Your collections data is not collateral. Can it become bankable?

A strong collections dashboard is not collateral. Banks lend against legally identifiable receivables, contracts or loan portfolios when they can test the payments, control where money arrives and recover value under stress. The difference is not more data. It is a structure a lender can trust when things go wrong.

In February 2024, d.light and Chapel Hill Denham closed a ₦10 billion securitisation in Nigeria, then approximately $7.4 million. The question behind the deal was not whether d.light had a clever pay-as-you-go payment algorithm. It was more uncomfortable than that.

If d.light stopped servicing its customers, who would keep collecting the money?

The answer was built into the structure. The Nigeria Infrastructure Debt Fund provided the funding. African Frontier Capital structured the transaction and became both master and backup servicer, meaning it could administer collections and step in if the originator failed. The customer receivables were funded in naira, the currency in which those customers paid.

That is a lender decision moment. Before money moved, the deal had to answer who owned the payment claims, where the cash would go and what happened if the company at the centre of the operation could no longer perform.

Most founders begin somewhere else. They open the dashboard.

Collections data records who paid, how much and when. It can be excellent evidence. Yet it is not, by itself, an asset a bank can seize or sell. A receivable, meaning the legal right to collect a specific payment, is closer to the thing being financed.

Imagine a delivery app's order history. It may prove that customers often order dinner. But it does not give you a legal right to next Tuesday's payment. For a lender, that distinction is the whole game.

The revenue number that disappears

The conventional belief is straightforward: once a startup has revenue, debt should follow.

It often does not. Revenue can be cancelled, disputed, concentrated in one customer, collected in the wrong currency or swallowed by operating losses. A lender therefore creates a borrowing base, the share of receivables it considers eligible for lending. It is a smaller, harsher version of the revenue figure on the founder's slide.

US bank-supervision guidance on receivables finance focuses on whether payments are eligible, overdue, diluted by credits or disputes, concentrated in a small number of customers, protected by first-priority security and collected through controlled routes. Common advance rates are 70-85% of eligible receivables. That is not 70-85% of turnover.

Consider the lender looking at a company with £1 million of reported invoices. If half sit with one buyer, some are disputed and others are late, the usable lending pool may be a fraction of the headline number. Accounts commonly become ineligible when they are three times normal terms overdue, typically 90 days on a 30-day invoice. A single customer is commonly limited to 10-20% of the borrowing base.

Here is what the numbers do not tell you: a business can grow quickly while its financeable asset pool shrinks.

A lender will also examine the debt-service coverage ratio, or DSCR, which compares available cash with scheduled principal and interest. Fast collections do not help much if operating losses consume the cash before the debt payment is due.

HSBC Innovation Banking makes the point in its recurring-revenue lending product. It requires independent collateral and “harvest” audits, first-ranking security, minimum-liquidity or recurring-revenue conditions, and collections through dedicated HSBC accounts. The facilities typically run for 12-36 months. The bank is not lending because a chart looks persuasive. It is building a way to test, monitor and capture repayment.

The three things data cannot do alone

Put d.light, M-KOPA, PayJoy and Zilch side by side and a pattern appears that none of the announcements says outright.

Each company had payment data. None was financed on data alone.

M-KOPA, the Kenya and Uganda business providing financed products through customer repayments, received its first commercial debt from Standard Bank, the lender that began financing the company in 2016. In May 2023, Standard Bank led and arranged more than $200 million of sustainability-linked debt alongside IFC and other development lenders.

IFC, a development lender participating in the financing, described its portion as senior secured loans of up to $50 million in dollars or Kenyan shillings and $15 million equivalent in Ugandan shillings. M-KOPA's 2025 accounts show borrowing-base tests for eligible receivables and inventory, all-asset debentures, restricted cash and debt in operating currencies. External bank rates ranged from 9.61% to 30.25% in 2025.

The interesting part is what M-KOPA did not get to ignore. High-frequency mobile payments did not make collateral, security or currency risk vanish. Its multi-year repayment history became useful because it was converted into secured, monitored facilities, partly matched to operating currencies.

M-KOPA’s debt story rests on secured, monitored receivables and operating-currency funding, not mobile payments alone. Photo: Nicholas Githiri / Pexels, Pexels licence (free commercial use).

Then there is PayJoy, the consumer-credit company operating across Mexico, Brazil, South Africa and other Latin American markets. In September 2023, it raised $210 million of debt led by Citi, alongside $150 million of equity. It had already provided more than $2 billion of credit to over eight million customers.

PayJoy's financed smartphones served as collateral. Its technology could restrict device functionality after non-payment.

That may sound severe. It is also revealing. For a lender, data that predicts default is useful. Data paired with a practical response to default is more useful still. PayJoy's case was not merely about knowing which customers might miss payments. It was about having a credible route to recover value when they did.

PayJoy paired behavioural payment data with a controllable smartphone recovery mechanism. Photo: Israel Torres / Pexels, Pexels licence (free commercial use).

Zilch, the UK consumer-finance business, offers the legal-structure version of the same lesson. Its audited FY2025 accounts record a £100 million Deutsche Bank securitisation facility established in June 2024, expanded to £150 million in October. Collateralised loans were transferred into a dedicated structure.

That detail matters. The facility did not simply rest on Zilch's corporate growth story. It financed an isolated pool of loans through a payment waterfall, meaning pre-agreed rules for where collections go first. The bank had a defined asset pool and a defined claim on the resulting cash.

Zilch’s facility financed a legally isolated pool of loans with a defined payment waterfall. Photo: Myotus / Wikimedia Commons, CC BY 4.0.

Permission is not protection

Legal assignment is where many otherwise impressive collections stories become thinner than they first appear.

Kenya's Movable Property Security Rights Act 2017 expressly includes receivables and outright transfers. Nigeria's 2017 secured-transactions regime recognises accounts receivable and provides a collateral registry. Those legal frameworks make it possible for receivables to support security interests.

Possible is not the same as done.

A lender still needs to establish priority, register its interest where required, test whether the debtor can be required to pay, control the collection account and understand how recovery works in practice. A legal right that cannot be enforced or a payment route that can be diverted is of limited comfort when a borrower is distressed.

That is why d.light's Nigerian deal mattered. Chapel Hill Denham closed the transaction with d.light, but the resilience came from the wider architecture: naira receivables, naira funding and African Frontier Capital's role as master and backup servicer. The structure was designed not only for routine payments, but for the possibility that routine would end.

The UK offers a different kind of help. The 2018 Regulations generally invalidate contractual bans on assignment for qualifying post-2018 business contracts. But the rule has exclusions, including financial services, large suppliers, special purpose vehicles and several project contracts. A lender cannot safely assume that every invoice can be transferred just because the business operates in the UK.

In India, the Trade Receivables Discounting System shows another route. Reserve Bank of India rules allow banks and approved non-bank factors to finance buyer-accepted MSME receivables, generally without recourse to the seller. In September 2026, the five platforms financed 178,431 of 190,471 uploaded factoring units, or 93.7%, worth ₹408.2 billion.

The twist is that the startup's dashboard is not the centre of the credit case. Buyer acceptance is. The lender is largely assessing a confirmed obligation of a corporate or government buyer, while platform controls reduce the danger that the same invoice is financed twice.

When clever structures fail

Greensill Capital is the case that should make every founder and investor pause before treating securitisation as a magic word.

The company entered administration in March 2021. The UK government's Boardman review recorded that BaFin, Germany's authority examining Greensill Bank's claimed receivable purchases, could not verify receivables the bank said it had acquired from GFG Alliance companies. Some structures involved future receivables, concentration and reliance on trade-credit insurance.

The failure was not a shortage of financial engineering. It was an absence of reliable proof that the underlying assets existed independently and could be trusted.

This is where most people stop looking. They see insurance, a structured vehicle or a large facility and assume somebody else has checked the assets. Greensill shows why that assumption can be dangerous. Sophisticated packaging cannot turn unverified or highly concentrated receivables into ordinary invoices.

The evidence does not prove that every early-stage company must wait years before borrowing. Nor does it prove that every receivables pool needs securitisation. A simple facility can work where receivables are clear, customers are diversified, cash is controlled and security is enforceable. But the more uncertain the assets, the more the lender needs another source of comfort: stronger controls, a credible recovery path, local-currency funding or a better-capitalised participant.

For a related example of this underlying logic, see warehouse receipt finance. The asset changes. The lender's questions do not.

What founders should build before the pitch

Do not spend the next twelve months making a prettier revenue dashboard. Build an underwriting file.

First, reconcile every contract, invoice or customer loan to actual bank or mobile-money collections. A lender should be able to trace what was due, what arrived, what is overdue and why.

Next, separate the eligible pool from late, disputed, cancelled or highly concentrated payments. Measure payment regularity, ageing, defaults, recoveries, customer concentration and the currency in which each payment arrives.

Then test the legal route. Can the receivable be assigned? Can security be registered? Who has priority? Can collections be directed into a controlled account? Kenyan and Nigerian law can support the starting structure, while UK assignment rules may help in qualifying contracts. None removes the need for transaction-specific diligence.

Finally, build a downside repayment model. Include principal, interest, fees and the cost of currency risk. If customers pay in naira but debt is in dollars, saying the company will “manage” the gap is not a repayment plan.

GI Network's view: Treat debt readiness as an exercise in proving what survives when the founder's optimism is removed. The asset must still be verifiable, the cash must still be controllable and the lender must still have a recovery route.

For uncertain product development or early growth, equity or a grant may be more honest than fixed repayments. How small startup grants work explains why the purpose of the money matters as much as its price. Debt extends runway only when the asset it funds produces the cash to repay it. Otherwise, it can shorten runway with impressive efficiency.

GI Network would begin by testing the actual receivables pool before investor outreach: reconciling collections, mapping concentration and currencies, identifying the security and assignment questions, and modelling the base and downside repayment cases. From there, GI Network would assess whether the cash flows can support a borrowing-base facility, require a more resilient local-currency structure, or point instead to equity, grant funding or a smaller non-bank pilot.

What experienced investors check first

First-time founders often see debt as less dilutive equity. Experienced investors see a fixed claim with deadlines.

They ask whether the company is using debt to finance assets that turn back into cash, or to conceal operating losses. They want the borrowing-base assumptions, arrears profile, concentration limits, collection-account controls and downside DSCR case. They also ask what happens if servicing fails, as d.light's back-up arrangement did, or if receivables prove hard to verify, as Greensill's did.

The investor's concern is not anti-debt. It is anti-self-deception.

Use the VCR Test: Verify, Control, Recover.

1. 1.Verify: Can an independent party confirm what is owed, by whom, when and in what currency?
2. 2.Control: Can the lender hold enforceable security, establish priority and direct collections through a controlled route?
3. 3.Recover: If the originator fails or customers default, is there a practical servicing, collateral or legal recovery path?

If one answer is weak, collections data is still intelligence. If all three are strong, it may be the beginning of a financeable asset.