GARCH, Implied Volatilities and Implied Distributions: An Evaluation for Forecasting Purposes

Volatility implied in option prices reflects the market participant's beliefs about future volatility and incorporates information that is not historical. Implied volatility is therefore widely believed to perform better as an indicator of future volatility than other forecasts based on historical time-series. In this study, I investigate the information content and predictive power of implied volatility from currency options traded on the OTC-market. Furthermore, I evaluate implied volatility both against other forecasts based on option prices and against volatility forecasts from models that are strictly historical by nature such as different GARCH models. I find that implied volatility has predictive power in forecasting future volatility, at least for shorter forecast horizons, although in most cases the forecasts are not unbiased. Furthermore, for some currencies GARCH volatility forecasts outperform implied volatility.