Salary Credits vs Business Income: Why the Distinction Matters for MSME Lending

When a credit analyst looks at a bank statement for an MSME borrower, they're usually looking for income. The question sounds simple — how much money is coming in? — but for the self-employed, the answer is structurally complicated. MSME business accounts receive inflows from customers, from family transfers, from loans being credited, from investment redemptions, and sometimes from salary income if the proprietor also holds a job. Treating all of these as 'income' leads to significant overestimation of credit capacity.

Indian bank statements present transaction descriptions in a largely unstructured format. A ₹2.5 lakh NEFT credit described as 'TRANSFER FROM RAMESH ENTERPRISES' could be a business receipt, a related-party transfer, a loan advance, or a director's capital infusion. A ₹45,000 credit described as 'UPI/9876XXXXXX' is even more ambiguous.

For salaried borrowers, this is manageable — salary credits are typically tagged with employer names, payroll processing company descriptions (HDFC PAYROLL, NETSALARY, etc.), and arrive on consistent dates. For MSME borrowers, there's no such structure. Income arrives from multiple counterparties, on irregular schedules, in varying amounts, and often mixed with non-income flows.

Getting income classification wrong has direct downstream consequences for credit quality. Overcounting income produces a lower FOIR, which may push a borderline file over the approval threshold. It also produces a higher Loan-to-Income ratio, which affects pricing on risk-based models. In aggregate, systematic income overcounting leads to a portfolio that's more leveraged than it appears — which becomes visible in defaults during stress periods.

The subtler risk is undercounting income. An MSME borrower whose genuine business receipts are being flagged as 'transfers' or 'other credits' may be assessed as too low-income to qualify for the loan size they actually need. This produces unnecessary rejections and contributes to the credit gap that still affects India's MSME sector.

A well-trained classification model looks at multiple signals simultaneously: transaction description patterns (employer name databases, payroll processor identifiers), credit amount stability and regularity (salary varies little month-to-month; business income varies a lot), counterparty diversity (salary comes from one source; business income typically comes from many), and timing patterns (salary on a specific date; business receipts cluster around billing cycles).

Corroboration matters too. A GST-registered business should show GSTR-1 outward supply that roughly matches its banking inflows. A salaried borrower should have an employer that appears in available databases. When the primary income signal is corroborated by secondary sources, confidence in the classification goes up substantially.

The practical output of good income classification is an income estimate that lenders can actually rely on: not just the gross inflow number, but a categorised breakdown — verified salary income, business receipts, related-party transfers, investment inflows, and unclassified credits. Each category carries a different weight in a responsible income assessment.

For MSME lending at scale, this classification needs to happen automatically, consistently, and with explainability — so that credit analysts can review the categorisation and override it when local context requires, rather than doing the classification from scratch on every file.