Generalised additive modelling of the credit risk of Korean personal bank loans

Peter Moffatt, Young-Ah Kim, Simon Peters

Research output: Contribution to journalArticlepeer-review


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 non-linear impact of loan and borrower characteristics. In particular, the likelihood of default is high for both low income borrowers as well as high income borrowers. Our results are robust to a range of different tests, and highlight the usefulness of the GAM framework, especially the graphical presentation of non-linearities.
Original languageEnglish
JournalJournal of Credit Risk
Publication statusAccepted/In press - 13 Sep 2021


  • Generalised additive models; B-spline; credit scoring; loan defaults; signal detection theory; mis-classification costs

Cite this