Why Many Lenders Lose Money After Disbursement

The numbers the industry likes to quote are all about money going out. Nigeria's digital lenders pushed out around $865 million in loans in 2025, with transaction growth running past 45 percent a year and credit reaching more than 50 million adults the banks never served. Disbursement, by every available measure, is a solved problem. We have gotten very good at getting money out the door, fast, at scale, in minutes, from a phone.

Which is precisely the problem, because disbursement was never the hard part. Getting the money back is. And the industry keeps grading itself on the easy half while quietly bleeding on the half that actually decides whether a lending business lives or dies.

The Loss happens after the Decision, Not at it

Most people assume a lender's risk is concentrated at the moment of approval. Pick the wrong borrower, lose the money. That is real, but it is not where most losses come from. The credit decision is a single event. Everything that determines whether that decision was right or wrong happens in the weeks and months after the cash leaves, in collections, monitoring, and day-to-day operations. A book can look pristine at origination and rot quietly afterward, and most lending operations are built to disburse and badly underbuilt to recover.

The cost of that imbalance is now measurable in trillions. When the CBN withdrew its pandemic-era forbearance in mid 2025, the losses that rollovers had been hiding finally surfaced on the books. Nigeria's eleven largest banks booked roughly 3.4 trillion naira in loan impairment charges across their 2025 accounts. Ten of them alone set aside about 1.99 trillion naira in just the first nine months of the year, a 44.5 percent jump on the same period in 2024. A single tier-one bank provisioned more than 800 billion naira; another missed its dividend entirely after taking over 300 billion naira in provisions to reclassify loans it had been allowed to carry as healthy. The industry NPL ratio, around 5.1 percent in early 2024, climbed past 8 percent by January 2026 once the loans were classified honestly.

And this is not new money lost to a single bad year. AMCON, the bad-loan vehicle the government created after the last credit blow-up, is still chasing more than 4 trillion naira in unrecovered debt more than a decade later, and its levy quietly shaves an estimated 100 to 150 basis points off the return on equity of every bank in the country. The losses do not disappear. They compound, they get socialized, and they sit on the system for years. None of that is a disbursement failure. It is a failure of everything that comes after.

The Disbursement Bias is Built into the Org hart

The root cause is structural. Growth metrics reward loans out the door: book size, customer acquisition, gross volume. Those are the numbers in the board deck and the press release. Collections is unglamorous, gets treated as back-office, and is resourced accordingly. So capital and talent flow toward acquisition while recovery runs on a skeleton crew and a spreadsheet.

The asymmetry underneath is simple and brutal: it is far easier to give money away than to get it back. As one Nigerian lending executive put it, it has become harder to be a bad lender than a bad borrower. A whole informal economy has grown up around not repaying, with online groups trading tips on dodging lenders, fabricating identities, and even selling the service of scrubbing a borrower's name from a lender's database. The lender who optimized purely for speed of disbursement walked straight into that.

Three gaps where the money leaks

Underwriting treated as a one-time event. The score is calculated once, at origination, then the file goes static. But risk is not static. A borrower's circumstances move after you lend, and if nothing is watching, you find out about trouble only when a repayment is already missed, which is the most expensive possible moment to find out. Lenders who survive monitor continuously and score dynamically, so a deteriorating account raises a flag before it becomes a default rather than after.

Collections that are reactive, manual, and slow. The first missed payment is a fork in the road, and the odds of recovery fall sharply with every week that passes after it. Yet most lenders treat every delinquent account the same way, send the same generic messages, and escalate far too late. By the time a file reaches a recovery agent it has usually hardened into a write-off. The borrower who simply forgot and the borrower who never intended to pay need completely different responses, and undifferentiated, late collections gets both wrong.

Operational leakage and the consequence vacuum. Money disappears through broken payment rails when a willing borrower cannot actually complete a payment, through reconciliation gaps, through data silos that prevent a single view of the customer, and through fraud that only surfaces after disbursement, including first-payment defaults and multi-app borrowing rings. Layered on top is a structural hole: until recently the Global Standing Instruction, the mechanism that lets a lender recover from a defaulter's other accounts, was confined largely to commercial banks, leaving most fin techs and microfinance banks outside it. Serial defaulters learned to exploit exactly that gap. The CBN is now extending GSI to fin techs and MFBs in phases, but lenders who built no recovery infrastructure of their own were relying on consequences that, for them, did not yet exist.

The easy exits are closing

For years, two shortcuts hid these gaps. The first was hidden rollovers, quietly refinancing bad loans so they never showed up as defaults, which only defers the reckoning and inflates the eventual loss. The 3.4 trillion naira that hit the banks in 2025 is, in large part, exactly those deferred losses arriving all at once. The second shortcut was brute-force recovery, the shaming and harassment route. Both are now expensive. The FCCPC's rules attach fines of up to 100 million naira or one percent of annual turnover, five-year director bans, and delisting to deceptive pricing, hidden rollovers, and abusive collections, and more than a hundred non-compliant apps have already been removed. The market is expected to shrink in number and strengthen in structure, with survivors moving from short-term loan churn toward sustainable underwriting and longer customer relationships. The lenders who leaned on the shortcuts are running out of room.

The fix is to treat lending as a lifecycle, not a transaction

The operators who come out of this intact are the ones who stop thinking of a loan as a moment and start treating it as a relationship managed from disbursement to final repayment. In practice that means dynamic scoring that updates after the money goes out, early-warning monitoring built on behavioral and transactional signals, collections that are segmented and automated and act at the first sign of trouble rather than the last, repayment channels reliable enough that a willing borrower can always pay, a single view of each customer, and recovery rails like GSI plugged in as they open up.

That is a serious infrastructure build, and it is the reason many lenders never get there: doing monitoring, dynamic scoring, and collections well from scratch is enormously expensive, which is why so much capital keeps going into the disbursement side instead. It is also the gap credit-infrastructure platforms now exist to close, and it is the specific problem VeendHQ built Vida AI to solve. Rather than optimizing for faster disbursement, Vida's recovery service connects directly to a borrower's bank account so repayment is collected automatically and reliably instead of depending on the borrower to remember or on a payment rail that fails at the wrong moment. Around that sit the parts that close the other two gaps: continuous customer insights that track repayment ability after the loan is live, and a risk-officer co-pilot that uses predictive analytics to surface accounts slipping toward default early enough to act. The effect is to move a lender from reacting to missed payments to preventing them, which is where recovery is actually won. The strategic shift worth noticing across the market is that the smart money is moving from competing on how fast you can lend to competing on how reliably you can get it back.

Recovery is the business

Disbursement is a vanity metric. It feels like progress, it photographs well, and it tells you almost nothing about whether you are running a viable lending business. Recovery is the business. The 3.4 trillion naira the banks just wrote down is the price of forgetting that. The lenders still standing in five years will not be the ones who disbursed the most. They will be the ones who got the most back, because they built for the part everyone else treated as an afterthought. The money was never made when it left the building. It is made, or lost, on the way home.