Forecast Comparisons in Unstable Environments

We propose new methods for comparing the relative out-of-sample forecasting performance of two competing models in the presence of possible instabilities. The main idea is to develop a measure of the relative ìlocal forecasting performanceî for the two models, and to investigate its stability over time by means of statistical tests. We propose two tests (the ìFluctuation testî and the test against a ìOne-time Reversalî) that analyze the evolution of the modelsí relative performance over historical samples. In contrast to previous approaches to forecast comparison, which are based on measures of ìglobal performanceî, we focus on the entire time path of the modelsí relative performance, which may contain useful information that is lost when looking for the model that forecasts best on average. We apply our tests to the analysis of the time variation in the out-of-sample forecasting performance of monetary models of exchange rate determination relative to the random walk.