Are Risk-Neutral Variance and Skewness Good Proxies for their Physical Counterparts?

Abstract

Many empirical papers use VIX, risk-neutral variance, or risk-neutral skewness obtained from option prices to proxy for the physical moments. I investigate whether risk-neutral moments are good proxies for physical moments in the time series and in the cross-section of stock returns. I find that risk-neutral variance is a good proxy while risk-neutral skewness is a bad proxy in both the time series and cross-section. In the time series, physical skewness is a stronger predictor of business cycle fluctuations than risk-neutral skewness. In the cross-section, a high-minus-low physical skewness portfolio carries a positive premium that is unexplained by risk-neutral skewness.

Keywords: Option implied moments, physical moments, risk-neutral moments

JEL Classification: G00, G11, G12, G13