A chief financial officer asked to hold a tokenised treasury fund, or a general counsel reviewing a stablecoin's reserves, will eventually hit the same question: is this thing rated, in the sense a bond or a money market fund is rated, or does it carry something that only looks like a rating? The answer depends on which agency, which instrument and which year you ask. Moody's, S&P, Fitch and KBRA have each published on digital assets, and they have not converged on the same answer.
A rating and an assessment are not the same document
S&P Global Ratings launched its Stablecoin Stability Assessment (SSA) on 12 December 2023, and its launch release states that "this report does not constitute a rating action." The SSA scores a stablecoin's ability to hold its peg on a five-point scale, from 1 (very strong) to 5 (weak). It starts with an asset assessment covering credit, market-value and custody risk in the reserve, allowing for any overcollateralisation, then looks at five further areas: governance, the legal and regulatory framework, redeemability and liquidity, technology and third-party dependencies, and track record. Those can lower the final score below the asset assessment but not raise it above. Eight stablecoins were scored at launch, from 2 for USDC, USDP and GUSD to 5 for FRAX and TUSD. In a release dated 4 August 2026, S&P said it covered eleven stablecoins, six of them at adequate or above, and that two of the eleven had been revised to a weaker level over the previous three quarters.
Moody's took a different route: a rating methodology rather than a parallel scale. It published a request for comment on a proposed stablecoin methodology on 12 December 2025, took comments until 26 January 2026, and published the final methodology on 17 March 2026, which it describes as the first such framework for fiat-backed stablecoins. The scope is narrow. A stablecoin is eligible only if it is fully collateralised, convertible into fiat on demand, and its reserves are effectively segregated from the issuer's balance sheet, so that they remain available to holders even in the issuer's bankruptcy. Algorithmic stablecoins are outside it. Within scope, the analysis centres on the credit quality of the reserve pool, and names blockchain risks such as smart contract bugs, network forks and 51% attacks as operational considerations for which the issuer, having chosen the infrastructure, is responsible.
Where a rating already sits on tokenised instruments
Tokenised money market funds have moved furthest toward ordinary rating treatment. Moody's assigned Aaa-mf, its top money market fund assessment, to BlackRock's USD Institutional Digital Liquidity Fund (BUIDL) and to Fidelity International's USD Digital Liquidity Fund in May 2026, and to Franklin Templeton's Franklin OnChain U.S. Government Money Fund on 26 August 2026. S&P published a credit FAQ, reported on 10 July 2026, setting out how its existing Principal Stability Fund Rating and Fund Credit Quality Rating scales apply to tokenised vehicles; at that point it rated three: OpenEden's TBILL Fund, the Janus Henderson Anemoy Treasury Fund, and the Delta Wellington Ultra Short Treasury On-Chain Fund. The underlying portfolio of short-dated Treasuries and repo is what earns the rating. Tokenisation adds a separate layer that S&P says it examines: smart contract integrity, oracle accuracy, which stablecoins are used for subscriptions and redemptions, wallet security, and the resilience of bridging mechanisms. How those vehicles are structured in the first place is covered in our piece on digital asset fund structures.
Tokenised debt has an example rather than a published framework. Shenzhen Futian Investment Holdings, a state-owned issuer, sold a two-year, offshore RMB 500 million bond with a 2.62% coupon on the Ethereum blockchain, listed on the Shenzhen and Macau stock exchanges, in September 2025. We found no Fitch criteria document for on-chain bonds generally. For the wider credit market moving on-chain, see private credit on-chain.
Where agencies have said not yet
The clearest published hesitation is old. In a report dated 22 April 2019, KBRA said, as reported at the time by Modern Consensus, that issuers had approached it about rating fully on-chain securitisations and that it "has not agreed to do so yet." Its concern was that a permissioned blockchain introduces a new party, the blockchain administrator, and that a securitisation depends on every party being replaceable. It wanted backup conventional service providers identified, contracted and priced before issuance, so a structure could keep running if the blockchain environment failed. That is KBRA's 2019 position. We found no later KBRA statement either repeating or revising it, so it should not be read as KBRA's view today.
KBRA has since rated something that sounds similar but is not. On 2 April 2026 it assigned BBB issuer ratings to Ripple Prime CIV US BD HoldCo LLC and its operating subsidiary, Hidden Road Partners CIV US LLC, a US broker-dealer and futures commission merchant owned by Ripple Labs, which KBRA said had injected about $500 million of capital. That is an entity rating of a regulated intermediary, not a rating of a token or an on-chain instrument. The difference between rating the intermediary and rating the asset it holds is the one that matters when reading how crypto prime brokerage works: the first tells you about a balance sheet, not about whether your assets would survive that balance sheet failing.
Fitch has written mainly from the bank side. In commentary dated 8 December 2025, as reported by Investing.com and Cointelegraph, it said that US banks' digital asset activities bring reputational, liquidity, operational and compliance risks even in lower-risk businesses such as trust and custody, and that it "may negatively re-assess the business models or risk profiles of US banks with concentrated digital asset exposures." Either way, it concerns the risk a crypto business poses to a bank's credit, not a method for rating crypto assets.
The questions the methodologies keep returning to
Smart contract and operational risk
Both S&P's tokenised fund FAQ and Moody's stablecoin methodology name smart contract failure specifically, rather than leaving it inside a general operational-risk heading. A coding error or a bad price feed can move value instantly and irreversibly, which a back-office reconciliation cannot catch in time.
Custody
S&P's SSA scores custody risk within the asset assessment. Moody's makes reserve segregation a condition of being rated at all. Underneath both is the question institutional custody arrangements have to answer: whose insolvency would leave a holder short of the asset they thought they held.
Legal enforceability of token holder rights
The clearest recent statement is from a regulator, not an agency. In a joint staff statement dated 28 January 2026, the SEC's Divisions of Corporation Finance, Investment Management, and Trading and Markets said that a tokenised security issued by the issuer itself may be treated as the same class as the traditional security where it carries substantially similar rights. Where a third party holds the security and issues a token, the holder may have only an indirect interest through a security entitlement, and may be exposed to risks such as that third party's bankruptcy. The statement says it has no legal force and creates no new obligations; it is staff guidance on existing law.
Settlement finality
An article in the ECB's Macroprudential Bulletin of April 2026 on tokenised money market funds notes that "permissionless blockchains often settle transactions probabilistically, leaving settlement finality ambiguous," and that parallel off-chain registers "can dilute DLT immutability and increase operational complexity." It is analysis, not a rule, but it states plainly why "the ledger shows it" and "ownership is legally final" are two different claims.
What is settled, and what is not
Settled, for now: short-dated tokenised Treasury and money market funds get top-tier fund ratings when the portfolio would earn one anyway, with the on-chain layer examined alongside. Not settled: whether an S&P stability assessment and a Moody's stablecoin rating can be read as comparable; how any agency would treat algorithmic or under-collateralised stablecoins, which the published frameworks exclude; and how a rated on-chain instrument fares when its chain, bridge or oracle fails, which none of these methodologies has yet had to work through in a live default. This is not investment, legal, tax or accounting advice.