摘要: | 本研究以經發會召開之前後,台股加權股價指數(以下稱股價指數)、台灣期貨指數(以下稱期貨指數)、台灣電子類股指數(以下稱電子指數)、台灣金融類股指數(以下稱金融指數)等類股之間的關聯性,並試著以計量方法來探索其之間是否具有共整合與因果關係,研究期間則自1998年7月24日至2006年12月29日的週資料,其中以2001年8月經發會召開期間為分界點,將本研究分為1998年7月24日至2006年12月29日(以下稱全區間),1998年7月24日至2001年8月31日(以下稱前半段),2001年9月7日至2006年12月29日(以下稱後半段),以此三段時間來做探討,實證結果歸納如下結論: 1、在單根檢定方面:所有序列在1%的顯著水準下皆無法拒絕單根之虛無假設,然而都拒絕一階差分序列有單根之虛無假設。因此,我們可以確定台灣股價指數、台灣期貨指數、電子類股指數、金融類股指數,其序列為相同整合級數I(1)的型態,符合隨機漫步的走勢,與預期相同。 2、在共整合檢定方面:(1)選取研究期間不論時間長短(全區間、前半段或後半段),各變數間除了股價指數與期貨指數之外,其餘股價指數與電子指數、股價指數與金融指數、期貨指數與電子指數、期貨指數與金融指數以及電子指數與金融指數彼此間均無共整合關係,顯示對投資人而言,並無法藉由這些指數間指標共移的特性來獲取超額報酬,也充份反應了期貨指數與股價指數的領先落後關係。 3、在誤差修正模型方面:由檢定結果可以得知,股價指數與期貨指數短期間確實存在失衡的現象,但長期上則會經由誤差修正項的調整而回到均衡的狀態,由誤差修正項的係數可發現,隨著各期變動的狀況,期指與股價指數正負向皆有之。 4、因果關係檢定方面:不管選取研究期間的長短,股價指數與期貨指數的因果關係較顯著,惟前半段的股價指數成為期貨指數的因較明顯,後半段則股價指數與期貨指數互有因果關係,即具有雙向回饋的影響;而其他金融指數、電子指數與期貨指數和股價指數的因果關係,則以落後期數拉長之後,較具顯著的變化。 The purpose of this study is to find out the relationships among the four indexes of pre-and-post Economic Advisory Conference (EDAC) : Taiwan Weighted Stock Index (TAIEX) , Taiwan Index Futures (TX), Taiwan Stock Exchange Electronic Sector Index and Taiwan Stock Exchange Finance Sector Index. We would try to investigate the Cointegration Test and Granger Causality Test by Econometric methodology. The data was from weekly data of July 24, 1998 to December 29, 2006. With the Economic Advisory Conference held in August 2001, this study divided the data into three sets: the whole data (July 24, 1998 to December 29, 2006), pre-EDAC (July 24, 1998 to August 31, 2001) and post-EDAC (September 7, 2001 to December 29, 2006) to examine. The empirical results are described as follows: 1. Unit root results: The null hypothesis of unit root can’t be rejected at the 1% significant level for all indexes series. The unit root analysis implies that the index series are individually integrated of order one, I(1). Therefore, we ascertain that the series of TAIEX, TX, Taiwan Stock Exchange Electronic Sector Index and Taiwan Stock Exchange Finance Sector Index are same as I(1) which matches the random walk. 2. Cointegration results: No matter which period we chose, the variables of TAIEX and TX are with cointegration relationship. It presents that the investors can’t gain the extra profits from the index co-movement as well as reflects the lead-lag relationship between two indexes. 3. Error correction results: TAIEX and TX exist the non-balance effect in the short run. However, after the adjustment from the error correction model, it becomes balance. 4.Granger Causality Test results: No mater which period we chose, it is significantly with Causality relationship between TAIEX and TX. While Taiwan Stock Exchange Finance Sector Index and Taiwan Stock Exchange Electronic Sector Index are more significant in Causality relationship while the number of lag is extended. |