房價結構性改變之檢測—以台北縣、市房價為例

Abstract

本文利用Bai and Perron（1998）多次結構性改變檢測法對台北縣、市房價進行\\r 檢測，檢測結果得出台北縣、市房價在1981年至2001年皆發生三次結構性改變，其\\r 中台北市房價結構性改變之發生時點為『1988年第2季』、『1994年第1季』、『1998年\\r 第1季』，結構性改變信賴區間為『1987 年第4季~1988 年第3季』、『1993年第3季\\r ~1996年第2季』、『1996年第4季~1998年第4季』。台北縣房價結構性改變時點則為\\r 『1988年第3季』、『1992年第3季』、『1998年第4季』，結構性改變信賴區間為『1988 年第1季~1988年第4季』、『1991年第3季~1994年第4季』、『1998年第3季~1999年第 2季』。此結果明顯推翻了過去房價在1981年至2001年期間只發生一次結構性改變之 說法。此外，台北縣房價結構性改變之時點大致是落後台北市房價結構性改變的時 點，但在第二次結構性改變時則領先台北市，這說明儘管兩縣市的房地產市場關係 密切，卻可能隨時間的改變而導致兩者房價結構性改變時點有所差異。

This paper applied the multiple structural changes testing models of Bai and Perron (1998) to detect housing prices of Taipei city and Taipei county. We found that there were three structural changes in both Taipei city and Taipei county during 1981 to 2001. This result broke the saying of only one structural change in Taiwan real estate market after 1980. The structural changes times of Taipei city were 1988Q2, 1994Q1, and 1998Q1, and their confidence intervals were 1987Q4 to 1988Q3, 1993Q3 to 1996Q2, and 1996Q4 to 1998Q4. The structural changes times of Taipei county were 1988Q3, 1992Q3, and 1998Q4, and their confidence intervals were 1988Q1 to 1988Q4, 1991Q3 to 1994Q4, and 1998Q3 to 1999Q2. Although the structural change times of Taipei county often lagged behind Taipei city, it was not the truth for the second structural change. It implied the close relationship between real estate markets of Taipei city and Taipei county still would change over time.