Loan underwriting in non banking financial companies depends on a clear view of applicant risk and repayment capacity. Teams need data that is timely, reliable, and explainable to satisfy board, regulator, and customer expectations. A robust checklist keeps decisions fair and scalable across geographies and products. This guide outlines the Top 10 Data Points for NBFC Loan Underwriting to help credit teams focus on signals that matter most. Each point balances prudence with speed, showing what to collect, how to interpret it, and where to cross verify. Use these sections to standardize policies, improve model features, and strengthen audit trails while keeping journeys fast.
#1 Income stability and sources
Assess primary and secondary income with attention to volatility, seasonality, and employer stability. Capture salary credits from bank statements, tax returns, and verified payroll records. For self employed customers, analyze sales ledgers, GST filings, and expense ratios to estimate normalized cash flows. Compare declared income with inflows and investigate gaps or spikes that signal informal earnings or stress. Record industry risk, tenure, and employment type, because formal salaried work behaves differently from gig or contract work. Use simple ratios like income to EMI, and maintain income verification rules by ticket size and product. Refresh estimates after any recent job change or business pivot.
#2 Banking behavior and cash flow patterns
Evaluate account conduct across twelve months to reveal spending discipline and liquidity. Parse recurring inflows, mandated outflows, and minimum balance breaches that indicate pressure. Flag returned transactions, frequent cash withdrawals, and circular transfers that may hide obligations. Identify seasonality in working capital for small businesses and quantify net operating cash. Build monthly free cash after fixed obligations, then stress test it under a moderate downside. Map commitments to lenders and build a complete EMI schedule. Use automated statement analyzers to reduce manual errors and to standardize flags across portfolios. Cross verify merchant settlements, wallet loads, and UPI inflows with invoices or sales data where relevant.
#3 Credit bureau history and score depth
Pull multi bureau reports to capture score, delinquencies, utilization, and trade lines. Look beyond the headline score to examine inquiries, recent openings, and the vintage of the oldest account. Calculate behavioral metrics such as utilization trend, proportion of unsecured exposure, and performance after any past settlement. Reconcile bureau data with declared obligations to locate missing loans or unreported cards. Segment thin files, new to credit profiles, and mature borrowers differently to avoid unfair declines. Blend bureau features with internal repayment history to build a stable, explainable risk view. Track score volatility over recent months to detect unstable patterns before approval.
#4 Existing obligations and indebtedness
Construct a full view of debt using bureau trades, bank statement EMI picks, and self declarations. Quantify fixed obligations to income, debt to income, and unsecured to secured exposure. Pay attention to recent balance transfers or top ups that can mask real leverage. Examine co obligations such as guarantor roles that could create hidden strain. For business borrowers, model contingent liabilities like supplier credit, lease payments, and tax arrears. Set guardrails for peak EMI to free cash and forecast payment cliffs. Require proof for any claimed soon to close loan before granting relief in ratios.
#5 Collateral quality and legal enforceability
When lending against property, gold, or equipment, validate title, ownership, and encumbrances through independent search and valuation. Check marketability in the borrower location and liquidity in a time bound sale. Ensure loan to value is calibrated by asset type, condition, and appraisal method. Validate storage, insurance, and custody arrangements for movable collateral. For property, examine approved plans, completion status, and local authority clearances. Confirm enforceability under contract and local law, including notice periods and auction processes. Tie margin requirements and monitoring frequency to the volatility of the pledged asset. Revalue periodically and capture independent photos with geotags to deter substitution risk.
#6 Business model and segment risk
For MSME lending, map the borrower business model, supply chain position, and customer concentration. Estimate variable versus fixed costs and breakeven throughput to understand sensitivity. Use GST data, e way bills, and invoices to reconcile revenues and seasonality. Evaluate vendor dependencies, inventory cycles, and payment terms to judge working capital needs and cash conversion. Score sector health using external data like input prices and regional demand proxies. For consumer lending, assess employer category and attrition risk. Tie pricing, tenure, and covenants to segment risk grade so incentives remain aligned. Document alternative revenue streams and shock absorbers that can sustain obligations during downturns.
#7 KYC integrity and fraud risk controls
Verify identity, address, and beneficial ownership using authoritative sources and biometric checks where permitted. Cross check photographs, signatures, and geotagged field visits to reduce impersonation or synthetic profiles. Screen against sanctions, watchlists, and politically exposed person databases using up to date lists. For businesses, confirm directors, shareholding, and ultimate beneficial owners through registries. Use device fingerprinting, contact network analysis, and application velocity to detect collusion. Maintain maker checker segregation and audit trails for all overrides. Calibrate escalation rules for mismatches so genuine customers are not blocked but risks are contained. Incorporate face liveness and document forensics to catch tampering before credit approval.
#8 Spousal, household, or co borrower dynamics
Household level analysis improves accuracy where incomes and obligations are pooled. Capture spouse income, dependents, and shared expenses to estimate true free cash. Confirm co borrower alignment on tenure, collateral pledges, and repayment mode to avoid disputes. Review household credit history to surface correlated risk, including repeated delinquencies or stacking across lenders. For informal incomes, triangulate with utility bills, rent receipts, and community references. Model the effect of life events such as parental leave, relocation, or medical costs on cash flow. Align total exposure limits with household resilience rather than a single applicant view. Capture savings buffers and insurance coverage to improve resilience estimates.
#9 Repayment behavior and channel readiness
Choose repayment instruments that fit the customer cash cycle and reduce slippage. Assess historical mandate success rates, NACH return codes, and bounce reasons carefully. Review digital payment readiness, such as authentication capability and device stability. For micro businesses, align due dates with revenue peaks to lower roll rates. Provide fallback modes like UPI or cash pickup only where operationally secure. Track cohort level cure rates after soft and hard buckets to refine strategies. Well matched channels and nudges reduce cure time, lower collections cost, and improve customer experience. Capture consent quality, mandate frequency, and autopay participation to predict friction risk.
#10 Early warning signals and monitoring
Underwriting should plan for change by defining measurable early warning triggers. Track days past due movement, declining bank end balances, rising inquiry velocity, and employer level stress events. For MSMEs, monitor GST dips, invoice delays, and inventory spikes that hint at demand shocks. Combine internal and external feeds to refresh risk ratings and review limits. Define action ladders that escalate from friendly nudges to restructuring pathways. Document playbooks for disasters or regional disruptions so teams respond consistently. Use post disbursal data to update models, then feed lessons back into policy and scorecards. Close feedback loops shorten decision time and prevent repeat underwriting mistakes.