An Improved Composite Estimate for Realized Volatility

The purpose of this study is to determine whether a superior estimation for security volatility can be derived by finding a balance between historical data, the implied volatility and an empirical implied distribution placed on the options chains of four exchange traded funds. Data is collected for 30-day option contracts expiring on the first trading day of every month over a minimum of three years and analyzed to see whether a weighted combination of the three estimates provides an estimate with a higher correlation of realized volatility. A linear optimization solution was formulated to determine the best possible composite volatility estimate. The results of a hypothesis test showed that there is statistically significant evidence in which the composite volatility estimate is preferred at a 95% confidence level. With a better predictor for security volatility, this optimization process could be applied to the creation of portfolios that better meet investor risk preference. The final version of this paper is published in the Journal of Statistical and Econometric Methods.