TBW - Simon Letort (Digital Asset): "Ethereum is years ahead of retail, Canton ahead of institutional"

The Big Whale: To begin with, could you remind us how the Digital Asset adventure came about? We're talking about a company that's already a decade old, which is an eternity in this ecosystem.

Simon Letort: That's right, Digital Asset has been around for about ten years. It all started at DRW (owner of Cumberland, one of the pioneers of crypto market making) where the founders, Yuval Rooz and Eric Saraniecki, met. They realized very early on that there was no "backbone," no market infrastructure suited to the institutional processing of blockchain transactions.

We started by building the foundations for clients such as Goldman Sachs, HSBC and BNP Paribas. Initially, the idea was simple: the tokenisation of assets for use cases specific to each bank. Then we moved into repo and interbank refinancing. But we have to be honest: for a long time, it was a slow process. Each bank had its own 'Canton', its own closed garden. The networks were not connected.

Can you give us an order of magnitude on current activity. Volumes, transactions?

The biggest use case today is interbank repo operated by Broadridge: around $400 billion nominal per day, for 8,000 to 10,000 daily transactions. For a blockchain, that's huge. For global finance, it's still small: Tradeweb, for example, processes around 2.5 trillion per day (globally).

Our goal for success is not to stay at a few hundred billion: it's to move towards trillions per day, because the potential for migration is massive.

>> Analysis of the Canton Network blockchain

What triggered the creation of what we now call the Canton Network?

It came from the users themselves. A customer said to us, "I'd like to be able to issue a bond on the Goldman Sachs platform and immediately refinance it on the Broadridge platform to do a repo." That's where Canton Network was born.

Unlike many projects that write a white paper, raise funds, and then look for clients, we did the opposite. We first created the "Layers 2" (the banking applications), made sure they had a real use, and only then launched the "Layer 1" to connect them together, before listing the corner.

You call it a "network", but it sounds more like an assembly of smart contracts, creating sub-domains. Is that right?

We say "network" because it's easier to understand, but in reality, a network materialises via smart contracts and access rules.

All the nodes that have deployed the smart contract for a given asset form a sub-domain. For example, all those that interact with the Circle asset form a subnetwork around Circle. And then these sub-domains can communicate when the standards are met.

The analogy is the Internet: you get the impression that everything is "a single network", but you don't have the whole of the Internet on your hard drive. You access parts of it depending on how you use it. Canton follows a similar logic: the data is with the participants concerned, not replicated globally.

"The value of these systems lies in replacing the back office"

You were at Société Générale for 20 years, and you sat on the board of R3, named after the consortium of banks responsible for exploiting the possibilities offered by blockchain. What is your view of these historic projects like Corda, which now seem to be marking time?

The value of these systems lies in replacing the back office. It's not glamorous. It's about optimising costs. Since 2008, banks have been constantly cutting costs. For a transformation of this scale to succeed, you have to decommission the old system.

The problem is that if you keep your Swift rails while adding blockchain rails (whether Corda, Canton, Polygon or whatever you want), your costs go up. The CEO ends up saying, "I was promised savings, and it's costing me more." It's the same difficulty as migrating to the cloud. Corda's vision was correct, but it was the go-to-market that sinned.

What is the fundamental difference with your current approach?

Incentive. What Bitcoin and Ethereum have proven is the power of the token incentive model. For years, people have been saying, "Blockchain is great, crypto is bad." That's a mistake.

When we launched Canton Coin two years ago, we combined institutional-grade technology (privacy, control) with tokenomics. Today, players like Citadel or hedge funds are joining the network, not just for back-office optimisation, but because there is a financial upside to the coin. The token is the lubricant of adoption.

We are still seeing competing networks emerge: Kinexys (JP Morgan), Partior, and SWIFT, which is getting busy. Are you all going to connect, or are we heading for a war of standards?

At JP Morgan, there has been a major change: historically, they were very focused on Quorum. Then they expanded: they issued a Deposit Token on Base recently, and they announced that Canton would be another supported channel. It's a significant shift.

Partior is a consortium in which JP Morgan plays a role. Is this shift having an impact? I can't say, I don't know their stack exactly.

On SWIFT, they're not a blockchain developer per se: they'll probably choose technologies. They have enormous potential (massive onboarding, key management, etc.), but they also have a strategic dilemma: blockchain threatens their business model. So they're torn between speeding up and stalling.

DTCC, it's the same thing: a central database threatened by tokenisation. Their movement is to end up adopting the wave rather than undergoing it. They're now involved in Canton, which is a strong signal.

"It's the image of the nightclub: for it to work, you need connected people inside and a velvet cord at the entrance"

When we look at your partners, we get the impression that the entire gotha of global finance is there. How do you explain this power of attraction?

It's the image of a nightclub: for it to work, you need connected people inside and a "Velvet Rope" at the entrance. We convinced the big players in the right order: first the banks, then the asset managers and the crypto market makers. This naturally attracts Prime Brokers.

Today, we are integrating the major custodians. BNY Mellon has just joined us, Taurus is live, and another particularly important player will arrive in Canton in the second quarter of 2026.

Once the custodians are fully live, we will be able to unlock the traditional asset managers, and then behind them, the pension funds - where the mass of capital is. And as these institutions work through their custodians, the strategy is not to convince each one individually: it's to get the infrastructure upstream.

If you had to convince the Head of Digital Assets of a TradFi group that was still hesitating, what would be the killer argument?

Often, the best seller isn't even me. It's platforms like Tradeweb, which these companies already use on a daily basis, and which are already members of Canton.

The key argument is settlement: being able to do delivery-versus-payment much faster, including outside traditional cut-offs. Today, we can trade "out of hours" in certain cases, but cash often moves with a 24 to 30 hour delay. Here, we can have settlement in a matter of minutes, against the asset purchased.

And above all, the real difference is not just 'going faster'. It's optimising cash flow when you do multi-leg operations: spreads, arbitrages, cross trades between markets and currencies.

If you pay in dollars and receive yen, the cash flow lag costs you. If you can synchronise the two legs - even at a precise time - you reduce a real cost of funding. The game changer is optimising cash and collateral.

On Canton, each institution manages its node and its rules. Give me a concrete example of a rule that can be parameterised.

It all happens at the level of smart contracts, and more specifically at the level of assets issued.

Simple example: waitlisting vs blacklisting. On certain assets (a bond, for example), we can require each holder to be explicitly authorised: this is a waitlist-type model, with KYC/pre-validation. Conversely, a stablecoin often operates on a "blacklist" model: you can receive and transfer by default, but the issuer must be able to blacklist certain addresses if necessary.

Another dimension: who sees what. Canton Coin is public: everyone sees the transactions. But a stablecoin issued on Canton can be configured so that transactions are only visible to the buyer, the seller and the issuer - not to the whole network. This is confidentiality encoded in the smart contract: unlike a public blockchain where the data is replicated everywhere, here the data is segmented.

We can also imagine "observer" nodes: a regulator could receive certain information, but only if the issuer allows it at the smart contract level (and typically because it is obliged to do so in a reporting framework).

"When Circle issues USDC on Canton, only the stakeholders see the flow"

One of the big debates in on-chain finance is confidentiality, and Canton is using this as a strong argument to attract institutional investors. Most projects rely on Zero-Knowledge Proofs (ZKP). But not you. Why this choice?

It's a fundamental point. Shaul Kfir, our CTO and co-founder, is one of the researchers who wrote the ZK-SNARKs library (forms of Zero-Knowledge Proofs used by blockchains) ten years ago. He knows the subject inside out. His observation is pragmatic: for the volumes of institutional finance (tens of billions a day), Zero-Knowledge is not yet 'production grade'. It's too complex and can create unmanageable situations for a risk manager.

So how do you guarantee the secrecy of transactions?

By segmenting the data. On Ethereum, each node stores the entire blockchain. On Canton, nobody has everything on their hard drive. Transactions are only visible to the buyer, the seller and, possibly, the issuer. If Circle issues USDC on Canton, only the parties involved see the flow. It's peer-to-peer. So we don't need the complexity of ZKPs because the data is simply not replicated everywhere.

Doesn't this run the risk of creating "private gardens" unable to communicate with each other?

That's the whole point of the protocol. We operate like a common operating system - say Linux - installed on all the servers. To ensure interoperability, we have created a standard (CIP-56, the equivalent of ERC-20). If you comply with this standard, integration takes five minutes. The network is a collection of smart contracts that communicate via a "Global Synchronizer".

Fairly, let's talk about this Global Synchronizer. Isn't it a difficult black box for the compliance department of an institutional group to audit?

The most apt analogy is that of an internet service provider. In a bank, you audit certain critical suppliers, but you don't audit all the world's operators. What counts is confidence in a subset of operators and control of supplier risk.

Here, the Global Synchronizer is made up of around thirty "super-validators". For a transaction to be validated, two-thirds of them must confirm receipt.

The strength of Canton is that these validators are regulated entities: Euroclear (which chairs the Canton Foundation), DTCC, Circle, Cumberland, etc. For a European bank, the systemic risk is already accepted: if Euroclear or DTCC fall, the global financial system has a much bigger problem than the Canton network anyway.

"For a regulator to see a transaction, the issuer of the smart contract (the bank) has to add it as an "observer"

Can a regulator, such as the SEC, observe everything that happens on the network even on transactions that do not concern the United States?

This is technically impossible. At Digital Asset, we are simply the technology provider. For a regulator to see a transaction, the issuer of the smart contract (the bank) has to add it as an "observer". If Société Générale issues a bond regulated by the SEC, it can give the SEC access to that particular contract. But nobody, not even Digital Asset, has a 'universal pass' to see all the flows.

In a tense geopolitical context, could the US government 'cut off' European banks' access to Canton?

There is no central operator. There are super-validators in several jurisdictions. Just because a US regulator says something doesn't mean that a Japanese or European entity is going to change its stance.

What could happen is that some of the entities under US influence would have to comply and could refuse certain flows. But that won't bring down the network: you need two-thirds of the super-validators. The analogy is Bitcoin China mining: when China "cut off", Bitcoin continued. However, some localised players were impacted.

So the risk exists at the margin on some participants, but it's not a centralised cut.

Let's turn to the token. It is listed on crypto exchange platforms, which came as a surprise for such an institutional project. Why this listing?

It was necessary to give the token a market value. But be careful, we have "institutionalised" tokenomics.

We have introduced a fee mechanism expressed in dollars, with settlement in Canton Coin, but with an indirect payment option for institutions that do not wish to hold crypto-assets.

In practical terms, the cost of processing a transaction is defined in US dollars, depending on the size of the package (kilobytes). Taking the example of a smart contract that would cost around a dollar, a bank can estimate "if I do 100,000 transactions a month, it will cost me around $100,000". This gives total budget visibility.

Banks have regulatory constraints (Basel III) that penalise holding cryptos on their balance sheet. How do you get around this obstacle?

We have provided a "token-free" option. A bank can pay a service provider (a super-validator) in traditional currency (euro, dollar, etc.) to buy bandwidth. It is this provider that will be responsible for using the tokens needed to process the transactions. The bank will therefore never have any crypto-assets on its balance sheet if it doesn't want to.

How are Canton Coin tokenomics designed?

On Canton, remuneration is not only directed towards those who operate the infrastructure. A large part of the value is directed towards those who build applications. It's a logic akin to an 'app store': the more your app is used, the more it is paid for. This is a fundamental difference from Bitcoin, where most of the revenue goes to miners, or Ethereum, where gas fees don't naturally go to apps.

What will give the token value in the long term?

There are two main mechanisms.

The first is the mint curve: it is deterministic, with halving-style reductions in rate over time (and a lengthening duration). On the other hand, there is no strict supply cap: this is a model where issuance tends asymptotically towards low inflation.

The second is the burn linked to usage. As fees are expressed in dollars, the more transactions the network processes, the more burn. And if the burn exceeds the mint, then supply enters a deflationary dynamic.

After that, we have to be honest: the price is also driven by short-term speculative flows, as everywhere else. And one particular point: tokens are not "easy" to obtain other than by participating in the network (via apps, volumes, commitments).

"There will be bridges between Canton and Ethereum"

Ethereum is establishing itself as the DeFi settlement layer. Can Canton really compete?

Ethereum is years ahead of retail, Canton is years ahead of institutional. I don't see Ethereum as a competitor, but rather the "status quo". 99% of financial flows are not yet on-chain. Our job is to get that figure up to 5%, 10% and then 20%.

There will be bridges between Canton and Ethereum, just as there are between Solana and Ethereum. But for heavy finance, which requires confidentiality and parallel processing (not being blocked by an NFT success that congests the network), Canton is better equipped.

Are public blockchains doomed to remain "retail" due to a lack of confidentiality and parallelisation?

It's hard to be categorical, but for the institutional, full transparency is a hindrance. And full replication makes certain optimisations (particularly parallelization) more difficult. Canton works more like segmented peer-to-peer: if two clusters of players interact, they don't mobilise the whole network.

Once again, our interest is that blockchain succeeds globally. But the institutional sector needs a different tool: institutional finance is not designed to be "censorship resistant" in the Bitcoin sense. It is designed to be compliant, auditable, controllable - and to protect information.

On Canton, will we have a dominant "settlement currency"? Stablecoins, tokenised deposits, CBDC wholesale... how is it structured?

Our strategy is to have the major players. We already have Circle with USDC, which is a leader in stablecoin. We also have tokenised deposits, like those of major banks. And you also have European players who are credible references on the euro.

The crucial point is that people need a common denominator: a stable, on-chain form of cash to settle asset transactions.

Then we have a conversion and netting issue: if one participant receives cash in the form of a tokenised bank deposit, and the other in the form of stablecoin, how do you manage the clearing between these forms of cash?

We're working on stablecoin clearing building blocks, but at a more systemic level, there's a model taking shape: tokenised deposits in commercial banks, and a conversion/clearing mechanism at central bank level - a wholesale CBDC or regulated liability register logic. We have already worked on projects such as the Regulated Liability Network (UK/US). The Banque de France is also interested. It will take time, but it's a coherent trajectory.

And there's an important point on the banks' side: stablecoin represents a "deposit flight" risk. To defend themselves, they have an interest in offering their own tokenised deposits, to stay in the cash-on-chain game.

>> Canton Network blockchain analysis

‍