本研究重點放在台灣總體經濟指標變數與電子類加權股價指數之探討,分析彼此間是否存在著關聯性。由於景氣指標的發布,目的是反應目前經濟的狀況並能同時預測未來,且經濟的繁榮與衰退直接或間接地影響大眾投資的決策和意願,兩者間存在的關聯性為本研究所欲深入探討之主題。 本文採之台灣景氣指標中之總體經濟變數,包括SEMI半導體接單出貨比(B/B值)、海關出口值及匯率等重要指標,透過向量自我迴歸模型 (VAR) 分析以及 Granger 因果關係檢定等方法進行研究,探討所選定之變數自2001年1月至2012年12月期間,其與電子類股指數之領先落後情況,再進一步利用衝擊反應分析以及預測誤差變異分解,以了解各變數之間的影響互動關係。 經本研究實證結果顯示:台灣景氣指標中所選之變數皆會受電子類股指數報酬之影響,且由 Granger 因果關係檢定得知,電子類股指數報酬對於其他三者變數之影響均為領先之關係,而匯率與海關出口值有互相回饋之關係,可相互預測。由預測誤差變異數分析亦再次得知電子類股指數報酬本身解釋力強,其他三者變數則容易被其他變數干擾。電子類股指數的確可作為市場之先行指標。
This study investigates the relationship between the stock index return and the economic indicator variables, with a focus on the electronic stock index in TAIEX. Given that economic indicators are used to measure the current state of the economy and to predict the prosperity of the future economy as well, they affect investment decisions directly or indirectly and in turn have impacts on the stock markets. The economic indicator variables considered in the empirical study of the thesis include SEMI Semiconductor book-to-bill (B/B) ratio, the value of exports and the foreign exchange rate, sampled from January 2001 to December 2012. The relation between electronic stock index return in Taiwan and the indicator variables are constructed through the vector autoregression (VAR) model, followed by the Granger causality test, the impulse response analysis and the forecast error variance decomposition method to further illustrate the interactions among the variables across time. The empirical results of this study show that the dynamics of the electronic index returns are highly correlated with the economic indicator variables incorporated. The Granger causality test suggests that the stock returns granger cause the exchange rates, the value of the export, and the B/B ratio during the sample period. Also, a feedback relation is identified between the exchange rate and exports. Furthermore, the impulse response analysis indicates that the indicator variables have negative responses to the shock from the stock returns except the exchange rate. The variance decomposition analysis finds that the variance of stock index returns can be explained by itself for over 95% consistently across time while the exchange rate has an increasing explanatory power over that of the export as time passes. The study concludes that the stock index serves as a leading indicator properly.