The report rightly highlights that tokenized markets are fragmented.
But it doesn’t fully shed light on why.

The IMF’s October Global Financial Stability Report includes a chapter on tokenization, “Scaling Tokenization: New Efficiencies, New Vulnerabilities.” Much of it is right. Tokenized markets are fragmented. Continued growth depends on legal certainty, regulatory clarity, interoperability between networks and safe settlement assets. Policy should be technology-neutral: same activity, same risk, same regulatory outcome.

One area that didn’t get much attention, and is worth expanding on, is the pivotal role of privacy and how it addresses some of these challenges in the process.

Collateral mobility has found product-market fit

ValueExchange research highlights why institutions are increasingly moving business onchain. The average financial institution is pledging and managing up to $74Bn in collateral, yet 25% of a firm’s collateral ($25Bn) is either excess or not remunerated overnight. 7% of firms post excess collateral for buffering ($4Bn, with 35% of firms posting more than half of their collateral overnight.

When collateral can move atomically against cash, in minutes and at any hour, firms can hold less idle liquidity, meet margin calls in real time and put the same securities to work several times a day.

The IMF highlights intraday repo as a high priority use case, evidenced by the $350+ billion a day in tokenized repo facilitated by Broadridge’s Distributed Ledger Repo (a platform that runs on Canton). With DTC-custodied treasuries, equities, and ETFs also now live on Canton, and platforms like Tradeweb, Talos, and Temple enabling firms to use those assets to access financing and meet obligations around the clock, those numbers are primed to scale.

The next step is policy, not technology. Central banks and CCPs now need to clarify when tokenized assets are eligible collateral, so the efficiency gains in bilateral markets carry through to cleared markets and central bank facilities.

Legal and regulatory certainty is critical

Legal certainty has to cover the asset, not just the token. The IMF is right that legal certainty is a precondition for scale. We would go even further. Holders need certainty on two things:

1. The token must confer ownership, not just exposure.

The IMF highlights the ongoing risks of synthetic tokens that convey price exposure without ownership.

A wrapper is a promise of the security, held through an intermediary, with added counterparty risk and no guarantee of the rights that come with ownership. When a CSD issues on Canton, the token is recorded on the CSD’s own books. It carries the same rights, the same entitlements and the same counterparty profile as the security it represents, and there is no second record to reconcile.

DTCC’s tokenization service now making $30T+ of eligible US securities available for use on Canton, including large-cap stocks from the Russell 1000 index, is a prime example.

2. Transfer on ledger must be a full transfer of ownership

The ledger must be recognized as the official books and records of the asset. It also requires settlement finality and CSD rules that treat settlement on DLT the same as settlement anywhere else. In the United States, this is being advanced through the SEC’s No Action Letter to the DTCC, recognizing that Canton serves as the ledger for the asset, and through proposed rules on Transfer Agents. This clarity is now needed across other key jurisdictions.

What it missed

Where we diverge in opinion is on scale and the future potential of tokenization. The report concludes that “the current trajectory points to limited scaling and adoption.” There are reasons that some networks have started to hit a scale ceiling for financial markets, but on Canton, tokenization is not stalling; it is accelerating where it has found real market utility.

The issue the IMF overlooked is the fundamental need for control over privacy onchain. Across nearly thirty pages, privacy is mentioned twice: once as a risk, once as a supervisory headache. Issuer control over a tokenized asset, that is who may hold it, move it and see it, is hardly discussed. Those are some of the first questions every bank, asset manager and market infrastructure provider ask us about before moving anything onchain.

As a result, the report describes the use cases and the policy outlook, but overlooks privacy as a critical enabler. Two things, in particular, deserve a closer look.

1. Privacy is not an outlier risk: privacy trade-offs have increased fragmentation.

What the IMF says: “Public ledgers could improve transparency, but there is a risk that participants will choose private networks or privacy technologies instead, limiting the benefits of public ledgers. Meanwhile the market is stuck in a “fragmented equilibrium” of private platforms run by individual institutions or consortiums.”

What it missed: The IMF treats privacy as a risk to be managed and fragmentation as a separate problem. In reality they are two sides of the same problem. Institutions did not retreat into private platforms to protect turf. They did it because most public blockchains broadcast every balance, counterparty and trade to the world. No bank will put its repo book on a block explorer. No asset manager will show competitors its positions in real time. No treasurer will publish the firm’s cash balances.

Given a choice between radical transparency and building private islands, many originally built the island. On the other end of the spectrum, privacy techniques on top of fully transparent chains exist for the right reasons, but have led to opacity for regulators. Institutional (and increasingly retail) preference for privacy is not an outlier. It is a universal design requirement. But fragmentation should not be the cost, and the approach shouldn’t be to hide everything.

Privacy is about defining who is entitled to see what. Embed those entitlements into smart contracts that can interoperate across applications without breaking the rules, and everyone sees what they need, and nothing more. Regulators included.

Without a network that solves this at its core, no amount of legal clarity or interoperability standards will bring institutional liquidity onto shared rails at scale.

Canton is the only network today that is proven to deliver the full composability benefits of public blockchains, without the trade-offs: unified liquidity, not fragmentation. Selective disclosure, not all-or-nothing privacy.

2. Private, but auditable: the supervisory tech the IMF wants already exists

What the IMF says: “Supervisors will need to build capacity in new areas, privacy technologies among them, and blockchain analytics firms could help authorities monitor markets. Ideally, real-time reporting would shift oversight “from periodic reconciliation toward more continuous supervision and monitoring.”

What it missed: The report points to the opportunity of real-time reporting, but that is not a future state, it exists today. On Canton, control over privacy is built into the protocol and then defined by each application’s smart contracts. Transaction data is only visible to entitled parties; regulators can see required information where appropriate and data is not exposed to everyone with a block explorer.

That mirrors how regulated markets already work: a buyer and seller see the trade, the settlement bank sees only the cash leg, the CSD records only the transfer of title. A regulator can be named as an observer on exactly the contracts within its remit and receive authoritative data on them in real time, with nothing exposed on a public ledger unless it’s been agreed. There’s no need for analytics firms to work out who is behind a wallet. Regulators can reach a point where, rather than requiring firms to report market data hourly or daily, the reports are submitted in real time and regulators have an ongoing understanding of market dynamics.

Independent control doesn’t need to mean fragmentation

The two themes the report spends most of its pages on: fragmentation across networks, and the risks of round-the-clock trading, look very different when privacy and legal control over the asset are solved.

The IMF sees two possible end states: one shared platform that concentrates risk, or many separate ledgers glued together by bridges that are fragile and split liquidity.

But there is a third way. And it is how the internet (and Canton) work today: a public network at the base, with independently governed applications on top, each with its own privacy and compliance rules. Canton adds the blockchain innovation of atomic value transfer across these applications, while ensuring each party only sees its own leg of the transaction. No bridges or global data broadcast.

In July 2026, that model was once again demonstrated in production when 30+ firms used DTC-tokenized Treasuries, equities and ETFs in live trades: delivery-versus-payment against onchain cash, intraday treasury repo lasting minutes, securities lending and real-time margin delivery.

On Canton, the DTCC’s tokenization service and cash issuers’ (whether stablecoin or tokenized deposits), trading and financing applications (like Tradeweb, Temple and Broadridge) all operate on one network. They can settle atomically. BNP Paribas lent tokenized Treasuries intraday. Marex ran repo for hours, not days. Every participant keeps control over who sees their transactions, collateral and inventory. No bridges, no islands.

On the issue of trading hours, the institutional question is not whether markets can trade around the clock but whether they can settle when they need to. As Tradeweb’s chief technology officer, Justin Peterson, put it: “We’ve built 24/7 trading, but it doesn’t mean anything unless you can settle 24/7.”

The IMF is right that adoption needs the coordination that individual institutions have often struggled to provide on their own.

Our point is that the coordination is already happening, and it became possible when privacy and control stopped being treated as obstacles to shared open networks, and started being treated as the conditions for them.

The institutions the IMF most wants to see on shared infrastructure are already there. The question is no longer what is needed for tokenization to deliver. It is how quickly the rest of the market connects.