Suit Up · Asia's institutional desk on tokenisation

The conventional cross-border wholesale-payments architecture is the system every named tokenisation initiative on the cross-border lane is implicitly asked to displace. Correspondent banking, the Society for Worldwide Interbank Financial Telecommunication (SWIFT) messaging layer, and the chain of nostro and vostro accounts that intermediate a payment between the originating bank in one jurisdiction and the beneficiary bank in another have been the operating standard for the better part of half a century. The system works. It also incurs multi-day settlement, opaque per-corridor fees, fragmented compliance handling, an FX-risk window between value-date confirmation and final settlement, and working-capital trapped in pre-funded nostro positions. The named tokenisation programmes (mBridge, Project Agorá, Partior, Ant International's Whale, Kinexys cross-currency settlement) each address a subset of those frictions, with the structural promise of atomic FX delivery-versus-payment, 24/7 operability, and programmable conditional payments built in. This page maps the conventional friction set, the tokenisation-side rewire patterns, and which live programme attacks which piece.

What the conventional system actually does

The standard cross-border wholesale payment runs through a chain of intermediating banks. The originator's bank holds an account at a correspondent bank in the originator's jurisdiction or in an intermediate jurisdiction; the beneficiary's bank holds an account at a correspondent bank that the originator's correspondent can transact with. The payment instruction transits the SWIFT messaging layer; the actual cash movement happens through debits and credits across the chain of correspondent accounts. The FX leg, where the payment crosses currencies, is typically handled at one of the correspondent banks in the chain, with the FX rate set bilaterally and the FX settlement running on the underlying-currency settlement cadence (usually T+2 for major currencies through CLS).

Five frictions get attached to the conventional flow consistently across the central-bank, BIS, and consultancy literature on cross-border payments improvement.

Settlement latency. A standard cross-border wholesale payment can take several days to clear from originator to beneficiary, with the value-date convention absorbing the FX leg's T+2 cadence and the correspondent-chain processing absorbing additional time for messaging confirmation and reconciliation. The Bank for International Settlements has been explicit that the conventional latency is materially out of step with what current technology could deliver.

Opaque per-corridor fees. The fee structure is bilateral between each pair of correspondents and the customer-facing fee is the cumulative total across the chain. The originating bank typically does not know the full all-in cost until after the payment clears, and the beneficiary often receives less than the gross amount sent because of intermediating-bank deductions.

Fragmented compliance handling. AML, sanctions screening, and customer-identification obligations apply at each intermediating bank in the chain, with each bank running its own compliance stack against the same payment. The result is duplicated compliance work and an elevated rate of payment delay or rejection when one of the intermediating banks flags an issue that would not have been flagged by another.

FX-risk window. Between the time an FX rate is locked and the time the FX trade actually settles, the parties carry market risk on the rate. For institutional-grade flows the risk is mitigated through CLS settlement (Continuous Linked Settlement, the multilateral PvP settlement system for FX trades), but CLS does not cover all currency pairs and does not eliminate the latency between trade and settlement.

Working capital trapped in pre-funded nostro positions. To support cross-border payment activity, banks pre-fund nostro accounts at correspondent banks in foreign currencies. The pre-funding ties up working capital on the bank's balance sheet and represents one of the larger structural inefficiencies in the conventional system. The aggregate nostro pre-funding across the global banking system is a multi-trillion-dollar number, and the structural cost is borne ultimately by the corporate and institutional customers using the system.

How tokenisation rewires the flow

Three structurally consequential rewire patterns, each addressing a different subset of the conventional frictions.

Atomic FX delivery-versus-payment. A tokenised wholesale settlement primitive (whether tokenised wholesale CBDC on mBridge, tokenised commercial bank money on Project Agorá, or tokenised deposits on Partior or Kinexys) can deliver atomic FX-DvP across two currency legs on a single ledger or across linked ledgers using hashed time-locked or equivalent constructions. The settlement primitive eliminates the FX-risk window: the two legs settle simultaneously or not at all. Foundations Chapter VI Part 2 covers the mechanics; the cross-border applications are the natural extension.

mBridge is the worked example of multi-CBDC atomic FX-DvP at production-style scale. The platform's cumulative throughput crossed USD 55.5 billion across more than 4,000 cross-border transactions, with the e-CNY accounting for approximately 95 percent of platform volume (Cointelegraph, Ledger Insights). The structural design (each participating central bank operates a node and issues its own wholesale CBDC on the platform; commercial-bank participants settle cross-border payments and FX trades by exchanging tokenised central-bank money across currencies without an intermediating correspondent bank) is the cleanest architectural illustration of the atomic FX-DvP pattern.

24/7 operability. The conventional cross-border payments system is constrained by the operating hours of the intermediating banks, the operating hours of the FX market settlement infrastructure (CLS), and the operating hours of the underlying-currency RTGS systems. The result is that cross-border payments effectively operate during the overlap of the intermediating jurisdictions' business hours, with off-hour payments queued for the next business cycle.

Tokenised wholesale settlement infrastructure can operate 24/7 because the participating central-bank and commercial-bank nodes can run continuously rather than in business-hours windows. Partior, the Singapore-headquartered consortium-operated tokenised-deposit bridge with DBS, JPMorgan, Standard Chartered, and adjacent banks as participants, has positioned 24/7 operability as one of its structural design points. Kinexys Digital Payments has run continuously as a 24/7 institutional-payment rail since its 2024 rebrand from Onyx (details). The 24/7 operating capability is not a marginal convenience; for corporate treasurers managing cash positions across multiple time zones, it is a substantive operating change.

Programmable conditional payments. The conventional cross-border payments system delivers an unconditional value transfer: the originator instructs, the chain processes, the beneficiary receives. There is no native mechanism for a payment to be conditioned on an external trigger (delivery confirmation, regulatory clearance, dispute-resolution outcome). Conditional payments in the conventional system are handled through escrow arrangements, letters of credit, or other off-system structures.

Tokenised wholesale settlement infrastructure can carry programmable conditional logic at the contract level. The Singapore-coined "purpose-bound money" (PBM) construction (covered in Project Guardian) is a wrapper around a payment instrument that carries programmable conditions on use, expiry, recipient, or purpose. The construction has been tested at the Project Guardian FX workstream level for atomic FX swap settlement with conditional release based on receipt confirmation, and conditional payment release tied to delivery in a goods context. The programmability is not a separate product layer; it sits at the settlement-asset level and is one of the structurally distinctive features of the tokenised approach.

The named live programmes

Five named live cross-border tokenisation programmes anchor the regional and global maps. Each one targets a different subset of the conventional friction set, with overlapping but non-identical participant rosters.

mBridge. The multi-CBDC platform with HKMA, Bank of Thailand, PBoC Digital Currency Institute, Central Bank of UAE, and Saudi Central Bank as participants. Graduated from BIS Innovation Hub governance in late 2024 to post-graduation governance led by the participating central banks. Cumulative throughput USD 55.5 billion across 4,000-plus transactions through 2025-2026, with e-CNY accounting for roughly 95 percent of volume. The structural read is that mBridge is the most operationally mature multi-CBDC platform in the world, anchored on the APAC-Middle-East trade corridor.

Project Agorá. The cross-border tokenised-commercial-bank-money project led by the BIS Innovation Hub Switzerland Centre with seven central banks (Banque de France for the Eurosystem, Bank of England, Bank of Japan, Bank of Korea, Bank of Mexico, Federal Reserve Bank of New York, Swiss National Bank) and a private-sector cohort of more than forty firms convened by the Institute of International Finance (Project Agorá BIS reference page). Structurally complementary to mBridge: Agorá addresses the tokenised-commercial-bank-money layer where mBridge addresses the wholesale-CBDC layer. The conceptual release was published in 2025; the live experimentation phase continues through 2025-2026 without a committed production-platform timeline.

Partior. The Singapore-headquartered consortium-operated tokenised-deposit bridge with DBS, JPMorgan, and Standard Chartered as founding participants and the SBI Shinsei JPY corridor as the most recent expansion. Structurally different from mBridge and Agorá: Partior is a consortium-operated tokenised-deposit ledger rather than a multi-CBDC or unified-ledger platform, with the participating banks operating their own tokenised-deposit positions on the shared ledger and settling cross-border payments and FX trades through the consortium-operated settlement primitive. The 24/7 operability and the multi-currency tokenised-deposit settlement are the structural design points.

Whale, operated by Ant International. The shared-platform tokenised-deposit infrastructure that supports the multi-currency (HKD, CNH, USD, SGD) tokenised-deposit programme with Standard Chartered as the production-launch counterparty (18 December 2025, details). Different shape from Partior: Whale is a shared platform operated by a non-bank Singapore-headquartered fintech with bank participants holding tokenised-deposit positions on the platform, rather than a consortium-operated bank ledger. The Ant International + ISDA July 2025 Project Guardian report co-authorship is the architectural-framework piece.

Kinexys cross-currency settlement. The JPMorgan-operated permissioned EVM-compatible rail with the blockchain deposit account (BDA) construct, and the JPMD natively-issued USD deposit token on Base. Cumulative notional past USD 3 trillion and average daily volume above USD 5 billion as of late 2025 (the largest single-bank tokenised-cash programme by published volume globally). The cross-currency settlement extension and the DBS-Kinexys interoperability framework (announced 11 November 2025) are the structurally distinctive cross-border applications.

What is structurally complementary, what is competitive

The five programmes are not all addressing the same problem. mBridge and Agorá are structurally complementary: mBridge handles the wholesale-CBDC layer, Agorá handles the tokenised-commercial-bank-money layer, with overlapping participating central banks (Bank of Japan and Bank of Korea on the Agorá side; the APAC-Middle-East central banks on the mBridge side). The two platforms together cover most of the global cross-border wholesale-payments flow.

Partior, Whale, and Kinexys are competitive at the tokenised-deposit cross-border settlement layer, with overlapping but non-identical participant sets. The DBS-Kinexys interoperability framework is the closest the major commercial-bank tokenised-deposit programmes have come to a coordinated cross-rail interoperability layer, and it is in development rather than live as of the announcement.

The SWIFT-led tokenised-payments interoperability work is the institutional-incumbent alternative to all five, positioned as a non-unified-ledger pattern that uses the existing messaging layer to carry tokenised-payment instructions across non-tokenised settlement infrastructure. The SWIFT approach addresses the messaging-and-routing layer rather than the underlying-settlement layer.

What is still missing

Three structural pieces that the cross-border tokenisation infrastructure has not yet resolved.

The cross-platform settlement question. mBridge handles wholesale CBDC; Agorá handles tokenised commercial bank money on a unified ledger; Partior, Whale, and Kinexys handle tokenised deposits on different platform architectures. A cross-border payment that needs to traverse two of these platforms (for example, a Partior tokenised-deposit leg connecting to an mBridge wholesale-CBDC settlement leg) requires interoperability between the platforms. None of the named programmes has shipped a production cross-platform interoperability mechanism.

The agentic-commerce posture. As AI-agent-controlled wallets become a credible counterparty category, whether the cross-border tokenisation platforms admit agent-controlled holders at the commercial-bank tokenised-deposit layer is the structural question. The Singapore Project Guardian perimeter has engaged publicly on the question; mBridge, Agorá, and the consortium-operated platforms have not addressed it substantively.

The retail extension. All five named platforms operate at the wholesale tier with commercial-bank or central-bank participants only. The retail cross-border payments friction set is structurally similar (latency, opaque fees, FX risk) but the regulatory and operational architecture is different. The bridging from wholesale-tier tokenisation infrastructure to retail-tier cross-border payments has not been comprehensively addressed by any of the named programmes, and is the natural extension of the structural rewire that the wholesale work has begun.

Open questions

Whether mBridge and Agorá converge into a unified architecture, remain structurally separate parallel rails, or one absorbs the other. The two platforms have overlapping participating central banks and structurally complementary architectures; the integration question has not been publicly addressed.
The post-November 2025 production trajectory of the DBS-Kinexys interoperability framework, and whether it becomes a published standard the wider GSIB set can adopt or remains a closed bilateral.
Whether SBI Shinsei + Partior + DCJPY moves from MoU stage to production cross-currency settlement on the JPY corridor, and on what timeline.
The interaction between Partior, Whale, and Kinexys cross-currency settlement on the SGD-USD-HKD-CNH corridor where the three platforms have overlapping commercial-bank participants. Whether the three converge on a common interoperability layer or remain competitive infrastructure is one of the consequential APAC-cluster questions.
The Federal Reserve Bank of New York's continued participation in Project Agorá. The FRBNY participation was confirmed at announcement; subsequent political-economic developments around US central-bank engagement with multilateral cross-border tokenisation work have been actively debated in trade press, but the formal participation status is unchanged in current public material.

[Suit Up]

Cross-border payments: the problem tokenisation is asked to solve