台北市不動產景氣循環週期之頻譜分析

Abstract

本文研究方法採用頻譜分析(Spectral Analysis)，探討台北市房地產景氣循環週期，並從台北市2003年至2017年間各樣本變數資料，進行頻譜連動性 (Spectral Coherence)分析，利用所估計的頻譜連動性及相位(phase)，來分析各個變數間具有關聯的週期和領先與落後關係的情況，最終整理找出頻譜主頻率，以及「領先指標」與「落後指標」來作為本文之研究結果。 實證研究結果發現，台北市2003年至2017年為區間，頻譜主頻率為第15頻，週期以12個月為主，也代表著台北市房地產波動較台灣整體的為快速。 領先指標與落後指標分別如下:(1)領先指標:市立圖書館外借圖書冊、交通類指數、刑事發生率、居住 類指數、食物類指數、消費者物價總指數、教養娛樂類指數、新公司登記設立家數、新商業登記設立家數、電影戲劇觀賞人次、臺北市營造立家數、新商業登記設立家數、電影戲劇觀賞人次、臺北市營造工程物價指數、醫藥保健類指數。(2)落後指標:人口密度、人口增加率、戶量、出生率、平均每日捷運人次數、社會增加率、空氣污染測定、粗結婚率。

This research studies the business cycle of Taipei city’s real estates via spectral analysis. We measured and analyzed the parameters from Taipei city’s statistics during the period from 2003 to 2017 through spectral coherence analysis. The analysis is assessed the cycles associated with parameters and the lead-lag relationship. The results are major cycle, leading indicator and lagging indicators. The empirical results show that the city of Taipei from 2003 to 2017 is the interval, the major cycle of the spectrum is the 15th frequency, and the cycle is mainly 12 months, which also represents the rapid fluctuation of real estate in Taipei than in Taiwan. The leading indicators and the lagging indicators are as follows:(1)Leading indicators: numer of borrowing books in municipal library, transportation&communication index, criminal incidence rate, housing index,food index, consumer price index, education&entertainment index, the number of new companies registered, the number of new business registered, visitor count for movie and drama, Taipei city construction cost index,and medicines & medical care.(2)Lagging indicators: population density, increase rate of population, household numbers, birth rate, Taipei metro system daily ridership, social increase rate, air pollution measurement, and crude marriage rate.