Generalized additive modeling of the credit risk of Korean personal bank loans

Young-Ah Kim, Peter G. Moffatt, Simon A. Peters

Research output: Contribution to journalArticlepeer-review

12 Downloads (Pure)

Abstract

We analyze consumer defaults in a sample of 64 000 customers taking personal loans from a Korean bank. Applying a generalized additive modeling (GAM) framework, we show a nonlinear impact of loan and borrower characteristics. In particular, the likelihood of default is high for both low-income borrowers and high-income borrow-ers. Our results are robust to a range of different tests, and they highlight the useful-ness of the GAM framework, especially the graphical presentation of nonlinearities.

Original languageEnglish
Pages (from-to)77-103
Number of pages27
JournalJournal of Credit Risk
Volume18
Issue number3
Early online date26 Aug 2022
DOIs
Publication statusPublished - Sep 2022

Keywords

  • Generalised additive models; B-spline; credit scoring; loan defaults; signal detection theory; mis-classification costs
  • misclassification costs
  • generalized additive models
  • signal detection theory
  • credit scoring
  • loan defaults
  • basis splines (B-splines)

Cite this