本研究旨在建立一個可預測未來房地產景氣的指標，提供一個包含多個房市面向的參考指標。參考 Stock and Watson (1998) 提出的擴散指數 (Diffusion Index) 方法，本文選擇了多個與房地產市場相關的變數，並透過主成份分析 (Principal components analysis) 將這些變數合成一個綜合指標，以代表整體房地產市場的景氣狀況。同時，本研究也借鑑了國發會發布景氣對策燈號的方式，將合成的房地產景氣指標以燈號的形式呈現，使景氣狀況更加直觀可視化。最後，本研究檢驗了該指標的預測準確度，將樣本期間切割為訓練集和測試集。結果顯示，該指標對未來六個月的預測表現最佳，呈現出相對較高的預測能力。本研究成果的目的在於提供一個客觀的房地產景氣資訊，以供使用者在決策過程中獲得更多的參考依據。;The aim of this study is to establish a predictive indicator for future real estate cycle, providing a comprehensive reference indicator that encompasses multiple aspects of the housing market. Drawing on the Diffusion Index methodology proposed by Stock and Watson (1998), this research selects various variables related to the real estate market and employs Principal Component Analysis to synthesize these variables into a composite indicator that represents the overall state of the real estate market. Additionally, this study adopts the approach of using signal lights, similar to those used by the National Development Council, to visually present the synthesized real estate market indicator, enhancing the intuitiveness and clarity of the market conditions. Finally, the study examines the predictive accuracy of the proposed indicator by dividing the sample period into training and testing sets. The results demonstrate that the indicator performs best in forecasting the upcoming six months, exhibiting relatively higher predictive capability. The objective of this research is to provide objective information on real estate cycle, offering users additional reference points during the decision-making process.
