Asymmetric Information about Collateral Values

I empirically analyze credit market outcomes when competing lenders are differentially informed about the expected return from making a loan. I study the residential mortgage market, where property developers often cooperate with vertically integrated mortgage lenders to offer financing to buyers of new homes. I show that these integrated lenders have superior information about the construction quality of individual homes and exploit this information to lend against higher quality collateral, decreasing foreclosures by up to 40%. To compensate for this adverse selection on collateral quality, nonintegrated lenders charge higher interest rates when competing against a better-informed integrated lender. What is the impact on equilibrium credit market outcomes when competing lenders are differentially informed about the expected return from making a loan? I analyze this question by studying lender competition in the residential mortgage market. Empirical studies of the impact of asymmetric information in credit markets usually face the challenge of identifying the relative information of the different parties (Petersen and Rajan (1994), Ausubel (1999), Karlan and Zinman (2009)). I address this challenge by focusing on the competition between lenders to originate mortgages used to purchase newly developed properties. In this market, property developers regularly provide home buyers with financing offers through vertically integrated mortgage lenders (Agarwal et al. (2011), Gartenberg (2011)). These integrated lenders have potentially better information than nonintegrated lenders about two factors that affect the expected return from making a mortgage: the value of the house used to collateralize the mortgage, and characteristics of the borrower.1 For example, ∗Stroebel is at the New York University Stern School of Business. I am indebted to Caroline Hoxby, Monika Piazzesi, Martin Schneider, and John Taylor for their encouragement and guidance. I thank the Editor, Michael Roberts, and two anonymous referees for constructive comments and helpful feedback. Seminar participants at Stanford, Kellogg, Wharton, Princeton, MIT, Harvard Business School, UCLA Anderson, Chicago Booth, MIT Sloan, LSE, LBS, Michigan, Berkeley Haas, NYU Stern, Stanford GSB, Cornell, Bonn, Oxford, EUI, and the Chicago Fed as well as conference participants and discussants at the NBER Summer Institute, Wisconsin HULM, and the Philadelphia Fed Workshop on Consumer Credit and Payments provided insightful comments. I thank Trulia, Buildfax, Sumit Agarwal, and Chenxi Luo for providing data. Financial support through SIEPR and the Hoover Institution is gratefully acknowledged. I thank Miguel de Faria e Castro for excellent research assistance. 1 The existing literature on information asymmetries in mortgage lending focuses on information about characteristics of borrowers, such as their income prospects (e.g., Keys et al. (2010), Elul DOI: 10.1111/jofi.12288