OpenFX Releases Its Playbook Arguing For Modern Rails For A Modern World

A payment leaving São Paulo for Bogotá might “travel” something like 7,000 miles. On the way, it will pass through at least three countries and a half-dozen or more different financial institutions. It will need to survive several, separate regulatory screenings, and will be debited a meaningful fraction of its total value in fees.

Time and money that might be measured in days and hundreds of thousands of dollars.

This doesn’t just happen in Latin America, but for most of the world outside of certain economies in the G20.

The mechanism driving this is SWIFT, a messaging service from 1973 that in many ways still carries constraints from the telex-era. CHIPs and Fedwire, other portions of the machine, are also designed for a world that works much differently than the one that exists today.

Underneath all of it sits the nostro account, a pre-funded pool of foreign currency parked at a correspondent bank, a structure Florentine merchants were using to settle Mediterranean trade in 1383.

Nobody designed this. It accreted. Each layer solved the most pressing problem of its era, and each patch became a permanent feature that now serves trillions in cross-border FX.

What’s The Problem?

There really isn’t one, unless you happen to live outside of the major economies.

If you are moving money in the G20, most of the system's clunkiness is invisible. The major corridors have deep liquidity, direct banking relationships, and competitive pricing.

Everywhere else, FX moves through a patchwork system that is both slow and expensive.

And it’s a system that is actually getting worse.

The Bank for International Settlements counted a 22 percent decline in active correspondent banking relationships between 2011 and 2019. In North America, the decline was 13 percent. In Latin America, 34 percent. In small island developing states, 41 percent. The World Bank found the Pacific Islands lost 60 percent of their correspondent relationships between 2011 and 2024.

Without correspondents, the cost and time to complete settlements only rises.

The obvious response is reform. Modernize the rails, mandate faster settlement.The problem is that the institutions capable of doing this are paid not to.

A payment that takes three days to settle is three days of float. A corridor that requires pre-funded nostro accounts is trapped capital earning a return for the institution holding it. Instant settlement doesn't upgrade this business, it eliminates a source of revenue.

The incentives are bad, and unlikely to change anytime soon.

This is why OpenFX is not trying to fix correspondent banking, we are building its replacement.

The full argument about what we’re building, and why we believe it wins is laid out in our Playbook.