Abstract:
Recent evidence highlights that commodity price changes exhibit a short-lived, yet robust contemporaneous effect on commodity currencies, which is mainly detectable in daily-frequency data. We use MIDAS models in a Bayesian setting to include mixed-frequency dynamics while accounting for time-variation in predictive ability. Using the random walk Metropolis-Hastings technique as a new tool to estimate our class of MIDAS regressions, we find that for most of the commodity currencies in our sample exploiting this short-lived relationship yields to statistically more precise out-of-sample exchange rate point and density forecasts relative to the no-change benchmark. Further, the usual low-frequency predictors, such as money supplies and interest rates differentials, typically receive little support from the data at monthly forecasting horizons. In contrast, models featuring daily commodity prices are highly likely.
