Statistical and Actuarial Sciences Publications

Document Type

Article

Publication Date

1-6-2021

Journal

The Journal of Credit Risk

Volume

16

Issue

4

First Page

119

Last Page

156

URL with Digital Object Identifier

https://doi.org/10.21314/JCR.2020.272

Abstract

The main aim of this study is to analyse the joint effects of customer segmentation, borrowers' characteristics and modelling techniques on the classification accuracy of a scoring model for agribusinesses. To this end, we used data provided by a Chilean company on 161,163 loans from January 2007 to December 2013. We considered random forest, neural network and logistic regression models as analytical methods. Regarding the borrowers' profiles, we examined the effects of socio-demographic, repayment-behaviour, agribusiness-specific and credit-related variables. We also segmented the customers as individuals, SMEs and large holdings. As the segments show different risk behaviours, we obtained a better performance when we estimated a scoring model for each segment instead of using a segmentation variable. In terms of the value of each set of variables, behavioural variables increased the predictive capability of the model by double the amount achieved by including agribusiness-related variables. The random forest is the model with the best classification accuracy.

Creative Commons License

Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.

Find in your library

Share

COinS