Optimal Dynamic Hedging Strategy of Energy Futures Markets: Application of Asymmetric Multivariate Models

• Main Authors: Wan-Ling Chao, 趙婉伶
• Other Authors: Teng-Tsai Tu
• Format: Others
• Language: zh-TW
• Published: 2008
• Online Access: http://ndltd.ncl.edu.tw/handle/4qw4zp

碩士 === 銘傳大學 === 財務金融學系碩士班 === 96 === This study proposes employing various asymmetric multivariate models to formulate optimal dynamic hedging strategy. To construct optimal dynamical hedging ratio, two types of the asymmetric models, namely the conditional correlation model and the Copula-GARCH model, are utilized in this study. The former explores static and dynamic correlation models, respectively, applying asymmetric constant conditional correlation (A-CCC) model with structural breaks and asymmetric generalized dynamic conditional correlation (AG-DCC) model to capture the asymmetric correlation process between futures and their corresponding underlyings. To simultaneously consider possible asymmetric volatility effects between futures and their corresponding underlyings, the GJR-GARCH model, asymmetric conditional autoregressive range model (A-CARR), and multiplicative error model for asymmetric realized volatility model (AMEM-RV) are further applied to estimate the volatility of futures and their corresponding underlyings. The latter Copula-GARCH model employs dynamic mixed copula (M-Copula) function to investigate the time-varying bivariate dependent structure since the dependent structure of spot and futures prices possibly exhibit asymmetric structure among spot and futures markets. To achieve the robustness of inference, we compare the hedging performance of the proposed asymmetric multivarite models with those of the symmetric conditional OLS models for energy spot and futures markets. The empirical results of this study indicate that using the A-CARR model to estimate volatility of futures in the short-term could obtain better hedging performance, while using the GJR-GARCH model in the medium and long-term. Under consideration of the dynamic basis effect, the hedging performance of conditional OLS model will be generally superior to that of the asymmetric conditional correlation models. However, if further consider the asymmetric structure of spot and futures correlation, the hedging strategy constructed with M-Copula function is relatively effective and flexible for energy futures, and also can reduce the risk with enormous fluctuation of energy prices.