Serial-Dependence and Persistence Robust Inference in Predictive Regressions

TL;DR

This paper presents a new robust testing method for predictive regressions that effectively handles serial dependence, persistence, and heteroskedasticity without adjustments, supported by theoretical and simulation evidence.

Contribution

Introduces a novel family of test statistics for predictive regressions that are robust to predictor persistence and serial correlation, eliminating the need for corrections.

Findings

Test statistics follow a chi-square distribution under null hypothesis.

Method maintains size and power in finite samples as shown by simulations.

Robust to serial correlation and heteroskedasticity without adjustments.

Abstract

This paper introduces a new method for testing the statistical significance of estimated parameters in predictive regressions. The approach features a new family of test statistics that are robust to the degree of persistence of the predictors. Importantly, the method accounts for serial correlation and conditional heteroskedasticity without requiring any corrections or adjustments. This is achieved through a mechanism embedded within the test statistics that effectively decouples serial dependence present in the data. The limiting null distributions of these test statistics are shown to follow a chi-square distribution, and their asymptotic power under local alternatives is derived. A comprehensive set of simulation experiments illustrates their finite sample size and power properties.