Template-Type: ReDIF-Paper 1.0
Series: Tinbergen Institute Discussion Papers
Creation-Date: 2000-11-10
Number: 00-090/4
Author-Name: J.S. Cramer
Author-Email: cramer@tinbergen.nl
Author-Workplace-Name: University of Amsterdam
Title: Scoring Bank Loans that may go wrong: A Case Study
Abstract: A bank employs logistic regression with state-dependent sample selection to identify loans thatmay go wrong. Inspection shows that the logit model is inappropriate. A bounded logit model witha ceiling of (far) less than 1 fits the data much better.
File-Url: https://papers.tinbergen.nl/00090.pdf
File-Format: application/pdf
File-Size: 220514 bytes
Handle: RePEc:tin:wpaper:20000090