台灣房價變動與總體經濟變數關係之實證研究

本研究主要探討總體經濟變數對台灣房地產價格之關聯性，其中台灣房地產價格包含台灣、台北市、新北市、桃園市、新竹市、台中市及高雄市，使用的研究變數有台灣及六大都會區信義房價指數、五大行庫房屋貸款利率、國內生產毛額、消費者物價指數及貨幣供給增長率，選用之研究變數均為季資料，共68筆季資料，資料期間為2001年Q1至2017年Q4。本研究透過單根檢定及共整合檢定分析，發現台灣及六大都會區房價指數、五大行庫房屋貸款利率、國內生產毛額、消費者物價指數及貨幣供給增長率變數間存在長期穩定之均衡關係，接著由向量誤差修正模型顯示，台北市、新北市及台灣房價指數受前一期本身房價有顯著的正向影響，新竹市及高雄市房價指數受前一期本身房價則有顯著的負向影響，表示各都會區之房價指數短期受到外在因素衝擊而偏離長期均衡時，會由本身的均衡誤差修正項與前一期本身的房價調整回長期均衡狀態。房屋貸款利率對台北市房價有顯著負向影響、高雄市為顯著正向影響，另國內生產毛額對新竹市房價有顯著負向影響，則桃園市與台中市房價指數與四項總體經濟變數影響均不顯著。最後使用因果關係檢定，分析研究房價指數與總體經濟變數間之因果關係，發現不同的經濟變數之變動，確實與各都市房地產價格波動呈現領先、落後或雙向回饋等不同對應關係，檢定結果顯示各變數間互動緊密，可作為推測未來趨勢之參考。因此，銀行可針對房地產價格波動做預測，作為控管授信風險之指標，管控銀行對不動產的授信額度，降低承做該產業之授信風險，穩定銀行健全發展。

關鍵字

房地產價格指數；貸款利率；總體經濟變數；單根檢定；共整合檢定；向量誤差修正模型；因果關係檢定

並列摘要

This study is base in Taiwan house prices index from Sinyi Realty Inc, retrieving 68 samples from 2001Q1 to 2017Q4. The research discusses how macro variables influence the real estate market and the connections and causal relations among them. Chosen variables are the Taiwan house prices index, Taipei City house prices index, New Taipei City house prices index, Taoyuan City house prices index, Hsinchu City house prices index, Taichung City house prices index, Kaohsiung City house prices index, loan rate, gross domestic product growth rate, consumer price index and M2 balance. This study was analyzed by unit root test and cointegration test, empirical results show that there is a long-term equilibrium relationship among Taiwan house prices index, six metropolitan areas house prices index, loan rate, gross domestic product growth rate, consumer price index and M2 balance. Then, the vector error correction model shows that the house price index of Taiwan and Taipei City have a significant positive impact on the house price of the previous period. the house price index of New Taipei City, Hsinchu City and Kaohsiung City have a significant negative impact on the price of the previous period. The impact indicates that the house price index of each metropolitan area is affected by external factors in the short-term and deviates from the long-term equilibrium. It will be adjusted back to the long-term equilibrium state by its own equilibrium error correction term and the previous period's own house price. The home loan interest rate has a significant negative impact on housing prices in Taipei, and Kaohsiung City has a significant positive impact. The gross domestic product has a significant negative impact on housing prices in Hsinchu City. The price index of Taoyuan City and Taichung City and the impact of four general economic variables are not significant. Finally, using the Granger causality test check, the causal relationship between house price index and economic variables is analyzed and different populations are found. The changes in economic variables do indeed have different correspondences with the trend of real estate prices in various cities, such as leading, backward or two-way feedback. The results of the verification show that the variables interact closely and can be used as a reference for speculating future trends. Therefore, the bank can make predictions on real estate price fluctuations, as an indicator to manage the credit risks, manage the bank's credit line for real estate, reduce the credit risk of undertaking the industry, and stabilize the sound development of the bank.