指數基金上市以來,學者以不同的股價指數的資料研究期貨、現貨與指數基金三者的價格發現關係,結論多半是期貨的價格領先於現貨;而指數基金的部分則因為資料及型態不同的關係,無法得出明確的價格領先與落後的關係。因此本研究僅針對指數基金與大盤股價指數,而以美國S&P 500指數和指閟基金SPDRs為樣本,進行價格發現的研究。
本研究以2004年1月2日至2006年4月18日的交易資料來進行研究,共578筆的日資料,以共整合檢定、Granger因果關係檢定驗證其價格發現關係,再使用衝擊反應函數分析當指數基金或大盤股價發生變動時,彼此被影響的程度。最後再利用預測誤差變異數分解來探討當數列發生變動時,有多少比例是從自身的衝擊所產生,及多少比例是由另一變數所產生。
結果顯示兩者之間具有共整合關係,顯示兩者長期之間價格會維持相同的方向,在Granger因果關係的檢定上,以時差落後六期,兩者之間具有互為領先落後的關係。在預測誤差變異數分解方面,不論衝擊來自於S&P 500指數或是SPDRs,S&P 500指數其可以解釋的變異程度幾達90%。
Many researchers has used different ETFs data to research the price discovery rela-tionship between ETFs market index. Most conclusions are the price of futures lead to market index. Because of the different sample and ETFs, the ETFs didn’t have exactly price discovery relationship. The article focuses on S&P 500and SPDRs to research price relationship. The research period is during Jan.2 2004 to Apr.18 2006.there are 578 daily data except for holidays.
We used the method of cointegration test and Granger causality test to exam for their price discovery relationship. And then we use the impulse response function to analyses which the ETFs and the market index fluctuating each other. The last , we use the variance decomposition to analyses the same thing as above.
The conclusion is the cointegration relationship in S&P 500 and SPDRs. It maens that their prices will go the same direction in the long run. The Granger causality test results show the price discovery relationship is a feedback relationship with six lag pe-riods. The result of variance decomposition can interpret about 90%variance of the S&P 500index, whether the variance came from S&P 500 index and SPDRs.