Can housing liquidity help forecast subsequent house price appreciation: Evidence from the US and the Netherlands

Abstract

The literature has documented and rationalized a positive correlation between volume of sales and appreciation rates in the housing market. Moreover, home appreciation rates have shown to be persistent and predictable over time. In this study, we test the predictive power of variables that measure market tightness on future home prices. A stylized search-and-matching model is used to illustrate that indicators that measure market tightness, such as sale probabilities and seller's bargaining power, can be associated with future home price appreciation. The empirical analysis uses Multiple Listing Services data from the Netherlands and from Fairfax County, VA, that contain all residential units offered for sale through a real estate broker over a 15 year period. The individual records are used to construct quarterly aggregate measures of housing conditions in about 40 regions in the Netherlands and in 41 zip codes in Fairfax County. Besides home price indices, the indicators include an index that measures seller's bargaining power and the (quality adjusted) probability that a home sells in less than 2 weeks. Conventional time-series models are then used to show that observed changes in sale rates and bargaining power can significantly reduce home price appreciation forecast errors.