Taming the Turbulence: Modelling and Forecasting Crude Oil Price Volatility with GARCH and SARIMA Models

Abstract

The volatility is a signal of risk, either good or bad, to investors. Crude oil price is a commodity of global concern because its volatility significantly affects the economic stability of many countries. Due to the lack of comparative studies in forecasting performance, this study aims to model and forecast crude oil prices and their volatility by comparing the Generalised Autoregressive Conditional Heteroskedasticity (GARCH) model and the Seasonal Autoregressive Integrated Moving Average (SARIMA) model. The monthly crude oil price data from January 2000 to December 2019, obtained from the U.S. Energy Information Administration, are used for this analysis. The results show that the GARCH(3,1) model is the most suitable model compared to EGARCH, TGARCH, and SARIMA in capturing the volatility of crude oil prices. This indicates that the GARCH(3,1) model offers superior forecasting performance and can serve as a reliable tool for investors and policymakers to anticipate market fluctuations and manage associated risks effectively.