The four issuers in scope
We picked four issuers covering most of the supply that an ordinary user might hold:
- Tether Limited — issuer of USDT. Approximately 150 billion in circulation as of mid-2026. The dominant stablecoin globally and the deepest cross-chain footprint.
- Circle Internet Group — issuer of USDC and EURC. Approximately 60 billion combined circulation. US-listed parent (June 2024 IPO) with parallel EU subsidiary for MiCA-aligned issuance.
- Paxos Trust Company — issuer of USDP (own brand) and PYUSD (PayPal partnership). Combined circulation around 1.5 billion. NYDFS-supervised trust company. Also issued the now-wound-down BUSD.
- First Digital Labs — issuer of FDUSD. Approximately 3 to 4 billion in circulation. Hong Kong trust company subsidiary. Primary use as Binance settlement asset.
The four are not the only stablecoin issuers worth attention — DAI / MakerDAO, EURI / Banking Circle, USDP via Gemini, and several smaller issuers each have their own profiles — but the four cover roughly 95% of the dollar-equivalent supply that an ordinary user might hold in a working balance.
The scoring methodology
Each dimension is scored A through D. A means "best-in-class, no meaningful caveats". B means "solid with minor gaps". C means "acceptable but with material limitations". D means "below market standards for a major issuer". The grades are our judgement based on publicly available disclosures as of May 2026. A different reasonable reader could grade differently; we publish the reasoning underneath each dimension so the grades can be argued with.
We do not weight the dimensions. A composite score would imply a precision the underlying data does not support. We give a single letter grade at the end of each dimension as a comparison anchor, not as a portfolio recommendation.
Summary scorecard
| Dimension | Tether (USDT) | Circle (USDC) | Paxos (USDP/PYUSD) | First Digital (FDUSD) |
|---|---|---|---|---|
| 1 · Attestation cadence | B | A | A | B |
| 2 · Auditor identity / independence | C | A | B | C |
| 3 · Asset-class breakdown | B | A | A | B |
| 4 · Custodian / bank disclosure | C | A | B | C |
| 5 · Reserve segregation | B | A | A | B |
| 6 · Redemption history at par | A | A | A | B |
| 7 · Supervisory framework | C | A | A | B |
| 8 · Incident handling track record | B | B | B | B |
Dimension 1 · Attestation cadence
Monthly attestations are now the market standard. Quarterly is acceptable for a smaller issuer; less than quarterly is below standard. Real-time or daily attestations are theoretical at present — no major issuer publishes daily — and the operational cost makes them unlikely.
Tether (B): Monthly attestations by BDO Italia since 2022. Some criticism that BDO is mid-tier rather than Big Four, but the cadence is standard. Quarterly assurance reports are also published.
Circle (A): Monthly attestations by Deloitte for USDC and EURC. Weekly transparency dashboard with daily reserve summary. Post-IPO, also subject to SEC quarterly 10-Q filings that include reserve disclosures.
Paxos (A): Monthly attestations by WithumSmith+Brown for both USDP and PYUSD. NYDFS supervision adds an additional reporting layer not in the public attestation but reviewed by the regulator.
First Digital (B): Monthly attestations by Prescient Assurance. The auditor is smaller than the others (specialised attestation firm); cadence is standard.
Dimension 2 · Auditor identity and independence
An attestation is only as strong as the firm doing it. We weight Big Four firms (Deloitte, EY, PwC, KPMG) at the top, large-tier global firms next, then specialised attestation firms. We also flag whether the firm has SEC PCAOB registration (relevant for US-listed issuers).
Tether (C): BDO Italia. Member of the BDO global network. Mid-tier in global accounting; Tether has stated repeatedly that no Big Four firm will accept the engagement, which is plausible given the audit scope and legal exposure. The cadence is regular and the format has improved over time, but the auditor identity is the weakest among the four.
Circle (A): Deloitte. Big Four. Engagement structure includes both monthly attestations and the SEC-required annual audit for the parent Circle Internet Group. The post-IPO disclosure regime is stronger than any other major stablecoin issuer.
Paxos (B): WithumSmith+Brown. Top-25 US accounting firm with PCAOB registration. Solid but not Big Four. The NYDFS supervisory framework adds another layer of independent review.
First Digital (C): Prescient Assurance. Specialised attestation firm based in Canada. The firm has SOC 2 and other compliance specialisation. Independence is not in doubt; scale is meaningfully smaller than the other auditors in this comparison.
Dimension 3 · Asset-class breakdown
The minimum standard is category-level breakdown: cash, T-bills, money-market funds, repos, other. Better is CUSIP-level disclosure (which specific securities). Best is real-time on-chain auditability via Proof-of-Reserves (cryptographic proof for the on-chain portion, with auditor attestation for the fiat portion).
Tether (B): Category-level breakdown with percentages for each category. Some sub-categorisation (secured loans, gold, Bitcoin reported separately). Not CUSIP level for T-bills.
Circle (A): Most of USDC's reserves sit in the BlackRock USD Circle Reserve Fund, which publishes CUSIP-level holdings on a regular basis. Effectively the reserve composition is auditable down to the security level. The cash residue is reported by counterparty bank in aggregate.
Paxos (A): NYDFS-supervised structure requires detailed reserve composition reporting. Published attestations are category-level; supervisory submissions are more granular. The framework constrains opacity even where the public-facing disclosure is summarised.
First Digital (B): Category-level breakdown with percentages. Not CUSIP level. Disclosure quality has improved through 2024-2025 but does not match Circle's BlackRock-fund-driven transparency.
Dimension 4 · Custodian and bank disclosure
The 2023 SVB episode demonstrated that "cash at a bank" is not equivalent to "cash". Knowing which specific institutions hold reserves matters because it lets holders price counterparty risk. Disclosure quality on this dimension varies sharply.
Tether (C): Banking partners are not named publicly. Tether has stated that the cash residue is held across "a number of regulated banks" but does not list them. The opacity is partly defensive (named banks would themselves face regulatory pressure) and partly historical.
Circle (A): Cash custodian (BNY Mellon) is named. The BlackRock fund's custodian is named. Operational banking partners are named in SEC filings post-IPO. This is the strongest disclosure in the comparison.
Paxos (B): The trust company structure means reserves are held by Paxos itself in segregated accounts at named banks (typically named in NYDFS filings, though not always in public-facing attestations). The framework constrains opacity even where the named-bank disclosure is not always public.
First Digital (C): Custodian banks not named at line-item level. The attestation references "regulated banks" without enumeration. Disclosure quality is similar to Tether's on this dimension.
Dimension 5 · Reserve segregation
Segregation is the legal mechanism that protects holders in an issuer-insolvency scenario. Trust company structures (USDC, USDP, PYUSD) provide the strongest segregation under common law. Tether's structure is contractual; segregation is asserted but the legal mechanism is less clean than a trust company.
Tether (B): Reserves held in accounts separate from operating capital but not in a trust structure. Solvency is supported by the structure but the bankruptcy-remoteness in a hypothetical Tether insolvency is less tested than for trust-company-issued stablecoins.
Circle (A): USDC reserves held in the BlackRock fund and in segregated bank accounts. The fund itself is a separate legal entity from Circle. EURC reserves segregated at French custodian banks under MiCA. Strongest segregation among the four.
Paxos (A): Trust company structure under NY law. Reserves are held in trust for the benefit of token holders; in a Paxos insolvency, holders have a priority claim against trust assets. Same applies to PYUSD.
First Digital (B): Hong Kong trust company structure. Reserves held in segregated client-asset accounts. The HK trust framework is well-established but less tested than NY trust law in a stablecoin insolvency context.
Dimension 6 · Redemption history at par
This is the most important dimension. An issuer can have any number of disclosure issues but if it has processed every redemption at par across every stress event, the system has worked. We weight actual track record heavily.
Tether (A): Has processed every redemption at par across every stress event since 2018. The May 2022 stress (Luna contagion) saw 7 billion of redemptions processed at par within a week. The 2018 banking event saw delays but no loss. Track record is best-in-class.
Circle (A): Processed every redemption at par. The March 2023 SVB weekend saw redemptions paused operationally for two days but resumed at par on Monday. No holder lost capital.
Paxos (A): Processed redemptions at par across both USDP and PYUSD. The BUSD wind-down was orderly with redemption support through February 2024 at par.
First Digital (B): No stress event yet. Routine redemptions process at par. The grade is B rather than A only because the track record is shorter and untested under sustained adversarial conditions.
Dimension 7 · Supervisory framework
The supervisory framework determines what happens when something goes wrong. A strong framework (NYDFS trust charter, MiCA EMI authorisation, HKMA stablecoin licence) means a regulator can intervene before the situation deteriorates. A weaker framework means holders have less procedural protection.
Tether (C): Domiciled in El Salvador since 2024. The El Salvador framework is bitcoin-friendly but the supervisory infrastructure is thinner than US, EU or Hong Kong. Multi-jurisdiction inquiries continue (DOJ, Italian authorities) without prudential consolidation. This is the weakest of the four on framework.
Circle (A): US Circle Mint as a state-money-transmitter-licensed issuer plus parent listed on NYSE. EU Circle Mint France as ACPR-authorised EMI under MiCA. Possibly the strongest dual-jurisdiction structure of any stablecoin issuer.
Paxos (A): NYDFS limited-purpose trust charter. Same framework that produced the BUSD wind-down — which is uncomfortable in some respects but reassuring in others (the framework works). Singapore MAS supervision for the Asia subsidiary.
First Digital (B): Hong Kong trust company licence currently, with HKMA stablecoin licence pending under the new ordinance. The B grade reflects "in transition" — once the formal HKMA licence is issued, the grade would move to A.
Dimension 8 · Incident handling track record
Every issuer of any size has had at least one stress event. The question is how the event was handled — communication speed, transparency under pressure, willingness to disclose adverse information, redemption operations during stress.
Tether (B): The 2018 Crypto Capital event and the 2021 NYAG settlement were not handled cleanly — communication was slow, disclosure was forced rather than offered. The 2022 Luna stress was handled visibly better; redemptions were processed in public view. Mixed record.
Circle (B): The 2023 SVB weekend communication was reasonable but operational decisions (the redemption pause framed as procedural, the Coinbase USDC-USD pause framed similarly) were read as solvency signals. The lesson appears to have been internalised — disclosure cadence has improved.
Paxos (B): The BUSD wind-down was orderly but the framing (Paxos publicly stating it had "no relationship" with Binance during the NYDFS event) was awkward and read as defensive. Substantively the wind-down protected holders.
First Digital (B): No major stress event yet. Some early concerns in 2024 about specific transfer patterns were addressed through clarifications. Grade B because the track record is shorter; could be A or C with a real stress event.
What the scorecard does and does not say
The scorecard reads, at first glance, like Circle wins. That is a fair reading. The structural advantages — US-listed parent, MiCA-aligned EU subsidiary, Deloitte audit, BlackRock fund reserve structure — make Circle the cleanest issuer on disclosure metrics. That does not translate directly into "USDC is the safest stablecoin to hold". A few important caveats:
Tether's redemption history is unparalleled. The system has worked through every stress event, at par, since 2018. Disclosure improvements over the same period have been substantial. The framework gap (El Salvador as primary jurisdiction) is real but it has not produced a redemption failure.
Paxos's BUSD wind-down is recent precedent. A trust company issuer can lose its primary product on regulator-driven oversight grounds, not on reserve-quality grounds. The same framework that protects PYUSD holders against insolvency does not protect them against a NYDFS oversight action. This is a real risk for any Paxos-issued token.
FDUSD's Binance-concentration is a different kind of risk. Reserves are clean, framework is solid (and improving), but 80% of supply lives on a single exchange. A Binance-specific event would propagate immediately to FDUSD demand. Diversification of issuer does not equal diversification of venue.
Diversification across issuers is more important than any single grade. Holding USDC and USDT and a small FDUSD position is structurally safer than holding the same total value in any single token, even the highest-graded one. Issuer-level idiosyncratic risk is meaningful enough that the cross-issuer hedge is worth the modest convenience cost.
How the scorecard should be used
For an ordinary holder making a holding decision: use the scorecard as one input alongside use-case fit (which token works best for what you do) and venue concentration (which exchanges or wallets do you already use). A token with a lower transparency grade can still be the right choice for a specific use case; a token with a higher grade can be wrong if it does not fit how you transact.
For a treasury or corporate decision: the scorecard is a starting point for a deeper diligence process that should include direct counterparty discussions, custody contract review, and stress-scenario modelling. The grades are not a substitute for institutional-level analysis.
For policy and journalistic use: the scorecard makes the differences between issuers legible without collapsing them into a single number. We strongly oppose composite scores that hide the underlying dimensions; the dimensions are where the actual risk lives.
What we want to change about this scorecard over time
Three improvements we hope to make in subsequent versions:
Real-time data. The current scorecard is mid-2026 point-in-time. A version that tracked rolling grades quarter by quarter would be more useful and would catch grade changes (Tether's framework grade moving up if a substantive prudential regime came online, Circle's grade moving on the next major regulatory development, First Digital's grade moving when the HKMA licence is issued).
Issuer interviews. The current grades are based on public disclosures. Direct interviews with issuer compliance teams would surface practices that are not public but are reflective of underlying culture. We have not pursued this for the May 2026 edition because the conversation would invite a different kind of bias; future versions might.
Reader pushback. Several of the grades — particularly Tether on supervisory framework, First Digital on custodian disclosure — are arguable. We publish the reasoning so readers can dispute the grade. The corrections log on the corrections page is the place to do it.
If you want to act on this
The narrow action is to look at your own stablecoin holdings and ask which issuer you are exposed to and on which dimension that exposure is weakest. For most holders, the recommendation is to split between at least two issuers (USDT and USDC) and to avoid concentrating more than 60% of stablecoin holdings in any single token. The desk uses Binance for spot trading; the Binance referral link uses code BN16188 and does not change your fees. For more on the comparison between USDT and USDC specifically, see our USDC vs USDT piece.
Methodology and source links
- Tether quarterly attestations and assurance reports, BDO Italia, 2022 through May 2026 (tether.to/transparency).
- Circle monthly transparency reports, Deloitte, USDC and EURC, 2024 through May 2026; quarterly Circle Internet Group SEC filings post-IPO June 2024.
- Paxos monthly attestations, WithumSmith+Brown, USDP and PYUSD, August 2023 through May 2026; NYDFS supervisory reporting (not public).
- First Digital Labs monthly attestations, Prescient Assurance, FDUSD, June 2023 through May 2026.
- NYDFS Consumer Alert, "Paxos Trust Company / BUSD", February 13, 2023 — context for the framework-grade discussion.
- OAG settlement order, "In the Matter of iFinex Inc., Tether Holdings Ltd, et al.", February 17, 2021; CFTC settlement October 15, 2021.
- Regulation (EU) 2023/1114 (MiCA) and HKMA Stablecoin Issuer Licensing Ordinance materials, 2023-2025.
- DeFiLlama, CoinGecko, Token Terminal — historical circulating supply and exchange concentration data.
If you disagree with any of the grades or have a public disclosure we should incorporate, write to privacy@vsccex.com; the corrections log is on the corrections page.