Comment on Optimal Exchange Rate Policy

Abstract

Transportation costs aside, under most circumstances, when markets are complete, when perfect competition prevails, and prices adjust instantaneously, an economy's choice of the exchange rate may not be that significant. However, when one of these features is missing, changes in the nominal exchange rate may have real effects because they can affect the behavior of relative prices over time. Charles' hunch is that changes in the nominal peso-dollar exchange rate may affect the real exchange rate because of violations to the law of one price in the conditions of trade between Mexico and the United States. This paper examines the optimal exchange rate policy when, because of some underlying price stickiness and local-currency pricing effects, the law of one price fails. As a way of motivation, in section I of the paper, Charles starts with an expression that decomposes consumer prices into tradeables and nontradeables. He also obtains an expression that describes changes in the real exchange rate in terms of deviations from the law of one price and changes in the relative prices of tradeables to nontradeables. Applying the insight of some of his earlier work, Charles tries to disentangle the alternative sources of changes in the real exchange rate. He focuses on the commodities component of the consumer prices for the U.S. and consumer prices of traded goods for Mexico. He reports that his analysis suggests that it is violations of the law of one price that dominate real consumer exchange rates in Mexico. There could be other reasons for his observation that the law of one price fails, such as the presence of nontraded marketing services. Charles reports that his analysis, at the level of disaggregation he picks, suggests that marketing factors are not likely to be very relevant in explaining movements in the U.S./Mexico real exchange rate. The question remains; why is the law of one price violated? It is important to answer this question to better construct an appropriate model in which to evaluate the optimality of alternative exchange rate regimes. He reports that there are two pieces of information that may help in resolving what accounts for violations to the law of one price. The first piece of information is Rogers (1999). As Charles reports, Rogers extends Rogers and Engel (1996) by looking at aggregate consumer prices for cities