Why Nomba’s Sub-1% Default Rate is a Win for Product People, Not Just Bankers
I was just reading about Nomba and Globus Bank pulling off a sub-1% default rate in Nigeria, and it hit me—this isn't just a finance story, it's a data engineering flex.
I’ve spent enough time in the Lagos tech scene to know that lending in Nigeria is usually a "vibes and prayers" business. You give out money, hope the borrower doesn't disappear into the depths of Ikorodu, and pray the macro-economy doesn't do a backflip. Most banks are currently bleeding out—we’re talking trillions in bad loans—yet here come Nomba and Globus Bank, casually announcing a default rate of less than 1% on a 21.3 billion Naira portfolio.
When I saw that number, my developer brain didn't go to the "billions." I went straight to the "how." In a market where people struggle to pay back even small "quick loans," keeping defaults that low is basically like trying to write bug-free code on your first try. It shouldn't happen, but somehow they made it work.
The Paperwork Lie vs. Live Streams
For years, getting a business loan in Nigeria meant printing out mountains of bank statements, most of which were probably "doctored" by some guy in a business center. Traditional banks look at the past—what you did six months ago. But if you’re running a restaurant in Yaba or a pharmacy in Ikeja, six months ago is ancient history.
Nomba’s edge is that they aren't looking at PDFs. They are looking at the live stream of transactions. Since they provide the PoS terminals and the payment software, they see every kobo entering the merchant’s account in real-time. It’s the difference between reading a status report and watching a live server monitor. If sales dip on Tuesday, they see it. If a merchant hits a record Friday, they see it.
As a builder, I love this because it’s a pure "data-over-documents" approach. They aren't guessing if a business can pay; they are calculating it based on the actual packets of data moving through their system.
Building the "Auto-Recover" Feature
The second thing that caught my eye is how they handle collateral. Most small businesses don't have a house in Lekki Phase 1 to drop as security. Nomba turned their own platform into the collateral.
They call it a "digitised collateral framework." In plain English? They’ve baked the repayment into the business flow. Instead of waiting for a merchant to remember to wire the monthly repayment (and we all know how "forgetful" people get when money is tight), the system handles it. It’s structurally embedded. If the money is flowing through the pipes you built, you can easily divert a small stream to cover the debt before it even hits the merchant's main balance.
It’s like setting up a recurring cron job for debt collection. It removes the human element—and the human temptation to skip a payment—from the equation.
Why This Matters Beyond the Hype
We’ve seen the horror stories lately. Moniepoint is in court with Alerzo over billions. Access Bank is chasing billions lost to internal fraud. The "old way" of throwing money at big names and hoping for the best is hitting a wall.
Nomba is proving that if you build the infrastructure first—the payments, the settlements, the business tools—the lending part becomes a lot less scary. They want to scale this to 500 billion Naira. That’s a massive jump, but if the underlying engine is built on real-time data instead of "trust," they might actually pull it off without the wheels falling off.
For me, the takeaway is simple: stop trying to solve Nigerian problems with "legacy" logic. If you want to lend in this climate, you have to be the one holding the pipes. When you own the data, you own the risk. And when you own the risk, you don't have to spend your weekends chasing people down Lagos-Abeokuta Expressway.
Related from Venture
Let's build your next big product.
Accepting project-based freelance, remote engineering roles, and hybrid positions.