Can the variance risk premium be predicted? Evidence from major U.S. indices
KeywordsVariance swap ; Variance risk premium ; Realized variance ; Index implied volatility ; In-Sample analysis ; Out-of-Sample analysis
In this dissertation, we investigate whether the variance risk premium (VRP) in the three U.S indices, particularly S&P 500, Dow Jones and Nasdaq, can be predicted. To this end, we conduct both in-sample analysis and out-of-sample analysis. To quantify the VRP we use the realized variance and the squared index implied volatility as an approach for the variance swap rate. We find that in all three markets the average VRP is negative. A number of factors which are considered indicators of the trading activity and the stock market conditions can predict the VRP. This is confirmed by the in-sample analysis and the out-of-sample analysis.