DFI “approval” isn’t money: Nigeria housing’s quiet gatekeepers

Most housing pitches still obsess over the price of money. The capital is often waiting behind a different door: collectible offtake and controlled, document-driven disbursement.

On 14 December 2023, a Nigerian developer called Modern Shelter Systems and Services Ltd got a kind of backing most housing projects never see.

InfraCredit announced it was putting an initial ₦3 billion debt financing guarantee behind a larger ₦12.5 billion programme to deliver 370 affordable housing units at Nasarawa Technology Village.

The press release is tidy. The human stake is not.

For the people who will end up living in those 370 units, the difference between a guarantee-backed programme and a “we’re still raising” project is not a spreadsheet nuance. It is whether blocks rise out of the ground at all, or whether the site becomes another promise that survives only in a brochure.

It mattered because it was one of those quiet Nigerian finance moments where the argument shifts. Not “can we get a cheaper rate?” but “can we make repayment believable?” Modern Shelter paired a domestic guarantee wrapper (InfraCredit) with a co-financier, Shelter Afrique Development Bank, and moved the conversation from price to proof.

Modern Shelter’s programme shows how structure, not cheap money, can unlock local-currency housing finance. Photo: Sesan Osanyinbi / Pexels, Pexels licence (free commercial use).

You can almost hear the reflex question from any boardroom: “Fine. But what’s the interest rate?”

That question is where most people stop looking.

The question underneath the rate question

If Nigeria’s affordable-housing pipeline is blocked mainly by expensive debt, high inflation and FX volatility, then cheaper money should be the unlock. Rate buydowns. Concessional lines. More government funding.

So why do so many “affordable” projects still struggle to reach first drawdown, while a structured deal can get institutions to lean in without Nigeria’s macro conditions turning into a fairy tale?

What surprised us, reading the World Bank housing finance restructuring paper tied to Nigeria alongside other World Bank legal and financing documentation that spells out escrow and conditions precedent mechanics, is how rarely the decisive gate is pricing.

The myth: affordable housing is mainly a price problem

The conventional view is clean: rates are high, so projects cannot carry the debt. Bring rates down and the system moves.

It is also the easiest story to tell because it has a single villain.

DFIs and credit enhancers worry about a different villain: not the rate, but whether the cash will leak before the homes are finished, and whether “demand” is demand you can really collect.

The twist: DFIs fund controls, not optimism

Put the Modern Shelter deal next to World Bank housing finance documentation tied to Nigeria, then contrast both with India’s RERA ring-fencing rule and Ghana’s Saglemi dispute cycle, and a pattern appears that none of the individual reports states plainly.

DFIs will tolerate a tough interest-rate environment more readily than they will tolerate weak collections proof and weak cash controls.

Why? Because those two things decide whether the cashflow is real under stress. They decide repayment performance and, when something goes wrong, how big the losses get.

People assume a DFI is a cheaper lender. Often, the DFI behaves more like a referee with a checklist.

And the checklist is about enforceability.

Case 1: The Nigerian deal that moved when the structure did

Modern Shelter’s programme at Nasarawa Technology Village was framed as guarantee-backed local currency debt with Shelter Afrique Development Bank as co-financier.

Two clarifications matter if you do not live inside Nigerian capital markets.

InfraCredit is a Nigeria-based credit enhancement institution. In plain terms, it provides a guarantee that helps make local-currency debt feel “investable” to cautious lenders.

Shelter Afrique Development Bank is a pan-African housing development finance institution. It matters here because it co-financed the programme, signalling that the deal was being built to pass a development lender’s bankability tests, not just a developer’s sales pitch.

The Nasarawa site where a guarantee-backed programme made 370 units financeable under Nigerian conditions. Photo: Audu Samson / Pexels, Pexels licence (free commercial use).

InfraCredit’s announcement was explicit: the initial ₦3 billion guarantee sits inside a ₦12.5 billion programme.

The detail that matters is not just the 370 units. It is the decision to use a domestic guarantee wrapper to make the repayment story credible for institutions.

A guarantee is not “cheap money”. It is a credibility transfer. It changes who can take the risk, and under what governance expectations.

Here is the twist that tends to land with founders: you do not have to wait for the naira to be “stable enough” to raise long-term capital. You have to give capital a structure it can survive in.

The wrong assumption the deal overturns is simple: “If the rate is low enough, the capital will come.” In reality, capital came when repayment risk was structurally mitigated.

Case 2: The paperwork DFIs keep putting between you and your money

World Bank programme documentation tied to Nigeria’s housing finance market development returns to the same operational reality: “conditions precedent to disbursement” and documentary checks before money moves.

This is not rhetorical caution. It is how DFIs operationalise risk: disbursement is treated as a compliance event, not a celebratory milestone.

Separately, World Bank legal and financing documentation in the brief goes as far as defining an “Escrow Agent” and requiring satisfaction of conditions precedent before first disbursement, including escrow-account related requirements.

Translate that into plain English and the whole fundraising timeline changes.

Approval is a door you are allowed to walk towards. Disbursement is another door entirely.

If you have ever wondered why a project can be “in advanced discussions” for months, this is usually why. The DFI is not negotiating your rate. It is asking whether the control architecture is installed and provable.

If you want a deeper feel for how capital committees actually think, the closest parallel in our archive is The deal doesn’t die in the pitch, it dies in the IC memo.

Case 3: India’s RERA rule and the industry’s admission of risk

India’s Real Estate (Regulation and Development) Act, the RERA regime, contains a blunt response to a familiar problem: diversion.

It requires that 70% of amounts realised from allottees be deposited into a separate project account. Withdrawals are allowed only in proportion to project completion and supported by certification, as described in practitioner and regulatory materials.

At first glance, this reads like paperwork. But it is actually a confession. The rule does not say: “Make loans cheaper.” It says: “Stop moving buyers’ money around.”

RERA exists because the market learned, painfully, that pre-sales without ring-fencing enabled developers to pull cash from one project to prop up another until everything stalled.

So India wrote the controls into law: ring-fenced collections plus certified, progress-linked withdrawals.

Notice what this implies for Nigeria. When DFIs fear diversion and slippage, they do not negotiate sentiment. They install discipline: controlled accounts, verification, and drawdowns linked to progress.

Case 4: Ghana’s Saglemi and why “escrow” can still leak

Ghana’s Saglemi Affordable Housing Project was designed for 5,000 units, with about $200 million of government financing, according to Ghana’s Ministry of Works and Housing.

Ghana’s housing ministry sits at the centre of the Saglemi story: the cost of disputed triggers and governance. Photo: أمين / Wikimedia Commons, CC BY-SA 4.0.

The project became a long dispute cycle.

Now we have to be careful with the popular shorthand version of Saglemi. One Ghanaian media report alleges the contractor was paid $179m for $64m work and describes an EPC structure with a 40% advance payment tied to receipt of funds in an Escrow Account. That allegation is not corroborated by the Ghana ministry statement in this brief, so treat the specific $179m versus $64m claim as an unverified contention rather than a settled fact.

Saglemi became a reminder that “escrow” is only as strong as the release rules and verification behind it. Photo: John Bravar / Pexels, Pexels licence (free commercial use).

Still, even without hanging the whole lesson on one contested number, the structure described in reporting is the warning worth keeping: large, front-loaded advance payments and weakly enforceable triggers are exactly how projects end up in courtrooms and parliaments.

Saglemi’s broader lesson for Nigeria is brutally practical. An escrow can still be a chute for cash if release conditions are loose, front-loaded, or politically contested.

So when DFIs ask Nigerian developers about escrow, they are not impressed by the noun. They are asking about the verbs:

• Who can authorise releases?
• What triggers them?
• What evidence proves progress?
• Who can stop payments when the project is off-track?

This is where most people stop explaining. It is also where most capital raises quietly fail.

Case 5: Mexico’s payroll deduction, and what “real offtake” looks like

INFONAVIT is Mexico’s national worker housing fund, and the brief’s evidence here comes from an INFONAVIT help-centre page describing how repayments are collected for one specific category of borrower: salaried workers.

It states that if you are salaried, the employer is obligated to withhold (retain) credit payments from salary and remit them to INFONAVIT.

That is not marketing demand. That is collectible demand, at least for the salaried segment described.

In housing finance, “offtake” is meant to answer a childishly simple question: who will pay, and how do you know they will pay?

Most projects stop at “we have interest” or “we have pre-sales”. INFONAVIT’s mechanism, as described on that help page, goes further. It embeds collections into payroll infrastructure.

A surprise, if you have never sat through a credit committee: the cleanest demand proof is not a queue of excited buyers. It is a system that does not require them to stay excited for 15 years.

The world tour in one paragraph: the same fear in different accents

In Nigeria, the winning move in the Modern Shelter programme was to wrap local currency debt with a guarantee and DFI co-finance so institutions can believe the repayment story.

In India, regulators forced a ring-fenced account and certified withdrawals because pre-sales were being used as fuel for diversion.

In Ghana, the escrow label did not, by itself, prevent a governance and payment dispute spiral.

In Mexico, collections became bankable because repayment is withheld at source for salaried borrowers, according to INFONAVIT’s help-centre guidance.

Different countries, same anxiety: money disappears long before it becomes housing.

The pattern DFIs rarely say out loud

Here is the principle you can repeat at dinner without knowing a single finance acronym:

Affordable housing finance is not unlocked by cheaper money. It is unlocked by proving that money cannot wander off and that buyers’ payments can be collected.

Developers often treat financing as a negotiation over terms. DFIs treat financing as a governance test.

And there is an uncomfortable implication in that. If your controls are weak, a lower interest rate can make you worse, not better. It can encourage bigger early disbursements under looser discipline, which increases the size of the eventual mess.

If this sounds familiar beyond housing, it should. The same dynamic shows up in other sectors where cash is promised but not controlled. See How Nigerian fintechs die: not out of cash, out of usable cash.

When the thesis breaks (and a Nigerian counterexample worth facing)

There are situations where pricing really is decisive, even when controls are strong.

Nigeria has a clean illustration in the very deal we opened with. Modern Shelter did not fix the macro environment. It used structure, via a guarantee and co-financing, to make local-currency debt possible.

That is also the limit.

Even with strong controls and reputable wrappers, a project can still be killed by pricing if the eventual buyer cannot afford the monthly payment. This brief does not provide Nigerian affordability data, mortgage statistics, or construction-cost sensitivity assumptions, so we cannot quantify that boundary. But the logic is hard to escape: you can perfectly control a cashflow that never arrives.

The useful takeaway is not “rates don’t matter”. It is “rates are rarely the first gate”. If you cannot prove collections and stop diversion, DFIs will not even reach the pricing argument with you.

For businesses: what to build before you talk to DFIs

Think of preparing your project as if disbursement is a courtroom, not a coffee meeting.

1) Turn “interest” into enforceable offtake

Do not just collect expressions of interest.

Bring offtake evidence that survives enforcement: who the buyers are, how they qualify, and how payments are actually collected.

INFONAVIT shows the cleanest mechanism in this brief: payroll withholding for salaried workers, as described on its help-centre page. If you cannot replicate that, spell out what your collection mechanism is, who controls it, and what happens under stress.

2) Treat escrow as a rulebook, not a container

If your pitch says “escrow”, assume the next question is: “Under what release conditions, with what verification?”

Saglemi’s dispute cycle shows the danger of loose triggers and front-loaded payments. The label does not protect the money.

3) Build verification into your construction plan

RERA’s core logic is progress-linked withdrawals backed by certification. You do not have to be in India to learn the lesson: if cash release is not tied to verifiable progress, you are asking funders to finance trust.

4) Assume disbursement is gated by conditions precedent

World Bank documentation underscores that DFIs operationalise risk through conditions precedent and evidence packages.

Treat CP satisfaction as a critical path item, not a legal footnote.

If you want a plain-English mirror from another market on gated capital and monitoring, UK Development Bridge Finance 2025-2026: Exit Risk and Monitoring is a useful comparison, even though it is not Nigeria.

For investors: what experienced committees check first

Investors like to talk about “demand” because it sounds commercial. They like to talk about “collateral” because it sounds safe.

In affordable housing development finance, experienced committees often start with two quieter questions.

Can we control where cash goes?

RERA is a system-wide answer: ring-fence buyer cash and release it only with progress-linked verification.

Modern Shelter’s guarantee-backed structure is another answer: shift repayment risk into a framework institutions can underwrite.

Saglemi is the warning: an escrow label and public financing do not prevent value destruction if release conditions are weak or disputed.

Is the offtake collectible under stress?

INFONAVIT’s payroll deduction mechanism, as described for salaried workers, is a reminder that the best offtake is built on enforcement infrastructure.

A credit committee cares less about whether buyers like the units and more about whether receipts arrive on time when the economy turns.

That is why investors obsess over controlled accounts and payment waterfalls, the governed order of who gets paid first. It is not bureaucracy. It is how they stop a project becoming a cash buffet.

GI Network's view: If your affordable housing raise is stuck, assume the objection is not “your rate is too high” but “your cash can leak and your offtake won’t collect”. Fix those two and the pricing conversation usually becomes possible.

What GI Network would do in this exact situation

GI Network would treat the raise as a bankability build, not an introductions exercise. Concretely, we would (1) map the project’s disbursement path into a conditions-precedent checklist consistent with DFI practice shown in World Bank documentation, (2) stress-test the offtake evidence to separate “interest” from collectible demand and identify what additional enforceable documentation is needed, and (3) align the escrow and controlled-account mechanics with progress-linked verification and a receipts waterfall so an investment committee can underwrite governance, not just hope.

The output is a data room built to survive a credit committee.

The takeaway tool: the Bankability Stack

If you remember one thing, make it this four-layer model. It is built from the cases above and it explains why so many raises stall.

The Bankability Stack (in order)

1. 1.Collectible offtake: demand you can enforce and collect (INFONAVIT-style withholding for salaried borrowers is the cleanest example here).

1. 2.Ring-fenced cash: project inflows kept in controlled accounts, not mingled and movable at will (RERA’s separate account logic).

1. 3.Verified disbursement: money released only against clear, evidence-backed progress (the intent behind RERA, and the opposite risk implied by disputed advance-payment structures).

1. 4.Credit comfort: enhancements that make local-currency debt investable for institutions (Modern Shelter × InfraCredit × Shelter Afrique).

Build from the bottom and you will keep arguing about interest rates.

Build from the top and you give DFIs what they are actually buying: enforceability.