Essays on Real Estate Finance

Abstract

The three chapters of this dissertation investigate housing cycles and the residential housing investment in the U.S. The first chapter presents the non-monotonic relationships between the housing cycles and housing supply elasticities. Chapter Two explains this puzzling phenomenon by building up a dynamic model and emphasizes the roles played by the housing investors. The last chapter, coauthored with Wenli Li, explores how residential housing investment could predict the home price dynamics both in the short run and long run.

The first chapter systematically summarizes a puzzling phenomenon that occurred during the recent U.S. housing cycle in which areas with relatively elastic supplies, such as Phoenix and Las Vegas, experienced a more pronounced cycle than areas with the most elastic and inelastic supplies. My empirical analysis further highlights that the fraction of investment home purchases (i.e., non-owner-occupied homes) during the housing boom of 2004-2006 is also significantly higher in the middle elasticity areas and that the fraction of investment home purchases is significantly correlated with several key aspects of the housing cycle across different counties, such as the housing price increase during the boom of 2004-2006, the "over-building" in 2004-2006 that is unexplained by local economic fundamentals, and the housing price drop during the bust of 2007-2009.

Then in the second chapter I explain all these observations by developing a stylized model based on an interaction between time-to-build, an important attribute inversely associated with supply elasticity, and extrapolative home investors with a downward sloping demand curve. This model predicts that during a housing cycle, the housing price increase and over-building during the boom and the housing price drop during the bust all exhibit non-monotonic relationships with respect to supply elasticity. I empirically confirm these non-monotonic relationships across different counties during the recent housing cycle.

The last chapter looks into the predictability power of residential housing investment on the home price dynamics using comprehensive micro data on mortgage applications and performance. Across different zip codes, investment home purchases display significantly positive predictability for home prices within one or two years but significantly negative predictability for home prices after three to five years.