Every payment operator has an employee that earns nothing, never shows up to work, and cannot be dismissed. It is not a person. It is capital, millions of it, sitting dead in foreign bank accounts across Lagos, Nairobi, and Johannesburg, parked in naira, shillings, and rand for the sole purpose of making an instant payout look instant. It is the single largest cost most operators have never counted, and understanding why it exists, what it costs, and how it ends is, I would argue, the most important conversation in African financial infrastructure today.
The mechanics are simple. A payment service provider moving value from Europe into Nigeria, Kenya, and South Africa cannot settle in real time. The correspondent banking rails it depends on do not clear instantly, and the local markets it pays into demand funds be available before a transaction arrives. So the operator pre-funds: it locks working capital in local banks to cover daily volume before a single payment clears. That capital is not working. It is held hostage to the requirement for speed.
Hostage capital is expensive in ways that compound. It earns nothing. It cannot be deployed into the business. And it sits exposed to the depreciation of every currency it is forced to hold, currencies that, across much of the continent, have a long history of losing value against the dollar. On a single corridor, the difference between a lean operation and a bloated one can be the gap between tens of millions in trapped pre-funding and a modest operational float.
Your most expensive employee earns nothing, never shows up, and you cannot fire it.
Three symptoms, one disease
The pre-funding trap has evaded serious treatment because it does not present as one problem. It presents as three, and the industry has spent years treating them as unrelated. They are not. They are three symptoms of a single capital-structure condition.
Consider foreign-exchange leakage. There is no deep, direct market between the naira and the shilling, so value moving between them is routed through the dollar, two conversions and two spreads, and by the time the money lands, somewhere between five and eight percent of it has quietly evaporated. Operators experience this as an FX problem. It is, in fact, a direct consequence of pre-funding: because capital is pre-positioned in local currency rather than held in a neutral asset, every cross-currency movement must traverse the dollar bridge, and every traverse leaks.
Consider settlement latency. Payments still crawl through correspondent chains at three to five days, and the industry treats this as a speed problem to be solved with better technology. But the deeper cause is capital, not code. The liquidity required to settle faster is the same liquidity the operator has locked away in pre-funded accounts. The latency and the trap are the same phenomenon observed at different moments.
Consider corridor expansion. Every new market an operator wants to enter demands another pre-funded account, another balance-sheet commitment, another reason for the finance function to say no. Growth that should be a commercial decision becomes a capital-allocation constraint. Operators experience this as a strategy problem. It is a liquidity problem wearing a strategic mask.
The treasury lens the industry has been missing
These three symptoms have never been unified because the industry has approached them through an operations lens rather than a treasury one. Seen operationally, FX leakage, latency, and stalled expansion look like distinct workstreams for distinct teams. Seen through treasury, they are a single question about the efficiency of capital, and treasury has the tools to price precisely what operations can only feel.
The first tool is classification. Not all pre-funded capital is the same, and treating it as one undifferentiated pool is the original error. It separates into three. There is, working float, capital actively clearing payments. There is buffer float, the safety margin held against volume spikes and settlement timing, necessary but frequently oversized because, absent settlement certainty, treasury teams overfund it to avoid failed payouts. And there is dead float, the remainder, sitting in the account earning nothing, protecting against nothing, simply trapped. In most operations, dead float runs between twenty and forty percent of the total pre-funded balance. On a corridor holding twenty-five million dollars of float, that is five to ten million dollars doing nothing while it depreciates.
No operator can optimise capital they have never classified, and most are managing a single number where they ought to be managing three. Much of what is labelled a buffer is, on inspection, a dead float wearing a safety label. A useful corollary follows: when settlement certainty improves, the required buffer shrinks, which is itself a source of released capital, before a neutral asset does any further work.
The second tool is the honest calculation of carrying cost, which is not one number but two: the depreciation the currency suffers while held, plus the risk-free yield the capital forgoes by sitting idle rather than being deployed. Run corridor by corridor, float held multiplied by that currency’s real depreciation, plus forgone yield, summed across every corridor, and the result is a number most operators have never calculated, and almost always the most expensive figure in the business.
Why pre-funding was never a law of physics
Pre-funding was never inevitable. It was a workaround, a rational response to the absence of a neutral, instantly settling reserve asset that every corridor would accept on demand. The nostro and vostro model that underpins correspondent banking was built to bridge that gap: hold balances abroad so payouts need not wait for clearing. It solved for speed by sacrificing capital efficiency, and for decades there was no alternative to that trade.
But a workaround is not a permanent truth. If a neutral, hard-backed settlement asset existed, one that every corridor would accept, that consolidated the fragmented pools into a single mobile reserve, and that could be redeemed into local currency faster than the corridor’s own payment cycle, the entire logic of pre-funding would collapse. The trapped capital would come home, not as a payments upgrade but as working capital released into productive deployment. A nostro balance is trapped per corridor and can serve no other; a single neutral reserve serves every corridor at once, so aggregate pre-positioning collapses even where any individual unit still moves through a cycle.
That is a treasury conversation, not a payments conversation, and it is the conversation around which I believe the next decade of African financial infrastructure will be built. The shift is from asking how much capital to pre-position and where, to asking how quickly it can be redeemed on demand.
The operating system beneath the asset
An asset alone, however well designed, does not move money. It requires an operating layer around it, orchestration that connects payment providers, banks, liquidity venues, foreign-exchange execution, and settlement into a single system rather than a web of point-to-point integrations. Conventional payment orchestration stops at authorisation, routing a transaction to the rail most likely to succeed. The layer this problem demands goes deeper, into settlement and capital, so that the routing decision and the treasury decision become the same decision. For an operator running many corridors, the practical effect is to stop managing capital corridor by corridor and start managing it as one intelligently routed pool.
Why a hard-backed, neutral asset, and why now
If the settlement asset is the bridge, its properties matter. A dollar-denominated instrument settles quickly but re-exposes the holder to a single monetary policy, to issuer concentration, and increasingly to local regulatory restriction on dollar-pegged instruments. A hard-backed, politically neutral reserve asset, gold being the oldest and most universally accepted example, avoids that concentration and resists inflation by its nature. Neutrality, in that light, is not a footnote. It is a feature.
For most of history, gold was a store of value one held and rarely moved, vaulted, static, settling nothing. Making it a working settlement instrument required three things to become true at once: genuine one-to-one physical backing with auditability, so a unit is a claim on real vaulted metal rather than a synthetic exposure; local on- and off-ramps, so it can be minted from and redeemed into local currency on demand; and institutional-grade rails, fast, secure, and observable enough that a regulated treasury can rely on them. Gold was never illiquid. It was unmobilised. The difference is infrastructure.
The condition that keeps the thesis honest
It would be easy, and dishonest, to end there, with the clean promise that a neutral asset dissolves the trap. It does not, not on its own, so here is the condition, stated plainly.
A neutral settlement asset only escapes the pre-funding trap if it settles faster than the corridor’s payment cycle. If redemption is slower than the cycle, the operator is forced back into pre-positioning the asset in advance, and a neutral asset that must be pre-positioned is simply the old trap in new clothing. A nicer waiting room, but a waiting room still. The asset is never the answer. Capital velocity is, and the asset merely makes the velocity possible.
This is what I have come to call the velocity test, and it separates genuine solutions from sophisticated repackaging. It requires measuring three things: the corridor’s payment-cycle time, the redemption speed of the neutral asset into destination fiat, and, the genuinely binding variable, the depth of local off-ramp liquidity at the destination. That last dependency is the hard engineering problem hiding beneath the elegant thesis, and any operator or provider who waves it away is not to be trusted with your balance sheet.
Choosing the liquidity that makes it real
Because the whole model rests on destination liquidity, how an operator selects its providers is not a procurement detail. It is central to whether the thesis survives contact with a volatile Tuesday. Most operators negotiate a single number, the spread, because it is the visible one. But a liquidity relationship has five dimensions, and four only announce their importance under stress: pricing, depth, reliability, corridor coverage, and counterparty risk. The dimension most consistently underweighted is reliability under stress, and cheap liquidity that vanishes in a squeeze is the most expensive liquidity there is.
The number to run today
For any executive who runs a corridor, a desk, or a treasury, the pre-funding trap stops being an abstraction the moment it is measured. Before anything else, calculate your holding cost. Holding cost equals your total pre-funded balance, multiplied by the depreciation rate of every currency you are forced to hold, plus the risk-free yield you forgo on that idle capital. Segment the float first, applying the calculation to the dead float and the oversized buffer, not to genuinely working capital. Use real depreciation, not headline inflation. Sum it across your corridors, and the figure that results is what the silent tax costs you every year.
Most operators have never run that number. The largest cost on the balance sheet is the one nobody has been asked to own. And once you have it, the question becomes who owns it going forward, which is a governance question most organisations have never answered. The remedy is a discipline of re-underwriting rather than review, asking not whether a corridor decision is still being followed but whether the assumptions that justified it still hold, challenged from first principles on a defined cadence, with a single accountable owner. In most operations, the honest answer to when a position was last re-underwritten is never.
The future of cross-border payments in Africa will not be won by whoever moves money fastest. Speed is necessary, but it is not the prize. It will be won by whoever makes capital work hardest, by whoever finally releases the most expensive employee no one ever hired, and puts it back to work building the commerce of a continent.
- business a.m. commits to publishing a diversity of views, opinions and comments. It, therefore, welcomes your reaction to this and any of our articles via email: comment@businessamlive.comÂ
Nigeria elite: Agents of future destruction or development? (2)