市場間的訊息傳遞效果可藉由跨市場波動性之關聯性來加以說明,若我們能精確的估計出市場間的關聯性,則可以有效的將此資訊運用在市場的交易行為上。本研究以Engle和Lee (1999) 提出之component GARCH模型為基礎,一方面進行台股指數與台股指數期貨市場波動性的長短期效果分析以及討論不對稱長短期效果的表現,此外我們亦分析非經濟事件對長短期波動性的影響,另一方面將component GARCH模型結合雙變數GARCH模型,進一步討論台股指數與台股指數期貨市場波動性長短期效果的關聯性。實證結果顯示,台股指數與台股指數期貨市場之長短期波動性表現顯著異於零,我們用波動性長期效果的估計參數來計算兩市場的半衰期,結果發現台股指數期貨市場波動性具有較快的均數回復特性;兩市場皆存在短期槓桿效果但不存在長期的槓桿效果;非經濟事件對兩市場短期波動性的影響不一致,其對兩市場長期波動性的影響有限;在跨市場長短期波動性的分析中,我們證實前期市場未預期的訊息會影響另一市場的條件變異數,此亦表示市場間具有訊息傳遞的效果;我們亦證明主要影響台股指數與台股指數期貨市場條件變異數的變數為過去累積的訊息流量;長期波動性只影響個別市場的條件變異數,其在跨市場的交互影響上表現並不顯著;跨市場條件共變異數的相關係數相當高,表示市場間確實存在相當程度的關聯性。
The volatility linkages across market are based on the relation between volatility and information flow. We apply to the component GARCH model by Engle and Lee (1999) for the TAIEX and TAIFEX markets. We also combine component GARCH model and bi-variate GARCH model, trying to discuss complete volatility linkages. The empirical results represent the trend and the transitory components are significant. And we also used the estimates of trend component to calculate the half-lives of two markets. We consider that the TAIFEX market has quick mean-reverting. The leverage term is significant in the short-run component for both of markets. The unexpected information flows will affect the cross conditional covariance. It shows that the volatility linkages between markets are really potent. We also find that the most important variable of the conditional variance equation is the accumulative information flow. Moreover the correlation coefficient of the conditional covariance is quite high. Given this, it can be inferred that highly linkages between TAIEX and TAIFEX markets are indeed strong.