Does market sentiment push up China’s housing prices? - An empirical study based on the data of 45 mainstream cities in China

This paper reports a series of empirical tests that scrutinised the potential effect of market sentiment on China’s housing prices at the city level. The analyses employed unbalanced panel data from 45 large- and medium-sized cities in China for the period of 2011 to 2017. We first constructed the housing market sentiment index using principal component analysis, following which the index was applied in a system GMM estimation to analyse its impact. The results of the dynamic GMM estimation indicated that market sentiment plays a significant role in increasing housing prices in the 45 selected cities. Subsequently, general fixed effect regressions, placebo tests, and Poisson regressions were performed to test the robustness of the dynamic GMM estimation results. All the robustness checks confirmed the positive impact of market sentiment on housing prices. Additionally, we investigated the moderating effects of mobile network coverage, wage rate, and education on the positive relationship between market sentiment and housing prices. It was revealed that mobile network coverage has a positive moderating effect on this link, while wage rate and education have negative moderating effects. Lastly, this study explored the heterogeneity of market sentiment’s effect on housing prices, concluding that although this effect is positive in both first- and second-tier cities, it is significantly stronger in first-tier cities. The research findings are useful for the Chinese government in regulating housing prices by stabilising market sentiment.