Tests for m-dependence based on Sample Splitting Methods

Abstract

This paper develops new test methods for m-dependent data. Our approach is based on sample splitting by regular sampling of the original data at lower frequencies, so that standard techniques for testing independence can be used for each individual subsample. We then propose several alternative statistics that aggregate information across subsamples and investigate their asymptotic and nite sample properties. We apply our methods to test the predictability of excess returns in foreign exchange markets. We also illustrate how our serial dependence tests can provide useful information for identifying particular economic alternatives when testing the expectations hypothesis in foreign exchange markets.