The wrong kind of gap

Africa's economy grew faster than the global average last year. Real GDP across the continent strengthened from 3.4% in 2024 to 4.5% in 2025, outpacing a global economy that slowed to 3.4% over the same period. Merchandise trade expanded by 6.1%, reaching approximately $1.5 trillion. By most measures, African trade is moving in the right direction.

‍

The infrastructure financing that trade, however, is moving in the opposite direction.

‍

Africa's trade finance gap, the difference between what businesses need to fund cross-border trade and what the financial system actually provides, stands at approximately $74 billion in 2025. The correspondent banking relationships that have historically underpinned that financing are declining. And the global banks that built those relationships are, quietly and systematically, pulling back.

‍

This isn't a case of a broken system staying broken. It's a case of a broken system getting worse, at exactly the moment African trade needs it most.

‍

‍

Why global banks are leaving

The term for what's happening is de-risking. It sounds technical and neutral, but it describes something with real consequences: major international banks terminating or scaling back correspondent banking relationships in markets they've decided are more trouble than they're worth.

‍

The decision is driven by a simple and brutal compliance calculation. Maintaining a correspondent banking relationship in a frontier market requires ongoing due diligence, anti-money laundering checks, KYC processes, and regulatory reporting. These costs are fixed regardless of transaction volume. The revenue generated from those markets, particularly smaller African economies, rarely justifies the compliance overhead. And crucially, the downside risk is asymmetric: a missed compliance violation can result in fines that dwarf years of revenue from the relationship. The rational response, from a compliance department's perspective, is to exit.

‍

Correspondent banking relationships dropped approximately 15 to 20% globally between 2010 and 2023, with the decline concentrated most heavily in Africa, South Asia, and the Middle East. The IMF, World Bank, and multiple financial standard-setting bodies have all acknowledged that this trend harms financial stability and creates the opposite of what regulators intend: by pushing transactions out of the formal banking system, de-risking pushes them toward informal channels with even less oversight.

‍

The acknowledgment hasn't reversed the trend.

‍

‍

The names pulling back

The withdrawal isn't abstract. It has names attached to it.

‍

Société Générale divested its subsidiaries in Chad, Mozambique, Morocco, and Madagascar in 2024 as part of a deliberate strategy to simplify its model and improve its risk profile. What remains of its Africa presence is a leaner, more selective footprint, focused on markets where it can combine acceptable returns, controlled risk, and strategic relevance. French banks more broadly are pivoting away from local retail branches toward sovereign debt and strategic corridors, a fundamental shift in how they engage with the continent. BNP Paribas is in the process of exiting Morocco, with its sale of BMCI expected to close in late 2026.

‍

These aren't banks that stumbled into Africa by accident. Société Générale and BNP Paribas built significant retail and commercial banking presences across the continent over decades. Their exit isn't a correction of an earlier mistake. It's a reassessment of where frontier market risk fits in a world of tightening Basel capital requirements and rising compliance costs.

‍

For African businesses and financial institutions, the consequence is a shrinking pool of international banking relationships available to anchor their cross-border transactions.

‍

‍

Who actually pays

The withdrawal of correspondent banking doesn't affect every business equally. Large multinationals have alternatives: they can source foreign exchange through parent companies, access credit lines from international lenders directly, and navigate around gaps in the correspondent network. For them, de-risking is an inconvenience.

‍

For small and medium-sized businesses, it's a structural barrier.

‍

SMEs account for at least 80% of businesses across Africa and more than half of GDP. Yet banks approved only 63% of SME trade finance applications on average, compared to an 80% overall approval rate for all applicants. The rejection rate isn't primarily a credit risk problem either. SME default rates on trade finance, while higher than those of larger clients, remained stable and predictable over the study period, suggesting the problem isn't that SMEs are bad credit risks. It's that the compliance cost of serving them, relative to the transaction sizes involved, makes them commercially unattractive to the banks still in the market.

‍

In ECOWAS, trade finance supported only 25% of goods trade, well below the African average of 40% and the global average of 60 to 80%. Nine out of ten local banks in the region reported difficulties meeting the requirements of foreign correspondent banks. The demand is there. The financing isn't.

‍

The consequence is a continent where trade is growing, but where the financial infrastructure to support that growth is rationing access, and rationing it most severely to the businesses that drive the most employment and economic activity.

‍

‍

Who is filling the gap

The retreat of European and North American banks is real. But the vacuum isn't entirely unfilled.

‍

Something significant has shifted in who the dominant correspondent banks for African institutions actually are. Where international surveys once showed only two African banks in the top ten confirming banks for African issuers, the most recent data covering 2020 to 2024 shows six of the top seven are now African institutions, led by UBAF, Banque Centrale Populaire, Absa, Crown, EBI, and BMCE. The retreat of European banks and the growing capacity of homegrown institutions are happening simultaneously.

‍

Mobile money is also expanding far beyond its original use case. West Africa alone processed approximately $498 billion in mobile money transactions in 2025, supported by more than 517 million registered accounts. Platforms like MTN MoMo and Orange Money have moved from simple person-to-person transfers into merchant payments and early forms of trade-related finance. The first wallet-based cross-border payment corridor between Nigeria and Ghana launched in early 2026 through a partnership with Onafriq and PAPSS, offering fast, low-cost settlement in local currencies.

‍

Stablecoins are also emerging as a meaningful part of this new architecture. Unlike traditional cross-border transfers that route through multiple correspondent banks before reaching a beneficiary, stablecoins enable value to move almost instantly across networks. For payment providers, this reduces settlement times and improves liquidity management. For businesses, the underlying technology is largely invisible. What matters is that payments arrive faster and capital stays available for longer. Stablecoins are unlikely to replace banks. They are far more likely to become another settlement layer beneath existing financial institutions, much like cloud computing became infrastructure beneath modern software rather than a replacement for it.

‍

These are not complete replacements for the correspondent banking infrastructure that's being withdrawn. But together, they represent a different architecture for how cross-border transactions can move, one built around connectivity rather than correspondent networks, and one that doesn't depend on a shrinking pool of international banks willing to maintain African relationships.

‍

‍

This isn't a banks-versus-fintech story

The conversation around financial services often frames banks and fintechs as competitors. In reality, the direction of travel is increasingly collaborative.

‍

Banks continue to provide regulatory oversight, deposits, lending, and the institutional trust that businesses and regulators require. Fintech infrastructure providers contribute connectivity, automation, developer tools, and direct access to multiple payment networks. Increasingly, businesses don't choose between them. They use both. The bank remains the financial institution. The fintech becomes the infrastructure connecting businesses to multiple financial systems simultaneously.

‍

That distinction matters because it reframes what "fixing" the cross-border payment problem actually looks like. It isn't about replacing the banking system. It's about building the connective layer that lets businesses move across it more directly, without depending on a correspondent network that is, by design, getting smaller.

‍

‍

The window that's opening

There is an uncomfortable irony in this moment. The withdrawal of global banks, driven by compliance cost calculations made in London and Paris, is creating space for African-owned institutions and fintech infrastructure to build something designed specifically for how African trade actually moves, rather than adapted from infrastructure built for different markets and different eras.

‍

That opportunity is real. It is also time-sensitive. A trade finance gap that stands at $74 billion today could widen to between $86.6 and $102.6 billion by 2027 under moderate to severe geopolitical stress scenarios. Every year the gap persists, businesses that should be trading across borders aren't, deals that should be happening don't, and growth that should be compounding stays flat.

‍

Cross-border payments are no longer simply an operational function. They have become a competitive advantage. Businesses that settle suppliers faster build stronger commercial relationships. Companies that reduce foreign exchange costs improve margins. Platforms that can collect and pay out locally across multiple African markets expand more efficiently than those still routing everything through a correspondent network built for a different time.

‍

The cross-border payment problem in Africa isn't just broken. It's broken and being actively withdrawn. The question for any business trading across African borders is whether their financial infrastructure is built for what's replacing it, or for what's leaving.

‍