Predicted Repayment Behaviour of Borrowers in Urban Cooperative Banks in Assam: A Case Study of Nagarik Samabay Bank

Parnakshi Bayan, Kandarpa Kumar Barman

University of Science and Technology

The purpose of this study is to analyse the predicted repayment behavior of sample borrowers of Nagarik Samabay Bank, an Urban Cooperative Bank in Assam for the period 2023-24 by employing the Artificial Neural Network (ANN) model. The results indicate that eleven socio-economic factors were statistically significant in predicting the repayment behavior of sample borrowers, with the ANN model achieving a predicted default classification accuracy of 83.3%. However, in comparison, the logistic regression (LR) model demonstrated higher predictive performance, correctly classifying defaulters with an accuracy of 88.2%, thereby outperforming the ANN model. However, the ANN model predicted regular borrowers more accurately than logistic regression, achieving an accuracy of 98.9% compared to 97.6%. Based on these findings, it is suggested that financial institutions, particularly Urban Cooperative Banks, adopt robust risk assessment measures to strengthen their position in the banking sector.

Article link: https://www.asianinstituteofresearch.org/JEBarchives/predicted-repayment-behaviour-of-borrowers-in-urban-cooperative-banks-in-assam%3A-a-case-study-of-nagarik-samabay-bank