Var estimation of oil price based on clustering brownian motion with drift

Value-at-Risk (VaR) is an essential tool for risk management in financial markets. A new model for oil markets, the Monte Carlo simulation based clustering Brownian motion with drift (MCSCBMD), is developed in this paper, which considers the dynamics of oil prices evidently characterized by clustering, mean reversion and asymmetry. We evaluate predictive performance of a selection of VaR models for WTI crude oil spot price including proposed MCSCBMD approach and several traditional VaR models such as the variance- covariance (VC), the historical simulation (HS), and the Monte Carlo simulation based geometric Brownian motion (MCSGBM) methods. The results show that the MCSCBMD approach offers a more flexible VaR quantification, which fits the continuous oil price movements better and provides an efficient risk quantification.