For some borrowers, it means constant phone calls, frozen accounts, workplace visits, threats, and even public shaming. In response to these practices, the Federal Competition and Consumer Protection Commission (FCCPC) in 2025 introduced stricter penalties, including fines of up to ₦100 million or 1% of annual turnover for lenders that rely on harassment to recover loans.
Behind the scenes, these defaults accumulate into non-performing loans (NPLs), quietly eroding lenders’ balance sheets and limiting their ability to extend new credit. Against this backdrop, Nomba and Globus Bank say they have built a lending model designed to prevent such outcomes and early results suggest it is working.
The fintech and the tier-3 commercial bank report that their 18-month credit partnership has disbursed ₦21.3 billion to Nigerian businesses, with less than 1% classified as non-performing. The loans span sectors including wholesale and retail, professional services, hospitality, oil and gas, and fast-moving consumer goods.
This performance stands in contrast to broader industry trends. In Nigeria, loans that remain unpaid for more than 90 days are classified as non-performing, and rising NPL levels typically force banks to tighten lending. Industry-wide, NPLs stood at 4.2% in early 2023 and are projected to climb to around 7% by the end of 2025, driven by inflation, currency volatility, and mounting economic pressure on businesses.
Nomba and Globus Bank argue their model diverges sharply from traditional lending. Rather than relying on financial statements or fixed collateral, their approach centres on real-time business data and continuous monitoring.
According to Nomba CEO Yinka Adewale, the low default rate is a product of deliberate design. He points to a system built on live transaction data, structured collateral, and borrowers with what he describes as “skin in the game.”
Access to credit under this model is tightly controlled. Of more than 600,000 businesses using Nomba’s platform, only about 20,000 are considered eligible for loans. These businesses must be formally registered, show consistent transaction volumes, and have a track record on the platform. Even then, only a fraction roughly 10% actually receive credit.
Unlike traditional lenders, Nomba does not require applicants to submit financial statements. Instead, it evaluates businesses based on the transaction data flowing through its infrastructure, giving it direct visibility into revenue patterns, cash flow cycles, and operational activity.
Loans are typically capped at around 1% of a business’s annual revenue, a structure intended to keep repayments manageable. After disbursement, borrowers are monitored continuously, with systems designed to flag early signs of financial stress before they translate into missed payments. When risks emerge, the response may include restructuring terms or engaging the borrower before default occurs.
Collateral is another key component of the model, but in a digitised form. This can include inventory linked to the loan’s purpose, digital assets such as stocks or stablecoins, and certain semi-liquid physical assets. Borrowers are also required to provide a 30% cash collateral buffer to hedge against volatility, particularly where digital assets are involved. In the event of default, these assets form the primary recovery pathway.
Still, the model’s strengths also highlight its limits. Its effectiveness depends heavily on visibility into a merchant’s financial activity, which means businesses operating largely outside Nomba’s ecosystem or those that are cash-heavy are less likely to qualify.
The tight selection criteria also shape its performance. With only a small subset of merchants eligible, and an even smaller group receiving loans, the model is applied to a relatively narrow segment of the market. Defaults, while low, are not entirely eliminated. When they occur, restructuring is typically the first step, followed by recovery through pledged assets and cash collateral if necessary.
