Chapter 4 Origins of the Crisis in Asia
- International Monetary Fund
- Published Date:
- January 2001
It has proven difficult to find a consistent pattern of fundamentals leading up to recent crises in emerging markets. The simple story of a government living beyond its means until the private sector cuts off its access to credit does not seem to fit the Asian tigers. Moreover, the apparent spread of trouble from one emerging market to the next seems unrelated to common problems arising within the individual economies. The inability to understand the disease or its contagion has led many economists and policy-makers to guess that some inherent weakness in private international capital markets accounts for these events. This interpretation has always found a sympathetic audience among policy makers that naturally prefer to believe that the markets are behaving badly rather than the governments they are sometimes called upon to defend. These sentiments have been given very substantial intellectual support by the development of multiple equilibrium models of crises. In these analytic frameworks costly bouts of illiquidity and crisis can be driven by shifts in investors’ expectations that are unrelated to economic fundamentals. The paper circulated with my remarks for this conference (Dooley 1997) swims against this stream in that it argues that it is not necessary to abandon a traditional first-generation model of speculative attacks in order to understand recent crises in emerging markets. In my remarks I want to focus on what is at stake in determining which approach better fits the facts.
2. WHY DO WE CARE?
To prevent future crises, we need to understand how and why crises have occurred. If the origins of crises are shifts in expectations, perhaps generated by “contagion,” straightforward adjustments in governments’ balance sheets and the composition of capital flows might provide a relatively costless and effective defense against future crises. In contrast if the origins of recent crises lie in rational reactions of private investors to incentives created by governments, the origins of crises, and the factors that make crises costly, will be much more difficult to eliminate from the international monetary system.
The multiple equilibrium framework suggests that countries should reduce their vulnerability to shifts in expectations but does not explain why countries have chosen to be vulnerable in the first place. The problem generated by short-term debt, or more precisely debt that is frequently reindexed to market interest rates, is now familiar. Expectations about devaluation in the case of domestic currency denominated debt, or default in the case of foreign currency denominated debt, generates risk premiums on any debt that has to be renegotiated. The rise in spreads in turn increases the real debt service costs of the governments’ debt. Even in countries with modest government debt, spreads paid by domestic banks and nonfinancial firms will also rise calling into question the solvency of the banking system. Since the government is expected to step in to support asset values, the line between private domestic debt and the government’s debt is blurred. In the end the devaluation or default feared by the private sector becomes the optimal choice for the government confronted by this hostile market environment.
The message from these stories is that governments must insulate their own balance sheets from arbitrary shifts in expectations but this is not nearly sufficient to avoid trouble. Governments must also monitor risks taken on by their banking systems either directly or through nonfinancial firms that can bring the banks down with them. It is very hard to argue with the idea that governments should manage their risks carefully or that vulnerable banking markets are a clear threat to governments’ finances. But what I find missing from these stories is a sense of why private investors get into these markets when they know that their own behavior can bring the countries and themselves to ruin.
Moreover, why was there such a general inflow into emerging markets after 1989? The hypothesis that after 1989 investors were too exuberant and then became too sober is dangerously close to a description of the facts that cannot be refuted. Surely there are naive investors but why are there more naive investors at certain times in certain places? For reasons developed below it is important to go deeper into the reasons behind capital inflows and crises.
The alternative “insurance” interpretation is that a change in the comfort level offered to investors by governments after 1989 generated the sequence of capital inflows and crises we have observed. The important implication of this model is that private investors anticipated the crisis and positioned themselves to earn high yields before the crisis and avoid losses at the crisis. For example, this framework suggests that the shift in the composition of capital inflows toward short-term credit to banks in the year or so before recent crises was an endogenous reaction of investors that saw crises coming. Simply put these are the positions the government is most likely to liquidate without loss during the crisis.
This, in turn, suggests that policies that merely alter the composition of capital inflows will not delay or prevent a crisis. Investors might prefer short-term claims on banks but if this form of capital flow is cut off they will find another to take its place. Indeed, if there are no short-term claims to be liquidated at the time of crisis, governments’ resources can be used to liquidate other claims on the debtor country. The insurance model suggests that even direct investment will be undertaken with the intention of reversing the flow when the crisis comes. Policies that discourage all capital inflows might be effective but would presumably also generate important administrative and efficiency costs for the developing country. Moreover, there is substantial evidence that while control programs have been successful in altering the composition of capital inflows, their ability to limit aggregate private inflows is limited and temporary.
The fact that some investors lost money in recent crises is not sufficient to show that the insurance model is not an important part of the story. We argue below that governments did offer sufficient insurance to allow all the new investors in emerging markets after 1989 to exit without loss during the crises that followed. The fact that some particularly slow investors were left behind does not mean that the important private decisionmakers were also naive about the risks of investing in emerging markets.
3. POLICY CHOICES FOLLOWING CRISES
It is important to think long and hard about preventing future crises but it is even more important today to deal with the residual effects of crises that have already occurred. It is not surprising that alternative interpretations of the origins of crises also have important implications for policies designed to minimize the losses in output that have followed. The multiple equilibrium interpretation suggests that changes in market psychology force vulnerable countries to liquidate assets at fire sale prices. This creates a real economic loss but one that is avoidable. Quick action by a lender of last resort can, in principle, stop the unnecessary and costly liquidation of investments and ensure a quick economic recovery.
The insurance view suggests that avoiding output loss will be much more difficult. There are two reasons for this. First, the crisis is triggered by an accumulation of a real unrecoverable loss in insured domestic banks and corporations in the debtor country. If this loss is not quickly recognized and credibly allocated among creditors, new investment in the debtor country will be discouraged. I suspect that high lending rates in post-crisis economies are due more to lingering contamination of new credits by default risk than to excessively tight monetary policies. Moreover, as long as this overhang is unallocated, a substantial share of new domestic savings will also be lost through capital flight.
The second reason why it will be difficult to avoid losses in output is that creditors understand that their de facto seniority when the crisis comes depends on the cost to the government in not paying them relative to other creditors. It is a fact of life that moral hazard arises from the government’s incentives to bail out creditors not from the bailout itself. Credits are designed so that negotiations for workouts favor creditors. In particular, creditors try to limit the debtor’s ability to withhold payment and then use these resources to retire debt on favorable terms.
In practical terms this means that if an expected bailout is not forthcoming creditors will find it very difficult to renegotiate credits in a way that does not impose considerable economic hardship on the debtor country. My interpretation of the 1982 international debt crisis and the seven years of negotiations that followed is that private creditors had little incentive to restructure debt until the official sector offered substantial subsidies to the agreements (Dooley, 1995).
In contrast, if crises are the result of unexpected shifts in expectations, propagation of declines in output can be truncated by planning ahead so that debts are easily and quickly restructured. New contractual arrangements for cross border lending, for example, changes in risk sharing, voting rules for restructuring and other reforms might be helpful. If the threat of losses in output is an integral part of the reason investors are attracted to the country in the first place, there will be strong resistance to changes in contracts that blunt debtors’ incentives to pay.
Finally, there are important differences in interpretations of moral hazard in the two approaches. One view is that international rescue packages have been exogenous shocks to the system that have generated moral hazard. The insurance model suggests that governments bail out investors because the alternative is a debt overhang that generates large costs in terms of lost output. The source of moral hazard in this case is not the bailout per se but the cost of not bailing out. Indeed we argue below that recent capital inflows are largely motivated by government insurance and that insurance determines the structure and consequences of capital inflows and crises. But refusing to honor implicit insurance merely postpones the eventual Brady Plan type restructuring that, given the debt overhang, is the rational policy response.