How Data-Driven Underwriting is Transforming SME Lending in India

1. The traditional SME lending problem in India

India has roughly 63 million MSMEs (Micro, Small, Medium Enterprises) — the largest employer outside agriculture. Yet SME credit has been chronically under-supplied. The credit gap is estimated at ₹25-30 lakh crore.

The reason wasn't lack of capital or willingness — it was the cost of underwriting. Traditional SME lending required:

• 3 years of audited financials (most small businesses don't have these)
• Collateral worth 1.5-2x loan amount (most don't have)
• Personal guarantees from promoters
• Detailed bank statements, GST returns, business profile documentation
• Branch visits, manual verification, in-person interviews

The cost of underwriting a ₹50 lakh loan was nearly the same as a ₹5 crore loan, but the revenue was 10x lower. Banks rationally focused on bigger ticket sizes, leaving SMEs to expensive informal lenders or no credit at all.

2. What's changed: GST, account aggregator framework, UPI data

Three structural reforms have transformed Indian SME lending:

GST (2017): Created a digital trail of every business's revenue and expenses. Lenders can now verify revenue figures by querying the GSTN, eliminating the audit-financial requirement for many loan sizes.

Account Aggregator framework (2021-2026): RBI-licensed Account Aggregators allow customers to share their financial data (bank statements, mutual fund holdings, insurance) with lenders digitally and securely, with consent. What used to take 2 weeks of paperwork now happens in minutes.

UPI transaction data (2017-now): A small business that uses UPI for receipts has a verifiable revenue stream visible to lenders. Income that was previously informal becomes underwritable.

Together, these create what's called the "Digital Public Infrastructure" — a foundation that makes digital, low-cost SME lending possible at scale.

3. The new underwriting stack in 2026

Modern SME lenders combine multiple data sources for underwriting decisions in seconds rather than days:

Data Source	What It Tells the Lender
GSTN return data	Revenue trend, business cyclicity, sectoral risk
Bank statement (via AA)	Cash flow, balance trends, EMI behaviour
UPI transaction history	Daily revenue patterns, customer base diversity
CIBIL Commercial & Personal score	Past credit behaviour
MCA/ROC filings	Corporate structure, related party transactions
ITR data (via Income Tax APIs)	Income consistency, tax compliance
Industry-specific data	Demand trends in the borrower's sector
Borrower's mobile/digital footprint	Engagement patterns, business activity signals

What used to be an underwriter manually piecing together a story is now an algorithm processing 50+ data points in seconds.

4. Cash flow underwriting vs collateral-based lending

The biggest shift in modern SME lending: from collateral-based to cash flow-based underwriting.

Old model (collateral-based):

• Loan size determined by collateral value (typically 50-70% of property value)
• If borrower defaults, lender liquidates collateral
• Excludes asset-light businesses (services, professional firms, traders)

New model (cash flow-based):

• Loan size determined by serviceability — can the business cash flows support the EMI?
• Underwriting focuses on revenue stability, customer concentration, growth trends
• Includes asset-light businesses
• Faster approvals (often 24-48 hours)

For most service businesses (consultancies, professional firms, IT services, small clinics), cash flow lending is the only viable option — they don't have enough collateral but they have stable revenues.

5. How AI/ML models read SME credit risk

Modern SME credit decisions are made by algorithmic models that ingest dozens of variables and output a default probability. The patterns these models surface:

• Revenue volatility — businesses with stable monthly revenue are lower risk than highly cyclical ones
• Customer concentration — businesses where 1 customer is 50%+ of revenue are higher risk
• Banking discipline — minimum balance breaches, returned cheques, overdraft usage
• Promoter financial behaviour — credit card utilisation, personal loan history of the promoter
• GST filing pattern — late filings, abrupt changes in turnover declared
• Industry-level signals — businesses in declining sectors face higher rates regardless of individual quality

The honest reality: these models are powerful but imperfect. A great business with poor digital footprint may be misclassified. A struggling business with clean digital records may pass through. Human review still matters for marginal cases.

6. New-age lenders disrupting the segment

The lending landscape has expanded dramatically:

• Lendingkart, FlexiLoans, Indifi — pioneer SME-focused fintechs
• NeoGrowth, U GRO Capital — focused NBFCs with sectoral specialisations
• Razorpay Capital, PayU LazyPay — embedded lending tied to merchant payment data
• Paytm Lending, BharatPe — small-ticket merchant cash advances
• HDFC Bank, ICICI, Axis SME-Smart — traditional banks with digital SME products

The competitive pressure has compressed interest rates for prime borrowers from 18-22% (informal sources) to 12-16% (digital lenders) to 10-13% (banks for the best). For SMEs that qualify, credit access in 2026 is the easiest it has ever been.

7. What this means for SME borrowers

If you run an SME, the key shifts:

• Digital footprint matters as much as financials — a business with strong digital records (GST, UPI, bank statements) often gets better terms than one with audited but offline financials
• Speed has gone from weeks to days — digital lenders can disburse in 48 hours for standard cases
• Collateral is no longer mandatory for many loan types — cash flow lending opens doors
• Multiple competing offers are normal — borrowers can compare multiple lenders simultaneously through aggregator platforms
• Pricing is risk-based — same loan size, two different businesses can pay 11% vs 19% based on data signals

8. Preparing your business for data-driven lending

For SME owners who want better access to credit:

1. Use digital banking actively — minimum cheque transactions, more bank transfers and UPI
2. File GST returns on time, every time — late filings tank your underwriting score
3. Maintain clean banking discipline — no minimum balance breaches, no returned cheques
4. Diversify customer base — avoid having one customer above 30% of revenue
5. Build CIBIL Commercial score — if you've never borrowed, take a small business credit card and pay it off systematically
6. Maintain ITR consistency — declared income should match GST turnover within reason
7. Use Account Aggregators wisely — sharing data with one lender can fast-track approval significantly
8. Build relationships with 2-3 lenders — diversify so you're not dependent on a single source