Modeling energy price dynamics: GARCH versus stochastic volatility

- We compare a variety of GARCH and SV models for fitting nine series of energy prices.
- We find that SV models generally compare favorably to their GARCH counterparts.
- The SV model with moving average innovations is the best model for all nine series.

We compare a number of GARCH and stochastic volatility (SV) models using nine series of oil, petroleum product and natural gas prices in a formal Bayesian model comparison exercise. The competing models include the standard models of GARCH(1,1) and SV with an AR(1) log-volatility process, as well as more flexible models with jumps, volatility in mean, leverage effects, and t distributed and moving average innovations. We find that: (1) SV models generally compare favorably to their GARCH counterparts; (2) the jump component and t distributed innovations substantially improve the performance of the standard GARCH, but are unimportant for the SV model; (3) the volatility feedback channel seems to be superfluous; (4) the moving average component markedly improves the fit of both GARCH and SV models; and (5) the leverage effect is important for modeling crude oil prices-West Texas Intermediate and Brent-but not for other energy prices. Overall, the SV model with moving average innovations is the best model for all nine series.