Comments: On Dysfunctional Adjustment and Financing
- Alexander Swoboda, and Peter Kenen
- Published Date:
- December 2000
As I was involved in devising the program for this conference, it is perhaps unfair for me to criticize the agenda for this session. I feel obliged to do so, however, because we may risk missing the fundamental, qualitative issue. We have been asked whether the balance between adjustment and financing has shifted unduly toward adjustment and, if so, how a better balance might be achieved. In my view, we should be concerned primarily with the character of the substantial adjustment that has occurred in the crisis-stricken countries, not the amount of adjustment, and with the manner in which some countries used the financing available to them.
There has, I believe, been ample financing even after allowance is made for the remarkable increase in the volume and volatility of international capital flows. But much of it was used improvidently to defend pegged or quasi-pegged exchange rates. There was at the same time massive adjustment, but it must be described as dysfunctional adjustment.
To explain these assertions, let me define financing and adjustment in ways that should not cause controversy but will be quite helpful to me: By financing, I mean the use of reserves plus the use of official funding supplied by the IMF, by other multilateral institutions, and by individual foreign governments. By adjustment, I mean the change in the sign or size of the current account balance. On these definitions, the Asian crises of 1997-98 clearly involved large-scale financing as well as substantial adjustment.1
Some of the crisis-stricken countries in Asia drew down their reserves very rapidly. At the start of 1997, Thailand’s foreign exchange reserves stood at $37.2 billion. By the end of July, they had fallen to $28.9 billion, and they fell by $4.5 billion in August, despite the first infusion of IMF credit. In addition, the central bank sold dollars forward in amounts that encumbered a very large fraction of its remaining reserves. At the start of 1997, Korea’s foreign exchange reserves stood at $33.2 billion, and they did not fall at all during the next six months. By the end of October, however, they had fallen to $30.5 billion, and they fell by $10.1 billion in the next two months, despite a large infusion of IMF credit. As in the case of Thailand, moreover, some of Korea’s remaining reserves were not freely available; they had been deposited with foreign branches of Korean banks to cover those banks’ foreign currency exposure. The amounts of foreign official financing obtained by Indonesia, Korea, and Thailand are shown in Table 1 below. They totaled $111.7 billion. It must be emphasized, however, that these large packages were not front loaded, and actual disbursements were fairly small in the first months following the dates on which the packages were assembled.
|Country and Source||Original Original Package||Disbursed as of March 1999|
|World Bank and Asian Development Bank||8.0||2.5|
|World Bank and Asian Development Bank||14.2||9.6|
|World Bank and Asian Development Bank||2.7||2.0|
Includes $1.0 billion under Miyazawa Initiative.
Before deducting $4.8 billion repaid by Korea.
Includes $1.0 billion under Miyazawa Initiative.
Before deducting $4.8 billion repaid by Korea.
Too often, however, and most egregiously in the case of Thailand, reserves were used to defend what Calvo and Reinhart call soft pegs. Like Mexico in 1994, Asian governments clung to the belief that good news, good luck, or dogged determination would lead to a quick revival of capital inflows. When that did not happen and they were compelled to abandon their pegged rates, their remaining reserves and the foreign official financing readily available to them were insufficient to help them keep their currencies from going into free fall. In short, financing was used improvidently to postpone adjustment rather than to achieve orderly adjustment.
Turning from financing to adjustment itself, a number of the crisis-stricken countries moved swiftly from large current account deficits to large current account surpluses, and some were even able to rebuild their reserves without a revival of capital inflows. In the first six months of 1998, Korea’s foreign exchange reserves rose by $21.0 billion, and foreign official financing accounted for only a small part of the increase. No one can claim that there was too little adjustment. But it was dysfunctional in that it reflected sharp reductions in output, income, and imports.
Ordinarily, large currency depreciations like those that occurred in the early phases of the Asian crisis might be expected to have strong expenditure-switching effects—effects that should improve the current account and raise the demand for domestic goods, stimulating output. These effects, however, are usually long lagged. In Asia, by contrast, the current account improvement took place rapidly, and the demand for domestic goods fell abruptly. The adjustment was thus due to expenditure-reducing effects rather than expenditure- switching effects.
Many economists attribute the sharp reductions in output and imports to the tightening of monetary policies mandated by IMF programs, and the policy tightening did play a role. But the currency depreciations had direct expenditure-reducing effects because they had devastating consequences for banks and corporations with large foreign currency debts. I have not yet seen any serious attempt to measure the relative sizes of the interest rate and balance sheet effects on output. I am, nevertheless, confident that careful research will ascribe a large share of the fall in output to the freezing up of credit flows that occurred when lenders and borrowers alike were driven into insolvency by the balance sheet effects of the currency depreciations.2
Calvo and Reinhart caution against blaming pegged exchange rates for all of the bad things that happened in Asia. They argue convincingly that some of those bad things would have occurred under flexible exchange rate arrangements because of the large foreign currency debts of banks and corporations. The size of those debts, however, can likewise be blamed on the exchange rate policies of the Asian countries. They were, in effect, the victims of their own success in persuading borrowers and lenders that their countries’ currencies would stay stable. There is no other way to explain the accumulation of unhedged foreign currency debt by Asian banks and corporations.
More flexible exchange rates by themselves might not have prevented excessive foreign currency borrowing, but exchange rate pegging made matters worse. In fact, it had three unfortunate consequences. In the run-up to the crisis, it encouraged the accumulation of foreign currency debt. With the onset of the crisis, it encouraged an excessive drawdown of reserves, which then made it impossible for the crisis-stricken countries to keep their currencies from going into free fall. In the wake of the crisis, the large depreciations of the Asian countries’ currencies combined with their large foreign currency debts to produce what I have described as dysfunctional adjustment, in which the expenditure-reducing effects of insolvency dominated the expenditure-switching effects that we normally associate with large exchange rate changes.
Years ago, economists debating the relative merits of fixed and floating exchange rates disagreed about their catalytic effects on politicians. Which was more likely to attract their attention—a large loss of reserves or a large depreciation? We do not know how Asian governments would have responded to currency depreciations if exchange rates had been more flexible, nor do we know how far their countries’ currencies would have fallen if they had not tried to defend their exchange rates and run down their reserves in the process. It is hard to believe, however, that the outcome would have been worse than the actual experience of 1997–98.
I agree with Calvo and Reinhart on two important points. First, the liquidation of domestic currency debt can exhaust a country’s reserves just as quickly as the liquidation of foreign currency debt. Second, exchange rate flexibility may not suffice to deter imprudent foreign currency borrowing. But I draw somewhat different conclusions from these observations.
On the first point, the exchange rate regime determines the degree to which both forms of debt represent potential claims on official reserves. While both forms of debt represent potential claims on reserves, the strength of the claims depends on the strength of a government’s commitment to draw down its reserves in defense of a pegged exchange rate. Furthermore, there is one clear difference between the two forms of debt. It was the large stock of foreign currency debt that produced the dysfunctional adjustment characteristic of the Asian crisis.
On the second point, it would be imprudent to rely entirely on exchange rate flexibility to deter foreign currency borrowing. Additional safeguards may be needed, especially in countries where banks are not adequately supervised and have not developed their own risk-management regimes. I agree strongly with U.S. Secretary Rubin’s views on this matter: “Where financial systems and supervisory regimes are underdeveloped, steps are probably needed to limit banks’ foreign currency exposures, especially their short term borrowing” (Rubin, 1999).
I would tie that recommendation to another safeguard. High-risk weights should be attached to interbank lending when the borrowing countries do not have adequate supervisory regimes and fail to impose appropriate limitations on their banks’ foreign currency debts. The capital adequacy rules proposed by the Basel Committee, which would end the preferential treatment of interbank lending, may not go far enough.
I began by criticizing the agenda for this session, and I conclude by doing it again. It is impossible to discuss the appropriate balance between adjustment and financing without examining the role that has been played—or should be played—by the private sector. In other words, it is artificial to separate the subject of this session from that of the next session.
This leads me, in turn, to pose a question I have asked repeatedly without obtaining a satisfactory answer. How can we explain the remarkable disjuncture between the rhetoric of the official community and its actual behavior during the Asian crisis?
The Rey Report of 1996, endorsed by the Group of Ten ministers and governors, led off with this warning:
… neither debtor countries nor their creditors should expect to be insulated from adverse financial consequences by the provision of large-scale official financing in the event of a crisis. Markets are equipped, or should be equipped, to assess the risks involved in lending to sovereign borrowers and to set the prices and other terms of the instruments accordingly. There should be no presumption that any type of debt will be exempt from payment suspensions or restructurings in the event of a future sovereign liquidity crisis.3
Much the same thing has been said repeatedly since, and recent formulations, such as the one in the G-22 report on crisis management (Group of Twenty-Two, 1998), have referred explicitly to private sector debt as well as sovereign debt. Incase after case, however, the official community has responded to crises by providing large-scale financing rather than encouraging or requiring the suspension and subsequent restructuring of debt-service payments. Efforts were made to achieve a rollover of Thai banks’ short-term debt to foreign banks, and more formal arrangements were made in the Korean case (see Lane and others, 1999). In both instances, however, these steps were taken only after a substantial rundown of the debt involved. In the Korean case, moreover, the restructuring of interbank debt was achieved with the aid of a guarantee by the Korean government—which was, in turn, backed implicitly by large-scale official financing.
We are told that any suspension of debt-service payments at the start of a crisis will intensify contagion. If one country interrupts its debt-service payments, creditors will flee other countries promptly. This warning, however, fails to distinguish between the first such case and all subsequent cases. The first case would signal a basic change in the rules of the game and could cause a pullback from other countries. That is what happened when Russia defaulted in August 1998, though that was not merely a change in the rules of the game, but was disruptive to the game itself. In subsequent cases, the contagion is apt to be much less severe. The Asian crisis, moreover, was acutely contagious despite the substitution of large-scale financing for a prompt suspension of debt-service payments.
There are at last signs that the official community is acting in a manner consistent with its rhetoric. They are discussed by Barry Eichengreen in the next chapter. I trust that the discussion of that paper will shed light on the reasons for the persistent disjuncture between official rhetoric and actual behavior and that it will help us devise better ways to involve the private sector in preventing, containing, and resolving future currency crises.
DornbuschRudiger1973“Currency Depreciation, Hoarding, and Relative Prices,”Journal of Political EconomyVol. 81 (July-August) pp. 893–915.
Group of Ten1996The Resolution of Sovereign Liquidity Crises: A Report to the Ministers and Governors Prepared Under the Auspices of the Deputies (Basel: Group of Ten).
Group of Twenty-Two1998Report of the Working Group on International Financial Crise (Washington: Group of Twenty-Two).
LaneTimothy and others1999IMF-Supported Programs in Indonesia Korea and Thailand: A Preliminary Assessment IMF Occasional PaperNo. 178 (Washington: International Monetary Fund).
Note that my definition of financing differs from one used in papers given at this conference. In the papers by Mussa and others, and by Calvo and Reinhart, the change in a country’s reserves is combined with the exchange rate change to gauge the severity of the crisis facing the country. This is an appropriate procedure for their purpose because a rundown of reserves is indicative of pressure in the foreign exchange market and thus an alternative to an exchange rate change. When comparing financing and adjustment, however, a drawdown of reserves should be viewed as a form of financing that can defer and even reduce the size of the change in the current account balance needed to cope with a capital outflow.
Rudiger Dornbusch (1973) provided a model in which a devaluation improves the current account by raising the domestic price level, reducing real balances, and thus reducing expenditure and imports. Because his model was so parsimonious, I accused him of slitting his own throat with Occam’s Razor. I can perhaps be accused of doing the same thing now by emphasizing the contribution of balance sheet effects to the current account adjustment that occurred in Asia. But it was the unique feature of the Asian crisis and a key reason for the attention being paid to foreign currency borrowing and exchange rate pegging in the debate about reform of the international financial system.
Group of Ten (1996), p. i.