Commodity Variance Risk Premia and Expected Futures Returns : Evidence from the Crude Oil Market

We develop an extended mean-variance model to investigate the relationship between variance risk premia (VRP) and expected futures returns in the commodity market. In the presence of stochastic variance, commodity producers trade both futures and options to hedge their exposure to commodity price and volatility risk; speculators provide liquidity and ask for risk premia. This model reveals a negative relationship between VRP and expected futures returns. Empirically, we measure VRP using options and high-frequency futures data in the crude oil market. Consistent with our model, we find that VRP negatively predict futures returns even after controlling for other predictor variables