The Study of Optimal Hedge Ratios and Utility Based Approach to the Crude Oil Hedging Strategies Using Multivariate GARCH

Various GARCH models have been applied to the research of financial time series. For example, studies of Myers and Thompson (1989), Baillie and Myers (1991) and Lien et al. (2002), and Chang, McAleer, and Tansuchat (2011) apply GARCH models to improve the research of optimal hedge ratios. Although multivariate conditional volatility models are often employed in the measurement of minimum variance hedging strategy, they are rarely applied to the research which involves both the optimal hedge ratio and the preference of investors. This dissertation evaluate four different multivariate conditional volatility models, they are the Constant Conditional Correlations (CCC) model, the Dynamic Conditional Correlations (DCC), the BEKK model, and the Diagonal BEKK model. These models are applied to the calculation of the optimal hedge ratio and the utility based hedge ratio of Brent and WTI crude oil markets. The main aim of this research is to examine whether the performance of hedge ratios under different multivariate volatility models could be employed by investors who are not infinitely risk averse. It is concluded that the diagonal BEKK model exhibits as the best model in terms of the measurement of hedging effectiveness, and the BEKK model appears to be the worse one among the four models measured in terms of the Brent crude oil market. Moreover, estimates of the utility based hedge ratio suggest that the DCC model and the diagonal BEKK model exhibit a similar level of utility based hedge ratios in terms of the Brent and WTI crude oil data series that examined in this dissertation. However, the utility based hedging approach suffers from several limitations and the ex-post utility measure may fail to provide better insights to the selection of hedging strategies. It should be noted that there are still other multivariate GARCH models or different utility functions which can be applied in the future research of utility based optimal hedge ratios (OHRs). In addition, the method illustrated in this dissertation can not only be used in the estimation of crude oil data but also be applied to other financial series such as commodities or securities. This would be of particular interest to those who involved in the management or hedging strategies of portfolio investment and risk management