How we source, normalise, and rate

We collect data from three classes of source, in this order of preference: (1) on-chain queries via verified node providers; (2) public issuer disclosures including attestation reports and fact sheets; (3) direct data partnerships with issuers, where available.

Every cell on the platform records the source class, the source identifier, and the retrieval timestamp.

Issuer disclosures use heterogeneous definitions. We normalise to a single schema:

• AUM is reported in USD at the most recent attestation or NAV close.
• Yield is reported as both 7-day and 30-day SEC-style annualised where applicable.
• Holders count is on-chain wallet count, deduplicated by canonical address.
• Top-N concentration is share of supply held by N largest addresses.

Plinch publishes four component scores — custody, attestation, liquidity, and smart-contract — plus a composite, all on the A+ to C- scale. Every score is computed deterministically from observable inputs. Each rated product page lists exactly which inputs contributed to each component.

Custody score

• Auditor (Big 4 / Tier 2 / Tier 3 / unranked) — 30%
• Regulatory wrapper ('40 Act / Reg D / Bermuda DABA / Cayman EF / etc.) — 35%
• Jurisdiction (US / Switzerland / Liechtenstein / offshore) — 25%
• Investor identification (KYC vs no-KYC) — 10%

Attestation score

• Cadence (daily / weekly / monthly / quarterly) — 70%
• Recency (within configured cadence vs late) — 20%
• Attesting auditor tier — 10%

Liquidity score

• Primary redemption mechanics (instant / T+0 / T+1 / T+2 / notice period) — 40%
• On-chain secondary market depth (active / limited / none) — 30%
• Holder distribution (top-10 concentration percentage) — 30%

Smart-contract score

• Smart-contract auditor tier (Trail of Bits / OpenZeppelin / Quantstamp / Hacken / internal-only) — 70%
• Contract age, no incident (>2y / 1–2y / 6–12mo / <6mo) — 30%

Composite

Weighted average: custody 30% + attestation 30% + liquidity 25% + smart-contract 15%.

Tier classifications (auditors, wrappers, jurisdictions) are editorial judgements, documented in the source. We accept they will not be universally agreed with. Every methodology version is tagged; historical ratings can be re-explained under the version they were computed under.

Ratings are opinions formed from public information, not investment advice. We invite right of reply from issuers and publish corrections promptly.

On-chain metrics update every 60 seconds. Issuer-disclosure metrics update on each new publication. Ratings recompute daily. The methodology document is versioned; this is v1.0 and changes will be announced in advance.

We post corrections publicly on the affected product page with a changelog entry on this methodology page. We are also happy to receive corrections via hello@plinch.io.