Abhigyan

Renu Arora ¹, Archana Singh ²

Affiliations
1 Mata Sundri College for Women, University of Delhi, New Delhi, IN
2 Delhi School of Management, Delhi Technological University, Delhi, IN

Abstract

This paper aims to develop a credit risk assessment model based on multivariate discriminant analysis (MDA) for predicting default risk in grant of business loans by Indian public sector commercial banks. The study uses a sample of 47 bank loans to small and medium enterprises (SMEs) and mid-corporates, by an Indian public sector bank, to design a three-group discriminant model, based on financial and non-financial factors. The results show that this model performs better than the Altman et al. (1995) - Emerging Markets Z-score model with re-estimated discriminant scores. The combination of quantitative and qualitative risk factors improved credit risk assessment and the model could accurately classify 97.5 per cent and 71.4 per cent in analysis and hold-out samples respectively. The findings confirm that by using both financial and non-financial characteristics of loan counterparties in multi-discriminant analysis, banks can predict credit risk in each loan transaction, and can map rating transitions to develop early warning signals of default which will ultimately help them to capture bad loans.

Keywords

Credit Ratings, Default Risk, Mid-Corporates, Inadequate Safety, Rating History.

Full Text