We build an on-the-house-search model and show analytically that the rentto-price ratio (or rental yield) and turnover rate, which are frequently used metrics for the housing market, are jointly determined in equilibrium. We therefore adopt a simultaneous equation approach on matched sale-rental pairs in our empirical investigation, as a housing unit cannot be owner-occupied and renter-occupied at the same time. Our empirical results confirm a higher turnover rate is associated with a lower rent-to-price ratio, as predicted by the model. Furthermore, our results suggest a form of "dichotomy" in the empirical determinants of rental yield and turnover at the real-estate-development (RED) level: the demographic structure, and past return performance affect its turnover rate, while popularity, human capital environment, mortgage burden, and long run rent growth determine the rental yield. The robustness of our results are established through a series of tests. In addition to these findings, our tractable search-theoretic model, a ranking of more than 130 RED in Hong Kong based on the popularity index we construct, and the estimated brandpremium for different major real estate developers may also carry independent research and practical interests.