Some more difficult issues concern ways of involving private creditors in forestalling and resolving financial crises. Here, the central problem is how to maintain private creditors' exposure—and impose losses on short-term creditors if appropriate—without unduly exacerbating contagion. Recent international discussions have concentrated on assessing the potential benefits and costs of preparatory steps that could be taken in normal times—for instance, modifying standard bond contracts to facilitate their restructuring in the event of a crisis and establishing contingent credit lines with private financial institutions. A related issue is the change in the IMF's policy on lending to countries with arrears to private sector creditors.

Another thorny issue is that of capital controls. Here, there are three key aspects: one is the sequencing of capital account liberalization, where the crisis has highlighted the pitfalls of liberalizing short-term flows while leaving restrictions on longer-term flows, as well as the need to keep the pace of capital account liberalization in line with the strengthening of the domestic financial system. A second relates to the possible merits of taxes to discourage short-term capital inflows, such as those implemented in Chile. A final issue is the effectiveness of controls over capital outflows in the event of a crisis: the central question for the international system is not whether controls could have alleviated a particular crisis but whether a regime in which controls tend to be imposed in the event of a crisis is characterized by more or fewer crises, of greater or lesser severity—given that the prospect of controls strengthens the incentive for capital to run at the first signs of trouble. The longer-term implications of the resulting limitations on access to international capital markets also need to be taken into account.

Another set of issues pertains to crisis management. With regard to monetary policy, the experience of the Asian crisis suggests that, by and large, tight policies worked: after a period of high interest rates, market pressures abated and interest rates fell below precrisis levels. In fact, an earlier firming of monetary policy might have been even more effective in containing the crisis. Nevertheless, important unsettled issues remain regarding the effectiveness of raising interest rates in a crisis characterized by widespread insolvencies. With regard to fiscal policy, the crisis highlights the need for countries to adapt their policy responses to changing macroeconomic conditions. With regard to structural policies, the crisis underscores the importance of decisive action but also raises some important—and, for the most part, still unsettled—questions about the optimal pace and sequencing of reforms.

Finally, the crisis has brought about a rethinking of the way official financing is provided to address a crisis, including the appropriate size and phasing of IMF support to countries facing market pressures. One important step in this regard is the IMF's recent introduction of the Contingent Credit Lines, which would provide large-scale financing to countries that might be affected by market contagion.

Timothy Lane is Chief of the Policy Review Division in the IMF's Policy Development and Review Department.