The newest payment rails still settle through the oldest part of the system. A stablecoin can move across the world in moments, a card can authorise in a second, an instant scheme can clear in minutes — and yet at the end of the chain something usually has to be settled by institutions holding accounts with one another, in the correspondent core that has changed comparatively little. This is not a failure of the new rails. It is a description of where finality lives.

New rails, old settlement

Finality requires an institution to accept an obligation. That acceptance rests on relationships and accounts, not on the speed of the message that preceded it. So new rails shorten the journey while leaving the destination intact: value is passed along, and somewhere a balance between institutions is adjusted. The correspondent core is where that adjustment happens.

That core has been under strain. The number of active correspondent banks worldwide fell by about 22% between 2011 and 2019, even as payment volumes rose; in Latin America, active relationships fell by about 30% since 2012. Fewer relationships carrying more traffic is precisely the contradiction at the heart of digital-asset payments.

The missing layer

The missing layer is not another token. It is the set of institutional relationships and rules that let new rails settle reliably: who is accountable for validation, what data must accompany an operation, and how institutions that are also competitors can trust one another. A neutral layer that holds correspondent accounts for institutions, validates each operation once, nets offsetting flows and routes the remainder is one architectural answer to that question.

What would change it

Two shifts would matter. First, reversing the direction of relationships: a hub holding accounts for institutions, rather than each institution opening accounts everywhere. Second, making trust a codified multilateral rule — reliance agreements signed on joining — rather than personal, bilateral and non-transferable trust. Neither is a technology problem, which is why neither has been solved by faster rails alone.

In short

  • New rails move value; the old core still settles it.
  • Finality rests on institutional relationships and accounts.
  • Correspondent relationships have been shrinking while volumes grew.
  • The missing layer is institutional, not another token.
  • Reversed relationships and codified reliance are the levers.

For the wider picture, see /about/