Cross-Border Settlement: The Next Institutional Battleground

Institutional cross-border settlement has become the sharpest battleground in financial infrastructure. Not because it is the biggest market — though it is, with global cross-border payment flows reaching an estimated $195 trillion in 2024 — but because it is where the incumbent rails are most visibly inadequate and where the new rails have the shortest path to enterprise adoption. The gap between what the system promises and what it delivers has become measurable, political, and increasingly expensive to ignore.

The Incumbent Rail Problem

The correspondent banking network that underpins most international payments was not designed for the speeds, volumes, or compliance expectations of 2026. It was designed for a world where a one-to-three-day settlement window was operationally acceptable and where information travelled separately from the funds it described. Both assumptions are now wrong.

The structural cost sits in the nostro-vostro system. To offer cross-border payments, banks must pre-fund accounts in foreign currencies held at correspondent banks around the world — capital sitting dormant, non-earning, simply waiting. Estimates of the total capital immobilised in nostro accounts globally reach approximately $27 trillion, generating what analysts describe as a $1.2 trillion annual liquidity tax on the banking system. In a 5% interest rate environment, the opportunity cost on idle foreign-currency balances is not a rounding error. It is a structural drag embedded into every international transaction.

The cost to end-users reflects this friction. As of Q1 2025, the global average cost of sending $200 internationally stood at 6.5% — far above the G20 target of 3% by 2027. In emerging-market corridors, the corridor premium is substantially higher. The Financial Stability Board’s 2025 progress report delivered an uncomfortable verdict: not only has improvement been minimal since 2023, but it is unlikely the 2027 G20 targets will be achieved on time for any of the principal metrics — cost, speed, access, or transparency. The policy framework exists. The on-the-ground improvement does not.

The speed problem compounds the cost problem. Roughly 50% of SWIFT wholesale transactions clear end-to-end within one hour; the 2027 target is 75%. For retail cross-border flows, only 46% reached recipients within an hour as of 2024. The FX exposure accumulated during the settlement window — the risk that exchange rates move between initiation and settlement — is not academic. For institutions running large cross-border books, it is a daily management problem.

The New Rails

The institutions building alternatives to correspondent banking have taken two distinct approaches. The first is the public-private multilateral: a consortium of central banks and major financial institutions building shared infrastructure on tokenised representations of central bank money. The second is the institutional network play: private-sector consortia creating interbank ledgers where settlement is atomic, continuous, and not dependent on pre-funded nostro positions.

BIS Project Agora sits in the first category. Announced in April 2024 and now in prototype phase, it brings together seven central banks — including the Federal Reserve Bank of New York, Bank of England, Bank of Japan, and the Eurosystem — alongside more than 40 private financial institutions including JPMorgan, Citi, HSBC, and Swift. The design goal is a unified multi-currency ledger where tokenised commercial bank deposits and tokenised wholesale central bank money settle atomically against each other, eliminating the correspondent hop entirely. As of late 2025, the project has moved from design to prototype build. A lessons-learned report is expected in H1 2026. No production timeline has been announced.

Partior operates in the second category and is the furthest along in commercial deployment. Founded by DBS, JPMorgan, and Standard Chartered out of Singapore’s Project Ubin, Partior runs a live blockchain network supporting 24/7 atomic settlement in USD, EUR, and SGD, with AED, GBP, JPY, and additional currencies in the pipeline. Deutsche Bank joined in 2024 and went live on the platform in 2025. In June 2025, Partior integrated with OSTTRA and Baton Systems to extend FX payment-versus-payment settlement — allowing institutions to exchange tokenised commercial bank funds on demand across currencies without pre-funding. The $60M Series B closed in July 2024. Partior is live, not in pilot.

JPMorgan’s Kinexys (formerly Onyx) is further evidence that the wholesale layer is past announcement phase. The platform processes more than $2 billion daily in notional value, has settled over $1.5 trillion in total since launch, and added on-chain FX settlement in USD and EUR in Q1 2025. Fnality, the Bank of England-regulated wholesale sterling settlement system, received settlement finality designation from the Bank of England in December 2024 — legally significant because it means transactions processed during financial stress or participant insolvency cannot be unwound. Fnality raised $136M in Series C funding in September 2025, led by WisdomTree, Bank of America, and Citi. It is the world’s first regulated DLT-based wholesale payment system to have gone live, having commenced controlled sterling operations in December 2023.

Where does the stablecoin rail fit in this picture? Honestly, ahead of where institutional consensus assumes. Stablecoin cross-border settlement volume reached $1.25 trillion monthly as of 2025. Banks are twice as likely to prioritise cross-border payments as a stablecoin use case compared to other applications, with speed — not cost — cited as the primary driver. The GENIUS Act, passed by the US Congress in July 2025, created a federal regulatory framework for payment stablecoins, which removed a significant adoption barrier for US-regulated institutions. The honest assessment: stablecoin rails are operationally deployed and moving institutional volume in corridors where the bank-to-bank networks are not live. The tokenised-deposit model (Partior, Fnality, Agora) has more regulatory legitimacy and central bank backing, but a narrower current deployment footprint. Both are real. Neither has won.

SWIFT’s ISO 20022 migration — the standardisation of payment message formats across the global correspondent banking network — completed its coexistence period in November 2025. The Fed migrated Fedwire in July 2025; CHIPS followed. The significance is often overstated in coverage: ISO 20022 does not fix the pre-funding problem or reduce settlement lag. It makes messages richer and more structured, which reduces compliance friction and increases straight-through processing rates. It is infrastructure maintenance on the incumbent rails, not a replacement. Worth knowing, but not conflating with the structural shift underway.

What Biptap’s Position Signals

Biptap is not Partior, not Kinexys, and not Fnality. It is not trying to be. The consortium model — tier-one banks, central banks, multilateral coordination — is a specific instrument suited to a specific problem: rebuilding the wholesale settlement layer between the world’s largest financial institutions. That problem is real and the solution is overdue. But it leaves a large portion of the market unaddressed.

The mid-tier institutions — regional banks, fintechs, enterprises, and financial operators in markets the large-bank consortia have not prioritised — still need cross-border infrastructure. Their access to the Partior network is indirect at best. Their ability to build their own settlement rails does not exist. This is the segment Biptap’s omnibanking model is positioned to serve: a unified platform integrating crypto, fiat, and Banking-as-a-Service under a single infrastructure, designed to be deployed by institutions that need the capability rather than those building the next-generation rails themselves.

The Al Fardan Ventures joint venture in Abu Dhabi is a legitimate geographic lever. The UAE sits at the intersection of Asia-to-MENA and Africa-to-MENA capital flows — corridors where correspondent banking concentration is high and the premium for using it is measurable. Abu Dhabi’s ADGM regulatory environment is competitive for digital banking licensing, and Al Fardan Ventures brings local financial network depth that accelerates market access. The CAR sovereign partnership demonstrates a willingness to operate in markets where correspondent banking is not merely expensive but structurally absent. These are not the same thesis, and they should not be conflated: one is about capturing a premium corridor, the other is about building primary infrastructure in a market with none. Both are consistent with an omnibanking platform that competes on deployment reach rather than on consortium membership.

To be direct about scale: Biptap’s deployments at 40+ institutions and its sovereign partnerships do not place it in the same conversation as the entities processing $2 billion daily. The relevant comparison is not scale — it is the segment being served and the problem being solved. At the mid-market and emerging-corridor layer of cross-border settlement, the large-bank infrastructure does not exist yet, and the omnibanking model has a structural argument.

What to Watch in H2 2026 and 2027

The BIS Project Agora lessons-learned report, expected H1 2026, will be the most significant indicator of whether the multilateral tokenised-deposit model is on a path to production or requires another design cycle. Watch whether central banks signal a production mandate or a continued research posture — the distinction matters for how quickly the wholesale layer moves.

Partior’s currency expansion into AED, GBP, and JPY will determine whether the network achieves the corridor coverage needed to reduce reliance on correspondent banking in meaningful volume. If it does, the competitive pressure on the incumbent SWIFT correspondent network in wholesale payments will become structural rather than marginal. Deutsche Bank’s live participation in 2025 is a data point; the next tier of bank joiners will be the signal.

The regulatory trajectory for stablecoins post-GENIUS Act will determine whether they become a genuine bridge instrument for corridors without bank-to-bank rail coverage — or whether regulatory complexity in non-US jurisdictions limits them to US-anchored flows. The Federal Reserve’s March 2026 working paper on stablecoins and cross-border payments noted both the benefits and the monetary policy transmission risks. That ambivalence is not resolved. It will be worked out market by market over the next 18 months.

The 2027 FSB deadline for G20 cross-border payment targets is now functionally a credibility test, not an achievable roadmap. The BIS stated plainly in 2025 that the targets are unlikely to be met. The more useful framing for operators: the gap between current performance and target performance defines the commercial opportunity. Infrastructure that can credibly close that gap — on cost, speed, or corridor access — is what institutional capital is pursuing. The battle for who builds that infrastructure is the story of cross-border settlement in 2026.