Liquidity follows utility, not listings

2 numbers describe the same market and appear to disagree. Japan's ODX runs START, the country's dedicated secondary exchange for security tokens, and against a listed market capitalisation of JPY 33.6 billion it was turning over roughly JPY 23 million a month as of the FY2025 Japan Security Token Market Report published in April 2026. That is well under a tenth of a percent of the market changing hands in a month, on a venue built for exactly this purpose. Broadridge's Distributed Ledger Repo platform moved an average of USD 365 billion a day in July 2026, which means one tokenised-collateral platform shifts more value in a single day than the entire USD 16.21 billion tokenised Treasury market holds, more than 20 times over, per rwa.xyz counts as of 9 August 2026.

These numbers do not disagree. They show that "liquidity" in digital assets is 3 different problems wearing one name, and that the industry keeps promising the one it cannot yet deliver while underselling the 2 it already can.

3 problems, one word

Conventional finance treats these as separate questions, which is why a money market fund with zero secondary trading is still called a liquid asset. It redeems at par daily, and it pledges beautifully. Nobody has ever demanded an order book for it.

• Exit. Can a holder get out at a fair price, on a known timetable, without begging the issuer.
• Depth. Can size trade continuously against a live two-sided market without moving the price.
• Velocity. Can the asset be used, as collateral or margin or a repo leg, without being sold at all.

Tokenisation commentary collapsed the 3 into one, and the collapse produces bad conclusions in both directions. Critics point at START-grade turnover and call the category a failure. Promoters point at 24/7 transferability and call it liquid. Both miss where the evidence actually points.

What the buyers say, and what the builders built

The buyer evidence is consistent. Sygnum's APAC survey of 212 wealthy and professional investors found secondary-market liquidity worrying 43% of respondents whether or not they already held tokenised assets, one of only 2 barriers that experience does not soften, a finding I wrote about in The barriers that don't decay. Exit risk is priced accordingly. What the successful products built in response was not exchanges.

BlackRock's BUIDL, at USD 2.68 billion as of 10 August 2026, has no secondary venue at all and roughly 113 holders. Its liquidity is a Circle smart contract, live since April 2024, that converts BUIDL to USDC around the clock. Ondo built Nexus to give third-party tokenised Treasuries the same instant-redemption exit, and when it launched tokenised equities it chose mint-and-redeem against underlying broker liquidity, not an order book. The growth stories track utility rather than tradability. BUIDL's inflows followed crypto venues accepting it as margin collateral, and the strongest liquidity number in the entire asset class is repo velocity on Canton. Even the market makers are telling us this with their feet: when B2C2, Cumberland, FalconX and GSR committed to Canton in July 2025, they signed up to a collateral initiative, not a listing venue.

From my perspective, exit plus velocity is not a consolation prize while the market waits for depth. For NAV-based products, which is most of what has tokenised so far, it is the whole prize. The GDF and ISDA workstreams, 120-plus firms concluding in July 2026 that tokenised MMFs work as US institutional collateral, and the CFTC admitting tokenised Treasuries and MMFs as derivatives collateral in its December 2025 pilot programme, are worth more to adoption than any exchange listing announced this year.

Where depth genuinely matters

Honesty requires the other half. Bonds before maturity, equities, private credit at scale, anything a holder cannot simply redeem, does need real secondary depth eventually, and the evidence there is thinner and more interesting than either camp admits. The HKMA's research on digital-twin bonds found tokenised issues pricing at lower yields with bid-ask spreads about 5.3% tighter than conventional twins, while the BIS's 2025 assessment found a much more modest premium and called the gains modest in a nascent market. Both can be right. A small liquidity premium exists, and it is nowhere near paying for the infrastructure yet.

The structural worry belongs to the officials. The FSB warned in October 2024 that non-fungible token versions of the same asset fragment liquidity and let prices diverge, and the BIS's 2026 annual report makes fragmentation across non-interoperable ledgers its central objection. Kaiko's 2026 read of tokenised equities says the same thing from the trading floor, with the same stock listed in pools that do not share order flow. Every chain, wrapper and permissioned venue that launches makes depth harder, not easier, because depth is the one liquidity that cannot be engineered per-issuer. It has to pool.

Asia is where the next 18 months of evidence will come from. 3 markets are worth watching closely, each running a different bet:

• Hong Kong. The SFC opened public secondary trading of tokenised authorised funds on licensed venues in April 2026, against roughly USD 10.7 billion of tokenised share classes as of March 2026, and no volumes have been published yet. Whether that framework produces real turnover or a second START is the cleanest experiment running anywhere.
• Korea. Security-token amendments take effect in February 2027, a fresh market designed after everyone watched Japan's.
• Singapore. The city-state keeps choosing the other road entirely, favouring daily-liquidity funds through regulated distributors and sgBENJI pledged into repo at DBS, a bet that exit and velocity are the product and depth can wait.

This argument has a failure mode worth stating plainly. If issuers market "liquid" tokenised products on the strength of instant redemption, and redemption depends on a single facility or a single market maker, thin depth becomes run risk rather than inconvenience. The FSB's transition-risk framing is the right lens here. Engineered exits concentrate risk, while pooled depth absorbs it. The honest position is that engineered exit is sufficient for money-like assets and a stopgap for everything else.

So when a tokenised product claims liquidity, the questions worth asking are not whether it is listed. They are what the redemption service level actually is and who stands behind it, which venues and clearers accept it as collateral, and how deep the stablecoin pair on the other side of the swap runs, because in practice nearly every exit clears into stablecoins, whose adjusted transfer volume hit USD 1.79 trillion in June 2026. Stablecoin market structure is the liquidity backbone the whole asset class borrows. Products that answer those 3 questions well are liquid in the way that matters. The order books can come later, and mostly, they will have to.