Improved Forecast Ability of Oil Market Volatility Based on combined Markov Switching and GARCH-class Model

This paper estimates and forecasts West Texas Intermediate (WTI) crude oil market volatility with GARCH, EGARCH models. We find that, in terms of Root Mean Squared Error (RMSE) criteria, EGARCH model outperformances GARCH model in both short term and long term forecast length. Then, we apply Markov Switching model on oil volatility term and find that we can improve forecast accuracy of oil market volatility further with MSGARCH model, when compared with normal GARCH and normal EGARCH model. Besides, empirical results confirm that two regimes consist of regime 1 with low volatility and regime 2 with high volatility exist in WTI crude oil market. Also, the smoothed probabilities indicate that before 2004, WTI crude oil market stays in a low volatility regime most of time, while after 2004, WTI crude oil market switches to high volatility regime and stays in it until present most of time.