TESTING LONG-HORIZON PREDICTIVE ABILITY WITH HIGH PERSISTENCE, AND THE MEESE–ROGOFF PUZZLE

A well-known puzzle in international finance is that a random walk predicts exchange rates better than economic models. I offer a potential explanation. When exchange rates and fundamentals are highly persistent, long-horizon forecasts of economic models are biased by the estimation error. When this bias is big, a random walk will forecast better, even if the economic model is true. I propose a test for equal predictability in the presence of high persistence. It shows that the poor forecasting ability of economic models does not imply that the models are not good descriptions of the data.