Welcome to the fourth edition of Common Thread.
No region has done more with mobile money than Africa. In 2025, the continent moved around 74% of the world's mobile-money transactions — some 92 billion of them, worth roughly $1.4 trillion. A trader in Nairobi, a shopkeeper in Lagos, a driver in Accra, all sending, receiving and settling in seconds, on a feature phone, without ever setting foot in a bank branch. On the narrow problem of moving money, Africa didn't just catch up to the developed world; it moved ahead of it.
Which makes it easy to claim that the continent has cracked financial inclusion. On the evidence of the rails, it has.
But I've spent enough time in these markets to know that moving money and lending money are different problems, and the second one is nowhere near solved.
The half that's missing
Consider the number that should give anyone in this business pause. Across low- and middle-income economies, just 24% of adults borrowed from a formal source in 2024 — a bank, a card, or a mobile money account — while 35% borrowed informally, from family and friends. In Sub-Saharan Africa, the tilt toward informal is sharper still: the World Bank's Global Findex finds most adults still lean on family, friends and savings groups rather than a formal lender.
For businesses, the gap runs into hundreds of billions. The IFC and MIT Sloan's Kuo Sharper Center put Sub-Saharan Africa's SME finance gap at around $331 billion, roughly four-fifths the size of South Africa's entire economy, in demand from small firms that simply can't get funded. In Nigeria, only about 5% of SMEs can access formal credit.
At the level of a single customer, the effect is stark: someone who moves money faster than most Europeans often can't borrow $200 against ten years of doing it. The rails are world-class. The credit that should sit on top of them isn't there yet.
This is the case Boston Consulting Group makes in its March 2026 report, Beyond Payments: Unlocking Africa's Second FinTech Wave. The first wave built the rails; the second has to make them productive, and at the centre of that, in BCG's words, is building credit markets anchored in risk-based lending. If you read one thing on this, read that.
Why the second half is the hard one
Moving money is, in the end, an engineering problem: build the switch, connect the wallets, secure the transaction, and it works the same way for everyone.
Lending is a problem of judgment, and judgment depends on something the rails don't hand you — a reliable read on whether a borrower will pay you back.
That read is exactly what most African borrowers can't provide in the form a lender expects. No payslip, no audited accounts, no collateral, no filed credit history. Under the old rules that makes you unlikely to get credit, not because you're a risky borrower but because the lender has no way to know what kind you are.
A shopkeeper can run every shilling of her business through mobile money for a decade and still walk into a bank a stranger, because none of that history reaches the person making the decision.
The data exists. It just hasn't been turned into creditworthiness.
The market isn't standing still
It would be lazy to claim the continent is asleep on this. Some of the most inventive credit work happening anywhere is happening in African markets right now, and I'd back a few of these teams against anyone in London or Singapore.
The pattern is now well established. A mobile-money operator, who holds the transaction data and the customer relationship, partners with a bank or licensed lender, who holds the balance sheet and the licence. The underwriting in between increasingly runs on alternative data: mobile-money flows, airtime, bill payments, device signals.
Instead of a pay slip, the model reads the digital footprint, and it has already put small, instant loans (merchant advances, device finance, working capital) in the hands of tens of millions who'd never have cleared a traditional credit check.
Kenya is at the frontier. In 2024, 32% of Kenyan adults borrowed from a mobile-money provider, the highest share in the region, and a quarter borrowed that way and no other. Across the region, mobile money now accounts for close to 60% of all formal borrowing.
This part works.
So why isn't it finished?
Because a model isn't a rail
A payment rail scales for free. Add a customer or a merchant, and the rail doesn't care who they are. A credit model is the opposite. It has to be built, tested, and tuned against the behaviour of a specific population, and it holds only for as long as that behaviour does.
So, the work is iterative and slow by nature. You launch a product in one market, watch who repays and who doesn't, find where the model was wrong, retune, and launch the next one. Then you start again in the next market, where the data, the rails and the regulator are all different, and last year's model may not travel at all.
And Africa is not one market but dozens, each with its own data, its own rails, its own rules. A model that works in Nairobi doesn't lift and drop into Lagos or Kinshasa. Run that build-tune-relaunch loop across that many markets and you can see why credit is taking a decade where payments took a few years. This isn't a shortfall of ambition; it's the shape of the problem.
And getting it wrong is expensive
Credit extended carelessly destroys opportunity instead of widening it. Push instant loans onto thin data and thinner affordability checks and you manufacture over-indebtedness: borrowers stacking short-term loans at punishing rates, defaulting, ending up worse off than before anyone "included" them. Kenya has already lived this.
Before regulators stepped in, hundreds of unlicensed digital lenders were operating, some charging effective rates as high as 400% a year, and default rates on mobile-bank loans reached nearly 51% in the central bank's own household survey, more than double the rate on traditional bank loans. It took the Central Bank of Kenya's Digital Credit Providers Regulations in 2022, which forced every digital lender to license, cut the market from hundreds of players to a few dozen, and reined in rates, collections and data practices, to begin cleaning it up. Findex is unsentimental about the loans themselves: typically small, short, and carrying high effective interest. Useful, but no one's idea of transformative.
So the goal was never simply more credit; it's sharper judgment about who can safely carry it. Speed without judgment is a debt problem wearing the costume of financial inclusion, and ours is an industry that has worn that costume before.
What would move it
If the constraint is the credit model — building it, tuning it, carrying it into the next market — then that's where the effort should go.
Two things matter. First, turning transaction data into credit infrastructure: making a customer's own history usable, with their consent, as the basis for a decision.
That's what the open-banking reforms now underway in Nigeria and Kenya are quietly about — portability, so that a decade of wallet history finally counts for something. Second, making the underwriting itself faster to build and adapt, so that standing up a responsible model in a new market takes months rather than years.
Get those two right and the many-markets problem stops looking like dozens of decade-long projects and starts to look like one repeatable capability. That's the real prize of the second wave: not another payment rail, but the layer of judgment that lets the rails do more than move money.
Where we come in
This is the problem we work on at FinBox. Not the rails, which are largely built and built well, but the decisioning layer above them: helping lenders turn transaction and alternative data into credit models they can deploy, monitor and adapt across markets, responsibly and at speed. It's the conviction that has run through all three of these editions, that the hardest and most valuable part of finance is almost always the last mile to the customer, the moment of the decision itself.
Africa won the first wave outright. The second, who gets trusted with credit and how safely, is the one I'd bet the next decade on.
More threads to pull next month.
Cheers,
Bindesh
