Variance Risk Premium Dynamics: The Role of Jumps∗

Using high-frequency stock market data and (synthetic) variance swap rates, this paper identifies and investigates the temporal variation in the market variance risk premium. The variance risk is manifest in two salient features of financial returns: stochastic volatility and jumps. The pricing of these two separate components is analyzed in a general semiparametric framework. The key empirical results imply that investors fears of future jumps are especially sensitive to recent jump activity and that their willingness to pay for protection against jumps increase significantly immediately after the occurrence of jumps. This in turn suggests that time-varying risk aversion, as previously documented in the literature, is primarily driven by large, or extreme, market moves. The dynamics of risk-neutral jump intensity extracted from deep out-of-the money put options confirms these findings. (JEL C51, C52, G12, G13)