Forecasting the Equity Risk Premium with Frequency-Decomposed Predictors

Abstract

We show that the out-of-sample forecast of the equity risk premium can be significantly improved by taking into account the frequency-domain relationship between the equity risk premium and several potential predictors. We consider fifteen predictors from the existing literature, for the out-of-sample forecasting period from January 1990 to December 2014. The best result achieved for individual predictors is a monthly out-of-sample R2 of 2.98% and utility gains of 549 basis points per year for a mean-variance investor. This performance is improved even further when the individual forecasts from the frequency-decomposed predictors are combined. These results are robust for different subsamples, including the Great Moderation period, the Great Financial Crisis period and, more generically, periods of bad, normal and good economic growth. The strong and robust performance of this method comes from its ability to disentangle the information aggregated in the original time series of each variable, which allows to isolate the frequencies of the predictors with the highest predictive power from the noisy parts.

Keywords: predictability, equity risk premium, frequency domain, discrete wavelets

JEL Classification: C58, G11, G12, G17