Tokenised treasuries and the yield you can actually verify

Tokenised treasuries are the rare crypto yield product where you can name the borrower. The fund buys short-dated US government debt, the government pays the coupon, and a token records your share. There is no utilisation curve, no perpetual funding rate and no counterparty who might have made a bad loan. The yield is the safest one on this site, and correspondingly it is also among the lowest.

The sector exists at its current size because of regulation rather than despite it. Once the GENIUS Act barred payment stablecoin issuers from paying interest for holding a stablecoin, and MiCA barred remuneration linked to holding period in the EU, a regulated fund paying a distribution became the compliant way to give an on-chain dollar balance a yield. That is the clearest regulatory-arbitrage story of 2026, and it explains why BlackRock, Franklin Templeton and Circle are all in this market at once.

What you are actually holding

Strip away the vocabulary and the structure is ordinary. A fund is formed, usually as a regulated vehicle with a transfer agent and an auditor. It buys Treasury bills, repurchase agreements and money market instruments. Investors subscribe and redeem, and the fund's income is distributed or accrued into the share price. All of that has existed for decades.

The innovation is where the share register lives. Instead of existing only in a transfer agent's book-entry system, ownership is recorded as a token on a public blockchain. That changes three practical things. Settlement runs continuously rather than on business days — OUSG advertises instant mint and redemption 24 hours a day, seven days a week, including weekends and bank holidays. Transfer becomes a blockchain transaction rather than an instruction to an intermediary. And the position can be used inside DeFi, as collateral or inside another product, without leaving the fund.

None of that changes the underlying credit. You are still holding a claim on short-dated US government debt, which is the reason these products pay what they pay and no more. The constraint on who may hold the token is also unchanged: it is enforced at the token level through allowlists, which is how a fund restricted to qualified purchasers stays restricted on a public chain.

The funds, side by side

Six products account for almost all of the market. The table below sets out what each one is, what it pays, what it costs to get in and — the column that actually decides the question for most readers — who is allowed to hold it.

The eligibility maze is the real story

Every comparison of this sector published so far ranks by yield. The yields are within 40 basis points of each other. What actually varies is whether you can buy the thing at all, and that variation is severe.

A US retail investor with $1,000 has exactly one plausible route in this table, BENJI, and it is the one product whose current terms we could not verify from the issuer. A US accredited investor who is also a qualified purchaser can access OUSG at a $5,000 minimum or USTB. A non-US individual gets USDY and is shut out of the products aimed at US residents. Everyone else is looking at a $5m minimum on BUIDL or $100k on USYC.

Qualified purchaser is a materially higher bar than accredited investor, and several of these products require both. The practical effect is that the sector's headline of "T-bill yield, on-chain, for everyone" describes almost nobody. That gap between the marketing and the eligibility terms is the single thing a reader should check first, before the yield.

Risks that do not disappear

Government credit risk is close to zero on a three-month horizon, and that is the risk most readers think they are evaluating. Three others remain, and none of them is priced into the yield.

The first is the wrapper. A token is a smart contract, and the allowlist, redemption logic and price oracle are code someone wrote. A fund whose portfolio is perfect can still have a bad afternoon in its contract. The second is the redemption path. Same-day redemption via USDC, as USTB offers, is only as good as the stablecoin and the bank rails behind it; a fund that promises instant liquidity is making a promise about an intermediary as well as about its own assets. The third is eligibility drift — the terms on which you may hold the token are set by the issuer and enforced at the contract level, and a change in those terms is not a market event you can trade around.

Set against a CeFi savings product, none of that compares badly. It is simply a different list, and worth reading alongside the one on crypto earn risks before treating any of these as a cash equivalent.

Why the market caps disagree between trackers

Superstate reports $746,576,347 of AUM for USTB. rwa.xyz shows $572.10m. Neither is wrong. The fund holds 71.75% of its shares on Ethereum, 0.60% on Plume, 0.26% on Solana — and27.39% in conventional book-entry form, exactly as an untokenised mutual fund would. A tokenisation tracker counts the tokenised portion; the issuer counts the whole fund.

The same effect, or something like it, explains BUIDL at $2.70bn on rwa.xyz against $3.515bn on DefiLlama across eight chains, and OUSG at $403.84m on rwa.xyz against Ondo's own $330.70m as of 15 September 2026. Anyone citing a single number for this market without saying which base it uses is overstating precision. We flag conflicts rather than resolving them, which is set out in our methodology.

The structural point: you are not buying extra return

Set the yields against the benchmark. A three-month US Treasury bill yielded 3.97% on 15 September 2026; the four-week bill was at 3.80%, the six-month at 4.07% and the one-year at 4.16%. The tokenised funds pay between 3.44% and 3.74%.

In other words the wrapper costs roughly 25 to 50 basis points against simply owning the underlying instrument. That is not a criticism — it is what a fund wrapper has always cost, and it is cheap compared with most things in crypto. But it does settle the question of what these products are for. You buy them for 24/7 settlement, for instant mint and redemption including weekends and bank holidays in OUSG's case, and for the ability to move the position into DeFi as collateral. You do not buy them to beat a Treasury bill, because by construction they cannot.

That comparison also reframes the rest of the market. Aave v3 paid 3.57% on supplied USDC on 16 September 2026, which is below a three-month bill for a position carrying smart-contract and utilisation risk. Ethena's sUSDe paid about 5.03%, down from roughly 27% at launch, for a delta-neutral basis trade with exchange and custody risk. Ledn pays 6.5% and 8.5% on stablecoins from a disclosed, overcollateralised bitcoin-backed loan book. A tokenised treasury at 3.44% is the floor of that ladder, and knowing where the floor is makes everything above it easier to judge — which is the whole argument ofstablecoin yield andcrypto interest rates.

What happens next depends on rulemaking

The regulatory pincer that created this market is still moving. In the US, the OCC released a 376-page notice of proposed rulemaking in February 2026 whose proposed §15.10(c)(4) would create a rebuttable presumption of violation where a stablecoin issuer contracts with affiliates or broadly defined related third parties — which would make exchange rewards programmes on branded stablecoins presumptively unlawful. The comment period closed on 1 May 2026, the Comptroller has said the aim is a final rule by November 2026, and GENIUS takes effect on the earlier of 18 January 2027 or 120 days after final rules. The CLARITY Act, which would have preserved transaction-based stablecoin rewards subject to Treasury authority, failed a Senate cloture vote 49–50 on 15 September 2026.

In the EU, the European Commission's targeted MiCA review consultation opened on 20 May 2026 with 86 questions, one of which asks directly whether the stablecoin interest prohibition should be kept or loosened. It closes on 30 September 2026. Nothing is decided and any legislative proposal is years away, but the possibility that the prohibition is relaxed is the main thing that could reduce the demand this sector currently enjoys. We track the detail on crypto yield regulation and the geographic consequences on availability by country.

For a reader deciding today, the honest summary is short. If you are eligible and you want a dollar yield you can audit, this is the most transparent thing available and it pays slightly less than the bill it holds. If you are not eligible, no amount of interest in the category changes that, and the alternatives — pooled lending,CeFi savings products — carry credit risk that Treasury bills do not. Sizing that decision sensibly is the subject ofpassive income with crypto.