日經225股價指數與指數期貨報酬率之動態關係DCC-GARCH模型分析

DCC-GARCH Model for the Dynamic Relationship of the Nikkei 225 Stock Index and Futures Returns

10.30171/JYUST.200612.0003

王怡文(Yi-Wen Wang)；李世昌(Shi-Chang Lee)；李彥賢(Yen-Hsien Lee)

日經225 ； 動態條件相關 ； 多變量GARCH ； 衝擊反應函數 ； 回饋效果 ； Nikkei 225 ； DCC ； Multi-GARCH ； Impulse response function ； Feedback effect

元培學報

13期（2006 / 12 / 01）

21 - 34

繁體中文

本文利用Engle^1提出的動態條件相關多變量GARCH模型來探討由東京證交所推出的日經225股價指數報酬率以及大阪證交所與新加坡交易所推出的日經225股價指數期貨報酬率三者之間的動態關係，並且使用衝擊反應函數來分析三者之間的跨期動態效果。實證結果顯示，透過概似比例檢定可得知在模型配適上t分配優於常態分配，並且DCC-GARCH-t模型顯著優於CCC-GARCH、CCC-GARCH-t、及DCC-GARCH模型。此外，股價指數現貨與期貨報酬率之間皆存在正向回饋效果，其中大阪證交所之指數期貨報酬率對於其他兩者而言影響較大。在衝擊反應的分析中，非預期的OSE期貨變動對Nikkei 225股價指數有較大的衝擊，且非預期的Nikkei 225股價指數亦對OSE-Nikkei 225有較大的衝擊。综上可知Nikkei 225指數報酬率與OSE-Nikkei 225期貨報酬率不管在動態關係以及跨期動態效果上都有明顯的關係，往後若投資Nikkei 225股價指數時可加以應用彼此的關係來進行避險，達到降低投資風險的效果。

This paper investigates the dynamic relationship among the Nikkei 225 Stock Index and Futures trading on the Tokyo Stock Exchange (TSE), the Osaka Security Exchange (OSE), and Singapore International Monetary Exchange (SIMEX) using the Dynamic Conditional Correlation (DDC) multivariate GARCH model, proposed by Engle (2002). Moreover we use Impulse Response Function to analysis the intertemporal dynamic effect among the returns. The empirical results show that the t distribution is better than the normal distribution for the Nikkei 225, and that the DCC-GARCH-t model is better one for the Nikkei 225 than CCC-GARCH, CCC-GARCH-t, DDC-GARCH, and DCC-GARCH-t model, by Likelihood Ratio test. In addition, the returns of stock index and futures exist positive feedback effect with each other. Among that the return trading on OSE has more influence to others. In the Impulse Response Function analysis, we find that the responses of the return’s innovation trading on TSE are relatively stronger than the others.