Title: Non-parametric testing of information asymmetry in the U.S. mortgage servicing market
Authors: Helmi Jedidi - HEC Montreal (Canada) [presenting]
Georges Dionne - HEC Montreal (Canada)
Abstract: The main objective is to test for evidence of information asymmetry in the U.S. mortgage servicing market. The main research question is: does selling the mortgage servicing rights by the initial lender to a second servicing institution unveil any residual asymmetric information? We investigate the link between the originators decision to sell the mortgage underlying MSR and the mortgage default likelihood using a large sample of U.S. mortgages that were securitized through the private-label channel during the 2000-2013 period. Our econometric methodology is purely non-parametric. We use Kernel Density Estimation technique to estimate the multivariate conditional density function. Our findings support the presence of a second-stage asymmetric information in the U.S. mortgage servicing market. Our empirical results show a significant positive relationship between the lenders decision to sell the underlying MBS and mortgage default. Our evidence suggests that originating lenders are indeed taking advantage of privileged information they obtain at the time of original underwriting. We also use parametric tests to corroborate our results after controlling for observable risk characteristics, econometric misspecification error, and endogeneity issues using instrumental variables approaches and simultaneous equations.