the state of tokenization

This edition maps the state of tokenization in 2026: how thirty billion dollars of real assets moved onchain, why some rushed in while others stalled, and what it changes for capital.

About thirty billion dollars of real assets now trade as tokens on public blockchains. Not test pilots or press releases, actual U.S. Treasuries, gold, and private credit that people buy, hold, and move onchain, and the total has multiplied sixfold in two and a half years. The debate about whether traditional assets could work on a blockchain is mostly over. What replaced it is a more useful question: once an asset is a token, what can it actually do that it could not before?

A token, on its own, is just a record of ownership on a shared ledger. What gives it weight is the apparatus around it: the legal registry that says you own the thing, the settlement that makes a trade final, the rules about who is allowed to hold it, the ability to post it as collateral somewhere else. Get that apparatus right and the token is a real claim. Get it wrong and you have a number on a screen that happens to track a price.

This is why the market grew so unevenly. A Treasury bill moved onchain almost immediately, because its terms are plain and its natural buyers were already holding stablecoins. A company share took longer, since it comes bundled with votes, dividends, and a shareholder register that has to stay legally valid. Real estate has barely moved, tangled as it is in local law and physical deeds. Wherever an asset's ownership was clean to reproduce in code, the money followed, and that single fact explains most of the map.

What follows is that map in full: the size, the contents, what works, what is stuck, and where it heads next. The place to begin is the thing the headlines skip, which is how much had to be rebuilt to get even this far.

The financial system that tokenization is reaching into is not waiting to be rescued. It runs on a few large central databases, the Depository Trust Company in the United States and Euroclear in Europe chief among them, which record the ownership of most of the world's securities and move enormous sums with something close to perfect reliability. Broker-dealers, exchanges, transfer agents, and clearing houses sit on top, and each of them is there because it once solved a real problem. The settlement delays and the closed-on-weekends rhythm that get mocked from the onchain side are mostly deliberate, not accidental.

The clearest example is the settlement cycle itself. The standard one-day delay exists so that big brokerages can net millions of trades against one another through the day and settle only the net difference at the close, which saves an extraordinary amount of capital. Instant settlement on a blockchain gives that up and asks every trade to be funded in full up front, which costs the highest-volume firms more, not less. A serious chunk of the market is in no hurry to abandon a system that is cheaper for them precisely because it is slower. That is one real reason this transition is measured in years rather than months.

So the work moves one regulated category at a time, and it began with the assets that were most straightforward to bring across and already had buyers waiting.

Tokenized money-market and Treasury funds were the obvious first category. A Treasury fund pays a predictable government yield, its terms leave little to interpret, and a large pool of onchain money was already looking for exactly that kind of return. The match was close enough that the category filled before any other got going.

That demand pulled the whole market up with it, and the breakdown shows how lopsided the result is. Of the roughly $31B tokenized today, tokenized U.S. Treasuries are the largest and most liquid part at around $13.5B, close to 43% of everything onchain. The next slice surprises people: commodities sit near $5.1B, almost all of it gold. Credit, grouped together, adds another $5.1B, while stocks, real estate, and private funds make up a long and still-thin tail.

What changed more than the numbers is who runs these funds now. The early market was built by onchain startups trying to be the technology provider, the fund manager, and the compliance desk all at once. Over late 2025 and into 2026 that settled into a clearer division of labor, with infrastructure firms running the rails underneath and established asset managers handling the money on top. BlackRock's tokenized fund runs on Securitize's infrastructure. Franklin Templeton operates its own. The people who manage capital for a living are not rebuilding blockchain plumbing, and the people who built the plumbing are stepping back from managing the capital.

The shift shows up most clearly when a large player rents that infrastructure rather than builds it. When Coinbase launched a stablecoin yield fund, it ran the fund on an existing tokenization platform from the start instead of constructing its own issuance and compliance stack. Several of the tokenized Treasury funds handed to managers like Invesco and Bitwise over 2026 sit on the same kind of shared infrastructure underneath. A firm with Coinbase's resources choosing to plug into the rails rather than own them says plainly that the rails are now a product.

Token Terminal made the same case for the chains hosting these assets:

The reason to hold a Treasury as a token rather than in a brokerage account has little to do with the yield, which is identical either way. It has to do with what the token can do the instant it exists. It can be pledged as collateral inside a lending market while still earning, moved across networks in seconds, and dropped into an automated strategy, without a broker's permission and without waiting for a settlement window to open.

> collateral that moves on a weekend

The venues for this are already running. Aave Horizon, a permissioned lending market that launched in August 2025 so that verified institutions could borrow stablecoins against tokenized securities, grew past $440M in deposits and averaged around $500M locked through early 2026, making it the largest real-world-asset lending market onchain. Bitwise's carry fund puts more than $100M of its own holdings to work as live collateral across protocols like Aave and Kamino.

The capability with no real equivalent in the old system is timing. Traditional bond markets and the banks behind them close on weekends and holidays, so a desk that needs to post or move collateral over a weekend has to wait until Monday. Tokenized collateral can be pledged, moved, or re-pledged through a smart contract while those markets are shut, which removes a class of risk that has nothing to do with the quality of the asset and everything to do with the calendar.

> how far along this really is

Set against the system it wants to rebuild, though, tokenized collateral is still small. Aave's total value locked sat above $42B in early 2026, almost all of it native digital assets, while the real-world-asset market accounted for somewhere around 1.1% to 1.3% of that. A 2025 study found that a good deal of tokenized Treasury supply simply sits in wallets, held for the yield rather than put to active use. The capability is built and proven, and most of the capital has not yet arrived to use it. Tokenization works. It has not yet won.

Centrifuge framed the utility question this way:

The clearest reason the big asset managers are bothering with any of this is the money already sitting onchain. Stablecoins have grown into a pool too large to leave on the table. Total stablecoin market value was about $318B on the first of June 2026, up from roughly $130B through much of 2023. The volume moving through them matters more than the balance. Stablecoins settled around $33T in 2025, more than Visa and Mastercard combined.

The GENIUS Act, signed in July 2025, made this official by requiring stablecoins to be fully backed by dollars and short-dated Treasuries and holding issuers to real capital and disclosure rules.10 Once that was settled, the logic for an asset manager became simple. A few hundred billion dollars of mostly idle cash sits onchain, wanting yield it cannot easily reach without leaving the chain. Whoever offers a compliant product that lets that money earn without off-ramping wins a distribution channel that did not exist before. Citi's base case has stablecoin issuers and tokenized funds driving an extra $1T of demand for U.S. Treasuries by 2030, which ties this whole experiment back to the financing of the federal government.

Stablecoins gave the onchain world its cash. Tokenization is the layer of real investments being built on top of that cash, and the products launching now are aimed straight at the money already sitting there.

A regulated security cannot move to just any wallet the way an ordinary token can. So a tokenized security is built with a gate: an allowlist of approved, identity-verified addresses, and a transfer to any address not on the list simply fails. To a DeFi native this reads like a sellout. The founding promise was permissionless composability, any contract talking to any other with no gatekeeper, money behaving like open building blocks. An asset that can only move between pre-approved wallets looks like it hands the gate back to the people the technology was meant to route around.

For a regulated security, though, that same perimeter is what makes the asset real. Jim Hiltner, who runs one of the firms building this infrastructure, put it plainly on the podcast behind this issue.

‘If somebody steals my keys and tries to move my security from my wallet to their wallet, one of two things happens. They are not on the allow list, so good luck, you are not going to take my assets. Or it goes to an individual where there is an audit trail, and that audit trail has repercussions in courts.’ - Jim Hiltner - Co-founder, Superstate

The permissioning that annoys the purist is the same thing that gives a tokenized security recourse, identity, and a legal record standing behind the cryptographic one. It is what lets an institution hold the asset in the first place. The SEC drew the same line on the 28th of January 2026, stating that securities law applies the same whether ownership lives in a database or on a ledger, and separating issuer-sponsored tokens, where the blockchain is the real record of ownership, from third-party synthetic tokens that only point at an asset held elsewhere. The issuer-sponsored model got regulatory backing. The synthetic version, the kind that carries a price and nothing else, was held to stricter limits.

There is a real cost to all this, and it should be named. Enforcing identity and transfer rules does limit composability, and the token standards that carry these compliance hooks, ERC-1400 and ERC-3643 among them, keep regulated assets out of the open automated markets where liquidity runs deepest. Newer designs like ERC-7943 aim to make compliance lighter without breaking DeFi flows, but adoption is scattered and no standard has won. For now, tokenized securities live in well-built enclosures rather than on the open network. The same perimeter that makes them legally portable is what keeps them partly fenced, and no one has squared that circle yet.

Tokenized stocks are a small part of the map, around a billion dollars, but they answer the question every other asset is really asking: can a token carry the actual rights of the thing it represents, not just its price? In September 2025 Galaxy Digital issued its real registered Class A stock on Solana, with a transfer agent reading the chain to keep the official shareholder register current. Then in April 2026 holders voted those shares at the company's annual meeting from inside their own wallets, with the votes recorded onchain. The vote and the legal standing, the things a price-tracking token could never carry, crossed over and counted.

The next step is the primary market, the moment a company first sells shares to the public. A fully onchain IPO has not happened yet, though the parts are arriving. Circle's June 2025 listing showed why it might matter, priced at $31 and closing its first day at $83.29, with most of that jump going to whoever held the pre-market allocation rather than to the company. Figure has filed to issue native equity on Solana, and Securitize won approval to settle tokenized stock against stablecoins inside a regulated broker-dealer. No company has run an offering through these rails end to end, partly because shares spread across competing chains make a clean cap table genuinely hard. Equity is tiny today, but it is the proof that ownership rights can travel onchain, which matters for every asset that follows.

> in practice

When a tokenized asset can be a yielding share, collateral, and liquidity all at once, the real question is not what to put onchain but when each holding should be which, as markets move.

aarnâ is built for that decision: it allocates treasury capital under enforceable rules, with concentration limits, liquidity-aware sizing, and every action logged onchain. More at aarna.ai

Put the whole map together and the moment is clear. The market is real and growing fast, concentrated in treasuries and gold and short-term credit and thinning out wherever ownership gets tangled in law. Tokenization has won the argument that assets can live on these rails. What it has not settled is whether being onchain actually changes how an asset behaves, and the evidence so far is split: the collateral use and the weekend settlement are real, but most tokenized assets still sit still.

What makes this more than a format change is the machinery underneath. A bond was never really about the paper, and a fund was never about the statement. They were about ownership, settlement, and a claim recorded somewhere trusted and enforced somewhere real. For a long time that lived in central databases and a market that keeps office hours. It is being rebuilt in code now, one regulated category at a time, starting with the assets whose rights are cleanest and working slowly toward the ones tangled in law. The day those assets stop sitting still and start moving, tokenization stops being a better way to store finance and becomes a different way to run it.

This issue grew out of a conversation with Jim Hiltner, co-founder of Superstate, on tokenized funds, onchain equities, and what it takes to bring a shareholder's rights onchain. The full episode is on Unhashed.

reflections-

Gain an edge in DeFi alpha with aarnâ’s AI-driven insights and DeFi vaults. Try the dApp now.