Determinants of real estate liquidity : a comparison of houses and apartment units in Hong Kong

Abstract

Both houses and apartments are part and parcel of the housing market. In general, houses are bigger in size, more exclusive, and located in lower density areas. At the same time, compared to apartments, houses are arguably more heterogeneous, ranging from less expensive village houses to luxury houses. In this dissertation, houses with a similar unit price to apartments’ are selected for study, in order to investigate and compare the liquidity of houses with that of apartments. By comparing 50 house developments with 50 apartment developments, it is found that the former (18%) has a lower turnover rate than the latter (21%). This confirms the intuition that houses are less liquid, though the difference seems to be small. A deeper question is whether the liquidity of houses and apartments is driven by the same set of factors. This is answered by cross-sectional analysis and time-series analysis. First, the cross-sectional analysis examines how different property attributes affect the liquidity of houses and apartments. The literature have proposed different determinants of liquidity such as capital constraints (Stein, 1995) and loss aversion (Genesove and Mayer, 2001). However, none of these studies has differentiated houses from apartments. Based on the literature, property price and price appreciation are two of the property attributes to be examined. Furthermore, two other determinants, namely building age and estate size, are also included to test Akelof’s (1970) lemon theory and Haurin’s (1988)’s atypicality theory. The empirical results confirm that building age, a proxy of information asymmetry, significantly reduces the liquidity in both markets, and the negative effect is slightly bigger on houses. However, estate size, a proxy of typicality, has an unexpected negative effect on the liquidity of houses, though its effect is insignificant for apartments. There is insufficient evidence that price, taken as a proxy of affordability, affects the liquidity of apartments more than that of houses. Similarly, price appreciation is insignificant in both markets, which may be due to the small sample size. Next, a time-series analysis is carried out to observe the response of the liquidity of houses and apartments to the policy changes in 2010. Results suggest that transaction volume and price had a positive correlation for both houses and apartments before 2010, and such correlation has sharply reduced after 2010, especially for apartments. This implies that the cooling policy implemented by the government in 2010, including the introduction of Special Stamp Duty and the reduction of maximum mortgage ceiling, has imposed a greater capital constraint on buyers. In particular, buyers of apartments have experienced a more significant impact of these policies than the buyers of houses. This dissertation has revealed an unexplored area of the residential property market – the liquidity of houses. With the strong repelling force of houses manifested in the pandemic, more investors would be interested in this mysterious area. Moreover, people have focused more on property liquidity under an unstable market atmosphere. Therefore, this dissertation could help investors achieve a better asset portfolio by knowing more about the determinants of liquidity and the difference in the impact of determinants on both markets. Furthermore, the government could take it as a reference to review the performance of the cooling policy. As the house experienced a smaller impact from the cooling policy, the government could reconsider whether a different policy should be implemented if she aims to suppress the price of houses as well.