Under active development. We're continuously fixing and improving the metrics and parameters, so scores and ratings may change. See changelog

RWABeat

Methodology

RWABeat asks what a lender or curator asks before accepting a tokenized real-world asset: what can go wrong, how much would I lose, and how fast can I get out? Every rating on the site maps back to this page. It is versioned in Git, and every change is listed in the changelog.

How it works

  1. 1Collect evidenceCurated facts with a source for each, admin powers read from the chain, live prices and lending markets, and measured liquidity.
  2. 2Rate seven risksEach category runs its checks and takes the worst result: green, yellow, red, or unverified when a fact isn't public.
  3. 3Place on a stageStages 0 to 2 say how much you have to trust the issuer, across all seven risks.
  4. 4Watch for changesContracts, filings, oracles and NAV are watched. A change raises an alert and re-runs discovery.

The seven risks

Each category runs a set of checks and takes the worst result. A known red check makes the category red. Otherwise, if a check couldn't be made because the fact isn't public, and the answer could make things worse, the category is unverified, shown with the worst color we could confirm ("unverified, at least yellow").

GreenNo known problem in this category.
YellowA weakness worth knowing about.
RedA serious problem: it can cause a loss or stop holders getting out.
UnverifiedA fact isn't public and could make things worse. Shown with the worst color we could confirm.
Not assessedNot rated yet. Only Legal Claim, until a legal reviewer joins.

Some checks depend on the asset type. Data and prices have to be at least this fresh:

Rated asFreshness that counts
T-bill or money market fundDaily
Short-duration creditWeekly
Private creditMonthly, plus defaults reported as they happen
Real estateQuarterly

Category rules

Open a category to see exactly what makes it green, yellow or red.

Underlying Asset

What backs the token, and how likely is it to lose value?
Green
At least 95% government debt, government repo or cash (including government money funds and stablecoins), and short-dated: a weighted average maturity of a year or less. A fund holding only bills, repo and cash qualifies.
Yellow
Private credit that is investment grade or mostly senior secured, spread across borrowers, with a year of history. Also government debt with a weighted average maturity over a year, since rate moves can cut the price.
Red
The largest borrower is over 20% of the pool, or there are fewer than 10 borrowers; mostly unsecured, sub-investment-grade loans; less than a year of history; or losses passed to holders in the last three years.
If the issuer goes bankrupt, do holders still own the assets?

Not assessed yet. It will check the structure (regulated fund, bankruptcy-remote SPV or direct issuer liability), what the token legally is, whether it is the official ownership record, the jurisdiction, an independent trustee or board, and regulatory status. The Legal tab shows the public records in the meantime.

Counterparties

Which offchain companies could fail, even if the asset is fine?
Green
An independent, regulated custodian and administrator, a recognized auditor (Big Four, PCAOB-registered, or a major international network), more than one settlement rail, and no other single point of failure.
Yellow
An unregulated custodian or administrator, a small auditor, one firm acting as both custodian and administrator, a single settlement rail (for example USDC only), or another company the product can't run without.
Red
The issuer's own group holds, administers or audits the assets; there is no audit; or the custodian or administrator isn't disclosed.

Control

Who can freeze, mint, pause or upgrade, and what stops them?

We read every privileged power from chain, follow proxy admins, timelocks and controlling contracts down to the account that signs, and rate the weakest holder of each power.

  • Critical powers: upgrade, mint, force-transfer, burn from a holder, set the price, allocate (move the backing assets) and grant roles.
  • High powers: freeze, pause and change compliance rules.
  • Single key: an EOA, an unrecognized contract, or a 1-of-n multisig.
  • Weak multisig: fewer than 3 required signers, or fewer than half of all signers.
Green
Powers limited and documented, held by independent multisigs behind a timelock, and audited in the last two years.
Yellow
A single key can pause or change compliance rules; a critical power sits with a weak multisig or one with no timelock; a Safe module can bypass the signers; multisig signers all work for the issuer; admin powers aren't documented by the issuer; the latest audit is over two years old; or the token depends on a bridge on some chain.
Red
A single key holds a critical power or can freeze; there is no published audit; or a contract's source isn't verified.

For regulated securities, freeze and force-transfer are often legally required. We rate who holds them and what checks exist, not whether they exist. Privileged functions no handler recognizes are listed for manual review and don't change the rating.

Transparency

Can you see what backs the token, and trust that data?

We rate the best public report on four ladders: who the data comes from (issuer, administrator, custodian, independent firm), how detailed it is (a single NAV, a breakdown, every position), how fresh it is for the asset type, and how it can be verified (self-reported, attestation, audit, cryptographic).

Green
Data from a third party, every position, fresh enough for the asset type, and at least attested. For private credit, defaults are also reported when they happen.
Yellow
A third-party attestation or audit, but only aggregate numbers, data from the issuer itself, or updates too slow for the asset. For private credit, defaults that only show up in the next periodic report.
Red
Self-reported only, or no meaningful disclosure.

Onchain data is a separate badge, not part of the rating: data published onchain helps automation, but public offchain data is still transparent to a person. Ladders adapted from Chronicle Labs, "The Transparency Gap in Onchain Finance" (Oct 2026).

Pricing

Is the onchain price right, fresh, and hard to manipulate?
Green
NAV from an independent administrator or custodian, delivered by an oracle with visible staleness, and losses reflected promptly.
Yellow
NAV published by the issuer with no independent check; an issuer multisig publishing the price; a feed running late (past one update cycle); write-downs reaching the price late; a traded token whose discount to NAV isn't tracked or has exceeded 2%; or a lending market hardcoding $1.
Red
No NAV feed (a stable token is then simply assumed to be worth $1); a single key writes the price or can write a vault's NAV down; updates slower than the asset's freshness bar; a feed older than three update cycles; or consumers can't tell when the price is stale.

Funds that strike NAV on business days only don't count weekends towards the price's age.

Exit

How fast can a holder, or a liquidator, get out?
Green
Redemption in a day or less with no discretionary gates, a real instant path (a $5M DEX sell within 1% of NAV, or an instant redemption facility that can pay $5M today), and any liquidator can hold the token.
Yellow
Redemption for allowlisted holders only, or with the issuer's approval each time; settlement in two to five business days; weekly dealing; a notice period of up to a week; a daily cap; liquidators limited to an allowlist; no instant path that can take $5M; or no redemption, so the only exit is selling.
Red
Redemptions can be suspended at the issuer's discretion; settlement takes more than five business days; redemptions only run monthly or quarterly; a notice period over a week; a lock-up; liquidators can't legally hold the token; or no redemption and no market.

For a curator, the real liquidation window is the exit time, not a few minutes. A rough rule: max LTV ≈ 1 − (worst price drop during the exit time + slippage + liquidation bonus).

Stages

Stages answer one question: how much do you have to trust the issuer? Each stage needs everything in the one before it.

0

Stage 0 · Disclosed

  • What the token holds is described publicly.
  • The issuer, custodian, administrator and auditor are named.
  • The legal structure (vehicle and jurisdiction) is documented.
  • Smart contracts are verified onchain.
1

Stage 1 · Protected

  • A bankruptcy-remote structure.
  • An independent custodian and administrator.
  • Holdings independently verified, at a pace that fits the asset.
  • Admin powers held by multisigs, never a single key, and documented by the issuer.
  • A documented redemption path with a stated settlement time.
2

Stage 2 · Trust-minimized

  • An independent trustee who can act for holders.
  • Holdings data that anyone can verify cryptographically.
  • Every critical power behind a timelock, so holders can exit first.
  • An exit that works without the issuer: a $5M sell within 1% onchain, and any liquidator can hold the token.
  • Pricing from an independent source: not the issuer's own data, and not written by a single key or the issuer's multisig.

Bankruptcy remoteness and the trustee are legal checks. Until they're assessed, no asset can reach Stage 1, and the asset page says when an asset is held back only by them.

Alerts

We watch every contract in the discovery snapshots, plus SEC filings, LEI records, oracle age and NAV per share.

SeverityWhat triggers it
criticalUpgrades and code changes, upgrade-admin and owner changes, new privileged roles, and core service changes
highRemoved roles; new code versions registered on a factory; Safe signer, threshold, module or guard changes; scheduled timelock or governance calls; pauses; stale oracles; a fall in NAV per share; a lapsed LEI
mediumFreezes and seizures; mints or burns above 1% of supply or $5M; new SEC filings; LEI changes; borrower and lender allowlist changes; oracle delay settings
lowInvestor allowlist entries added or removed

Each new alert also re-runs discovery for that asset, so its Control tab catches up with the change.

DeFi exposure

For each asset we list the Morpho and Aave (including Horizon) lending markets that accept it as collateral, with the loan asset, LLTV, the oracle used, collateral supplied and amount borrowed. Markets whose oracle hardcodes the price are flagged.

Data

KindExamplesWhere it lives
CuratedHoldings, providers, audits, redemption termsdata/assets and data/entities in Git, every fact with a source
DiscoveredProxy admins, roles, Safe signers, timelock delaysJSON snapshots in Git, written by the discovery tool
LiveSupply, prices, oracle age, DeFi exposure, alertsDatabase, refreshed by scheduled jobs
MeasuredTop holders, DEX sell quotesdata/measured, refreshed by bun framework:measure

Asset value (AUM) comes from DefiLlama when it tracks the token accurately. Otherwise it is onchain supply × price, summed over the chains we track. Each asset page names the source it used.

Scope

v1 covers tokenized Treasury and private-credit assets on EVM chains. Stablecoins, tokenized stocks and non-EVM chains come later.

Principles

  1. 1No single score.You see all seven ratings and the stage, so you can tell exactly which part is weak.
  2. 2Asset-class aware.A T-bill fund and a private credit pool aren't held to the same freshness bar.
  3. 3Unknown is its own state.If a fact isn't public, the rating says "unverified" rather than guessing.
  4. 4Every rating has evidence.Each check links to the document, contract or onchain data behind it.
  5. 5Disclosure constraints are context, not a penalty.Data kept private for a real legal reason is shown as such.
  6. 6Green doesn't mean safe.Each category describes one kind of risk, not overall quality.
  7. 7Independence.No issuer pays for a listing or a rating. Issuers may challenge a rating with evidence.

Disclaimer

RWABeat ratings are for information only. They are not investment, legal, tax, or financial advice, not a credit rating, and not an endorsement of any asset. Data may be incomplete or wrong; always verify with the linked primary sources. Each category describes one kind of risk and does not indicate overall safety.