Essays in Housing Markets and Finance

Abstract

In my research, I have tried to integrate spatial heterogeneity into new asset pricing theories of housing, emphasizing that the value of a house as an investment also depends on its location. In particular, I apply asset pricing tools and theory to the study of housing prices in a spatial framework, thereby identifying and quantifying sources of risk and returns and how they interact with location in housing markets. In the first chapter of my thesis, I focus on documenting stylized facts about the spatial distribution of returns to housing. In the second and third chapters of my thesis, I explore the drivers of housing returns by focusing on the roles of idiosyncratic housing price risk and liquidity. Finally, I examine how systematic regional differences in housing risk explain the growing regional gap in housing prices. In “Superstar Returns? Spatial Heterogeneity in Returns to Housing”, which is co- authored with Moritz Schularick, Martin Dohmen, and Sebastian Kohl, we document the spatial distribution of total returns to housing and provide supporting evidence for its drivers. To conduct this analysis, we have assembled a new long run city-level dataset covering annual house prices and rents in twenty-seven prominent (”superstar”) cities across fifteen OECD countries over the past 150 years. Our data reveals that, over the long run, superstar cities have experienced lower total returns on housing in comparison to other regions within the same country. While house prices have grown more rapidly in these larger cities, the rental returns are significantly higher in more remote locations, resulting in overall higher returns in other parts of the country. We show that our key finding can be explained within a standard asset pricing framework, where excess returns outside large cities serve as compensation for higher risk. We then test this mechanism empirically and find that housing investments are indeed riskier outside large cities. In the second chapter of my thesis, I use detailed transaction-level data to quantify the extent to which idiosyncratic risk impacts housing prices and returns. In Price Uncertainty and Returns to Housing, I present empirical evidence that residential properties with higher idiosyncratic price risk are, on average, sold at lower prices and yield higher total returns. I show that this result can be rationalized within a bargaining model, in which a risk-averse and non-diversified buyer faces future sales price uncertainty. Finally, I present empirical evidence that houses with higher idiosyncratic risk undergo a more uncertain trading process, thereby exposing their buyers to greater liquidity risk. In the third chapter of my thesis I explore in more depth the relation between liquidity, location and housing prices. In Urban Spatial Distribution of Housing Liquidity, which is co-authored with Mark Toth and Jonas Zdrzalek, we examine how location, liquidity and prices interact in housing markets. By combining real estate online listings data with transaction data, we introduce a novel dataset that pro- vides transaction-level measures of liquidity in large German cities over the past decade. Empirically, we find that both housing liquidity and prices decrease with distance to the city center. To explain our empirical findings, we build a spatial search model of a housing market within a monocentric city. We show qualitatively and quantitatively that increasing travel costs to the city center can explain the joint urban spatial distribution of prices and liquidity. Using our calibrated model, we structurally estimate a spatial liquidity premium gradient. In the fourth and final chapter of my thesis, I analyse the consequences of heterogeneity in housing risk across regions. In “Interest Rates and the Spatial Polarisation of Housing Markets”, which is co-authored with Moritz Schularick, Martin Dohmen, and Sebastian Kohl, we reexamine the causes of regional housing price inequality. We build a spatial housing valuation model to demonstrate how a fall in real interest rates at the national level disproportionately affects the valuation of housing in regions with lower housing risk.