Measuring the effect of characteristics of loan obligors on default risk in a portfolio of individual loans using generalized linear models
A large part of bank profits is eamed from the interest charged on the different loans they advance to their customers; therefore proper management of credit risk is crucial for commercial banks. This paper studies the characteristics of loan obligors in a portfolio of individual loans in a German Bank and how these characteristics contribute to the individual's probability of default. The study mainly seeks to determine the cross sectional dependence of default rates on various characteristics of loan obligors. The findings of this paper will enable banks to build credit scoring models based on the characteristics of loan obligors in order to better manage credit risk. Eighteen characteristics are considered in this study but only the five most significant attributes are analysed. The research found that loan obligors with the highest probability of default have the following set of characteristics: male, have other payment plans, have been employed for a longer period, have low loan repayment duration and have a low checking status.