本文運用多元指數模型探討總體因素與產業基本面因素和台灣電子類及金融類股指數的關連性。選取的總體因素包括國際因素的美國重貼現率、那斯達克股價指數、道瓊工業股價指數,貨幣政策的M1b餘額及M2餘額,財政政策的政府支出及公債餘額,金融市場因素的消費者物價指數、匯率、商業本票利率、放款利率。產業基本面因素包括電子業營業收入、製造業營業額、放款金額、逾放比率。另外,引進虛擬變數驗證重大事件發生對電子類股及金融類股指數的衝擊。研究期間自1995年至2007年,資料型態為月資料。實證結果如下: (一)貨幣餘額(M1b、M2)、匯率、國外因素(那斯達克股價指數或道瓊工業股價指數)對電子類股指數與金融類股指數均發揮顯著影響。電子業營業收入對電子類股指數發揮顯著影響;前一期金融類股指數對金融類股指數發揮顯著影響。 (二)引進交叉效果後,短期重大事件與匯率的交叉項對電子類股指數與金融類股指數均發揮顯著影響。那斯達克股價指數除直接影響電子類股指數,美國聯準會執行貨幣政策調整重貼現率與那斯達克股價指數交叉效果,對電子類股指數發揮顯著負向影響;政府支出亦會透過電子業營業收入而對電子類股指數產生正向影響,顯示擴張性財政政策對電子業是有幫助。公債餘額和長期放款利率交叉效果對金融類股發揮顯著影響,顯示公債餘額上升,資金需求上升,長期利率上升,有助於提升金融業盈餘,促使股價指數上揚。 (三)使用最完整且解釋能力最高的模型預測電子類股價指數及金融類股價指數,發現當市場有劇烈波動時,預測值和實際值差距就會變的很大,若市場屬於小幅波動時,預測值和實際值差距就會很小。
To investigate the correlations among the macroeconomic factors, the industrial fundamental factors, Taiwan Stock Exchange Electronics Index and Finance Index, the Multiple Index Model is employed in this thesis. The macroeconomic factors consist of four variables, which are related with the international phase (U.S. rediscount rate, Nasdaq Index, and Dow Jones Index), the monetary policy phase (Monetary Aggregate M1b and M2), the financial policy phase (expenditure and bonds outstanding for government), and the financial market phase (consumer price index, foreign exchange rates, commercial paper rate, and loan rate). The industrial fundamental factors includes the electronic industry revenue, the manufacturing revenue, the loan balance, and the over due ratio. Besides, the dummy variables are used to exam the impact of big events on Electronics Index and Finance Index. The monthly data are collected since 1995 to 2007. The empirical results are described as follows. 1. M1b and M2, the foreign exchange rates, Nasdaq Index, and Dow Jones Index affect Taiwan Stock Exchange Electronics Index and Finance Index significantly. Besides, the electronic industry revenue is another significant factor on Taiwan Stock Exchange Electronics Index. Furthermore, The earlier stage of the Taiwan Stock Exchange Finance Index affects the current financial index significantly also. 2. Considering the joint effect in regression model, the cross term of big event in short term and the foreign exchange rate affects for both Taiwan Stock Exchange Electronics Index and Finance Index significantly. Besides, the cross term of U.S. rediscount rate, which represents the U.S. Fed’s monetary policy, and Nasdaq Index has a negative impact on Taiwan Stock Exchange Electronics Index significantly. In addition, the cross term of the government expenditure and the electronic industry revenue has a positive impact on Electronics Index. That implies the expanded financial policy of government is helpful to the electronic industry. Finally, the cross term of the government bonds outstanding and long-term loan rate has a significant impact on Finance Index. It implies that the money demand increases if the government bonds increase. Thus, the long-term loan rate increases to a high level. That benefits the finance industry and stimulates Financial Index to rise. 3. Using the best model to predict Taiwan Stock Exchange Electronics Index and Financial Index shows that the difference between the predicted value and actual value becomes large if the market volatility varies tremendously. On the contrary, the difference would be small if the market volatility varies little.