Housing price Expectation and Propability to Sell in the Urban Housing Markets

A proportional hazard model is employed to exploit micro relationships between household housing price expectation and probability to sell (the latter is inversely related to the property owning spells). The primary data source is from 14 yearsí Singapore condominium transaction data. A homeownerís housing price expectation is measured by building-based housing price indexes derived from a two order spatio-temporal model. The results illustrate V-shaped threshold effects of housing price dynamics on probability to sell, meaning that a homeowner will have a higher probability to sell (by shortening his property owning duration) if his expected nominal equity gain from rising housing prices goes beyond a threshold and he will also have a higher probability to sell if his expected nominal equity loss from declining housing prices increases. The results also show that probability to sell increases faster if a homeowner experiences equity loss than if he experiences equity gain. The findings can be explained by both a homeownerís investment liking and his nominal loss aversion. Our results also imply that the observed positive aggregate price-volume relationship should be attributed to buyersí behavior.