Crypto & Digital Assets
PrivateThe only federally-chartered crypto bank — a genuine regulatory moat priced as a premium IPO candidate, but custody is a fee-compressing commodity where it is the AUC laggard (sub-$50B vs Coinbase's 80%+ ETF share), and a live Homeland Security AML probe is the single fact that can break the whole "trust" thesis. WATCHING into the 2027 IPO; the charter is real, the economics are not yet proven.
Research
The verdict
The only federally-chartered crypto bank — a genuine regulatory moat priced as a premium IPO candidate, but custody is a fee-compressing commodity where it is the AUC laggard (sub-$50B vs Coinbase's 80%+ ETF share), and a live Homeland Security AML probe is the single fact that can break the whole "trust" thesis. WATCHING into the 2027 IPO; the charter is real, the economics are not yet proven.
Anchorage Digital is the operating brand for Anchorage Digital Bank, N.A. — the first and (as of early-mid 2026) still the most distinctive federally-chartered crypto bank in the United States, supervised directly by the Office of the Comptroller of the Currency (OCC) as a national trust bank. It is not a deposit-taking commercial bank: a national trust bank has limited-purpose fiduciary/custody powers, takes no FDIC-insured deposits and makes no traditional loans. That charter is the entire identity of the company — it is what lets Anchorage call itself "an unequivocal qualified custodian" without the asterisks state-chartered rivals carry.
What it actually does — five revenue surfaces, all institutional-only:
Customers — blue-chip institutional, and the marquee logos are the moat-in-practice: primary custodian for BlackRock's spot Bitcoin (IBIT) and Ether (ETHA) ETFs, which together held ~$50B as of Q1 2025; custodian for the BlackRock BUIDL fund and one of the largest BUIDL holders; Cantor Fitzgerald selected it as custodian for its Bitcoin financing business. Suppliers are effectively cloud/HSM infrastructure and the underlying L1/L2 chains. Competitors — Coinbase Custody/Prime, BitGo, Fireblocks, Fidelity Digital Assets (Lens 3/7).
Contract structure — custody is a recurring AUC-based fee business (basis points on assets held) plus per-transaction and staking-spread revenue. This is not take-or-pay; it is asset-correlated and therefore highly levered to crypto price levels and ETF flows — a structural fragility flagged repeatedly below.
Founded 2017 by Nathan McCauley and Diogo Mónica. Offices: San Francisco, New York, Singapore, Sioux Falls (SD — the bank's chartered home), and Porto, Portugal.
Custody has no physical supply chain; the analogue is the chain of trust and the asset-flow rails. Mapping it end-to-end with named stakeholders:
Upstream (infrastructure & key-security inputs):
The company (the trust layer): Anchorage Digital Bank, N.A. — the OCC-regulated entity that holds keys, segregates assets, and provides the qualified-custodian wrapper. Sister regulated entities: Anchorage Digital Singapore (MAS-licensed) and Anchorage Digital New York (NYDFS BitLicense). The multi-jurisdiction license stack is part of the supply chain — it's what lets a global asset manager route assets through one provider.
Downstream (end customers / distribution):
Chokepoints / single-source dependencies:
Names present, chokepoints marked — this lens clears the "names or it didn't happen" bar.
The moat is regulatory, and it is real but narrow. Anchorage is the only federally (OCC) chartered crypto bank. Every competitor custodies under a state trust charter (BitGo Trust — South Dakota; Coinbase Custody Trust — NYDFS) or as a broker-dealer/parent umbrella. Federal charter delivers three durable advantages:
But the moat is being eroded from two sides:
Bargaining power — weak over customers, moderate over suppliers. Against BlackRock, Anchorage is the price-taker (BlackRock can and did add multiple custodians). Against token foundations and smaller institutions, the qualified-custodian status gives Anchorage pricing leverage. Insurance: Anchorage offers $350M+ coverage vs Coinbase Prime $320M, BitGo $250M, Fireblocks $30M+ — competitive but not a differentiator. Minimum client size $10M+ (vs Fireblocks/BitGo $1M, Coinbase $500K) — Anchorage deliberately plays only the top of the market, which is both a moat (high-touch, sticky) and a ceiling (small TAM of mega-institutions).
Verdict on moat: a real but shrinking regulatory edge layered on a commodity service where Anchorage is not the scale leader. The durable asset is the charter-plus-clean-track-record if the El Dorado probe resolves benignly; the durable risk is that the charter becomes table stakes.
n/a — private, not disclosed. Anchorage publishes no segment revenue, EBITDA, or geographic breakdown; there is no segments.csv data (file is header-only). Qualitatively, the revenue surfaces rank (by visible strategic emphasis, not disclosed dollars):
No segment numbers are sourceable; any split would be fabrication. Flagged as a genuine information gap an IPO S-1 would finally fill.
+private overlay applied. Lens 5 → Funding & valuation trajectory; Lens 7 → Cap table & secondary marks; plus a Traction & unit-economics note. SEC-grounded earnings analysis is not possible (no filings).
Round history, seed → latest disclosed:
| Round | Date | Amount | Post-money valuation | Lead | Source |
|---|---|---|---|---|---|
| Series C | Feb 2021 | $80M | n/a | GIC | |
| Series D | Dec 15, 2021 | $350M | >$3.0B | KKR (first digital-asset deal via NextGen Tech Growth Fund II) | |
| Current implied | 2025-2026 | — | ~$4.2B (company-cited "valuation of $4.2 billion") | — | |
| New round (in progress) | Jan 2026 | $200M–$400M sought | n/a — not yet priced | TBD |
Series D syndicate (Dec 2021) — exceptionally deep and crossover-heavy: KKR, Goldman Sachs, a16z, Apollo credit funds, BlackRock-managed accounts, GIC, Wellington, Thoma Bravo, PayPal Ventures, Kraken, Standard Investments, GoldenTree, Lux Capital, Blockchain Capital, Senator Investment Group, Alameda Research, Elad Gil. (Note: Alameda Research appears in the cap table — an FTX-affiliated holder; its stake is presumably in liquidation/transferred, a housekeeping flag for an S-1 but not a solvency issue for Anchorage.)
Valuation read: flat-to-modestly-up over a brutal cycle. $3B (Dec 2021 peak-of-cycle) → ~$4.2B cited 2025-2026 is a ~40% step-up across ~4 years — i.e. it roughly held through the 2022 crypto winter while many peers were marked down 50-80%. That durability is a quiet positive. Burn signal: the company is raising $200-400M "to scale core services and aggressively expand stablecoin" and to "double the stablecoin team" — consistent with a still-investing, likely not-yet-sustainably-profitable posture (an audited P&L would confirm; none exists).
On the "$14.7M revenue" figure circulating on one aggregator: I am not treating this as reliable. $14.7M annual revenue is implausibly small for a custodian with >$50B AUC — even at a conservative ~5-10bps custody fee, $50B AUC implies ~$25-50M of custody revenue alone before staking spreads or trading. The figure is either stale, mis-scoped, or an error; flagged, not cited as fact. Real revenue is undisclosed.
No earnings calls exist. The proxy is founder/CEO public commentary and strategic-message drift over time:
Recurring phrases: "first," "federally chartered," "qualified custodian," "institutional-grade," "compliance-first." What they stopped saying: the defensive "rebuilding our BSA/AML program" language of 2022-2024 has been replaced by "proven." The narrative arc is coherent and, post-consent-order-lift, credible — if the El Dorado probe doesn't reopen the compliance wound.
Syndicate quality is top-decile and a genuine IPO-proximity tell. The Series D cap table reads like a who's-who of crossover and TradFi capital — the exact investor base that precedes an IPO:
+private playbook flags as IPO-proximity signals.Secondary marks / mutual-fund markups: n/a — not disclosed. No public mutual-fund quarterly markup data was sourced for Anchorage specifically. The ~$4.2B "current" figure is company-cited, not a fund mark.
Peer "comps" — by IPO/valuation, since there are no public multiples for a private:
| Peer | Status | Valuation / mark | Scale signal | Source |
|---|---|---|---|---|
| Anchorage Digital | Private | ~$4.2B (cited) | >$50B AUC; BlackRock ETF custodian | |
| BitGo | Public (IPO'd Jan 2026) | ~$1.96B IPO val (11.8M shares @ $15-17, ~$201M raised, NYSE) | Broadest asset coverage | |
| Coinbase (Custody/Prime) | Public (COIN) | n/a — segment not broken out | 80%+ of US BTC/ETH ETF AUC | |
| Fireblocks | Private | n/a | >$10T cumulative transfers, 100+ chains | |
| Fidelity Digital Assets | Subsidiary (private parent) | n/a (parent $4T+ AUM) | Lowest modeled default risk (0.39%) |
The sharp read: Anchorage's cited ~$4.2B private mark is >2x BitGo's actual public IPO valuation (~$1.96B) despite BitGo carrying broader asset coverage. Either (a) the market is paying a premium for the federal charter + BlackRock relationship, (b) the ~$4.2B private mark is stale/optimistic and will reset at IPO toward BitGo's public comp, or (c) both. This valuation gap vs the one liquid comp is the single most important number to watch into the 2027 IPO — it is where private-mark optimism meets public-market reality. Multiples not fabricated; where unsourceable, marked n/a.
No public stock, so the "what moves the mark" history is funding + regulatory + mandate events:
Pattern: the franchise mark moves on (1) regulatory status events (charter, consent order on/off, El Dorado) and (2) marquee custody mandates (BlackRock). It does not move on token prices directly the way an exchange would — but its revenue absolutely does (AUC-fee model). The market reacts to trust and regulatory legitimacy above all for this name — which is exactly why the El Dorado probe is disproportionately dangerous.
Traction & unit economics (Phase B add): AUC >$50B all-time high (2025); primary custodian for ~$50B of BlackRock spot-ETF assets (Q1 2025); staking across 6+ networks; USDtb reserves "one of the largest BUIDL holders." Gross-margin/ARR figures undisclosed — custody + staking are structurally high-gross-margin (software + spread, minimal COGS), but absolute profitability is unconfirmed and the active capital raise implies ongoing burn. All unaudited.
CEO — Nathan McCauley (co-founder). Security engineer (ex-Docker, ex-Square/Block, ex-military-systems), patent-holder. Track record (quantified): co-built Anchorage from 2017 into the first crypto-native federal bank (2021) and to >$50B AUC with the BlackRock spot-ETF custody mandate — objectively a top-tier institutional-crypto franchise outcome. Skin in the game: co-founder, presumably significant equity (exact % undisclosed — no S-1). Tenure: 8+ years, founder-led throughout.
Co-founder — Diogo Mónica. Now Executive Chairman; became a GP at Haun Ventures in 2024. Partial step-back — drives strategy/recruiting/client relationships but no longer day-to-day. Both founders are the security-IP core of the company.
Capital-allocation history: mixed-but-rational. The Mountain Protocol acquisition (2025) was a buy-vs-build call to acquire stablecoin capability/team/licensing fast post-GENIUS — defensible and on-strategy, modest size, then sunset the acquired USDM token in favor of in-house USDtb (clean integration, no dual-token confusion). No buybacks (private). The repeated equity raises ($350M Series D; $200-400M new round) fund growth, not returns — appropriate for the stage. ROE/ROIC not computable (no audited financials).
Red flags (management-specific):
Archetype: founder-led, security-engineer-CEO — strong on product/security/regulatory positioning, historically weaker on the unglamorous compliance-operations execution that is, ironically, the core promise of a bank. For a regulated-custody business at IPO scale, that is the exact tension to underwrite.
Accounting forensics: n/a — no audited financials exist. As a private company with no SEC filings, the standard income-statement/balance-sheet/cash-flow forensic checks (revenue recognition, receivables vs revenue, SBC flattering non-GAAP, goodwill) cannot be performed. This opacity is itself the flag: a $4.2B-valued custodian asking the public to trust it with assets has published no audited financials — the IPO S-1 will be the first real look, and the absence of a number is a number. The Mountain Protocol acquisition will create the first goodwill/intangibles line to scrutinize when an S-1 lands.
Regulatory findings (required sub-section):
n/a — no 10-K exists (private).research/private-watch.json is absent, so there is no stored stage/ipo_readiness/catalyst seed and no write-back is possible — this lens is grounded entirely in web reporting.
Readiness assessment — LATE-STAGE / IPO-PROXIMATE:
Estimated window: 2026 (early/private round) → 2027 IPO, El-Dorado-resolution-permitting. Because no private-watch.json entry exists, the recommendation is to create one seeded: stage: late-private / pre-IPO, ipo_readiness: high (gated on El Dorado resolution + audited financials), catalyst: $200-400M round pricing (2026) → S-1 (2026-2027) → IPO (2027); BitGo public comp ~$1.96B. (Per the wave boundary, this dossier does not edit watchlist/private-watch files — flagging the seed for the human step.)
No EPS/Brier forecast logged — this is a pre-revenue-disclosure private; the --watchlist loop skips forecast.ts create, and the only scoreable binary here would be "Anchorage Digital files an S-1 / goes public by YE2027" — a candidate forecast for the human to log if conviction warrants, not logged here.
Bull case. Anchorage is the regulated, institutional, federally-chartered front-door to crypto custody at exactly the moment TradFi is forced through it. The spot-ETF mandate created a durable, non-speculative custody-demand pool; the GENIUS Act created a federally-regulated stablecoin lane Anchorage was first through; and the OCC consent-order lift converted "first" into "proven." Its anchor customer is BlackRock — the strongest possible reference. The cap table (Goldman, KKR, Wellington, GIC, Visa, a16z) is IPO-ready, and the valuation held flat through a brutal crypto winter ($3B 2021 → ~$4.2B 2025) when peers were halved. With AUM/AUC compounding as institutions allocate, custody + staking + stablecoin is a three-engine, high-gross-margin, recurring-revenue machine. Earnings surprise to the upside if the new round prices well above $4.2B and the IPO catches a crypto bull tape.
Bear case (permanent-impairment risks).
Pre-mortem (it's late 2027, the thesis broke — what happened?): The El Dorado probe escalated to a settlement or charge that re-implicated Anchorage's AML controls just as the S-1 was being drafted; the IPO was pulled or re-priced sharply below the $4.2B private mark (toward BitGo's ~$2B public comp); meanwhile Circle/Ripple/Coinbase competed custody fees toward zero and BlackRock multi-sourced more aggressively, so the revenue ramp that justified $4.2B never materialized.
Are multiples too high? Against the only liquid comp (BitGo ~$1.96B public), Anchorage's ~$4.2B private mark looks rich unless the charter + BlackRock + stablecoin premium is worth a 2x. That premium is plausible but unproven, and private marks are sticky-high.
Contrarian view (what the market refuses to see): The bull consensus treats the federal charter as a durable monopoly moat. The contrarian read is that the charter is becoming a commodity license the OCC is now issuing freely, and Anchorage's real, harder-to-replicate asset is the BlackRock relationship + the operational track record — both of which the El Dorado probe directly threatens. The market is pricing the charter; it should be pricing the compliance record.
Dismantling the bull case as a skeptical short:
A coal-plant bitcoin miner that re-priced itself into a $14B "AI landlord" by renting power to a venture-stage neocloud — the equity is a levered call on Fluidstack actually paying, with Google's $3.2B backstop the only thing standing between the multiple and zero.
A levered, perpetual-dividend-funded bitcoin holding company whose entire accretion engine — issue stock above NAV, buy BTC — has inverted to a ~0.85x discount, so it is now SELLING bitcoin to pay an ~$0.9B/yr preferred coupon; bullish only as a bitcoin call, bearish as a structure, and the discount is the tell.
The regulated front-door to crypto with a real moat and a strengthening regulatory tailwind, wrapped around an earnings stream so volatile it just printed back-to-back GAAP losses — a ~70x-NTM-P/E call on crypto cycle timing dressed up as a fintech compounder; own the franchise, not this price.