Variance Risk Premium and Cross-Section of Stock Returns

Authors:

Abstract
We use equity option prices and high frequency stock prices to estimate stock's variance risk premium (VRP), defined as the difference between the expected variances under the risk-neutral measure and under the empirical measure. Cross-sectionally, VRP is significantly related to stocks' sensitivities to common risk factors. In particular, stocks whose returns tend to be low when systematic volatility increases have higher variance risk premium. We find that stocks' expected returns increase with their variance risk premium. Stocks ranked in the top VRP quintile on average outperform those in the bottom quintile by 1.84% (resp. 1.44%) per month when portfolios are value-weighted (resp. equal-weighted). We reject explanations based on stock mispricing or informed trading in options. Our result is linked to another new finding: stocks with low beta with respect to change in the average idiosyncratic volatility or the market variance risk premium have higher expected returns.