CNBC’s 2026 fintech ranking, built with Statista, screened thousands of companies across seven segments and scored them on quantitative KPIs, growth metrics and industry influence. In the payments segment, one first-time entrant was XTransfer, a B2B cross-border trade payment platform.
The ranking itself is not the interesting part. The customer base underneath it is. XTransfer moves money for small and mid-sized exporters, which is the exact client type international banks have spent a decade removing from their books. A payments list that now recognises a company for serving that segment is telling you something about where the segment went.
The retreat is documented rather than anecdotal. BIS data shows active correspondent banking relationships fell by roughly 30% between 2011 and 2022, with a 4% drop in 2022 alone. Over the same twelve years, the volume and value of payments running through those networks rose by 61% and 37%. Fewer relationships carrying more money is concentration, not decline: the flows that survived went to the corridors and the clients large enough to justify the paperwork.
Banks have been fairly open about the reason. Compliance risk, not credit risk, is what they cite most often for cutting a correspondent tie. The consequence lands unevenly. A mid-sized importer in Rotterdam keeps its bank. A garment exporter shipping $60,000 invoices out of a jurisdiction with a poor AML reputation quietly discovers that no counterparty bank will carry the file.
It helps to be precise about which step is expensive, because it is not the payment. Clearing a wire is close to free. The cost is in onboarding: beneficial ownership checks, trade document review, counterparty and sanctions screening, then ongoing monitoring for as long as the relationship exists. Almost all of it is fixed per client rather than per dollar.
That single fact explains the whole pattern. A $60,000 invoice and a $60 million invoice cost about the same to verify, so the small exporter is structurally unprofitable at a bank’s cost base. No pricing cleverness fixes that, which is why a decade of “serving underbanked SMEs” pitches produced very little.
What does change the arithmetic is repetition. A platform that sees the same exporter’s invoices, shipping documents and buyers over and over is verifying an established pattern the second time, not a stranger. The bank behind it sees an anonymous wire and has to price the unknown. That asymmetry, not the user interface, is what makes these businesses viable, and it means the risk function is the product rather than a cost centre attached to it.
I would be careful about reading this as disintermediation. These platforms do not replace correspondent banking; they sit on top of it, holding licences and partner bank relationships that let them present many small clients as one well-documented flow. Their real asset is that stack of relationships, and a ranking built on KPIs and growth cannot measure its stability. One partner bank’s change in risk appetite can close a corridor faster than any volume chart would suggest.
Stablecoin rails are the other answer to the same problem, and they solve a different half of it. Moving settlement off the correspondent chain removes the intermediary, but the exporter still has to be identified, screened and monitored, and that obligation does not move anywhere. Whoever absorbs it ends up owning the customer either way.
The useful signal here is not that a fintech made a ranking. It is that compliance capability aimed at small exporters has become large enough to be scored next to card networks and processors, in a segment banks exited on the grounds that it could not pay for itself.
That leaves an open question rather than a conclusion. If servicing small exporters is now demonstrably a business, the banks that left are looking at a segment they wrote off, and they can come back through partnership or rebuild in-house. Which of those they choose over the next few years will decide whether platforms like this are permanent infrastructure or a well-run transition layer.
Who Ends Up Banking the Exporters Nobody Wants was originally published in Coinmonks on Medium, where people are continuing the conversation by highlighting and responding to this story.
