Abstract: This paper builds the theoretical model of the effects of real estate sentiment indices on housing prices. It draws lessons from the methods of constructing the investor sentiment index in capital market developed by Baker and Wurgler(2006), and selects suitable sentiment indicators from both sides of supply and demand. Meanwhile, this paper uses Baidu index as big data sentiment indicator. We use principal component analysis (PCA) to calculate real estate sentiment indices in China's 35 major cities, and further conduct empirical analysis of the impacts of market sentiment on urban housing prices. Additionally, this paper measures real estate sentiment indices in different areas, and horizontally compares the influences of market sentiment on housing prices through empirical analysis. The empirical results indicate, firstly market sentiment has a significant positive impact on real estate prices, moreover, the positive effects of market sentiment on real estate prices demonstrate regional heterogeneity, ie., the positive effects of market sentiment on real estate prices in the eastern regions are stronger than those in the central and western areas.

Key words: market sentiment , housing prices , sentiment index , principal component analysis (PCA)

摘要： 基于行为金融学理论建立了房地产情绪指数影响房价的模型，借鉴Baker and Wurgler（2006）提出的构建资本市场投资者情绪指数的方法，从供需两方选取合适的情绪指标；同时，运用大数据方法，选取百度指数作为情绪指标，运用主成分分析法构建了我国35个大中城市房地产市场的情绪指数，实证研究了市场情绪对房地产价格的影响。另外，还测度不同区域房地产市场情绪指数，通过实证分析横向对比东部和中西部城市市场情绪对房价的影响程度。实证结果表明，市场情绪对房地产价格有着显著的正向影响；市场情绪对房地产价格的正向影响呈现区域异质性，东部地区市场情绪对房地产价格的正向影响强于中西部地区。

关键词: 市场情绪 , 房地产价格 , 情绪指数 , 主成分分析