Narrow Results

This paper first confirms and extends findings in the previous literature that for major commodity exporters with market-based exchange rates, the world price of their primary commodity exports is an important and robust determinant for their real exchange rate values. However, despite inducing strong contemporaneous currency responses, commodity prices tell us little about subsequent exchange rate movements a quarter ahead. To further investigate real exchange rate predictability, we use Bayesian model averaging and least angle regression as mechanisms to address model uncertainty and select predictors. We show that while various combinations of macroeconomic fundamentals — including commodity prices at times — can help predict quarterly exchange rate changes, no single specification emerges as the clear winner across both countries and time periods.