Analyzing and Forecasting Crude Oil Price Based on Stochastic Process Model

The dynamic behavior of crude oil prices has become a hot issue in recent years because the increased oil prices worldwide are having a great impact on all economic activities. This paper aims to select the continuous-time stochastic model to describe and forecast the world crude oil price. The Maximum Likelihood Estimation method is implemented to fit the parameters of continuous-time stochastic processes. The result of unit root test shows that time series of the crude oil price is a stationary series. And the simulation of continuous-time stochastic processes and the mean error between the simulated prices and the market ones shows that the Geometric Brownian Motion is a very effective model for the world crude oil price.