Search

This study investigates the impact of real exchange rates on the trade balances between Thailand and its major trading partners. Previous empirical evidence gave mixed results of the impact of real exchange rates on trade balances. In this study, Augmented Dicky-Fuller and Phillips-Perron tests for stationarity followed by the cointegration tests are implemented. All variables in the model are nonstationary but cointegrated. In cointegrating regressions, biases are introduced by simultaneity and serial correlation in the error. The specification that deals with these problems is the non-linear specification of Stock and Watson [1989]. By using this non-linear model as modified by Reinhart [1995], the results show that the impact of real exchange rates (Thai baht/foreign currency) on trade balances is significant in most cases. Therefore, the generalized Marshall-Lerner condition seems to hold. Furthermore, the results show that the real exchange rates play a more important role in the determination of the bilateral trade balances than other factors. Since the real exchange rate variable plays a major role in this study, the policy recommendation is to prevent exchange rate misalignment. A policy that can neutralize the changes in nominal exchange rates and relative prices should be introduced to prevent further deterioration of the trade balance.