Investigating the puzzling intertemporal market risk-return relationship

Abstract

This paper investigates the conflicting results documented by the existing empirical literature on the intertemporal relationship between the expected market risk premium and the conditional market variance. We show that the previous tests are biased because they use the realized market risk premium to proxy the expected market risk premium, without accounting for the negative portion of the market risk premium distribution. The empirical evidence based on a new test, allowing up and down-market volatility to have different impacts on the market risk premium, indicates a consistent and significant risk-return relationship.