The issuer split that drives domestic capital and access (Part 2) becomes a multilateral problem the moment the flow crosses a border. Each wholesale central bank digital currency (wCBDC) is a liability of one specific central bank in one currency under one legal framework, and there is no shared infrastructure layer at scale that can move tokens between issuers. Tokenised deposits face a structurally different cross-bank problem at law. And the Fnality construction sits between the two categories as a deliberate hybrid worth naming.
Cross-border behaviour of wCBDC
wCBDC interoperability across central banks is an unsolved international standards problem. Two wCBDC tokens issued by two different central banks cannot be treated as fungible by either issuer, and the existing wholesale cross-border infrastructure (correspondent banking, the SWIFT messaging layer, regional RTGS interfaces) does not yet have a tokenised equivalent that scales beyond pilots.
Project Agorá at the BIS Innovation Hub is the most ambitious live attempt, bringing together seven central banks and a private-sector cohort to test tokenised cross-border wholesale settlement on a unified-ledger architecture (Project Agorá). Project Mariana, the AMM-based cross-border wCBDC design from the BIS Innovation Hub together with the central banks of France, Singapore, and Switzerland, demonstrated atomic FX settlement across three tokenised central bank moneys through an automated market maker (AMM); the questions of who operates the AMM, how access is gated, and how the design scales beyond three currencies remain open (BIS Project Mariana, 2023). mBridge, post-graduation from the BIS Innovation Hub, is the multilateral wCBDC initiative running with HKMA involvement and has produced live cross-border flows in pilot, though governance and scale questions remain open (BIS mBridge reference).
The point for a product team is that "settle in wCBDC across borders" is not yet a single operational option. It is a small set of bilateral and multilateral pilots, each with its own access rules, governance, and corridor coverage. The Bank for International Settlements (BIS) is the venue where the standards conversation is happening, and the BIS Annual Economic Report 2024 chapter associated with Hyun Song Shin sets out the unified-ledger thesis as the policy frame for that work (BIS AER 2024 Chapter III).
Cross-bank behaviour of tokenised deposits
Tokenised deposit cross-bank usage is operationally lighter than cross-border wCBDC, but legally trickier. A tokenised deposit at Bank A transferred to a holder who banks at Bank B is, in the simple model, a transfer of the claim from one party to another. Bank A still owes the balance, the new holder is now Bank A's depositor, and Bank B is not in the loop unless the structure routes through a settlement or substitution layer.
Practical tokenised deposit cross-bank schemes use one of two patterns. Either the receiving bank substitutes its own deposit liability for the original (Bank B credits the holder, Bank A pays Bank B in central bank money or an agreed settlement asset, and the original token is burned), or the chain accommodates direct holding of multi-issuer balances and the transfer is recorded as a change of issuer.
Partior runs the substitution pattern across DBS, JPMorgan, and Standard Chartered for cross-bank tokenised deposit settlement in multiple currencies. Kinexys runs primarily within JPMorgan's perimeter and uses correspondent flows for movement to non-participating banks. Project Ensemble's tokenised deposit layer is being designed to allow cross-bank settlement among participating Hong Kong banks, with the wCBDC layer providing the underlying interbank settlement (HKMA Ensemble Phase 2).
The point is that "tokenised deposit cross-border" is doing several distinct pieces of work depending on the architecture, and the legal characterisation of the resulting holding is not uniform across the patterns. A treasury team integrating a cross-bank tokenised deposit flow needs to know which pattern they are in, because the credit risk and the regulatory perimeter differ.
The Fnality variant
Fnality is the case that does not fit cleanly into either category and is worth naming, because readers who already know about it will ask. Fnality holds central bank reserves at the relevant national central bank and issues a tokenised settlement coin backed 1:1 by those reserves (Fnality). Legally, the holder of the settlement coin has a claim on Fnality, not directly on the central bank, but Fnality's reserves at the central bank are 1:1 backing and segregated.
Whether this counts as wCBDC, as tokenised deposit, or as a third category is a live debate. The conservative reading is that it is neither: it is a regulated payment system instrument issued by a private entity, with central bank reserve backing rather than central bank issuance. Under that reading, the holder has a claim on Fnality, with credit risk to Fnality, mitigated by the structural separation of the reserve account. The progressive reading is that the structural separation, combined with central bank operational involvement and the governing rulebook, gives the holder something close to wCBDC in economic substance.
Different jurisdictions have answered the question differently. The Bank of England (BoE) has worked closely with Fnality on the GBP variant and operates the omnibus account regime that holds the underlying sterling reserves (BoE Omnibus Accounts). The status of similar entities in USD, EUR, and JPY varies and the operational position has shifted across 2024 to 2026; treat the live status as something to verify before quoting a perimeter.
For the purpose of this chapter, the Fnality variant is the structure that demonstrates the binary "wCBDC or tokenised deposit" framing is not exhaustive. Hybrid intermediaries can sit between the two and inherit some properties of each, at the cost of a credit exposure to the intermediary. Whether such intermediaries become a permanent layer of the tokenised settlement stack or get absorbed into a broader wCBDC programme is one of the open questions that Part 4 picks up under the tiered-ledger thesis.