Why Payments to Foreign Suppliers Stall in Africa — Even When the Dollars Exist

A treasury team in Lagos, Lilongwe, or Luanda can be sitting on more than enough local currency to pay a foreign supplier and still miss the payment by months. The local money is there. The dollars, somewhere, exist. What's missing is the layer that converts one into the other and moves it across a border on the timeline international trade actually runs on.

This is the distinction most coverage of Africa's "dollar shortage" misses. Dollars are scarce and dollars won't move are two different problems. The first is a question of reserves — does the country have enough hard currency? The second is a question of plumbing — can this specific importer get dollars to this specific supplier this week? A country can fail the second test even when it passes the first. That gap is where FX controls, central-bank allocation queues, correspondent-banking de-risking, and trapped local-currency balances do their damage — and it is where most of the operating pain for businesses trading into Africa actually lives.

This guide explains the mechanism layer underneath the headline: why available dollars stall on a trade timeline, how FX controls differ from a true shortage, why correspondent banking can't fix it, what the picture looks like country by country, what treasury teams do in practice, and where cross-border settlement is heading.

If the dollars exist, why does my supplier payment still stall?

Because the constraint is the settlement and allocation channel, not the buyer's balance. A payment to a foreign supplier from an FX-constrained market runs a fixed sequence: the importer deposits local currency with a commercial bank, the bank submits a request for a foreign-exchange allocation, that request joins a queue — formal or de facto — for the central bank to release dollars, and only then does the payment move through the correspondent-banking chain to the supplier's bank. Each link can stall independently. The money sitting in the importer's account never touches the part of the system that is actually jammed.

The result is the pattern that defines trade in these corridors: businesses with cash and demand, unable to pay. The clearest evidence is from the firms themselves. The African Development Bank's 2025 Trade Finance Report, released in May 2026, found that 36% of African banks now cite limited foreign-exchange liquidity as the single biggest constraint on their trade-finance business — double the 18% of five years earlier. The same report put Africa's trade-finance gap at roughly $74 billion and noted that commercial banks now intermediate only about 23% of the continent's trade, down from around 40% a decade ago. Banks are retreating from the corridor, and FX liquidity is the reason they name first.

What changes in a crisis is not the demand for goods. Importers still need fuel, fertilizer, medicine, and manufacturing inputs, and they still have the local currency to pay for them. What changes is the system's ability to convert that local currency into hard currency fast enough to matter. That is a settlement problem wearing the costume of a shortage.

What are FX controls — and how do they differ from a dollar shortage?

FX controls are the rules and mechanisms a government uses to manage who gets access to scarce foreign currency, and when. A dollar shortage is the underlying condition — not enough hard currency relative to demand. The controls are the rationing system built on top of the shortage, and they are what an importer actually collides with.

A few terms recur across every constrained market:

• FX allocation queue — the line a bank's dollar request waits in for the central bank to release foreign currency. It may be a published policy or an unadmitted practice; either way, it is where payment time disappears. The IMF has repeatedly urged Sub-Saharan African central banks to let exchange rates adjust and avoid distortive administrative FX measures — guidance that exists because rationing is widespread.
• Authorized Dealer — the licensed bank through which exchange-control rules require trade payments and export proceeds to be routed. This is the catch for compliant firms: a business with audit obligations cannot use the informal channels that absorb the overflow, so it is forced into the slowest, most rationed path.
• Parallel rate — the unofficial exchange rate that emerges when the official rate can't clear demand. The size of the gap between official and parallel rates is a direct readout of how binding the controls are.
• Trapped cash — local-currency balances a business has earned or holds but cannot convert and move out, because the FX to externalize them isn't being allocated.

The distinction matters because the two problems have different cures. More reserves, more exports, or more aid ease the shortage. None of them, on their own, fix the allocation channel. A country can add reserves and still leave importers in the queue.

Why correspondent banking can't fix it

Because the correspondent network that moves dollars across borders has been contracting in Africa faster than almost anywhere else. Cross-border dollar payments depend on correspondent banking relationships — the accounts banks hold with one another to clear currency they don't issue. Over the past decade, global banks have shed these relationships to cut compliance cost and risk, a process known as de-risking. Bank for International Settlements data shows the number of active correspondents fell roughly 22% globally between 2011 and recent years — and de-risking has fallen hardest on regions like Africa, which lose dollar-clearing access disproportionately. The AfDB explicitly names "tighter correspondent risk appetite" as a force that could widen the trade-finance gap further.

When a local bank loses a correspondent, its clients don't just pay more — some can't transact at all, or get rerouted through longer, costlier chains via the UAE or Europe. Research on de-risking finds that firms whose banks lose correspondent access export significantly less, with smaller firms hit hardest. So the rail that is supposed to be the answer to a settlement problem is itself thinning out, on exactly the corridors that need it most. Correspondent banking is not a fix for FX-constrained trade; it is one of the places the constraint shows up.

How does the FX bottleneck differ across Malawi, Mozambique, Angola, Zimbabwe, and Nigeria?

The shortage is regional, but each market expresses it through a different failure mode — and the set of them, taken together, is the argument. What follows is the structural pattern, with each country as an archetype; the live readouts for each corridor — current reserve cover, wait times, parallel-rate spreads — move quarter to quarter and live in the cluster posts.

Malawi — the allocation queue at its most acute. A thin export base against an essential import bill — fuel alone costs the country roughly double its tobacco earnings — keeps reserves chronically below the safe buffer, so foreign exchange is released through a months-long queue and in tranches too small to clear an invoice. It is the clearest case of cash-rich, dollar-poor: firms hold the kwacha equivalent of a supplier bill while the dollars are rationed out. The full mechanics, with current figures, are in how Malawi sold gold to buy fuel.

Mozambique — stability that masks gated access. A managed-stable currency and healthy reserves can sit directly on top of a slow allocation queue; the calm headline rate is partly the rationing. And the export base that funds the dollars can thin out quickly — the idling of a single smelter that was close to a fifth of national exports is the kind of shock that tightens FX inflows well before any replacement revenue arrives. The operational layer is in paying foreign suppliers from Mozambique.

Angola — the reserves paradox. This is the cleanest proof that reserves and settlement are different problems: a country can hold many months of import cover, run an oil surplus, and still leave importers waiting months on a letter of credit — because the bottleneck is the allocation channel, not the reserve level. Ample reserves, jammed plumbing.

Zimbabwe — a stable local currency is not the ability to pay. Even a low-inflation, gold-anchored domestic currency doesn't settle a foreign invoice. With most transactions still dollarized and an interbank market short of willing dollar sellers, a business can hold sound local money and still be unable to source the USD an overseas supplier requires. The binding constraint is hard-currency liquidity, not the headline exchange rate.

Nigeria — what it costs to actually fix it. Nigeria is the counter-example: the exit from rationing works, but the price is steep. Floating the currency and clearing a multi-billion-dollar audited FX backlog (settled in 2024) collapsed the parallel-rate premium and restored predictable importer access — at the cost of a currency that lost roughly two-thirds of its value and a float the central bank still actively manages. Ending the queue removed the delay by repricing everyone's dollars.

The five together make the pillar's point: FX controls and currency pegs don't eliminate the dollar-payment problem — they relocate it, into parallel-market premiums, import-cover shortfalls, or settlement delays. Which form it takes is a function of the corridor; that the problem exists is not.

What importers and suppliers actually do about it

They stop waiting on the official queue as their only channel and build redundancy around it. In practice that means a mix of the following, in rough order of how much governance they preserve:

• Hold and pre-position FX. Where allowed, treasury teams keep balances in foreign-currency accounts and convert ahead of an invoice rather than at payment time, so a known payable isn't hostage to an allocation cycle.
• Restructure contract and entity terms. Suppliers shift from open-account and letters of credit toward upfront or cash-and-carry terms once confidence in settlement timing erodes — a rational response that quietly raises the working-capital cost of every import. On the buyer side, some structure payments through an offshore entity or treasury hub that holds dollars directly.
• Batch and time payment runs around correspondent cut-off windows to avoid the multi-day delays that batch processing and weekends introduce.
• Use compliant settlement rails that bypass the allocation queue. This is the fastest-growing option, and the one most often confused with crypto speculation. Done properly, it is the opposite — a regulated settlement path, covered below.

What businesses with audit obligations cannot responsibly do is route trade through the informal market to escape the queue. Exchange-control rules require proceeds to move through Authorized Dealers, and a clean audit trail is non-negotiable. That constraint is exactly why the credible alternatives are the ones built to sit inside the regulatory framework, not around it.

Where cross-border settlement is heading

Toward settlement layers that shrink or sidestep the hard-currency bottleneck without leaving the regulated system. Two developments matter most.

Local-currency rails for intra-African trade. The Pan-African Payment and Settlement System (PAPSS) lets businesses pay and get paid in local currency, netting positions across the network and settling only the residual imbalance in hard currency. It now connects on the order of 16 to 18 central banks and 150-plus commercial banks, and is built to cut the cost of cross-border payments well below the roughly 8% that has been typical for African corridors. The boundary is important and often misstated: PAPSS is powerful for trade between African countries, but it does not help an importer pay a supplier in China or the EU in dollars. It shrinks the hard-currency need; it doesn't remove it.

Compliance-first stablecoin settlement. For the dollar leg — paying an overseas supplier — a regulated settlement layer using dollar-denominated stablecoins has moved from the fringe to serious operator use. The mechanics are simple: the local-currency leg circulates domestically, between residents who need it; the dollar leg clears offshore against existing USD liquidity. No central-bank allocation queue, no parallel-market premium, and — when it is done through licensed Authorized Dealers and regulated payment service providers — a full audit trail. This is not a workaround bolted onto trade. It is settlement architecture designed to live inside existing exchange-control rules, which is the only version that a CFO with audit obligations can actually use.

The honest framing is this: the structural import-export gap that drives the shortage will take years to close. The settlement layer — the part that decides whether a willing buyer with local currency can pay a willing supplier this week — can be fixed in months. That is the layer worth building on.

Common questions

Is this a dollar shortage, or a settlement problem?
Both — and they aren't the same. There is a genuine hard-currency shortage at the macro level in many African markets. Underneath it sits a settlement problem: even when dollars exist, the system can't convert and move them on a trade timeline. More reserves ease the first. Only a better settlement layer fixes the second.

What exactly are FX controls?
The rules and mechanisms — allocation queues, mandatory conversion of export proceeds, routing through Authorized Dealers, official rates that don't clear the market — that governments use to ration scarce foreign currency. They're built on top of a shortage and are what importers collide with in practice.

Why can't correspondent banking solve cross-border payment delays?
Because the correspondent network is itself contracting in Africa — active relationships have fallen sharply over the past decade as global banks de-risk. The rail meant to clear dollars is thinning on exactly the corridors that need it, adding cost and delay rather than removing it.

Does a stable local currency mean payments flow?
No. Zimbabwe shows a relatively stable, low-inflation local currency coexisting with barely 1.5 months of import cover and an interbank market short of dollar sellers. A business can hold sound local money and still be unable to source USD to pay a foreign supplier.

Won't more exports or aid eventually fix it?
Not on the timeline businesses operate on. The import-export gap is structural and will take years to close. The settlement layer is the part that can change in months.

Is compliant stablecoin settlement legal for a business with audit obligations?
That's the entire point of the compliance-first model. It runs through licensed Authorized Dealers and regulated payment service providers, with a full audit trail — exactly what informal cash workarounds can't offer.

The bottom line

For CFOs and treasury teams with exposure to African markets, the operative question is no longer when the official queue improves. It's whether the business can keep absorbing the cost of waiting on it — in trapped cash, in stretched supplier terms, in payments that clear in months instead of days. The reserves story will keep moving; the settlement story is the one you can actually act on.

This is the gap Artoh is built to close. If you have payables or receivables aging in Malawi, Mozambique, Angola, Zimbabwe, Nigeria, or elsewhere in the corridor — and the dollars are not moving — let's talk.

Related analysis

• Malawi Sold Gold to Buy Fuel. Here's What That Actually Tells You — the FX-allocation queue, traced end to end through a single corridor.