A Risk-Assessment Model for Indian Non-Banking Financial Companies: A Multi-Risk Ratio and Stress-Testing Approach with Formal Statistical Validation

• Author(s): Syeda Bushra Ahmed; Dr. Manoj Kumar
• Paper ID: 1718347
• Page: 5070-5088
• Published Date: 02-06-2026
• Published In: Iconic Research And Engineering Journals
• Publisher: IRE Journals
• e-ISSN: 2456-8880
• Volume/Issue: Volume 9 Issue 11 May-2026

Abstract

The Non-Banking Financial Companies (NBFCs) are an integral part of the financial system, with outstanding assets of about ₹54 lakh crore (around 14.6 percent of GDP) as of March 2025. The crisis that has been recurring in recent years, mostly with Infrastructure Leasing and Financial Services (IL&FS) in 2018 and the follow-up collapse of Dwarkadhish Housing Finance Corporation Limited (DHFL) has revealed the glaring weaknesses of one-dimensional risk-monitoring systems for deposit-taking commercial banks. This study proposes and empirically tests a Composite Risk Score (CRS) incorporating the four types of risks (credit, liquidity, market, and operational) for four representative NBFCs (Bajaj Finance Limited, Muthoot Finance Limited, LIC Housing Finance Limited, and Shriram Finance Limited) during the panel period (FY2015–FY2025; n = 44 firm-year observations). The empirically calculated weights are very close to the theoretically calibrated weights and, in fact, obtained from Principal Component Analysis (PCA) with base weights (25.0%, Liquidity: 26.3%, Market: 23.4%, Operational: 25.4%) and very close ones with the five of six alternative weights (Spearman rank-order correlation: ρ = 1.000 between the base weight specification and five of six alternative weights), which confirms strong robustness. The incremental R² measure of the multidimensional model over a single-metric credit-only model is 0.841, which is significant at an F(3,39) = 827.71, p < 0.001, confirming Hypothesis 1 that a multidimensional model is superior. The large effect size (η² = 0.434) and the result of Kruskal-Wallis testing (H = 18.615; p < 0.001) support the assumption that systematic differences in CRS occur between the segments (H2). When macro-sensitivity analysis is conducted, retail finance CRS is significantly negatively correlated with GDP growth (ρ = −0.624 and p = 0.040) and gold loan CRS is positively correlated which is consistent with the counter-cyclical dynamics of gold price movement, so supporting H3. Results of the variance homogeneity test (Levene's test) for this study indicate the W value of 0.005 (p = 0.945), implying that the variance of the mean CRS scores among the segments is homogenous; collateral-backed segments have structurally lower mean CRS (0.315 versus 0.444), partially supporting H4. The differences in shock absorption across business models, measured as the risk sensitivity differential, are quantified through stress-testing, where the differences between the three calibrated macroeconomic scenarios range from +0.030 for gold loans to +0.087 for retail finance.

Keywords

NBFC Risk Assessment, Composite Risk Score, Panel Regression, Kruskal-Walli’s Test, PCA Weight Derivation, Stress Testing; Credit Risk, Liquidity Risk, India Shadow Banking, Financial Stability

Citations

IRE Journals:
Syeda Bushra Ahmed, Dr. Manoj Kumar "A Risk-Assessment Model for Indian Non-Banking Financial Companies: A Multi-Risk Ratio and Stress-Testing Approach with Formal Statistical Validation" Iconic Research And Engineering Journals Volume 9 Issue 11 2026 Page 5070-5088 https://doi.org/10.64388/IREV9I11-1718347

IEEE:
Syeda Bushra Ahmed, Dr. Manoj Kumar "A Risk-Assessment Model for Indian Non-Banking Financial Companies: A Multi-Risk Ratio and Stress-Testing Approach with Formal Statistical Validation" Iconic Research And Engineering Journals, 9(11) https://doi.org/10.64388/IREV9I11-1718347