房價指數及水泥、鋼鐵類股指數對台灣營建類股價指數影響之研究

Abstract

本文主要區分營收面房價指數，成本面鋼鐵、水泥類股指數，市場供需面因素（如貨幣數量與放款利率變動）探討如何影響營建類股指數，以期提供營建類股有效投資資訊。 因本文變數以時間序列資料為主而可能存在隨機趨勢，是以本文先對變數做單根檢定，將非定態變數處理成定態變數，再對各解釋變數對營建類股指數作共整合檢定探討長期趨勢、因果關係檢定確立變數間領先落後關係、誤差修正模型探討變數對營建類股指數短期影響，最後提出誤差修正模型上顯著變數建立複回歸模型以探討各種因素對營建類股指數的影響。本文實證結果發現水泥類股指數、H型鋼價呈正向長期趨勢、放款利率呈負向長期趨勢，且營建類股為各項解釋變數領先指標但與H型鋼價成同步指標。營建類股指數則受到前期水泥類股指數、H型鋼價呈正向短期變動調整，受到前期房價指數、房地產負擔率、M2貨幣數量、放款利率呈反向短期變動調整，最後複回歸實證結果發現，落後一期水泥類股指數對營建類股指數呈現正向顯著影響、落後兩期M2貨幣數量、落後一期放款利率對營建類股指數呈現反向顯著影響。

This thesis mainly distinguishes the revenue of housing price index, the cost of steel, cement stock index, market supply and demand factors(such as quantity of money and lending rate) to explore how to influence the construction of stock index, in order to provide information on investment in construction investment. Due to the fact that the variables in this paper are based on time-series data and may have stochastic trend, so we employ the unit root test and make them stationary. Then, explored the long term trend between each interpretation variables and construction index by Cointegration test, explored the leading or backward indicator between each interpretation variables and construction index by Granger Causality test, explored the short-term impact of variables on the construction stock index by Vector Error Correction Model, and finally proposes significant variables on the error correction model to establish a multiple regression to explore various factors for the impact of construction stock index. The result of our empirical research which is based on statistical methods are as follows as: (1) In the cointegration results, we find that Construction stock index, steel prices of H type and Cement stock index have positive long-term trends, Construction stock index and lending rate have negative long-term trend. (2) In the Granger Causality results, we find that Construction stock index synchronizes steel prices of H type, and leads rest of interpretation variables. (3) In the Vector Error Correction model results, we find that changes of the construction stock index is inversely affected by two-period-lag of Housing price index and M2 quantity of money, and positively affected by one-period-lag of cement stock index, steel price of H type, and real estate burden rate. (4) In the multiple regression results, we find that one-period-lag of cement stock index has positive influence on construction stock index, and one-period-lag of lending rate and two-period-lag of quantity of M2 have negative influence on construction stock index.