中古屋及預售屋房價指數之建立、評估與整合 ─台北市之實證分析Developing, Assessment and Integration of the Exist and Presale Housing Price Indexes—The Case of Taipei City

本研究首先觀察目前市場較常使用之信義房價指數(中古屋)及國泰房地產指數(預售屋及新成屋)，發現近年來信義房價指數無法適當反應房地產市場景氣繁榮的情形，而此二房價指數領先落後的關係亦不符合相關理論及過去實證。本研究利用政府(房地產交易價格簡訊)及業者(台灣不動產成交行情公報)資料以特徵價格法重新建立台北市中古屋價格指數，並與信義房價指數進行比較，結果發現在房價上漲或下跌幅度較多的時期，以及長期指數時間數列的尾端，信義房價指數對房價波動的反映幅度均顯著小於本研究的中古屋價格指數，顯示從長期使用的角度來看，以pooling model編製的房價指數極易受到特徵屬性價格的限制而產生結構性誤差，這也回應了為何近幾年信義房價指數的波動大幅小於市場平均價格的波動的原因。
最後，為求未來房地產資訊之應用，本研究將過去較具公信力之價格指數與本研究所建之房價指數銜接，並驗證國泰房地產指數領先本研究整合之中古屋價格指數一季，顯示此二價格指數相互間關係符合相關理論推導與過去實證結果。
關鍵詞：房價指數、特徵價格、領先落後、因果關係

We discuss the Sinyi housing price index (existing housing) and the Cathay real estate price index (new construction), and find that the Sinyi HPI poorly reflects the recent boom in the housing market. The Granger causality test of both indices is inconsequential to relevant theory and empirical studies. We then reestablish the new existing HPI for Taipei based on data compiled by the government and brokers. In comparison to the Sinyi HPI and the new HPI, the new HPI is more coherent with correlative market information. Finally, for future applications of housing market information, we integrate past empirical studies and examine the Sinyi HPI’s leadlag elationship with the Cathay index. The Granger causality test fits the relevant theory and empirical studies and, furthermore, we find that there is a one-season time gap between the Cathay index and the integrated existing HPI.
Key words: housing price index (HPI), hedonic price, lead-lag, Granger causality