Predicting Loss Severities for Residential Mortgage Loans: A Three-step Selection Approach

This paper develops a novel framework to model the loss given default (LGD) of residential mortgage loans which is the dominant consumer loan category for many commercial banks. LGDs in mortgage lending are subject to two selection processes: default and cure, where the collateral value exceeds the outstanding loan amount. We propose a three-step selection approach with a joint probability framework for default, cure (i.e., zero-LGD) and non-zero loss severity information. The proposed methodology demonstrates improved performance in out-of-time predictions compared to widely used OLS regressions.

[1]  João F. Cocco,et al.  A Model of Mortgage Default , 2011, The Journal of Finance.

[2]  Christophe Mues,et al.  Modelling LGD for unsecured personal loans: decision tree approach , 2010, J. Oper. Res. Soc..

[3]  Daniel Roesch,et al.  Downturn Credit Portfolio Risk, Regulatory Capital and Prudential Incentives , 2009 .

[4]  Steven Finlay,et al.  Multiple classifier architectures and their application to credit risk assessment , 2011, Eur. J. Oper. Res..

[5]  Kyle F. Herkenhoff,et al.  Unemployment, Negative Equity, and Strategic Default , 2013 .

[6]  Michael LaCour-Little,et al.  Risk Based Capital Requirements for Mortgage Loans , 2001 .

[7]  Dan Immergluck,et al.  The external costs of foreclosure: The impact of single‐family mortgage foreclosures on property values , 2006 .

[8]  J. Crook,et al.  Loss given default models incorporating macroeconomic variables for credit cards , 2012 .

[9]  Esa Jokivuolle,et al.  Incorporating Collateral Value Uncertainty in Loss Given Default Estimates and Loan-to-Value Ratios , 2003 .

[10]  Daniel Rösch,et al.  Cure events in default prediction , 2014, Eur. J. Oper. Res..

[11]  Jonathan Crook,et al.  Forecasting and explaining aggregate consumer credit delinquency behaviour , 2012 .

[12]  Ralph S. J. Koijen,et al.  Determinants and Consequences of Mortgage Default , 2010 .

[13]  Emily Johnston Ross,et al.  What Drives Loss Given Default? Evidence From Commercial Real Estate Loans at Failed Banks , 2015 .

[14]  Jonathan N. Crook,et al.  Recent developments in consumer credit risk assessment , 2007, Eur. J. Oper. Res..

[15]  Edward Kung,et al.  Estimates of the Size and Source of Price Declines Due to Nearby Foreclosures , 2013 .

[16]  Vincent W. Yao,et al.  Spillover Effects of Foreclosures on Neighborhood Property Values , 2009 .

[17]  Robert Jarrow,et al.  Default Parameter Estimation Using Market Prices , 2001 .

[18]  Yongheng Deng,et al.  Risk-Based Pricing and the Enhancement of Mortgage Credit Availability among Underserved and Higher Credit-Risk Populations , 2005 .

[19]  Lyn C. Thomas,et al.  Modelling LGD for unsecured retail loans using Bayesian methods , 2015, J. Oper. Res. Soc..

[20]  Sudheer Chava,et al.  Modeling the Loss Distribution , 2011, Manag. Sci..

[21]  Jonathan Crook,et al.  Enhancing two-stage modelling methodology for loss given default with support vector machines , 2017, Eur. J. Oper. Res..

[22]  Vincent W. Yao,et al.  Foreclosure externalities: New evidence , 2015 .

[23]  Andrea Resti,et al.  The Link between Default and Recovery Rates: Theory, Empirical Evidence and Implications , 2003 .

[24]  Gene Amromin,et al.  Comparing Patterns of Default Among Prime and Subprime Mortgages , 2009 .

[25]  B. Baesens,et al.  Benchmarking regression algorithms for loss given default modeling , 2012 .

[26]  Christophe Mues,et al.  The economy and loss given default: evidence from two UK retail lending data sets , 2014, J. Oper. Res. Soc..

[27]  Bart BaesensRudy.  Using Neural Network Rule Extraction and Decision Tables for Credit-Risk Evaluation , 2003 .

[28]  D. Glennon,et al.  What 'Triggers' Mortgage Default? , 2010 .

[29]  T Bellotti,et al.  Credit scoring with macroeconomic variables using survival analysis , 2009, J. Oper. Res. Soc..

[30]  Yuliya Demyanyk,et al.  Understanding the Subprime Mortgage Crisis , 2008 .

[31]  Patrick Bajari,et al.  An Empirical Model of Subprime Mortgage Default from 2000 to 2007 , 2008 .

[32]  Sumit Agarwal,et al.  Thy Neighbor’s Mortgage: Does Living in a Subprime Neighborhood Affect One’s Probability of Default? , 2012 .

[33]  Uday Rajan,et al.  The Failure of Models that Predict Failure: Distance, Incentives and Defaults , 2010 .

[34]  Daniel Rösch,et al.  Forecasting probabilities of default and loss rates given default in the presence of selection , 2012, J. Oper. Res. Soc..

[35]  Harald Scheule,et al.  Accuracy of mortgage portfolio risk forecasts during financial crises , 2016, Eur. J. Oper. Res..

[36]  Robert E W Hancock,et al.  Collateral damage , 2014, Nature Biotechnology.

[37]  Christophe Mues,et al.  Mixture cure models in credit scoring: If and when borrowers default , 2012, Eur. J. Oper. Res..

[38]  Yanan Zhang,et al.  Local Housing Market Cycle and Loss Given Default: Evidence from Sub-Prime Residential Mortgages , 2010, SSRN Electronic Journal.

[39]  John M. Quigley,et al.  Loan Loss Severity and Optimal Mortgage Default , 1993 .

[40]  Parag A. Pathak,et al.  Forced Sales and House Prices , 2009 .

[41]  W. Greene.  Sample selection in credit-scoring models1 , 1998 .

[42]  Anthony Pennington-Cross,et al.  Subprime and Prime Mortgages – Loss Distributions , 2003 .

[43]  Min Qi,et al.  Loss given default of high loan-to-value residential mortgages. , 2009 .

[44]  Jon Frye,et al.  Collateral damage detected , 2000 .

[45]  Sebastián Maldonado,et al.  Cost-based feature selection for Support Vector Machines: An application in credit scoring , 2017, Eur. J. Oper. Res..

[46]  C. Mues,et al.  Predicting loss given default (LGD) for residential mortgage loans: A two-stage model and empirical evidence for UK bank data , 2012 .

[47]  Michael J. Seiler,et al.  Mimetic Herding Behavior and the Decision to Strategically Default , 2014 .

[48]  Jonathan Crook,et al.  Dynamic consumer default models , 2010 .

[49]  Jun Zhu,et al.  The Effect of Mortgage Payment Reduction on Default: Evidence from the Home Affordable Refinance Program , 2015 .

[50]  Christophe Mues,et al.  A zero-adjusted gamma model for mortgage loan loss given default , 2013 .

[51]  T. N. Herzog,et al.  The Effect of State Foreclosure Laws on Loan Losses: Evidence from the Mortgage Insurance Industry , 1990 .

[52]  Eric Rosenblatt,et al.  Efficient Mortgage Default Option Exercise: Evidence from Loss Severity , 1995 .

[53]  Timothy J McQuade,et al.  How Do Foreclosures Exacerbate Housing Downturns? , 2019, The Review of Economic Studies.

[54]  Yong Shi,et al.  Credit risk evaluation using multi-criteria optimization classifier with kernel, fuzzification and penalty factors , 2014, Eur. J. Oper. Res..

[55]  Ke Yin,et al.  Negative Equity Trumps Unemployment in Predicting Defaults , 2010, The Journal of Fixed Income.

[56]  Leif B. G. Andersen,et al.  Extensions to the Gaussian Copula: Random Recovery and Random Factor Loadings , 2005 .

[57]  Dennis L. Hoffman,et al.  An econometric analysis of the bank credit scoring problem , 1989 .

[58]  Benjamin Bade,et al.  Default and Recovery Risk Dependencies in a Simple Credit Risk Model , 2010 .

[59]  Soner Akkoç,et al.  An empirical comparison of conventional techniques, neural networks and the three stage hybrid Adaptive Neuro Fuzzy Inference System (ANFIS) model for credit scoring analysis: The case of Turkish credit card data , 2012, Eur. J. Oper. Res..

[60]  Fredrik W. Andersson,et al.  Loss Severities on Residential Real Estate Debt during the Great Recession , 2014 .

[61]  Harald Scheule,et al.  A Theoretical and Empirical Analysis of Alternative Discount Rate Concepts for Computing LGDs Using Historical Bank Workout Data , 2017 .