Designing a proper model and software program to evaluate and predict credit risk of small and medium‑sized enterprises in commercial banks

Globalization and, consequently, intensifying the competition between banks and financial institutions in domestic and foreign financial markets increase significantly the importance and requirement of strengthening and modifying systems in financial enterprises. Banks are no exception. Credit risk assessment is one of the most significant components of the granted facilities process. Small and medium enterprises are the majority of customers of commercial banks; hence, it is possible that designing a credit risk system considerably helps banks to manage credit risk.This study aims to introduce a new approach to predict and assess the credit risk of small and medium-sized enterprises. To this end, we identified the indices effective on the credit risk of medium and small-sized enterprises and determined significant indices by selecting the feature. We selected 98 cases of medium and small-sized legal clients in the industrial sector for research data from one of the commercial banks during the years 2018-2020. We then implemented the logit regression models, artificial neural network, and hybrid model (fuzzy expert system, logit, and artificial neural network) in order to predict and assess customers' credit risk and also calculated the accuracy of the models. Ultimately, we have designed the program applying visual studio software. We calculated the customer using logit regression models, artificial neural network, and hybrid model according to the designed program by inserting each customer's information and we also determined credit rating and type of collateral of each customer based on customer risk.