Volatility Forecasts : Do Volatility Estimators and Evaluation Methods Matter?

This study investigates the volatility forecasting abilities of return-based and range-based estimators for two stock indices and two individual stocks in the U.S. stock market. The forecasting performances are evaluated by two robust statistical loss functions, and further by financial applications in risk management and option pricing. Consistent with previous studies, the range-based volatility forecasts outperform in terms of statistical evaluation, Value-at-Risk calculation, and option pricing. However, return-based volatility forecasts prove superior in the evaluation of market risk capital requirements