A robust test for predictability with unknown persistence

摘要

This paper provides a robust test that is a data-dependent weighted average of the regression-based test and covariance-based test. This new test allows for multivariate cases and yields chi-squared inference regardless of whether predictors are stationary, local-to-unity or I(1). The new test improves the covariance-based test proposed by Maynard and Shimotsu (2009) in stationary cases. Furthermore, similar to the covariance-based test, the new test does not force the dependent variable and predictors to share the same order of integration under the alternative hypothesis. This is very important because empirically the dependent variable usually appears to be stationary while predictors may be (nearly) nonstationary. The test shows good performance in simulations. (c) 2020 Elsevier B.V. All rights reserved.