NBER Reporter: Spring 2001

International Finance and Macroeconomics

The NBER's Program on International Finance and Economics held its spring meeting in Cambridge on March 23. Organizers Richard Lyons and Andrew Rose, NBER and University of California, Berkeley, organized this program:

Kathryn Dominguez and Linda Tesar, NBER and University of Michigan, "Exchange Rate Exposure"

Discussants: Bernard Dumas, NBER and INSEAD, and Kristin Forbes, NBER and MIT

Eric Parrado, New York University, and Andrès Velasco, NBER and Harvard University, "Optimal Interest Rate Policy in a Small Open Economy: The Case for a Clean Float"

Discussants: Paolo Pesenti, Federal Reserve Bank of New York, and Jaume Ventura, NBER and MIT

Pierre-Olivier Gourinchas, NBER and Princeton University, Rodrigo Valdés, Chilean Ministry of Finance, and Oscar Landerretche, MIT, "Lending Booms: Latin America and the World"

Discussants: Michael Dooley, NBER and University of California, Santa Cruz, and Carmen Reinhart, NBER and University of Maryland

Menzie Chinn, NBER and University of California at Santa Cruz, and Guy Meredith, International Monetary Fund, "Testing Uncovered Interest Parity at Short and Long Horizons"

Discussants: Geert Bekaert, NBER and Columbia University, and Karen Lewis, NBER and University of Pennsylvania

Margarida Duarte, Federal Reserve Bank of Richmond, and Alan Stockman, NBER and University of Rochester, "Rational Speculation and Exchange Rates"

Discussants: Kenneth Rogoff, NBER and Harvard University, and Olivier Jeanne, International Monetary Fund

Finance theory suggests that, in a world with integrated capital markets, exposure to foreign markets should have little influence on asset prices. Dominguez and Tesar find that in a pooled sample of eight (non-US) industrialized and emerging markets, between 12 and 23 percent of firms are exposed to exchange rate movements. They also find that: the choice of exchange rate matters; using the trade-weighted exchange rate is likely to understate the extent of exposure. The extent of exposure is not a result of a spurious correlation between random variables with high variances, though; exposure increases with the return horizon and within a country and an industry, exposure coefficients are roughly evenly split between positive and negative values. Averaging across the (absolute value of the) significant exposure coefficients in their sample of countries, the authors estimate the exposure coefficient to be about 0.5. The extent of exposure is not sensitive to the sample period, but the set of firms that is exposed does vary over time. The sign of the exposure coefficients changes across subperiods for about half of the firms in the sample. Thus exposure does not appear to be related to firm size, industry affiliation, multinational status, foreign sales, international assets, or industry-level trade.

Parrado and Velasco derive the optimal monetary and exchange rate policy for a small stochastic open economy with imperfect competition and short-run price rigidity. They conclude that the optimal policy depends crucially on the source of stochastic disturbances affecting the economy, much as in the literature pioneered by Poole (1970). Optimal monetary policy reacts to domestic disturbances but does not respond at all to foreign shocks. Consequently, under the optimal policy the exchange rate floats cleanly, and monetary policy is aimed exclusively at stabilizing the home economy.

Recent theories of crisis put lending booms at the root of financial collapses. Yet lending booms may be a natural consequence of economic development and fluctuations. Gourinchas, Valdés, and Landerretche investigate whether lending booms are indeed dangerous, based on a sample of episodes spanning 40 years, especially in Latin America. They find that lending booms often are associated with domestic investment booms, increases in domestic interest rates, a worsening of the current account, declines in reserves, real appreciation, and declines in output growth. The "typical" lending boom does not substantially increase the vulnerability of the banking sector or the balance of payments. Comparing countries, the authors find that lending booms in Latin America make the economy considerably more volatile and vulnerable to financial and balance-of-payment crises.

The hypothesis that interest rate differentials are unbiased predictors of future exchange rate movements has been rejected almost universally in empirical studies using short-horizon data; Chinn and Meredith test this hypothesis using interest rates on longer-maturity bonds for the G-7 countries. They find that the coefficients on interest differentials are of the correct sign, and almost all are closer to the predicted value of one than to zero. These results are robust to changes in data type and to base currency (Deutschemark versus U.S. dollar).

Models of exchange rates typically have failed to produce results consistent with the key fact that real and nominal exchange rates move in ways not closely connected to current (or past) macroeconomic or trade variables. Models that rely on the same shocks to drive fluctuations in macro variables and exchange rates typically predict counterfactually strong co-movements between them. Duarte and Stockman propose a new approach to exchange rates and implement it with a new-open-economy macro model. The approach focuses on the effects of speculation and the resulting changes in risk premiums on foreign-exchange markets. Exchange rates follow a forward-looking, first-order stochastic difference equation that includes terms involving risk premiums. Changes in risk premiums can affect the current exchange rate without necessarily creating large changes in current macroeconomic variables. Consequently, the approach has the potential to explain the Flood-Rose exchange-rate "disconnect" puzzle. However, the baseline model does not yet generate a sufficient degree of rational speculation to explain observed variation of risk premiums and exchange rates.