兩岸三地股票市場外溢效應研究—VARBEKKMVGARCH模型的應用

近十年來，兩岸三地通過制度設計，不斷提升經貿金融整合程度，然而經貿金融整合對三者股市外溢效應產生的影響仍未知。有鑒於此，本研究根據兩岸三地經貿金融關係整合程度，將資料分爲兩個階段，分別爲2004年1月2日至2010年9月10日的鬆散階段（階段I），以及2010年9月13日至2019年11月29日的緊密階段（階段II）。本文對滬深300指數收益率、恒生指數收益率和臺灣加權指數收益率建立三元VAR-BEKK-MVGARCH模型，分別檢定兩個階段中股指收益率的均值外溢效應和波動外溢效應。實證結果發現：整體上，三者的股指收益率將透過經濟基本面管道和金融市場傳遞管道產生均值外溢效應和波動外溢效應，且在經貿金融關係整合程度不同的階段中，外溢效應的方向、持續性和程度有所差異。均值外溢效應方面。階段I，存在滬深300指數收益率對恒生指數收益率，以及恒生指數收益率對臺灣加權指數收益率的單向均值外溢效應。同時，經貿關係的日益緊密，強化了三者股指收益率的均值外溢效應。階段II中，首先，滬深300指數收益率與臺灣加權指數收益率之間出現雙向均值外溢效應。其次，均值外溢效應的持續性更強。第三，陸股和臺股間的相互影響均有提高。同時，港股受陸股的影響程度有所提高。最後，港股對臺股的影響程度因陸股對臺股影響程度的提高而降低，即出現陸股對港股均值外溢的替代效應。波動外溢效應方面。階段I，兩岸三地存在恒生指數收益率和臺灣加權指數收益率對滬深300指數收益率，以及臺灣加權指數收益率對恒生指數收益率的單向波動和衝擊傳遞效應。在兩岸三地加強金融市場合作後，波動外溢效應得到加強。階段II，首先，出現了滬深300指數收益率對恒生指數收益率和臺灣加權指數收益率的波動傳遞效應，即陸股的風險可以外溢至港臺股市。其次，臺灣加權指數收益率對滬深300指數收益率僅存在衝擊傳遞效應。本文的研究結論為金融監管當局進行金融風險管理以及金融市場的對外開放提供了一定的政策參考，同時，也有助於投資人進行投資決策。

關鍵字

股市；外溢效應； VAR-BEKK-MVGARCH

並列摘要

In recent ten years, the Chinese mainland, Hongkong and Taiwan have continuously improved the integration of economy, trade and finance through system design. However, the impact of the integration on the spillover effect of these three stock markets is still unknown. In view of this, according to the integration levels among Chinese mainland, Hongkong and Taiwan, we define January 2, 2004 to September 10, 2010 as loose phase (Phase I) and September 13, 2010 to November 29, 2019 as tight phase (Phase II). In this study, we set up the ternary VAR-BEKK-MVGARCH model to test the mean spillover effect and the volatility spillover effect of return rate of CSI300 Index (CSI300), Hang Seng Index (HSI) and Taiwan Capitalization Weighted Stock Index (TAIEX) in these two phases respectively. The empirical results show that, in general, the return rates of the three indexes will generate the mean spillover effect and the volatility spillover effect through the economic fundamental channel and the financial market contagion channel. And the spillover effect has different directions, continuity and degree in different phases. In terms of the mean spillover effect, in Phase I, there exists an one-way mean spillover effect of the return rate of CSI300 on the return rate of HSI and the return rate of HSI on the return rate of TAIEX. Simultaneously, the increasingly close economic and trade relations strengthen the average spillover effect of the three stock index returns. In Phase II, the two-way mean spillover effect, first of all, appears between the return rate of CSI300 and the return rate of TAIEX. Second, the mean spillover effect is more persistent. Third, the mutual influence between the Chinese mainland stock market and the Taiwan stock market is improved. At the same time, the degree of impact of Chinese mainland stock market on Hong Kong stock market increase. Finally, the degree of impact of Hong Kong stocks on Taiwan stocks decreases with the increase of the degree of impact of mainland stocks on Taiwan stocks, that is, there is a substitution effect of the Chinese mainland stock market on the mean spillover of the Hong Kong stock market. In terms of the volatility spillover effects, in Phase I, there exists an one-way fluctuation and impact conduction effect of return rate of TAIEX on the return rate of HSI and return rate of TAIEX on the return rate of CSI300. With the strengthening of financial market cooperation between the Chinese mainland, Hong Kong and Taiwan, the volatility spillover effect has been enhanced. In Phase II: First, the fluctuation conduction effect of the return rate of CSI300 on the return rate of HSI and the return rate of the TAIEX appears, that is, the risk of the Chinese mainland stock market can spill over to Hong Kong and Taiwan stock markets. Second, the return rate of TAIEX has only an impact conduction effect on the return rate of CSI300. The conclusions of this paper provide some policy implications for financial regulatory authorities to manage financial risks and open up the financial market to the outside, and also help investors to make investment decisions.