Forecasting Crude Oil Price Volatility

We use high-frequency intra-day realized volatility to evaluate the relative forecasting performance of several models for the volatility of crude oil daily spot returns. Our objective is to evaluate the predictive ability of time-invariant and Markov switching GARCH models over different horizons. Using Carasco, Hu and Ploberger (2014) test for regime switching in the mean and variance of the GARCH(1,1), we find overwhelming support for a Markov switching model. A comprehensive out-of-sample comparison of different GARCH and Markov switching GARCH models suggests that the EGARCH-t performs better in forecasting the volatility of crude oil returns for shorter oneand five-day horizons. In contrast, the MS-GARCH-t tends to exhibit higher predictive accuracy at longer horizons. This result is estabilished by computing the Equal Predictive Ability of Diebold and Mariano(1995), the Reality Check of White (2000), the test of Superior Predictive Ability of Hansen (2005) and the Model Confidence Set of Hansen, Lunde and Nason (2011) over the totality of the evaluation sample. In addition, a comparison of the MSPE computed using a rolling window suggests that MS-GARCH-t model is better at predicting volatility during periods of turmoil.