Abstract
This paper is the first to investigate the informational content of model-free forward implied volatility (MF FIV) in energy markets, as previous studies mainly concentrated on model-dependent volatility estimates and used data on financial assets rather than on commodities. Using daily prices of crude oil and natural gas futures and options, we try to determine whether MF FIV provides unbiased forecasts of future IV and compare the performance of MF FIV with the performance of three other volatility forecast measures, such as model-dependent forward implied volatility (MD FIV) estimated using Black’s (1976) model, GARCH (1,1) and historical volatility (HV) based on a 250-day moving average. We find that in the crude oil sample MF FIV performs best in providing unbiased forecasts of IV among four volatility forecasting measures, although only at the nearby intervals. In the natural gas sample MF FIV seems to provide the best forecasts of MF IV, and GARCH takes second place, followed by MD FIV and HV. At forecasting MD IV, MD FIV is the most successful; GARCH is the second best performer, followed by MF FIV and HV. GARCH clearly outperforms HV in most of the tests, indicating that information about the 250 past daily volatility values contained in HV estimates does not add explanatory power. We also find support for informational efficiency of energy markets, as volatilities are difficult to predict far in the future. Hence, information contained in energy prices can be considered reliable, and options on energy products can serve as good hedging instruments to all market participants.