Nonlinear Causality between Exchange Rate and Trade: Empirical Evidence from Korea

Abstract

Despite the strong theoretical foundations of trade models, the empirical evidence in favor of causality between exchange rate and trade is relatively limited. Previous studies that adopt a linear framework typically fail to find evidence in favor of a relationship among the exchange rate, export and import. Unlike previous studies on the relationship among them, this study did not assume any specific functional form of linearity and extended the analysis by applying both linearity and nonlinearity to the specification of testing Granger causality. This study empirically investigates the link between exchange rate and trade through a new approach for general representation of causality relation without restricting a specific functional form to monthly data for Korea from 1990 to 2015. The empirical results show that there is a cointegration among exchange rate, exports and import, and the causality is not only linear but nonlinear as well. More importantly, they imply that allowing for nonlinearities increases the number of rejections of the null of no causality, which supports the presence of nonlinearities in the data generating process. The finding demonstrates agreement with theoretical predictions and significant improvement over previous efforts by the literature.