Nigeria pilot: SME loans underwritten by payment data
Nomba and Globus Bank are piloting SME loans in Nigeria that underwrite borrowers using merchants’ payment‑processing data rather than traditional credit files, aiming to broaden access and reduce defaults. The model tests whether real‑time transaction signals can support scaled lending to merchants, though the article flags that maintaining repayment discipline and recovery management will be the real test as it scales. (techpoint.africa)
A Nigerian payments company and a bank say they have already pushed out ₦21.3 billion in business credit using card-machine and transaction data, and the share of loans that went bad stayed below 1% over an 18-month run. The companies are Nomba and Globus Bank, and they announced the results on April 9 and April 10, 2026. (techpoint.africa) The unusual part is what they used instead of a normal credit file. Nomba watches the payment flows of merchants on its system, so it can see whether a shop is selling every day, whether sales are rising or falling, and whether cash is coming in steadily enough to support a loan. (techpoint.africa) That changes who can borrow. A lot of small and medium-sized enterprises in Nigeria do not have the kind of long paper trail banks usually want, but they do have live payment histories from point-of-sale terminals, online checkouts, and business accounts. (globusbank.com) Nomba is not a tiny test case. The company says it processes more than ₦3 trillion monthly for more than 600,000 active businesses, which means it already sits on a large stream of merchant data that can be turned into lending signals. (nomba.com) Globus Bank brings the balance sheet, and Nomba brings the merchant feed. In practice, that means the bank supplies the loan capital while the payments company helps decide who looks healthy enough to repay. (techpoint.africa) The money has gone into businesses in wholesale and retail, professional services, food and hospitality, oil and gas, and fast-moving consumer goods. Those are the kinds of merchants that ring up sales every day, which makes payment data more useful than it would be for a business that bills once every few months. (punchng.com) The early number everyone is staring at is the non-performing loan ratio below 1%. In banking, that is the slice of loans that have stopped paying as agreed, and it is far lower than the bad-debt levels many lenders fear when they move downmarket into smaller businesses. (technext24.com) But payment data is good at spotting movement, not character. A merchant can have strong sales in March and still decide to dodge repayment in May, which is why Techpoint’s reporting says the harder problem is not approval but keeping repayment discipline and handling recovery once the book gets bigger. (techpoint.africa) That is why Nomba’s executives are talking about controlled growth instead of a lending land grab. The model looks strongest when loans stay tied to merchants whose daily transactions Nomba can already see, score, and keep monitoring after disbursement. (techpoint.africa) If this keeps working at larger scale, it gives Nigerian business lending a different formula. Instead of asking a small merchant for years of formal records, a lender can look at the digital cash register and ask a simpler question: does this business actually sell enough, often enough, to pay us back? (techpoint.africa)
