The economic value of volatility forecasts: A simplified approach
|Speaker:||Nick Taylor, Cardiff University|
|Date:||Friday 5 November 2010|
|Time:||14:00 - 15:30|
|Location:||MBA Lecture Theatre|
This paper provides a simple to calculate measure of the economic value of asset return volatility predictability. In deriving this measure we find that it is proportional to the sum of squared proportional errors. Consequently, economic value can be further enhanced by the selection of an appropriate objective function in the estimation phase of the forecast construction process. We illustrate use of these results with an application to six US index futures markets. The results indicate that volatility timing investors are willing to pay around 3% per annum for forecasts based on appropriate volatility dynamics. Moreover, use of a pertinent predictor variable (the VIX volatility index) leads to an additional fee of around 0.5% per annum, though the magnitude of this benefit depends on the risk preferences of the investor, the Sharpe ratio, the realised volatility measure used, and the futures market considered.
Key Words: Economic value, predictability, volatility timing.
JEL Classification Code: C53, G11, G17.