本文以美國證券交易所(AMEX)掛牌的QQQ(追蹤Nasdaq 100 指數)與iShare EWT(追蹤MSCI台灣指數)兩檔ETF,選取QQQ自1999年3月10日至2003年5月30日止,以及iShare EWT自2000年6月25日至2003年5月30日止的日資料,探討ETF與股價指數間的價格發現與市場整合過程。 實證結果如下:根據共整合模型顯示,ETF與股價指數存在一共同長期趨勢,兩者形成共整合系統,相互參考其彼此的價格而調整當期的價格變化。而透過誤差修正模型的研究結果,則發現ETF與股價指數兩者皆具有價格發現的功能:在QQQ與Nasdaq 100指數中,是ETF領先股價指數,但在iShare EWT與MSCI台灣指數中,則是股價指數領先ETF,而衝擊反應分析亦證實了如上的結論,兩者並在七期(日)後即歸於收斂,具有正向關係。此外,由誤差變異數拆解可知,來自於股價指數的新訊息是預測誤差變異數的主要來源。 This study investigates the process of price discovery and market cointegration between ETFs and spot indices. We use daily closing price data of QQQ during 1999/03/10 ~ 2003/05/30 , and iShare EWT during 2000/06/25 ~ 2003/05/30 listed in American Exchange (AMEX). The major models applied are Vector Error Correction Model, Impulse Response Analysis and Variance Decomposition. The following conclusions have been drawn from empirical studies. 1. A long run relationship exists between ETF and Spot indices, but not for QQQ and Nasdaq 100 index or iShare EWT and MSCI Taiwan index. Two markets contribute a cointegration system, adjust their prices as a result of each other. 2. In the short term, QQQ tends to lead Nasdaq 100 index and iShare EWT tends to fall behind MSCI Taiwan index. Both samples possess price discovery functions and tend to converge after a seven-day period.