本研究係以台灣股票市場之台灣加權股價指數、股票分類指數和台灣總體經濟指標為研究對象,研究期間為1999年1月至 2009年12月之月調整後收盤股價及經濟指標數據做作為研究標的。資料來源取自於台灣證券交易所、行政院經濟建設委員會與中央銀行等,並結合灰色預測模型原理與向量自我迴歸模型(VAR)運用在台灣股市與經濟體系中,嘗試將灰色預測模型GM(1,1)導入VAR模型中,目的在於了解台灣股票指數與總體經濟指標間之互動結構。 研究實證結果發現:消費者物價指數和匯率與台灣加權股價指數、股票分類指數間存在「因果關係」最為顯著;並依據AIC準則決定自我迴歸落後期數,實證發現股價指數領先經濟指標8個月,再利用Granger因果關係檢定、預測變異數分解及衝擊反應分析了解台灣股市指數和總體經濟指標間存在「互動結構」,實證發現台灣的經濟指標與股價指數有密切的互動結構。
The study takes the Taiwan Capitalization Weighted Price Index and Stock Sector Indexes from Taiwan stocks market and Taiwan marcro-economic indices as an example. The monthly closing stock indices and marcro-economic indices from January 1999 to December 2009 are sampled, which were adopted from the Taiwan Stock Exchange Corporation,Council for Economic Planning and Development and Central Bank. I apply GM(1,1) on VAR into a GVAR to realize the dynamic structure between Economic Indices and Taiwan Stock Market Indices. According to the empirical results, I find interest Consumer Price Index and exchange rate have a Granger causality relationship with Taiwan Capitalization Weighted Price Index and Stock Sector Indexes.Based on AIC rule, stock market indices is a leading index to economic indices for eight months. And using by Granger causality, decomposition variance and the impulse response analysis, we realize the existance of the dynamic structure between economic indices and stock market indices in Taiwan. And we discovered this dynamic structure is interacted frequently.