Correspondent Banking Risk: The Quiet Map of Who Gets Cut Off

TL;DR: Correspondent banks have been quietly closing relationships and corridors in emerging markets for over a decade, and the firms that pay the price are not the banks themselves but exporters and importers who never chose which bank sat in the middle of their payment chain. The withdrawal pattern is measurable, regionally concentrated, and worth tracking as a leading indicator before a payment actually fails.

Key takeaways:

• Active correspondent banking relationships and corridors fell roughly 50 percent across Africa and 40 to 45 percent across Asia and Eastern Europe between 2011 and 2022, against 20 to 30 percent declines in Europe and North America.[^1]
• Firms that lost a correspondent banking relationship saw their probability of exporting fall 5.2 percentage points in the short term and 19.8 percentage points four years later, with export revenue down 57 percent relative to similar firms.[^2]
• Across 17 emerging European markets, export growth ran 8 percentage points lower and import growth 24 percentage points lower in the countries hit hardest by withdrawal.[^2]
• A 2016-17 survey of 306 banks across 92 emerging markets found 27 percent had experienced reductions in their correspondent relationships, with sub-Saharan Africa the hardest hit region at 35 percent.[^3]
• FATF's own guidance discourages wholesale "de-risking" and calls instead for case-by-case due diligence, a standard banks have not consistently followed.[^4]

A manufacturer in Sarajevo does not choose which foreign bank its local bank uses to clear a dollar payment. That choice sits one or two steps upstream, inside a correspondent banking relationship the manufacturer has never seen and cannot renegotiate, and when that relationship closes, the manufacturer's payment chain closes with it. This is geopolitical risk arriving through a channel most finance teams do not watch: not a sanction, not a tariff, but a quiet compliance decision made between two banks the company never dealt with directly.

Where the Corridors Actually Closed

Correspondent banking is the plumbing behind cross-border payments: a bank in one country holds an account with a bank in another to clear transactions in that second bank's currency, most often dollars or euros. When the second bank decides the relationship costs more in compliance risk than it earns in fees, it closes the account, and the first bank loses its route into that currency.

• The decline is a decade old and uneven by region. Between 2011 and 2022, the number of active correspondents and corridors fell by roughly 50 percent across Africa and 40 to 45 percent across Asia and Eastern Europe, against 20 to 30 percent declines across Europe excluding Eastern Europe and North America.[^1]
• The trigger is compliance cost, not client risk. Banks that withdraw are responding to the cost of anti-money-laundering and sanctions screening on a relationship, not necessarily to evidence that a specific corridor carries elevated risk.
• Smaller, thinner corridors go first. A bank with few correspondent relationships to begin with has less room to absorb one closure, and its customers have fewer alternative routes.
• The pattern predates the current geopolitical cycle but has not slowed. A 2016-17 baseline survey already found 27 percent of banks reporting reduced correspondent relationships, sub-Saharan Africa hardest hit at 35 percent, across 306 banks and 92 markets representing about five trillion dollars in combined assets.[^3] The multilateral bodies tracking the trend since have not reported a reversal.

The result is a map of who still has a reliable route to clear dollars and who does not, and that map moves independently of any sanctions list or credit rating a company might already be checking. It is also a map almost nobody outside the banks themselves keeps. A sanctions list is public and searchable. A credit rating is published and dated. A correspondent bank's decision to close one more corridor is disclosed to no one but the losing local bank, often without much notice.

What This Costs a Company That Never Chose Its Bank's Bank

The clearest evidence on what withdrawal costs a company, rather than a bank, comes from a study of firms in Bosnia and Herzegovina, Croatia, Hungary and Turkey whose banks lost correspondent relationships. In the short term, affected firms saw their probability of exporting fall 5.2 percentage points against similar firms whose banks kept their relationships intact. Four years out, that gap widened to 19.8 percentage points, and the export revenue of firms that kept exporting was 57 percent lower than their peers.[^2] Smaller and younger firms, and firms whose banks had thin correspondent networks to begin with, absorbed the worst of it.

The same research extended the pattern across 17 emerging European markets and found export growth running 8 percentage points lower and import growth 24 percentage points lower in the countries where withdrawal hit hardest, alongside broader revenue declines, job losses and firm closures.[^2] None of these firms did anything to cause the relationship termination. They were downstream of a decision made between two banks, neither of which they had a commercial relationship with directly.

What we infer from this pattern: a corporate treasury that has never mapped which correspondent bank actually clears its cross-border payments, and how many alternative corridors exist if that relationship closes, is carrying an exposure it cannot currently see. What the record does not show is any single company-level early warning system that catches a closure before it happens. FATF's guidance calls for banks to manage correspondent risk case by case rather than exit entire regions or client categories, a standard the aggregate decline suggests is not being consistently applied.[^4]

Reading Withdrawal as a Leading Indicator

Treating correspondent banking risk as a leading indicator rather than an operational surprise means asking a small set of questions before a payment fails, not after.

• Which bank actually clears the payment, not just which bank holds the account. Most corporate treasuries can name their bank. Fewer can name the correspondent bank two steps upstream that makes a dollar payment possible.
• How concentrated is that route. A single correspondent relationship for a given currency corridor is a single point of failure, whatever the underlying credit quality of the local bank looks like.
• Has the local bank's correspondent count changed recently. A shrinking correspondent network is a visible, askable fact, not a hidden one, and it precedes payment friction rather than following it.
• What is the fallback route, and has anyone actually tested it. A backup corridor that exists on paper but has never cleared a live payment is not a backup, it is an assumption.

These questions belong on the same cadence as a supplier audit, not as a one-time onboarding check. A relationship that was intact at contract signing can close eighteen months later without the counterparty doing anything differently, because the decision was never about the counterparty in the first place.

None of this replaces sanctions screening or the mechanism-by-mechanism accounting of how geopolitical events actually reach a balance sheet, which correspondent withdrawal is one channel among several. It adds a question most finance teams have not yet learned to ask their own bank. Companies building that questioning into a standing review can start from the Fortius Intel brief.

Fortius Intel note: Ask your bank, by name, which correspondent clears your largest cross-border payment corridor, and how many alternatives exist if that relationship ends. If the answer takes more than a phone call to produce, that is the finding.

Methodology: figures drawn from Bank for International Settlements Bulletin No. 87 (Garratt, Koo Wilkens and Shin, May 2024), the CEPR VoxEU column reporting Borchert, De Haas, Kirschenmann and Schultz's firm-level correspondent banking research (September 2024), the International Finance Corporation's 2017 correspondent banking survey, and FATF's published guidance on correspondent banking due diligence.

---