Estimating Estate-Specific Price-to-Rent Ratios in Shanghai and Shenzhen: A Bayesian Approach

The price-to-rent ratio, a common yardstick for the value of housing, is difficult to estimate when rental properties are poor substitutes of owner-occupied homes. In this study we estimate price-to-rent ratios of residential properties in two major cities in China, where urban high-rises (estates) comprise both rental and owner-occupied units. We conduct Bayesian inference on estate-specific parameters, using information of rental units to elicit priors of the unobserved rents of units sold in the same estate. We find that the price-to-rent ratios tend to be higher for low-end properties. We discuss economic explanations for the phenomenon and the policy implications.