Payment history as a predictor of credit risk

Abstract

A numerical scoring system to predict credit payment was developed from the payment records of customers of a large Los Angeles department store. The following types of items were used: number of payments of less than amount due, number of payments missed completely, number of collection notices previously sent, etc. Results showed good (r bi = .71) ability to discriminate between potentially good accounts and those likely to require special collection efforts. (PsycInfo Database Record (c) 2025 APA, all rights reserved)