1997-09

March 28, 1997

The Costs of Managing Speculative Capital Inflows in the Pacific Basin

Kenneth Kletzer

Pacific Basin Notes. This series appears on an occasional basis. It is prepared under the auspices of the Center for Pacific Basin Monetary and Economic Studies within the FRBSF’s Economic Research Department.


In recent years, many middle-income developing countries have experienced impressively large inflows of financial capital. The net capital inflow to developing Pacific Basin countries between 1990 and 1993 alone totaled $151 billion. Thailand and Malaysia each realized net capital inflows in excess of 13% of annual GDP. Net inflows of capital from abroad provide welcome resources for domestic investment during periods of rapid growth due to domestic technological progress; but they also can be disruptive if they are the result of external shocks, such as a decrease in world interest rates.

There are a variety of potential problems associated with externally motivated capital inflows. First, they may suddenly reverse to capital outflows, leading to liquidity problems. Second, the financial systems of emerging market economies may not be well-adapted to managing large inflows of capital efficiently. Third, capital inflows also may be associated with real appreciation of the domestic currency, reducing export competitiveness.

It has been demonstrated that a significant share of capital inflows into Pacific Basin countries since 1988 has been due to external factors. In response, many of these countries pursue nominal exchange rate targets to reduce the volatility of their export demand. To maintain these pegs during periods of capital inflow surges, central banks must purchase incoming foreign assets. In isolation, this would imply that the domestic money supply would expand with a capital inflow and defense of a nominal peg, leading to domestic inflationary pressure.

To avoid these inflationary pressures, the traditional policy response is to “sterilize” the capital inflow by combining an open market sale of domestic financial assets with the purchase of foreign financial assets by the central bank. This process rearranges the consolidated government portfolio. Governments of the East Asian countries have aggressively attempted to sterilize capital inflows. In 1993, for example, one-third of the $100 billion of net capital inflows to the member nations of the Asian Pacific Economic Cooperation forum were purchased by their respective central banks. However, it has been observed that sterilization can be costly. In this Economic Letter, we discuss sterilization costs and how they have affected the monetary policies of some Pacific Basin nations.

The costs of sterilizing capital inflows

The sterilization process can be costly because the interest rate the central bank earns on its holdings of securities issued by foreign governments, such as the United States, is lower than the market real rate of interest developing country governments must pay on their outstanding liabilities. Therefore, as the outstanding public debt held outside the central bank rises, the net debt service expenses of the public sector (including the central bank) increase. The difference between the interest rate that the sterilizing government pays on its outstanding liabilities and that which it earns on central bank reserves has been called a “quasi-fiscal” cost of sterilized intervention. The term “quasi-fiscal” refers to the discrepancy between the real cost to the government of interest owed to the central bank and that owed to the public at large.

Calvo (1991) has argued that sterilization programs designed to achieve low inflation may be self-defeating because the budgetary costs of exchanging domestic debt instruments for foreign assets in the portfolio of the central bank makes them unsustainable. This problem is exacerbated to the extent that real interest rate differentials between domestic financial claims and foreign assets are an underlying cause of capital inflows in the first place. Sterilization keeps domestic real rates of interest high, encouraging more capital to flow to the domestic economy and resulting in a need for even more intensive sterilization activity.

Measuring quasi-fiscal costs

In the literature, quasi-fiscal costs have been estimated as the product of interest rate differentials and the central bank’s holding of foreign assets. Using this method, estimates of the magnitude of quasi-fiscal costs of Latin American capital inflow surges have been reported at between 0.25 and 0.5 percent of GDP.

However, it is not straightforward that real interest rate differentials measure quasi-fiscal costs. In the case of developing countries, government debt typically pays higher real interest rates than that issued by the treasuries of advanced market economies because the potential holders of such securities believe that the risk of a sovereign default is greater for such debt. If this interest premium just compensates the investor for the higher risk of default incurred, then the net present value of the interest and principal payments that the government expects to pay will be just equal to the market value of the debt it issues. That is, the government that issues the debt realizes a benefit equal in value to the discounted stream of extra interest it pays. If this is the case, then the interest rate differential does not reflect true sterilization costs.

True sterilization costs require imperfections in domestic and foreign financial markets which lead to interest rate differentials that do not correspond to discounted differences in expected asset payoffs. For example, in situations where information asymmetries are present, investors may erroneously treat a number of heterogeneous countries as if they represent the same risk levels. If the government from a relatively safe nation knows that it is being “unfairly pooled” into this risk class, it will perceive such an interest premium as a “true cost.”

Nevertheless, we argue in Kletzer and Spiegel (1996) that measures based on real interest rate premia constitute “upper bound estimates” of quasi-fiscal costs. We estimate quasi-fiscal costs over a period as the product of the real interest premium and the magnitude of the capital inflow. This specification is an “upper-bound” because it treats all interest rate differentials as reflecting true costs to the central bank.

The estimates of quasi-fiscal costs obtained through this method reveal an interesting pattern, typified by the Korean case illustrated in Figure 1. During most of the period quasi-fiscal costs are small: the mean levels of these costs during periods of positive net foreign asset accumulation are 0.06% of GDP. But during the capital inflow surge of 1988, the costs rise suddenly to between 0.25% and 0.28% of GDP, depending on how expected exchange rates are estimated.

Do quasi-fiscal costs matter?

Having obtained upper-bound estimates of quasi-fiscal costs, we test for the importance of our quasi-fiscal cost estimates for governments’ abilities to defend pegged exchange rates and restrain monetary expansion. If quasi-fiscal costs are important, they should play a role in the timing of the abandonment of a sterilization program. In particular, central banks will tend to stop sterilizing capital inflows as these costs rise. The central bank will allow either the currency to appreciate or the money supply to expand as the contribution of additional sterilization to public sector debt rises.

Private sector behavior surrounding the abandonment of a sterilization episode will be analogous to a speculative attack on a pegged exchange rate regime. The market will anticipate the decision to quit sterilizing and respond according to its expectations of how and when the program will end. In contrast to the familiar circumstance of a run on the central bank’s reserve culminating in a currency devaluation, a speculative attack on an exchange rate peg resisting appreciation will present itself as a surge in capital inflows. At such a time, the rate of sterilization will peak and along with it, quasi-fiscal costs. If this rise in quasi-fiscal costs is sufficiently high, it can lead to abandonment of the sterilization program. This will manifest itself in the central bank allowing adjustment through either the exchange rate or the money supply. It is possible that surges in capital inflows experienced in recent years by developing countries reflected expectations of impending abandonment of a nominal exchange rate peg.

The budgetary costs of sterilization: the Pacific Basin experience

We study episodes of capital inflows and defense of a managed exchange rate regime in six Pacific Basin economies: Indonesia, Mexico, Korea, the Philippines, Singapore, and Taiwan, between 1980 and 1993. Our empirical tests are motivated by a simple theory of exchange rate determination in which the central bank seeks to maintain both low inflation and low unemployment. The central bank incorporates the effects of its actions on the future costs of sterilization in its monetary policy decisions: sterilizing capital inflows today implies larger future public sector budget deficits. Eventually, the implied levels of public debt will not be sustainable and sterilization must be abandoned. Therefore, sterilization today reduces the opportunity to sterilize tomorrow. In this framework outstanding privately held public debt and real interest rate differentials will influence the degree to which central banks offset foreign exchange reserve increases.

We find little evidence that quasi-fiscal costs affect monetary expansion and exchange rate policies over the entire period for any of the economies in the sample. However, a time-series analysis reveals that the spikes in the estimated quasi-fiscal costs during capital inflow surges may represent periods of speculative attack on nominal exchange rate pegs, where the central bank is actively attempting to prevent exchange rate appreciation. We find that the Philippines, Taiwan, and Mexico all responded to sudden surges of capital inflows and the accompanying rise in estimated quasi-fiscal costs by allowing domestic credit to expand. In all three cases, this was a temporary response and domestic credit expansion soon declined. For Indonesia, Korea, and Singapore, however, domestic credit growth did not respond. Singapore was the most dramatic case. Actual domestic credit for Singapore followed the path predicted by previous policy after a sudden rise in capital inflows, while the nominal and real exchange rate instead appreciated significantly.

Over the entire sample, the evidence suggests that quasi-fiscal costs do not influence monetary policy. This result may imply that these costs are incorrectly measured by real interest rate differentials, due to differential risks that are correctly priced by financial markets, or that monetary policy makers ignore such costs. However, during periods of surges in capital inflows, some nations do appear to respond to the costs of sterilization by allowing either the money supply or the exchange rate to adjust. The analysis therefore indicates that the budgetary costs of sterilized intervention only matter during episodes of serious speculative pressure.

Kenneth M. Kletzer
Professor, University of California, Santa Cruz, and
Visiting Scholar, Federal Reserve Bank of San Francisco

References

Calvo, Guillermo. 1991. “The Perils of Sterilization.” IMF Staff Papers, vol. 38, pp. 921-926.

Kletzer, Kenneth M., and Mark M. Spiegel. 1996. “Speculative Capital Inflows and Exchange Rate Targeting in the Pacific Basin: Theory and Evidence.” Center for Pacific Basin Monetary and Economic Studies, Federal Reserve Bank of San Francisco, Working Paper PB96-05 (September).

Pacific Basin Notes are published occasionally by the Center for Pacific Basin Studies. Opinions expressed in FRBSF Economic Letter do not necessarily reflect the views of the management of the Federal Reserve Bank of San Francisco or of the Board of Governors of the Federal Reserve System. This publication is edited by Sam Zuckerman and Anita Todd. Permission to reprint must be obtained in writing.

Please send editorial comments and requests for reprint permission to

Research Library
Attn: Research publications, MS 1140
Federal Reserve Bank of San Francisco
P.O. Box 7702
San Francisco, CA 94120