FSR 2025 (Pt. 2): How Stress Testing Should Shape the Lending Stack of NBFCs
- Sudip Chakraborty
- 4 days ago
- 5 min read

If you missed Part 1 of the FSR 2025 series, you can read it here.
Introduction
The RBI's December 2025 Financial Stability Report delivered both validation and a strategic signal for the NBFC sector. Stress tests confirmed strong capital positions at 22.8% CRAR, substantially above the 15% regulatory minimum. The system demonstrated clear resilience under severe stress scenarios.
But the FSR also revealed something more strategic for the sector. The stress testing methods NBFCs use today meet current requirements perfectly. Single-factor sensitivity analysis works well. Yet as Scale-Based Regulation drives convergence with bank-level norms, infrastructure matters as much as results. Banks already employ multivariate models. These integrate GDP, inflation, and interest rates across 18-month horizons. NBFCs will need to match this level.
For NBFC executives, the question is practical: Does your stress testing infrastructure enable strategic decisions, or does it just satisfy annual compliance requirements?
What the FSR Stress Tests Revealed
The NBFC Stress Test Framework
The FSR tested 174 NBFCs covering ₹30.74 lakh crore - 95% of sector advances. The approach applied standard deviation shocks to GNPA ratios. This assessed capital adequacy under stress over a 1-year period.
Results confirmed system strength:
Baseline:Â GNPA 2.3% to 2.9%, CRAR 22.8% to 21.7%
Medium stress (1 SD):Â CRAR 21.1%, only 11 NBFCs breach 15% minimum
Severe stress (2 SD):Â CRAR 20.9%, same 11 outliers below threshold
Concentration risk:Â Top 3 borrower defaults reduce CRAR by 223 bps

Capital buffers are strong and the system can absorb major shocks. Only a small number would face capital concerns under severe scenarios. But the tests revealed gaps in execution efficiency.
What Passing Actually Required
Beyond maintaining capital buffers, passing required specific working features:
Portfolio aggregation:Â Consolidating exposure data by segment and vintage
Scenario modeling:Â Building stress scenarios over 12 months
Risk calculations:Â Applying GNPA shocks
Capital assessment:Â Translating stress into CRAR movements
Concentration analysis:Â Modeling default scenarios for top borrowers
Regulatory reporting:Â Delivering consistent results
The Infrastructure Reality for Most NBFCs
The Annual Compliance Trap
Most NBFCs treat stress testing as annual compliance. Teams manually extract data from separate systems. They build scenarios in Excel and the process typically takes 2 to 3 weeks. This makes ad-hoc scenarios impossible.
This creates strategic blind spots during market stress. When NBFC-MFI credit costs surged from 4.4% to 15.5%, could you model the capital impact quickly? When trade disruptions hit export sectors, could you stress test portfolio exposure at once? Can credit committees get scenario analysis during meetings?
If answers reveal long timelines, a capability gap exists that needs addressing.
The Convergence Pressure Building
Today's requirements remain manageable for most NBFCs. Single-factor stress tests meet compliance and annual submissions work within existing processes. ICAAP requirements apply only to Upper Layer firms.
But the directional trend toward convergence with bank-level standards is unmistakable. Upper Layer NBFCs implementing ICAAP need multi-risk assessment systems that evaluate credit, market, and operational risks together. The Expected Credit Loss framework taking effect in April 2025 requires forward-looking provisioning based on scenario analysis.
Upper Layer NBFCs face immediate pressure to build sophisticated capabilities. Middle Layer firms should prepare proactively rather than waiting for regulatory mandates.

What Stress-Ready Infrastructure Looks Like
Unified Portfolio Intelligence
Modern stress testing requires a single source of truth. This spans all lending functions. Key features include:
Real-time aggregation by segment, vintage, and geography
Historical data for backtesting scenarios
No reconciliation delays between systems
Consistent data across risk, finance, and operations
This foundation enables immediate response. It serves regulatory requests and business needs. When RBI requests stress test submissions, accurate data is available instantly.
Scenario Execution Engine
Pre-built scenario libraries should cover standard tests:
GNPA shocks at multiple severity levels
Sectoral stress for key exposure concentrations
Interest rate movements affecting yields and costs
Liquidity scenarios testing funding resilience
Concentration tests for top borrowers or sectors
Custom scenarios work through configuration interfaces. No coding expertise required. Business users specify assumptions independently and execute tests on their own. The system automates CRAR calculations. It handles provisioning impacts and capital projections. Execution takes hours rather than weeks.
Parallel scenario execution enables comparative analysis. What happens if personal loan growth slows by 20%? What if gold loans increase by 30%? How does the portfolio respond if interest rates rise 150 basis points? What if GNPA increases at the same time? Infrastructure should deliver answers on demand.
Flexible Risk Parameters
Risk models must adapt to different scenarios:
PD, LGD, and EAD calculations adjust based on economic conditions
Segment-level and vintage-level modeling beyond institution averages
Support for single-factor analysis (current requirement)
Capability for multi-factor scenarios (future sophistication)
Integration points for macro-financial data feeds
As institutions mature their approaches, model flexibility becomes critical.
Embedded Governance and Reporting
Strong controls must be built into the platform:
Audit trails capturing all assumptions
Version control for scenario history
Regulatory reporting from same data source
Historical libraries for trend analysis
Tracking of which scenarios show stress
This intelligence informs risk appetite. It shapes portfolio strategy.
Strategic Actions for NBFC Leadership
Diagnostic Questions
Assess your institution's current stress testing capability with these questions:
How long does scenario execution take from specification to final results?
Can you test at segment or product level, or only institution-wide?
Do credit committees have scenario analysis available during meetings?
Is stress testing data consistent with management reporting and regulatory submissions?
Can you model combined scenarios testing multiple risk factors simultaneously?
If answers reveal long timelines or manual work, systems investment becomes a priority.
Layer-Specific Priorities
Upper Layer NBFCs:Â ICAAP compliance creates immediate needs for multi-risk assessment systems. Investment in bank-level scenario modeling is not optional. It's a regulatory must. View stress testing as strategic support driving capital choices.Â
Timeline: Act now.
Middle Layer NBFCs (₹1,000 crore+): Automate single-factor stress testing. Build full scenario libraries that can be run efficiently. Integrate stress results into risk dashboards. This gives continuous visibility preparing systems for potential Upper Layer status.Â
Timeline: 12 to 24 months.
Base Layer NBFCs:Â Focus on unified data systems that support future stress testing needs. Basic scenario tools work now. But invest in data quality today. Ensure platforms can scale as you grow.Â
Timeline: 24 to 36 months.

OneFin: A Leader in Lending Infrastructure
OneFin's unified lending platform embeds stress testing features designed for NBFC regulatory changes. The platform addresses the core infrastructure challenges that emerged from the FSR analysis:
(A) Unified Data Architecture: Single source of truth eliminating reconciliation gaps, enabling real-time scenario execution and regulatory submissions
(B) Configurable Scenario Engine: Pre-built and custom stress testing capabilities with automated regulatory calculations and multi-variate risk assessment
(C) Regulatory Readiness: ICAAP-ready infrastructure that scales from Base to Upper Layer as institutions grow, with seamless RBI reporting integration
(D) Proven Scale: 16 lakh+ loans processed with complete historical data, API-native design supporting macro-financial data integration
Conclusion
The FSR validated strong NBFC capital positions and confirmed system resilience under severe stress scenarios. Capital buffers are substantial. But the tests revealed capability gaps in stress testing infrastructure. As convergence with bank-level standards accelerates, the Upper Layer faces immediate pressure while the Middle and Base Layers should prepare ahead.
Foundations built today will determine response speed during stress. Those that successfully embed stress testing capabilities into their lending stack will capture opportunities while competitors stay stuck managing deadlines. OneFin's unified platform provides the infrastructure needed to transform stress testing from a compliance task to a strategic capability.
To know more about OneFin, schedule a Demo.
