Retiring the $81.8 billion figure: what the revised trade finance gap data says

We have cited the African Development Bank's estimate of an $81.8 billion annual African trade finance gap in our own published work, more than once. The figure has been the anchor number in almost every discussion of the subject for several years, including ours.

The measurement has been revised, and the revision is worth setting out carefully, because two of the three changes cut against the way the number is normally deployed, including by us.

What the revised measurement says

The updated work puts unmet trade finance demand in Africa between $74 billion and $92 billion for 2024. Expressed against the continent's total merchandise trade value, that is roughly 5.4 per cent.

Three things follow from stating it that way rather than as a single headline figure.

The range is honest about the measurement. Unmet demand is estimated from bank survey responses about rejected applications, and rejection is not a clean proxy for viable demand. Some rejected applications should be rejected. A range reflects that uncertainty in a way a single number does not.

The denominator matters. A gap of 5.4 per cent of merchandise trade is a different claim from an unanchored $81.8 billion, and it is a more useful one, because it can be compared across regions and tracked as trade volumes change. It also makes clear that the gap is a coverage problem concentrated in specific segments rather than a uniform shortfall across all African trade.

And the revised central estimate is lower than the figure everyone has been quoting.

The direction of travel was the opposite of the story

The finding that should change how this subject is discussed is that unmet demand fell by close to 10 per cent between 2019 and 2024.

That decline is attributed to sustained intervention: multilateral development banks expanding trade finance guarantee programmes, export credit agencies extending cover, governments capitalising trade finance facilities, and global banks selectively re-engaging in specific corridors.

This is not the narrative the sector tells about itself. The standard framing is of a large and worsening structural failure. The measured position between 2019 and 2024 is a large structural failure that was slowly improving, in response to the specific instruments deployed against it.

The interventions worked. That is a more useful finding than the gap being large, and it is the one that gets least attention.

We should have been clearer about this in earlier writing. Our position that the gap reflects four reinforcing structural causes, correspondent banking withdrawal, risk-weighting, foreign exchange controls and SME exclusion, remains supported by the data. Our implicit framing that the position was static or deteriorating does not.

Why it is forecast to widen again

The same work projects the gap reaching $102.6 billion by 2027. A decline through 2024 followed by an expansion to well above the previous headline figure requires explanation, and the components are identifiable.

The first is regulatory. The Basel III endgame package began taking effect from July 2025, bringing higher capital requirements, a revised leverage ratio framework and tighter liquidity standards. Trade finance has long argued, with reasonable evidence from default statistics, that its risk weights overstate actual loss experience. The endgame does not resolve that argument in trade finance's favour. Where capital becomes more expensive to hold against a given exposure, the marginal exposure that gets declined is the one with the thinnest margin, and in African trade finance that is the SME tier.

The second is trade growth itself. If intra-African trade expands as the continental free trade area intends, with projections of intra-African exports rising more than 20 per cent within a decade, the financing requirement expands with it. A gap measured as a percentage of trade can hold steady or improve while the absolute number rises, purely because the denominator is growing. Some of the projected widening is a success indicator wearing the costume of a failure.

The third is that the interventions responsible for the 2019 to 2024 improvement are not open ended. Multilateral guarantee capacity is finite and allocated across competing priorities, and trade finance competes for it against climate adaptation, infrastructure and social lending in the same capital allocation. Export credit agency cover expands and contracts with the export policy of the sponsoring state rather than with African demand. Global bank re-engagement in specific corridors is a commercial decision that reverses when the corridor stops paying, and several of the re-entries recorded since 2019 were selective rather than structural. An improvement built on those three inputs is not self-sustaining, and a forecast that assumes it continues would be the optimistic case rather than the base case.

What the rejection data shows

The most operationally useful part of the revised work is the breakdown of why applications are declined. Weak client creditworthiness accounts for 48 per cent of rejections. Insufficient collateral accounts for 39 per cent.

Nearly nine tenths of rejections are therefore attributed to characteristics of the borrower rather than to the bank's capacity or appetite. That is significant, because these two causes have different remedies and neither remedy is advocacy.

Weak creditworthiness at the borrower level is what credit enhancement exists to address. A receivable owed by a creditworthy buyer does not become bankable by improving the seller's balance sheet. It becomes bankable by routing the credit decision to the obligor that can carry it, through insurance cover, a fund risk-taker or an export credit agency framework. The borrower's own credit stops being the binding constraint.

Insufficient collateral is a structuring problem. Trade transactions generate their own security in the goods, the documents and the receivable. Where a bank is asking for balance sheet collateral against a self-liquidating trade exposure, the transaction has usually been presented as a loan rather than as structured trade.

For intra-African flows the constraint set differs: limited foreign exchange liquidity at 24 per cent, insufficient correspondent bank limits at 20 per cent, risk capital constraints at 18 per cent. Those are system constraints rather than borrower constraints, and they are not fixed by credit enhancement. They are the reason intra-African trade finance remains harder than the equivalent EU-Africa corridor transaction despite shorter distances and smaller tickets.

What this means for an advisor

Use the range and the percentage, not the headline. A gap of $74 billion to $92 billion, at 5.4 per cent of merchandise trade, is defensible in a credit committee. A worn single figure invites the response that the number has not been checked.

Take the improvement seriously. If unmet demand fell 10 per cent because guarantees, ECA cover and multilateral capacity were deployed against it, then the instruments work, and the case for using them is empirical rather than aspirational.

And treat the rejection breakdown as the design brief. If 87 per cent of declines come down to borrower credit and collateral, then the work is to move the credit decision to a party that can carry it and to structure against the assets the transaction already generates. That is the same conclusion we have argued from first principles. It is more persuasive now that it can be argued from the data.