滬、港、台、美四地股市指數與區域經濟成長關聯性及共整合之研究─以中、港CEPA實施前後期為例

吳其定(Chi-Ting Wu)

指導教授 ： 周冠男

國立中央大學/管理學院/財務金融學系碩士在職專班/碩士(2007年)

摘要

目前許多研究中都指出東亞國家之間股市的連動性，但極少以區域經濟成長對股票市場的連動性進行研究，然而貿易市場與資本之場本應存在相互關係，特定區域經濟合作事件對各地股票市場造成之影響，應為投資大眾持續關注之議題，於是促成本研究之動機，希望就中國與香港實施CEPA前後針對兩岸三地與美國股市之間的共移現象進行檢定，觀察CEPA實施之後，兩岸三地與美國間的股票市場之關係是否產生改變。

本研究的主要範圍是滬、港、台、美四地股市，因此以上海證券交易所A 股指數、香港恆生國企指數、台灣證券交易所發行之發行量加權股價指數與美國道瓊工業指數為標的。在資料取樣時間部份，以CEPA正式實施執行的2004年1月1日為資料期間的分割點，分成第一階段為CEPA實施前，期間為2001年1月3日至2003年12月31日；第二階段為CEPA實施後，期間為2004年1月3日至2007年7月13日。變數計算所使用的資料為四地股市之日收盤股價指數以計算各地股市之日報酬率，再以日報酬率來作為計算之基礎。

本研究使用的研究方法主要採用VAR向量自我迴歸模型（Vector Autoregression Model）、Johansen共整合模型Granger與因果關係檢定（Granger Causality Test）等，來檢驗四地股市報酬率在CEPA實施前後的互動或關聯性是否有所改變。經由實證分析發現：首先，在敘述性統計可知道四地股票市場報酬率不符合常態分配假設，其次透過單根檢定確認股市之報酬率皆符合定態之數列。而根據Granger因果關係檢定，在CEPA實施前後最大的差異在於中國大陸的上海A股於CEPA實施後會受到道瓊工業指數與香港H股的影響，VAR分析的結果也呈現一致的結論，並以衝擊反應函數與預測誤差變異數分解檢視各股票市場受其他股票市場在不同時間點的影響程度。

本研究結論認為：CEPA的實施對於中國大陸上海A股與國際股市的連動性有助益，在CEPA實施後上海A股的報酬率受到道瓊工業指數報酬率前1期之影響，表示上海A股正逐漸與國際市場接軌，未來關係可能會更密切。香港與台股均為與國際股市較密切之股票市場，且台股的報酬會受香港H股之影響，兩者並非為一區隔市場。香港仍為兩岸三地股市間的主要影響者，因香港相較於台灣與大陸市場，包括資金流動、外資投資、外匯管制皆較為開放。而近年兩岸企業紛紛往香港上市的趨勢亦加深了香港股市對台灣及大陸股市的影響程度。而中國QDII制度的推出強化不同資本市場間資金的流動，值得後續研究持續關注。

關鍵字

股市指數關聯性

並列摘要

Currently, a lot of studies point out the joint movement of stock market between East Asian countries, but very few studies would regard the joint movement of regional economic growth towards stock market for research and study, but trade market and capital market should have interrelations, the influence of specific regional economic integration for the stock market of each country should be the topic that the investors continued to pay close attention to; hence, this promotes the motive of this study, which hopes to carry out the test that is focused on the co-movement phenomenon between the cross-Strait regions and the US stock market before and after China and Hong Kong implemented CEPA, so as to observe after the implementation of CEPA, the relationship between cross-Strait and US stock markets has produced any changes or not.

The main scope of this study is the stock markets of U.S.A., China, Hong Kong and Taiwan, so it is based on Shanghai Stock Exchange''s A share index, Hang Seng China Enterprises Index (HSCEI), Taiwan Weighted Stock Index that is issued by Taiwan Stock Exchange Corporation and Dow Jones Industrial Average (DJIA). For the time of information sampling, the division point of information time will be January 1st 2004 that CEPA has been officially implemented, the time is divided into two stages, the first stage is before the implementation of CEPA, the duration is from January 3rd 2001 to December 31st 2003; the second stage is after the implementation of CEPA, the duration is from January 3rd 2004 to July 13th 2007. The information that is used for variable calculation is the daily closing quotation stock indexes of the four regions’ stock markets in order to calculate the stock return for each region, and then the daily stock return is served as the calculation basis.

The study methods that this study adopt mainly includes Vector Autoregression Model, Johansen’s cointegration model, and Granger Causality Test, which are used to examine whether the interactions or relationships between the stock returns of the stock markets in the four regions have changed before and after the implementation of CEPA. After case studies analysis, the following findings have been discovered. First of all, from descriptive statistics one can know the market returns of the four stock markets did not conform to the assumption of normal distribution. Secondly, through Unit root test, one can confirm that the stock returns are all conformed to the numbers of the stationary. And according to Granger Causality Test, the biggest difference before and after the implementation of CEPA lies in Shanghai Stock Exchange''s A share index form Mainland China, it is affected by Hang Seng China Enterprises Index and Dow Jones Industrial Average after the implementation of CEPA, the analytical result of VAR also showed an unanimous conclusion, and Impulse Response Function and Variance Decomposition are used to view the degree of influence that each stock market is affected by the other stock market in different times.

The conclusion of this study believes that the implementation of CEPA is contributive to the joint movement between Mainland China’s Shanghai A Share Index and international stock market. After the implementation of CEPA, the rate of return of Shanghai A Share Index is influenced by the previous period’s rate of return of Dow Jones Industrial Average, that means Shanghai A Share Index is integrating with the world market gradually, the relationship may be closer in the future. Hong Kong stock and Taiwan stock are the stock markets that are closer to the international stock market, and the return of Taiwan stock will be affected by Hong Kong H-share, they are not in the same distinguishing market. Hong Kong is still the main influence within the cross-Strait regions’ stock markets, because the fund flows, the foreign capital investment, and the foreign exchange control in Hong Kong are all comparatively opened when comparing to Taiwan and China markets. And in recent years, the trend of enterprises of cross-Strait regions went to Hong Kong to list their enterprises one after another have also strengthened the influence of Hong Kong stock market towards Taiwan and China stock market. And the introduction of QDII system strengthens the flow of capital between different capital markets, which is worthy of the continuous attention by the follow-up studies.

並列關鍵字

CEPA

參考文獻

[6] 孫國祥，「東亞區域整合之演進與願景：經貿、金融與安全面向的探討」，遠景基金會季刊，6(6)，119-173頁，民國95年。

[10] 楊踐為，賴怡洵，「美、日、香港與臺灣四地股價指數連動關係之探討」，台灣土地金融季刊，35(2)，1-15頁，民國87年6月。

[13] Agmon, T., “The Relationship among Equity Markets: A Study of Share Price Comovements in the United States United Kingdom, Germany and Japan”, Journal of Finance, Vol.27 (4), pp.839-55, 1973.

[14] Akaike, H. “Markovian Representation of Stochastic Processes and Its Application to the Analysis of Autoregeressive Moving Average Process”, Annuals of the Institute of Statistical Mathematics, Vol.26, pp.363-387, 1973.

[15] Arshanpalli, B. and Doukas, J., “International Stock Market Linkages: Evidence from the Pre- and Post-October 1987 Period”, Journal of Banking and Finance, Vol. 17, pp.193-203, 1993.