A tokenised bond, fund unit or deposit token is only as useful as the rails it can reach. Mint it on one chain and it trades with the wallets already there; try to place it with an investor whose custody sits on a different network, or move it into a settlement system that predates blockchains by decades, and the token needs help getting there. That gap is one reason so much tokenisation activity is still described as pilot rather than production.
It was also the subject of a panel at Proof of Talk's Paris 2026 edition at the Louvre Palace. "Scaling the Distribution Layer for Tokenised Markets" ran on the Hecto Main Stage on 3 June 2026, from 11:40 to 12:20. The programme billed Johann Kerbrat as SVP and GM of Crypto at Robinhood; Tom Zschach as independent (Cointelegraph quoted Zschach as Swift's chief innovation officer in August 2023); Nikhil Sharma as Digital Assets Director at BlackRock; and Stani Kulechov as founder and CEO of Aave Labs. Henry Jim, billed as ETF analyst at Bloomberg, moderated.
What follows is not a record of what was said on that stage. It sets out what the institutions represented there, and others, have actually shipped or tested.
What has to move, and what "interoperability" covers
Three separate things get bundled under the word. The first is moving an instruction: telling a system on chain B that an event happened on chain A, and having chain B trust that message enough to act on it. The second is moving the asset itself, or a claim on it, which usually means locking or burning a token on the source chain and minting a representation on the destination chain. The third is settling in money. This piece is about the first two. The third is covered in our examination of Project Agorá and settlement in central bank money.
Two adjacent problems are covered elsewhere too. A bridged or wrapped asset is a new object with its own custody and verification assumptions, and the diligence questions that follow are the subject of our threat model for institutional DeFi. And getting a message through is not the same as getting both sides to agree on what happened, which is the argument of our piece on how tokenised collateral moves through market infrastructure. Interoperability is the plumbing that makes those problems reachable, not a substitute for solving them.
Cross-chain messaging protocols
General-purpose messaging protocols differ mainly in whom you have to trust to relay a message honestly.
Chainlink's CCIP and the Swift experiments
Chainlink's Cross-Chain Interoperability Protocol is secured by decentralised oracle networks: nodes in a commit role observe and validate source-chain events, and nodes in an execution role validate pending messages for the destination chain. Chainlink's documentation says a separate Risk Management Network's automated role is no longer active in current deployments, so it should not be counted as a live second check.
Its best-known bank-network experiment is with Swift, the bank messaging cooperative. In 2023 Swift worked with more than a dozen institutions, including ANZ, BNP Paribas, BNY Mellon, Citi, Clearstream, Euroclear, Lloyds Banking Group, SIX Digital Exchange and the Depository Trust and Clearing Corporation, to test its network as a way to instruct transfers of tokenised value across public and private blockchains, with CCIP providing the cross-chain connectivity. When the results were published in August 2023, Zschach, then Swift's chief innovation officer, was quoted by Cointelegraph:
Our experiments have demonstrated clearly that existing secure and trusted Swift infrastructure can provide that central point of connectivity, removing a huge hurdle in the development of tokenization and unlocking its potential.
That was an experiment, not a product. According to Chainlink's own account, Swift, UBS Asset Management and Chainlink demonstrated settlement of tokenised fund subscriptions and redemptions over the Swift network in 2024, under Singapore's Project Guardian, and at Sibos in 2025 Chainlink introduced a Digital Transfer Agent technical standard, with UBS as the first global asset manager to adopt it.
On 9 July 2026 Swift declared its own blockchain ledger, built on Hyperledger Besu, ready for initial use, with 17 banks across six continents, among them BNY, Citi, HSBC, UBS, ANZ and Mashreq, preparing to test live tokenised cross-border payments. The ledger lets banks move customer funds overnight and at weekends, with final settlement still running through existing payment systems. Genfinity reports that CCIP connects the ledger to other networks; CoinDesk's report of Swift's announcement does not mention Chainlink. Read on its own terms, this is a pilot phase on live infrastructure, not full production.
Wormhole and BlackRock's BUIDL
Wormhole relies on a network of 19 validators, its Guardians, and a message is treated as valid once 13 of them have signed it. It has the clearer production case. In January 2025 Securitize said it had selected Wormhole as its exclusive interoperability platform, and BlackRock's USD Institutional Digital Liquidity Fund, BUIDL, which Securitize tokenises, uses it for cross-chain transfers. BUIDL launched in March 2024; by March 2025 it had added a share class on Solana, bringing it to seven chains (Aptos, Arbitrum, Avalanche, Ethereum, Optimism, Polygon and Solana), and in November 2025 it launched on BNB Chain.
Other messaging protocols
LayerZero and Axelar are other general-purpose messaging protocols, each with its own verification model. The marketing across this category reads much the same. The better diligence questions are who the protocol's regulated reference customers are, and whose honesty a given transfer finally depends on.
Decentralised finance met a version of the problem earlier. Aave's V3 release in March 2022 introduced Portals, a feature that lets cross-chain bridges mint and burn Aave deposit tokens across chains.
Building your own chain instead: Robinhood
Not every firm plugs into someone else's messaging layer. Robinhood launched Classic Stock Tokens for customers in Europe on Arbitrum One in 2025, then on 1 July 2026 launched the public mainnet of Robinhood Chain, its own layer-2 network built on the Arbitrum Platform. Its new Stock Tokens trade on that chain and are available through Robinhood Wallet in more than 120 countries, with availability varying by jurisdiction, and can be deployed into lending pools and used as trading collateral in DeFi. Rather than make one asset reachable across many chains, Robinhood controls the chain the asset lives on, and the interoperability problem becomes getting value into and out of that venue.
The trade-offs
- Reach against trust concentration. A general-purpose messaging protocol reaches many chains quickly, but every transfer then depends on that protocol's validator or oracle set behaving honestly, a dependency separate from the chains at either end.
- Familiar governance against speed. Routing through Swift keeps a bank inside a cooperative it already belongs to, but that route reaches only as far as Swift and its partners have built it, and it moved from a 2023 experiment to a 2026 pilot.
- Owning the venue against needing to reach one. A proprietary chain avoids depending on a third party's messaging protocol, at the cost of asking every counterparty, custodian or DeFi application to integrate with a venue it may not otherwise use.
Two constraints sit underneath all three. Who holds the keys and takes the fee for making an asset reachable is a question of market structure as much as technology, the subject of our piece on who controls institutional access on-chain. And the legacy side is itself still converging: on 26 February 2026 the Committee on Payments and Market Infrastructures at the Bank for International Settlements published updated harmonised ISO 20022 data requirements for cross-border payments, which it says are not regulatory requirements, allowing flexibility for adoption until the end of 2027. None of this is legal, tax or investment advice.
What is still unresolved
The International Organization of Securities Commissions published its final report on the tokenisation of financial assets on 11 November 2025. Developed by its Fintech Task Force, it found that interoperability challenges and the lack of credible settlement assets hinder scalability. That fits the record above: real production deployments, BUIDL's multichain distribution among them, alongside bank-network infrastructure that has only just reached pilot. Interoperability is neither solved nor vapourware. It is closer to electricity grids before a shared voltage standard: power moves, but whose plug you commit to still matters.