Why settlement networks do not talk to each other

Settlement networks do not talk to each other because they were never designed to: each one settles a different asset, admits a different membership under a different rulebook, and earns its keep by growing its own side of the wall. The result is a paradox worth staring at. A decade of innovation aimed at replacing the correspondent chain has produced faster, cheaper settlement inside each new network, and a new fragmentation problem between them.

The map itself is covered in the interbank settlement network landscape and alternatives to correspondent banking. This page is about the seams.

Reason one: they settle different assets

Money is not one thing. A balance on Partior is a claim on a commercial bank; Partior settles interbank flows in commercial bank money, live since 2023. A balance on Fnality is backed one-for-one by funds in an omnibus account at the Bank of England, about as close to central bank money as a private system gets. Value on the Circle Payments Network moves as USDC or EURC, a claim on the issuer’s reserve. Project mBridge settles in central bank digital currencies. Banks’ own deposit tokens are claims on the issuing bank.

Moving value from one network to another is therefore never a message hop, the way an email crosses providers. It is a conversion between different credits with different issuers, different legal regimes, and different risk. Someone must hold both assets, take the exchange risk, and stand behind finality on both sides. That is precisely the nostro-style intermediation these networks were built to eliminate, which is why none of them is eager to recreate it at their own edges. Central banks have spent years worrying about exactly this property under the banner of the singleness of money.

Reason two: membership is the product

Every settlement network is also a compliance perimeter. Partior admits banks under bank-grade diligence. Fnality is owned by the institutions that use it. Circle vets each CPN participant for licensing, regulatory compliance, and operational risk before admission, and its rulebook, the CPN Rules, governs conduct inside.

That vetting is not overhead; it is the product. Members transact with reduced counterparty anxiety precisely because everyone inside the wall passed the same checks. Interconnecting two networks means reconciling two rulebooks, two vetting standards, two dispute processes, and two liability models, so each network’s assurance now rests partly on a perimeter it does not control. APIs are the easy part; governance is the hard part, and it is why interconnection announcements are rare while membership announcements are weekly.

Reason three: the economics reward capture, not connection

A settlement network is a classic network-effects business: every added member makes membership more valuable, and volume that stays inside the network is volume the operator can see, govern, and monetize. A bridge to a rival network reverses that logic, commoditizing membership and handing growth to the other side. No operator is obliged to build one, and, revealed preference being what it is, none of the live commercial networks has.

SWIFT, which has watched this dynamic from the incumbent’s seat, named the outcome years ago: a world of “digital islands”, its phrase for tokenization platforms and CBDC systems that each work internally and do not connect. Its CBDC sandbox work on interlinking exists because the islands were forming even before most of the networks went live.

The official sector has moved from observation to warning. The FSB concluded in October 2025 that the G20’s end-2027 targets for cheaper, faster cross-border payments are unlikely to be met at the global level. And in May 2026, Fabio Panetta, Governor of Banca d’Italia and chair of the BIS Committee on Payments and Market Infrastructures, devoted a keynote to the problem, warning that a proliferation of parallel systems, “weakly connected or politically segmented”, would “reduce interoperability, increase costs and erode the efficiency gains” the G20 programme seeks, and that fragmentation “would deliver neither resilience nor sovereignty”. His prescription was in his title: interconnect to stabilize.

What connection would actually take

Three paths are visible, none complete.

Harmonized standards. ISO 20022 and the CPMI’s harmonisation work give networks a common data language, which removes one class of friction but does not touch settlement assets or governance.

Official-sector platforms. Project Agorá, convened by the BIS with eight central banks and over 40 financial institutions, is testing tokenized central bank reserves and commercial bank deposits on a single platform, with prototype results published in May 2026 demonstrating atomic multi-currency settlement. If it matures, it would dissolve the seam by putting both assets on one venue. The BIS is explicit that it is not building a finished product, and real-value operation is years away.

Interlinking. The CPMI frames two models, bilateral links between systems and multilateral hubs. Both exist for instant payment systems (Project Nexus is the multilateral case) and both remain unbuilt between the tokenized settlement networks.

The buyer’s dilemma

For an institution choosing infrastructure, fragmentation converts a technology decision into a wager. Commit to one network and you have bet on its corridors, its settlement asset, and its continued independence; hedge across several and your liquidity fragments across venues that do not connect, each with its own pre-funding, its own integration, and its own policy surface. Either way, compliance still runs in your own systems, bolted on before or after whichever network settles the value.

Where Frame fits

Frame is the settlement layer for global finance: one integration to orchestrate payments at scale across fiat rails, stablecoins, and tokenized deposits, with compliance enforced on every transaction and settlement in seconds instead of days.

Frame’s answer to the seams is to sit above them rather than add another island. Frame is rail-neutral: a payment enters through one integration, and Frame routes it across whichever rail fits the corridor, the counterparty, and the policy that governs it, a fiat network, a regulated stablecoin, or a tokenized deposit. Compliance is enforced inside settlement: Frame’s Rules Engine evaluates every transaction against its governing policies, and a transfer that cannot satisfy them does not settle, with verifiable evidence produced for every one that does, without exposing the underlying business data.

The networks on this page are, from that vantage point, rails: each excellent inside its own perimeter, none obliged to connect to the others. Interconnection may arrive from the official sector eventually. Institutions that need to pay across the islands this year do not have to wait for the bridges.