A credit scoring decision support system

Credit scoring models have become a standard tool for banks in assessing the credit worthiness of loan applicants. The aim of a credit scoring model is to establish the probability for a client with certain characteristics to default on a loan, i.e. to run into problems when repaying debt obligations. The paper presents a credit scoring decision support system based on a logistic regression model. Its specific feature is that it defines the financial indicators of the analyzed loan applicant as random variables whose values are simulated. The proposed decision support system envisages a larger number of simulations. The obtained results are used to determine the confidence interval for the mean probability of default, which is actually the basis for loan applicant assessment.