Rationality test in the housing market: Project-level evidence from China

We test whether investors rationally react to rent news when pricing houses. Previous literature on house pricing uses aggregated housing price and rent data to estimate the rent news. We analytically demonstrate that using aggregated data biases the rationality test because it forces researchers to estimate the rent news using the cross-project average dynamic relationship among the housing return, rent growth, and macroeconomic dynamics and, hence, misses the heterogeneity in the dynamic relationship. Our empirical tests using proprietary project-level housing data from China's megacities find that the bias from missing the heterogeneity is economically substantial.