Does the Volatility-Hedging Portfolio Underreact to Volatility Innovations?
85 Pages Posted: 25 Apr 2022 Last revised: 27 Apr 2022
Date Written: April 13, 2022
This paper proposes a new explanation of the negative correlation between VIX betas and expected stock returns documented by Ang et al. (2006). While the relation has been widely cited as the proof that market volatility risk is priced in the cross-section of stocks, we find this view highly implausible because stocks with high VIX betas perform particularly poorly when the VIX index spikes up during the holding period. Also challenging the risk-based explanation is the finding that VIX betas only negatively predict stock returns if the betas are measured using intraday returns. We argue that the beta-return relation is indicative of market inefficiency and develop four theoretical models revolving around the conjecture that stocks with high VIX betas are overpriced because their prices underreact to innovations in market volatility. Further empirical tests show strong support for the underreaction-based explanations.
Keywords: volatility, underreaction, expected return, mispricing, factor model
JEL Classification: G11, G12, G13
Suggested Citation: Suggested Citation