Uncertainty-risk and the Cross-section of Expected Returns

Abstract

I examine the pricing of uncertainty-risk in the cross-section of stock returns. Several theoretical channels can be used to argue why investors should be willing to accept lower expected returns on stocks with a high exposure to innovations in uncertainty. I replicate the seminal paper of Ang et al. (2006), who use the VXO implied volatility index as a proxy for uncertainty and confirm their results that, consistent with these theories, there is a significant negative relationship between stocks’ covariance with the VXO and their average returns over the period of 1986 to 2000. Subsequently, I update their study to 1986 to 2018 and find a similar, though weakened, relationship. I show the empirical support for an uncertainty-risk factor to have steadily increased until the early 2000s and steadily decreased since then. Finally, I examine the TYVIX implied treasury volatility index among post-crisis US banking stocks and find higher average returns for stocks that covary positively with this index. This reversed price of uncertainty-risk is consistent with government guarantees against interest-rate-uncertainty induced bank defaults.