原油價格波動受國際政經影響甚劇，針對原油價格波動進行避險已成為投資人的主要課題之一。由於原油價格與期貨價格可能皆會因稀少事件的發生而存著價格不連續現象。本研究先利用Chan(2003)提出的雙變量CBP-GARCH模型捕捉價格不連續的變動及現貨報酬與期貨報酬的共變異數關係。本研究以2010年至2011年英國布蘭特原油價格為主要研究對象，利用移動視窗（rolling window）法探討樣本外（out of sample）條件最小變異數避險組合之避險效益，比較未避險模型、雙變量GARCH(1,1)模型與雙變量CBP-GARCH(1,1)模型的條件最小變異數避險組合之避險效益。研究發現雙變量GARCH(1,1)模型與雙變量CBP-GARCH(1,1)模型存在著條件最小變異數避險組合之避險效益，且雙變量CBP-GARCH(1,1)模型較雙變量GARCH(1,1)模型的避險效益更好，因雙變量CBP-GARCH(1,1)模型能捕捉資產價格間動態跳躍與動態波動性，因而利用其條件最小變異數避險組合可得到較佳的避險效益，此結果可為投資人避險之參考。 The international political and economic effect the crude oil price volatility dramatically. One of the main topics is hedging for the crude oil price volatility of the investors. Crude oil spot and futures prices exist to discontinuously depend on rare events occurred. In order to capture the dynamic price jump and covariance between spot and futures returns, we use Chan(2003) to address bivariate the CBP-GARCH model. The discussions on this paper are using rolling window to investigate the out-of-sample hedging effectiveness for the minimum variance hedging portfolio. The data period probes Brent oil spot and futures price using daily data for the time span 2010 to 2011. The empirical results show that the bivariate GARCH (1,1) model and the bivariate CBP-GARCH (1,1) model have hedging effectiveness for minimum variance hedging portfolio. Moreover, hedging effectiveness of the bivariate CBP-GARCH (1,1) model better than the bivariate GARCH (1,1) model. The bivariate CBP-GARCH (1,1) model is able to capture the dynamic jump between the asset price volatility and dynamic correlation, thus the bivariate CBP-GARCH (1,1) model obtain is the better hedging effectiveness for minimum variance hedging portfolio. The results can be reference for investors.
