Bound and collapse Bayesian reject inference for credit scoring
Authors:
- 1 October 2012
- journal article
- Published by Taylor & Francis in Journal of the Operational Research Society
- Vol. 63 (10) , 1374-1387
Abstract
Reject inference is a method for inferring how a rejected credit applicant would have behaved had credit been granted. Credit-quality data on rejected applicants are usually missing not at random (MNAR). In order to infer credit-quality data MNAR, we propose a flexible method to generate the probability of missingness within a model-based bound and collapse Bayesian technique. We tested the method's performance relative to traditional reject-inference methods using real data. Results show that our method improves the classification power of credit scoring models under MNAR conditions.Keywords
This publication has 17 references indexed in Scilit:
- Credit Scoring and the Availability, Price, and Risk of Small Business CreditJournal of Money, Credit and Banking, 2005
- Does reject inference really improve the performance of application scoring models?Journal of Banking & Finance, 2004
- Sample selection bias in credit scoring modelsJournal of the Operational Research Society, 2003
- Bank lending policy, credit scoring and value-at-riskJournal of Banking & Finance, 2003
- Predictive Performance of the Binary Logit Model in Unbalanced SamplesJournal of the Royal Statistical Society: Series D (The Statistician), 1999
- An Efficient Semiparametric Estimator for Binary Response ModelsEconometrica, 1993
- Evaluating interventions with differential attrition: The importance of nonresponse mechanisms and use of follow-up data.Journal of Applied Psychology, 1993
- An econometric analysis of the bank credit scoring problemJournal of Econometrics, 1989
- Limited-dependent and qualitative variables in econometrics1983
- Sample Selection Bias as a Specification ErrorEconometrica, 1979