The Latin American exchange exposure of U.S. multinationals

This article examines whether there exists any relationship between individual U.S. multinationals' stock returns and fluctuations in Latin American exchange rates. By using a disaggregated dataset of weekly stock returns and real exchange rate movements, it appears that the apparent lack of currency exposure previously documented is mainly due to both the use of too aggregated economic variables and the ignorance of the intervaling effect. We find that about 4% (12%) of the firms experienced economically significant positive (negative) exposure effects for the period of January 1970 to December 2001. While there is time-variation in significant exposure effects, the overall extent of exposure is not sample dependent - a depreciating (appreciating) dollar against Latin American currencies has a net positive (negative) impact on U.S. multinationals. Individual firms in industry groups show high positive as well as negative exposure suggesting that exposure is not necessarily economically significant in the aggregate. The extent to which firms are exposed to Latin American currency fluctuations varies with return horizons; short-term exposure seems to be relatively well hedged, where considerable evidence of long-term exposure is found.