Volatility forecasting of crude oil futures: The role of investor sentiment and leverage effect

This paper explores the role of investor sentiment and leverage effect on the predictability of crude oil futures market volatility over daily, weekly and monthly horizons. Based on the existing five HAR-type models, we develop three kinds of new HAR-type models by incorporating investor sentiment and/or leverage effect in the corresponding original HAR-type models to examine this issue. We find that the investor sentiment and leverage effect have significant effects on volatility forecasting. In most cases, the leverage effect contains more in-sample and out-of-sample information than the investor sentiment, however, the investor sentiment seems to have more out-of-sample information when forecasting the long-term (i.e., 1-month) future volatility. Furthermore, taking into consideration both the investor sentiment and leverage effect in the original HAR-type models produces better in-sample fitting power and out-of-sample forecasting performance than the corresponding other three kinds of HAR-type models. The results suggest that investor sentiment and leverage effect should be taken into consideration when forecasting the volatility of crude oil futures market.