Out-of-sample predictability in predictive regressions with many predictor candidates

This paper is concerned with detecting the presence of out-of-sample predictability in linear predictive regressions with a potentially large set of candidate predictors. We propose a procedure based on out-of-sample MSE comparisons that is implemented in a pairwise manner using one predictor at a time. This results in an aggregate test statistic that is standard normally distributed under the global null hypothesis of no linear predictability. Predictors can be highly persistent, purely stationary, or a combination of both. Upon rejecting the null hypothesis, we introduce a predictor screening procedure designed to identify the most active predictors. An empirical application to key predictors of US economic activity illustrates the usefulness of our methods. It highlights the important forward-looking role played by the series of manufacturing new orders.

forecasting; high dimensional predictability; nested models; out-of-sample; predictive regression

Out-of-sample predictability in predictive regressions with many predictor candidates Articles