Can currency-based risk factors help forecast exchange rates?

This paper examines the time series predictability of bilateral exchange rates from linear factor models that utilize the unconditional and conditional expectations of three currency-based risk factors. Exploiting a comprehensive set of statistical criteria, we find that all versions of the linear factor models largely fail to outperform the benchmark random walk with drift model for the out-of-sample forecasting of monthly exchange rate returns. This holds true for both individual currencies and currency portfolios formed on forward discounts. We also show that the information embedded in the currency-based risk factors does not generate systematic economic value for investors.