Forecasting volatility and correlation between oil and gold prices using a novel multivariate GAS model

By Rongda Chen, Jianjun Xu. Energy Economics, 2019

75 citations recorded by the source corpus.

Forecasting the volatility and correlation among different kinds of assets has important applications in areas such as risk management, options pricing, and asset allocation. This paper mainly uses a novel multivariate Generalized Autoregressive Score (GAS) model to analyze and forecast volatilities and correlations between Brent, WTI and gold prices. The time-varying parameters of multivariate GAS model for a given distribution of crude oil and gold prices is observed which is supported by Doornik-Hansen test. The testing results of time-varying parameters based on LRT statistics reveal that the dependent structure between Brent and gold prices is more complex than those of WTI and gold. The estimation results show that the multivariate GAS method well captures the volatility persistence and nonlinear interaction effects between the crude oil and gold markets. In addition, we compare the forecasting performance of the GAS with the classical Dynamic Conditional Correlation Generalized Auto-Regressive Conditional Heteroskedasticity (DCC-GARCH) model, and find that the forecasting power of volatility and correlation in multivariate GAS model is better than the DCC-GARCH model.

Subfields: Economics and Econometrics, Renewable Energy, Sustainability and the Environment

View Paper: https://www.semanticscholar.org/paper/198b77f15204b16a62c31490c8e0d5251cec235b