Higher-Order Risk Premium, Stock Return Predictability, and Rare Event Dynamics
60 Pages Posted: 12 Feb 2018 Last revised: 16 Mar 2019
Date Written: March 14, 2019
We find that separating the higher-order risk premium from the pure variance risk premium can significantly improve the stock market predictability, with R-squared up to 14 percent for the 3-month horizon. This finding proves to be economically significant in an asset allocation exercise, becomes even stronger for the portfolio returns of the momentum factor, and survives a series of robustness checks. We show that a consumption-based asset pricing model with rare events can generate the predictability afforded by higher-order risk premium.
Keywords: Equity risk premium; Predictive regression; Higher-order risk premium; Variance risk premium; Rare events
JEL Classification: G12, G13, C22
Suggested Citation: Suggested Citation