Crypto & Digital Assets
PrivateThe best-regulated stablecoin issuer with the worst commercial control of its own product — Paxos owns the charter and the rails but PayPal, the GDN consortium, and Hyperliquid own the demand; it is a high-quality utility being commoditised from below (Bridge/M0) and disintermediated from above (partner-issued float), worth watching for an IPO the OCC charter now makes plausible, not chasing on a secondary mark.
Research
The verdict
The best-regulated stablecoin issuer with the worst commercial control of its own product — Paxos owns the charter and the rails but PayPal, the GDN consortium, and Hyperliquid own the demand; it is a high-quality utility being commoditised from below (Bridge/M0) and disintermediated from above (partner-issued float), worth watching for an IPO the OCC charter now makes plausible, not chasing on a secondary mark.
Paxos Trust Company is a regulated blockchain-infrastructure company whose core product is being the licensed issuer-of-record for other people's stablecoins. Founded 2012 as itBit (a bitcoin exchange) by Charles Cascarilla and Rich Teo; received the first NYDFS limited-purpose trust charter for digital assets in 2015. The business is best understood as a regulated utility layer sitting between traditional finance and public blockchains — it holds the licenses, the segregated reserves, and the mint/redeem infrastructure, and rents that regulatory perimeter to consumer brands who want a dollar token without becoming a trust company themselves.
Product portfolio (as of Q2 2026):
How it makes money: Two overlapping engines — (1) reserve interest on the T-bills/cash backing the float it controls (its share of PYUSD/USDG/USDP reserves, net of what it rebates to PayPal and GDN partners), and (2) infrastructure/service fees for issuance, custody, settlement, and tokenization delivered to institutions (PayPal, Interactive Brokers, Mercado Libre, Nubank, Mastercard, Bank of America, Societe Generale historically). Reserve income scales with (float × short-rate); fee income scales with enterprise integrations.
Contract structure — the central fact of the business: For PYUSD, Paxos is a contracted issuer, not the principal. The economics of the largest single token are governed by a commercial agreement with PayPal, and the brand/distribution/consumer relationship belong to PayPal. This is simultaneously Paxos's greatest asset (it captured the single most valuable consumer-payments stablecoin partnership) and its greatest structural vulnerability (its flagship revenue line is a contract that can be renegotiated or in-sourced).
The "supply chain" for a stablecoin issuer is the flow of fiat into reserves → tokens onto chains → into distribution channels → to end users, plus the custody/attestation stack. Named stakeholders along Paxos's chain:
Upstream (reserve + custody):
Midstream (Paxos itself — the regulated core):
Downstream (distribution — where the leverage leaks out):
Chokepoints / single-source dependencies:
The moat is regulatory, not technological — and it is real but narrowing.
Bargaining power — this is where the thesis gets uncomfortable. Paxos has strong bargaining power over its suppliers (reserve banks, chains are commoditised inputs) but weak bargaining power over its most important customer (PayPal). In the PYUSD relationship, PayPal needs Paxos less than Paxos needs PayPal — PayPal could re-tender the issuance mandate, in-source it under its own charter application, or split it. The USDG consortium was arguably built precisely to reduce this dependency by giving Paxos an own-brand token — but USDG's economics (reserve income shared away to distributors) mean Paxos deliberately gave up float margin to buy distribution, which is a moat-weakness dressed as innovation.
Net moat verdict: A genuine, hard-to-replicate regulatory moat wrapped around a commercially disadvantaged position. The license is the crown jewel; the revenue is a rented crowd.
No segments.csv data exists (private, empty seed) and Paxos does not publish audited segment revenue. Best reconstruction of the revenue architecture, all /, labelled as such:
| Revenue engine | Driver | Directional trend | Provenance |
|---|---|---|---|
| PYUSD reserve share | ~$3.5B float × short-rate × Paxos's contractual share | Growing fast (float +5× YoY) but margin diluted by PayPal rebate | supply; share n/a — not disclosed |
| USDG reserve share | ~$2.5–2.75B float × short-rate × Paxos's residual after GDN rebates | Growing, but structurally low-margin by design (income shared to 130+ partners) | supply; economics |
| USDP reserve income | ~$32M float — collapsing | Declining deliberately | |
| PAXG | AUM fees on tokenized gold | Stable niche | |
| Enterprise infra / Paxos Labs fees | Per-client issuance/custody/settlement fees | Growing; Paxos Labs now spun out (April 2026) |
Geography: US (PYUSD/PayPal core, USDP), Singapore + EU (USDG issuance entities), Abu Dhabi (FSRA). The deliberate offshore issuance of USDG (Singapore/EU rather than US) is itself a segment signal — it lets USDG operate a yield-sharing model that the GENIUS Act restricts for US issuers (see Lens 12).
The key segment story: Paxos is migrating its revenue base from high-margin own-brand float (USDP, dying) to low-margin partner/consortium float (PYUSD diluted by PayPal, USDG diluted by GDN). Top-line float is up massively; revenue quality per dollar of float is deteriorating. This is the single most important — and least appreciated — dynamic in the business.
Paxos has no earnings release (private, unaudited). The +private substitute is the round history and valuation path:
| Round | Date | Amount | Valuation | Lead / notable investors | Provenance |
|---|---|---|---|---|---|
| Series C | Dec 2020 | (undisclosed) | — | PayPal Ventures | |
| Series D | Apr 2021 | $300M | $2.4B | Oak HC/FT lead; Declaration Partners, PayPal Ventures, Mithril, Senator, WestCap, Liberty City Ventures | |
| Series D extension | Jul–Aug 2021 | (part of $300M+) | ~$2.4B | added Bank of America, Coinbase Ventures, Founders Fund, FTX | |
| Total raised | — | >$540M | — | — | |
| Paxos Labs (spin-off) | Apr 2026 | $12M | (separate entity) | Blockchain Capital lead; Robot Ventures, Maelstrom, Uniswap Labs |
Secondary-market marks (2025–26): implied ~$3.09B valuation per UpMarket's estimate (~$12.34/share on an assumed 250M shares) — explicitly caveated by the source as possibly stale. Also traded on EquityZen, Forge, Hiive. Interpretation: a $2.4B (2021) → ~$3.1B (2025–26) mark is a modest ~29% step-up over ~4–5 years — striking underperformance versus the stablecoin market's growth and versus Circle's public re-rating to ~$60B+. The secondary mark says the private market does not credit Paxos with Circle-like economics.
Revenue proxy: ~$100M (2024 est.), down from 2022 after BUSD's shutdown cut reserve-interest earnings. For scale, the S&P assessment places Paxos "between Tether and Circle" on transparency, but its revenue is an order of magnitude below Circle's $1.68B (2024). Cascarilla stated the company held >$500M on its balance sheet and was "in a very strong financial position" at the 2024 layoff.
Burn / cost signals: Two rounds of layoffs — ~65 staff / ~20% of workforce in late 2024, framed as "de-prioritizing adjacencies" to focus on stablecoins + tokenization. Combined with a strong balance sheet, this reads as discipline, not distress — but also as an admission that the diversified-crypto-infrastructure strategy (brokerage, securities settlement) did not pay and Paxos is retreating to the stablecoin core.
No earnings calls exist. The +private substitute is the public-communication trend from founders/execs:
Sentiment trajectory: cautious/defensive (2023, post-BUSD) → disciplined/focused (2024, layoffs) → confident/offensive (2026, post-GENIUS-Act + OCC charter). The arc tracks the regulatory environment turning from hostile to favourable.
Syndicate quality — mixed, with one scar:
Peer table — by business model (multiples are the closest listed proxy; Paxos is private so it has none of its own):
| Company | Ticker | Status | Model | ~Circulating float | Rev (latest) | Valuation | Provenance |
|---|---|---|---|---|---|---|---|
| Paxos | private | private | Regulated issuer-of-record + white-label rails | PYUSD ~$3.5B + USDG ~$2.75B (controls issuance, not brand) | ~$100M (2024 est.) | ~$3.1B secondary | |
| Circle | CRCL | public | Own-brand issuer (USDC) | ~$74–76B | $1.68B (2024); $578.6M Q1'25 | ~$60B+ mkt cap (peaked higher) | |
| Tether | private | private | Own-brand issuer (USDT) | ~$183–187B | >$10B net profit (2025) | n/a — private | |
| Bridge (Stripe) | — | subsidiary | Issuance-as-a-service API | (orchestration, not float) | n/a | acquired ~$1.1B by Stripe | |
| M0 | private | private | On-chain issuance protocol | (shared base token $M) | n/a | n/a — private |
No public stock; the +private substitute is the events that moved Paxos's valuation, float, or strategic standing materially:
Pattern: Paxos's value is driven overwhelmingly by regulatory events and partnership wins/losses, not by organic product metrics. The market prices this name on "which side of the regulator is it on" and "whose token did it win/lose" — a fundamentally different reaction function from an operating company.
(1) Track record: Built the first NYDFS-chartered digital-asset trust; won the single most valuable consumer-payments stablecoin partnership (PYUSD/PayPal); >$120B in stablecoins issued cumulatively over 5 years. Genuinely impressive regulatory-first execution. (2) Tenure & skin in the game: Founders since 2012 (14 years) — high tenure. Insider ownership n/a — private, not disclosed, but founder-led with significant equity presumed. (3) Capital allocation: Disciplined and self-correcting — cut the brokerage/securities-settlement adjacencies, executed two rounds of layoffs to refocus on stablecoins, wound down own-brand USDP in favour of partner float, and externalised Paxos Labs rather than funding it on the core balance sheet. Held >$500M cash through the downturn. The red mark: the pre-2023 "crypto infrastructure giant" diversification (much of which was later cut) represents capital and years spent on strategies that didn't pay — the focus is a virtue born of prior over-expansion. (4) Red flags: The BUSD/Binance AML failure is a management-oversight failure (Lens 10) — NYDFS explicitly cited Paxos's inadequate oversight of the Binance relationship. For a company whose entire brand is "the compliant issuer," a $48.5M AML settlement is a governance black eye, even if resolved. Taking FTX money in 2021 shows diligence lapses were not unique to Binance. (5) Founder vs professional: Founder-led, but with an institutional/regulatory temperament rather than a move-fast one. For a licensed-utility stage, this is the correct archetype — the risk is that the same caution that satisfies regulators is losing product/distribution races (Hyperliquid) to nimbler natives.
Acting as a forensic analyst — with the hard caveat that there are no audited financials, no 10-K, no cash-flow statement to interrogate. For a private issuer, the forensic surface is (a) reserve integrity, (b) regulatory/AML history, (c) disclosure quality. All findings ``, unaudited per public sources.
Reserve / accounting risk surface:
Regulatory findings (required sub-section):
regulatory/regulatory-findings.md (generated 2026-07-06): "Paxos has no CIK — it is private and not required to file with the SEC. No EDGAR enforcement search is possible.". Separately, the 2023 SEC Wells Notice alleging BUSD was a security was formally terminated with no enforcement action on Jul 9 2024 — a win, but it confirms Paxos operated a token the SEC scrutinised as a possible unregistered security."Paxos" (FTC OR DOJ OR FDA OR CFPB OR consent decree OR settlement OR fine OR penalty) enforcement): no additional material federal enforcement surfaces beyond the NYDFS/SEC-BUSD matter. Prior settled matters include a 2019 OCC/Fed-adjacent and 2023 SEC scrutiny all tied to the same BUSD/Binance episode.Forensic verdict: The books cannot be forensically stress-tested because they are private and unaudited — that opacity is itself the flag for anyone underwriting this at a secondary mark. The one hard, admitted problem is the BUSD-era AML failure and $48.5M settlement, which dents the central marketing claim. No evidence of reserve-backing fraud or attestation irregularity has surfaced.
IPO-readiness score: 3.5 / 5 (between late-stage and pre-IPO/secondary-active on the private-watch scale).
Milestones that unlock an S-1 — and where Paxos stands:
| Milestone | Status | Provenance |
|---|---|---|
| Federal charter / clean regulatory standing | ✅ OCC conditional approval Dec 2025 (major unlock) | |
| Resolved legacy enforcement | ✅ SEC-BUSD dropped 2024; NYDFS settled Aug 2025 | |
| GENIUS Act regulatory clarity | ✅ signed Jul 2025 | |
| A public comp that opened the window | ✅ Circle IPO'd Jun 2025 (proved appetite) | |
| Revenue scale / margin story | ⚠️ ~$100M rev, low/diluting margin — the weak leg | |
| Crossover-fund pre-positioning | ❌ no Fidelity/T. Rowe/Coatue on cap table |
Estimated window: plausible 2026–2027, gated not by regulation (now favourable) but by whether Paxos can present a growth-and-margin narrative that survives comparison to Circle. The binary that matters: can Paxos convert its regulatory pole-position into own-economics float — i.e. grow USDG (which it owns) fast enough, at acceptable margin, that the equity story isn't "a contractor to PayPal + a low-margin consortium." If USDG scales to Circle-adjacent size while retaining meaningful spread, the IPO is compelling; if it stays a shared-economics consortium token, Paxos IPOs as a utility at a utility multiple, well below the Circle re-rating.
Brier forecast (logged conceptually — not written to forecast.ts per --watchlist rule): "Paxos files an S-1 or announces a direct listing/SPAC by 2027-12-31" — p ≈ 0.45. "USDG circulating supply > $10B by 2027-12-31" — p ≈ 0.35.
Write-back note (respecting wave boundaries — NOT edited here): Paxos is not currently in research/private-watch.json. Recommend adding on the next non-boundary pass:
"paxos": { "beat": "crypto", "stage": "late-stage", "ipo_readiness": 3.5,
"lead_investors": "Oak HC/FT, Founders Fund, PayPal Ventures, Mercado Libre",
"catalyst": "OCC national trust bank charter (Dec 2025); Circle IPO opened window; USDG scale is the swing factor",
"dossier": "companies/paxos/deep-dive-2026-07-06.md" }
Bull case. Paxos is the best-regulated neutral issuer at the exact moment regulation became the moat. Post-GENIUS-Act + OCC charter, being a permitted, federally-supervised issuer is a scarce license — Paxos already holds the broadest stack (OCC, NYDFS, MAS, MiCA, FSRA). It has the most valuable consumer partnership in the category (PYUSD/PayPal, +5× YoY to ~$3.5B) and a fast-growing owned consortium token (USDG, ~$2.75B, 130+ partners incl. Mastercard/DBS). As a neutral white-label utility, it can be the issuer for many brands' tokens (the "AWS of stablecoins") without competing for consumers — a larger TAM than any single-brand issuer. The Circle IPO proved public-market appetite; the OCC charter makes Paxos the most obvious next stablecoin IPO. Secular tailwind: stablecoin market from ~$310B (early 2026) toward the $1–2T that bulls (and Paxos) project, on rails Paxos is licensed to run.
Bear case. Three risks that could permanently cap the value:
Pre-mortem (18 months out, thesis broke): It's early 2028. Paxos never IPO'd because the equity story couldn't escape the "PayPal contractor + shared-economics consortium" framing. USDG plateaued around $3–4B as distributors extracted most of the reserve yield; PYUSD growth slowed and PayPal opened a second issuer to hedge. Bridge/M0-style rails won the long tail of brand tokens on price and developer speed. The ~$3.1B secondary mark proved generous. Nothing broke catastrophically — Paxos simply stayed a healthy, well-regulated, sub-scale utility while the value in stablecoins accrued to whoever owned the float and the distribution.
Are multiples too high? There is no public multiple. The ~$3.1B secondary looks fair-to-full, not cheap, given ~$100M revenue at deteriorating margin — you are paying ~30× sales for a business the private market has refused to re-rate for four years. The contrarian view the market is refusing to see: the stablecoin bull case (huge TAM) and the Paxos bull case are not the same trade — the TAM can 5× while Paxos's share of the economics shrinks, because its entire strategy is trading margin for distribution it doesn't own.
Dismantling the bull case:
A toll-road compounder mispriced as a disruption victim — 60%+ margins and 16% top-line growth are intact while the market discounts a debit antitrust loss and a stablecoin bypass that the numbers (cross-border +17%, $7B stablecoin run-rate is on-network, not against it) do not yet support; structurally BULLISH, but the DOJ debit case and the interchange settlement's final approval are real, dateable downside.
A negative-book-equity Solana levered fund wearing a consumer-products costume; with the stock at ~0.7x NAV, ~$185M of SOL-repayable convert/credit debt against a ~$165M treasury, and converts struck 4–5x above the share price, the equity is a deep-out-of-the-money call on SOL where bondholders own the first ~52% of the coins — BEARISH on the equity, structurally distinct from "owning SOL.