Multivariate Analysis of Price Volatility in Petroleum Products Using the DCC and BEKK-GARCH Models

Abstract

The basic idea to extend univariate models to multivariate is its significance in predicting the dependence in the co-movements of asset returns in a portfolio. The main purpose of this paper was to model the dependence in the co-movement of petroleum products in Ghana using Dynamic Conditional Correlation (DCC) and Baba Engle Kraft and Kroner (BEKK) GARCH models and determine whether BEKK or DCC should be preferred in practical application. The results revealed that though the BEKK suffers from the archetypal curse of dimensionality whereas DCC does not. It was found that the models were adequate since the coefficient of both models were significant. Also, it was found that there was a co-movement in the petroleum products. The result also shows that BEKK was preferred to DCC in optimal model for the estimating conditional covariance regardless of whether targeting was used.