中國上海證券綜合指數與香港恆生中國企業股指數期現貨之連動關係及避險效果之研究-重大事件及VEC DCC GJR GARCH模型與VEC Copula GJR-GARCH-skewed-t模型之應用

The Study on Correlation and Hedge Effect among China Shanghai Securities Composite Index, Hong Kong Hang Seng China Enterprises Index and Hong Kong Hang Seng China Enterprises Index Future: The Application of Major Effect, VEC DCC GJR-GARCH Model and VEC Copula GJR-GARCH Skewed-t Model

陳威蓁(Wei-Chen Chen)

指導教授 ： 劉祥熹

國立臺北大學/商學院/國際企業研究所/碩士(2011年)

摘要

本文以上海證券綜合指數、香港恆生中國企業指數與香港恆生中國企業指數期貨當作標的，探討兩地之三市場在股市報酬率上是否存在連動關係。並瞭解次級房貸危機與金融海嘯之重大事件對中國與香港金融市場相關性所產生蔓延效果(contagion effect)之影響並加以探討三市場的避險效果。樣本期間為2003年12月8日至2011年2月28日。

實證結果證明在次級房貸危機與金融海嘯期間，中國及香港股市報酬率之相關係數呈現提升趨勢，三市場報酬率之相關係數呈現提升趨勢， 顯示三市場報酬率與波動性的動態相關性是受到重大事件蔓延效果(contagion effect)所影響，而不完全是跨市場間訊息共移所造成的外溢結果(co-movement effect)。且避險效果方面，在重大事件期間，恆生中國企業股指數與恆生中國企業股指數期貨間的直接避險效果有提升的趨勢，而上海證券綜合指數與恆生中國企業股指數期貨之交叉避險效果有減少的趨勢，顯示在重大事件發生下，採取直接避險效果較佳。

此外，由VEC Copula GJR-GARCH-skewed-t模型實證結果可知上海證券綜合指數與恆生中國企業指數期現貨間具有右尾相依之相關性結果，即當市場呈現多頭市場時，同時上漲的相關性較高；恆生中國企業股指數現貨及其期貨間具有雙尾相依之相關性結構，即當市場呈現多頭或空頭市場時，兩市場一漲一跌的相關性較高。避險效果方面，Copula模型估出來的避險比率與績效，皆比VEC DCC GJR-GARCH模型來得高。

關鍵字

VEC DCC GJR-GARCH模型、VECCopula GJR-GARCH-skewed-t模型 ； 蔓延效果 ； 外溢效果

並列摘要

This study investigates the correlations among China Shanghai Securities Composite Index(SSCI), Hong Kong Hang Seng China Enterprises Index(HSCEI) and its Futures(HSIF) under the crisis of subprime mortgage and financial tsunami by using VEC DCC GJR-GARCH Model and VEC Copula GJR-GARCH Skewed-t Model. It also discusses the contagion effects of the crisis of subprime mortgage and financial tsunami on the China Shanghai Securities Composite index, Hong Kong Hang Seng China Enterprises Index and its Futures. The sample period of this study is from December 8, 2003 to February 28, 2011.
The empirical results obtaining from the VEC DCC GJR-GARCH model verify that during the crisis of subprime mortgage and financial tsunami period, the correlation coefficients between SSCI- HCEI,SSCI- HSIF and HCEI-HSIF have increased. The results also indicated that the return and volatility correlation of these three markets are affected by the crisis of subprime mortgage and financial tsunami(contagion effect), rather than simply cross-market information transmission through the volatility spillovers between any two markets as mentioned above. Moreover, the estimated results signify that the hedge ratio and hedge performance of HSIF to their cash markets have increased during the subprime mortgage and financial tsunami period. The strategies effects of direct hedge are more than that of indirect hedge.
In addition, the VEC Copula GJR-GARCH skewed-t model signifies the highly tail-dependency structure between SSCI-HCEI and SSCI-HSIF, and the double tail- dependency between HSCEI-HSIF. We also found that the market dependency between those any two markets have increased during the period of subprime mortgage crisis and financial tsunami. The hedge ratio and hedge performance estimated by this Copula Model are higher than those estimated in the VEC DCC GJR-GARCH Model.

並列關鍵字

VEC DCC GJR-GARCH Model ； VECCopula GJR-GARCH-skewed-t Model ； Contagion Effect ； Co-movement effect

參考文獻

吳莉芳(2003) ，人民網-CEPA對台灣的影響。

王凱立、陳美玲（2002）,「美國與台灣股票期現貨市場之動態關聯：一般化多變量GARCH模型的應用」,中央研究院經濟研究所經濟論文,30：4,363-407。

Akaike, H. (1969), “Fitting autoregressive models for regression”, Annals of the Institute of Statistical Mathematics, Vol. 21, pp.243-247.

Ang, A. and G. Bekaert (2002), “International Asset Allocation with Regimeshifts”, Review of Financial Studies, Vol. 15, No. 4, pp.1137-1187.

Bartram, S. M. and Dufey, G. (2001), “International portfolio investment: theory, evidence, and institutional framework”, Financial Markets, Institutions & Instruments 10 (3), pp.85-155