Forecasting West Texas Intermediate Crude Oil Price: Stochastic Differential Approach

Uncertainty in oil markets has led economic researchers to the use of stochastic processes. The purpose of this paper, is the use of stochastic differential models to predict the crude oil price of West Texas Intermediate (WTI) and compare the forecasting performance of these models with ARIMA and GARCH models. In this paper, daily data of WTI crude oil prices from 2/01/1986 to 10/17/2016 has been used that the period 2/01/1986 to 29/08/2016 has been used for estimation in-sample and the rest of the observations have used for out of sample forecasting. The results of the comparison of prediction models using RMSE has shown that long-term memory models (Arfima-Figarch) and stochastic differential models are more accurate forecasting performance compared to ARIMA and GARCH models in in-sample and out of sample forecast for 5-days, 10-days, and 22-days horizons.