Interval Prediction Of Crude Oil Price Using V@R-Asymmetric-Garch Model To Optimize The Petroleum Production Sharing Contract In Indonesia Via Gross Split Method

Abstract

Forecasting crude oil prices is a critical factor in evaluating the potential risk of an oil and gas project. The price of crude oil tends to have volatile properties, so most investors are not confident in investing in oil and gas projects. A mistake in forecasting crude oil prices significantly impacted accuracy in evaluating a project proposed. One of the most used methods is point-prediction ARMA(p,q) model. However, this method could not capture the volatility of the crude oil price data, and point prediction is riskier to use because it is not robust, i.e., it tends to change due to the existence of extreme values. To solve these problems, instead of using the point prediction ARMA(p,q) model, we proposed an interval prediction called V@R-Assymetric-GARCH(p,q) model to predict the lowest and the most significant change in oil price probable to α-significant. The Asymmetric-GARCH is the modification of the conventional GARCH(p,q) model with the asymmetry distribution, and V@R (Value-at-risk) is the interval-prediction measure that is more robust than the point prediction. At the end of this study, the prediction values are used to calculate the economic quantities under the production sharing contract (PSC) scheme using Gross-Split Methods. The results presented in this paper can help investors or company management make better decisions in evaluating the potential of oil or gas projects.