Investors' perspective on forecasting crude oil return volatility : where do we stand today?

In this paper, we review studies of oil volatility prediction from a new perspective: that of investors who require economic evaluations of forecasting performance. Our results indicate that no single volatility model outperforms all of the competing models, of which GARCH and realized volatility models are the most popular. Most studies evaluate forecasting performance using two criteria: value at risk and hedging effectiveness. Parameter instability and model uncertainty are technical issues that affect out-of-sample performance. Most studies assess volatility forecasts from the perspectives of portfolio management and derivative pricing. Whether oil volatility can predict economic variables and the asset pricing implications of oil volatility for financial markets are important topics that require attention.