Risk-free rate effects on conditional variances and conditional correlations of stock returns

Abstract

This paper investigates whether the risk-free rate may explain the movements observed in the conditional second moments of asset returns. Original results are derived, within the C-CAPM framework, that attest the existence of a channel connecting these seemingly unrelated quantities. The empirical results, involving 165 time series of stock returns quoted at the NYSE, show that the risk-free rate does contain information that is relevant in predicting the 165 conditional variances and 13,530 conditional correlations. These findings are particularly pronounced at lower frequencies where the persistence of the conditional second moments is significantly weaker.