Tokenized Treasury products and cash stablecoins are often filed under the same sleeve: “on-chain dollars.” They fail in different places. Treating them as substitutes is how a cash book discovers, on a weekend, that it holds a security.
This is a risk note, not a product ranking. Hedgen does not issue buy, hold, or pass. The first named book in our public library carries a different problem — Ondo Finance: Leading position, unproven economics — and is not the subject here. The subject is the break between two cash-like objects.
What each object is
A cash stablecoin is usually an issuer liability: you hold a token that the issuer, if it honours the programme, will redeem for fiat or reserves under its rules. The economic promise is par in a unit of account. The legal promise is whatever the issuer’s terms, reserves, and banking chain actually support.
A tokenized T-bill or tokenized money-market interest is usually a security or a note: you hold a claim on a fund, a series, or a bankruptcy-remote vehicle that itself holds Treasuries or cash-equivalents. The economic promise is NAV plus a short-duration yield, minus fees. The legal promise is the offering document, not the explorer.
Those two sentences already split the operational stack. One is built to move. The other is built to sit, under an allowlist, and to be redeemed through an agent.
Where cash stables break
They break at confidence in the issuer and the reserve chain. Secondary price can leave par while primary redeem is still open — or while it is not. Banking partners can be concentrated. A freeze function can strand a wallet that did nothing except sit on the wrong side of a sanctions list. A chain outage or a bridge wrapper can isolate a balance that is still “fine” on the home issuance.
They also break at policy. A payment stablecoin may be forbidden from paying yield. The same desk that wants a cash instrument for settlement then parks idle balances in a different object and pretends the hedge is perfect. It is not. The idle object has a different redeem clock.
What usually does not break first is the Treasury market. If the stable is reserved in bills, the bills can be sound while the token is not transferable, not redeemable, or not accepted as collateral.
Where tokenized T-bills break
They break at the wrapper.
Eligibility. Primary mint and redeem are typically gated: KYC, accreditation, transfer restrictions. A wallet-to-wallet transfer, where allowed, does not create a redeem right for the buyer. A secondary holder who cannot pass CDD does not exit at NAV. They exit at whatever a thin, permissioned book will pay.
Clock. Underlying bills are liquid in their own market. The token is processed by a transfer agent, an administrator, and a banking day. T+0 or T+1 during hours is not 24/7 NAV. Weekend, holiday, and after-cut-off balances are inventory, not cash.
Oracle and collateral use. Once the token is posted into a lending market, the failure moves. A stale NAV, a restricted liquidation universe, or a buyer set limited to qualified wallets can turn a “cash-equivalent” into a gap the protocol socializes. The bill did not gap. The wrapper did.
Document risk. The token is not the CUSIP. Insolvency, administrator failure, or a change in the series documents is a claim process, not a redeem click.
Public explainers in 2026 make the same mechanical split: stables for balances that must move; tokenized funds for balances that can wait and accept a gate. See, for example, Eco’s August-era note on tokenized MMFs versus stablecoins (public support article, 2026), which frames the pair as complements — checking account versus sweep — rather than as interchangeable cash.
What to put in the cash memo
For each balance, write four lines.
- Legal form: issuer liability or fund/note interest.
- Who may redeem, on what T+n, and who may halt.
- Where the balance must be usable: payment, collateral, or idle reserve.
- What happens if the wrapper works and the rail does not — bank, agent, oracle, or allowlist.
If those four lines are the same for a cash stable and a tokenized bill, the memo is wrong.
Yield is the last line, not the first. A short-duration Treasury return does not repair a redeem that is closed, a secondary that is permissioned, or a liquidation that cannot find an eligible buyer. Par on a stable does not repair a reserve chain you have not read.
The break is not “crypto versus TradFi.” Both objects already sit on Treasury paper somewhere in the stack. The break is which failure you are paid to ignore. Settlement books that ignore gates will learn about securities law from an operations ticket. Yield books that ignore issuer and wrapper risk will learn about cash from a discount to NAV.
Write the failure first. The sleeve name can come after.
