In the last few decades, world financial markets have faced numerous volatilities and fluctuations. As a result active economic agents have become more intereted in how to understand and foresee these fluctuation. Forecasting developments of the oil market has became ever more important after the oil crisis of the 1970s. The objective was to reduce the potential risk of such fluctuations by having advance knowledge of expected changes. Review of time series data of the West Texas Intermediate (WTI) prices indicate the presence of clustering volatility, can not be ignored in forcasting. This has led us to focus in this study on the forecasting volatility of crud oil prices. We have used ARCH models to evaluate statistical errors of forecasts. The study results indicate continuous effect of shocks on conditional variance of crude oil prices. Future volatility of crude oil prices is dependent on oil price changes in the past. GARCH and TGARCH models perform better than other models of conditional variance in forecasting volatility of crude oil prices.
JEL Classification : F12