Sample Selection Bias in Mortgage Market Credit Risk Modeling

摘要: The mortgage crisis that started in the U.S. 2007 and lasted until 2009 was characterized by an unusually large number of defaults on subprime market. As a result, it developed into global economic recession placed stability world banking system jeopardy. Therefore, issues credit risk modeling showed shortcomings current practice. Truncation, or partial observability, simultaneous equations bias causes sample selection bias. parameter estimates are biased inconsistent. Firstly, we provide overview approaches literature to control for correction, such as Heckman model bivariate probit with selection. Secondly, review most significant studies discussing this problem is introduced. Specifically, different structural models, specific datasets empirical results regarded. In addition, discuss key determinants borrower characteristics, terms contract, macroeconomic conditions. Finally, conclude discussion possible research questions.