Asymmetric Roles of Macroeconomic Variables in the Real Exchange Rate: Insights from U.S.-Korea Data

This paper investigates the asymmetric out-of-sample predictability of macroeconomic variables for the real exchange rate between the United States and Korea. While conventional models often suggest that the bilateral real exchange rate is primarily driven by the relative economic performance of the two countries, our research highlights the superior predictive power of latent factors obtained from U.S. economic variables, while Korean factors fail to enhance predictability and often act as noise. We attribute the strong predictability of U.S. factors to significant cross-correlations observed among a panel of bilateral real exchange rates vis-\`a-vis the U.S. dollar, indicating a limited role for idiosyncratic factors associated with smaller economies. Our major findings are based on data from the pre-COVID19 era. We further explore how economic crises disrupt this relationship, resulting in temporary yet persistent disconnects between the real exchange rate and macroeconomic fundamentals.