Oil Strikes Back: Trend Factors and Exchange Rates

Abstract

A well-known puzzle in international finance is that fluctuations in exchange rate are very difficult to predict and existing predictive models often perform worse than the naive random walk model. In this paper, we construct an oil trend factor and find that it performs better than the naive random walk in predicting exchange rates out-of-sample and delivers superior economic significance. This result holds in both developed and developing countries, with different forecasting horizons and with different specifications of trend factors. Notably, we also show that the predictive power of the oil trend factor is closely related to macroeconomic fundamentals.