| Jul 30, 2026 | -10.59% | Q2 FY2026 | Read transcript briefingQuarter in one view- This transcript contains no financial results. The Q2 FY2026 call was an X-hosted Q&A only — no prepared remarks, no revenue, EBITDA, EPS, guidance, or balance-sheet figures appear anywhere in it. All P&L conclusions must come from the shareholder letter/filing, which is not supplied here. This is itself a disclosure event: the prior quarter's call at least contained management framing of results; this one contains none.
- The call was short (~10 questions, two rounds of five questioners) and ended early when Alexander Markgraff's (KeyBanc) second question failed on audio — management closed the call rather than substitute another questioner.
- The substantive disclosures: (1) the Circle contract has already met renewal conditions and will renew on the same terms — Haas explicitly said she wanted to "take away any ambiguity"; (2) Coinbase One hit an all-time high in paid subscribers in Q2, growing through a down market; (3) Coinbase joined the Onyx USD consortium as a deliberate multi-stablecoin strategy; (4) Bitcoin-related transactions are now ~12% of revenue (vs. >50% historically, per the questioner, uncorrected by management); (5) CLARITY Act is at the "1-yard line" heading to a Senate floor vote, with Armstrong optimistic but acknowledging ~30% prediction-market odds of passage.
- Management confirmed Q2 was a down market ("we saw crypto trading volumes down") — consistent with, but not quantified against, the prior quarter's -21% revenue decline.
What management is focused on- Regulatory endgame: Armstrong is "pretty optimistic" CLARITY gets a full Senate floor vote, citing last-minute negotiations and the August recess as a forcing function, plus >1M Stand with Crypto emails/calls. Crucially, he built a fallback case: SEC Chair Atkins and CFTC Chair Gensler (as named in the transcript) have "publicly said they are poised to pass clear rules whether CLARITY passes or not" — so failure means "business as usual" for Coinbase, with consumers as the real losers. This is a notable softening versus last quarter's "signed by end of summer" confidence.
- Multi-stablecoin platform strategy: joining Onyx USD (a perceived USDC competitor) is framed as customer-choice plus economics — "we want to strike good economic arrangements with every major stablecoin" (PYUSD, USDT already supported). Armstrong simultaneously reaffirmed USDC primacy: #1 in stablecoin transaction volume, #1 regulated stablecoin, #2 only to Tether on market cap, with explicit intent to keep sharing economics to reach #1 "across all categories."
- Coinbase One as the down-market proof point: Haas volunteered the ATH subscriber figure twice, framing subscription growth during falling volumes as evidence of product value, engagement, and retention.
- Base competitive moat vs. new L2s (Robinhood, Stripe): Armstrong's framework is fragmentation-then-consolidation, analogizing to stablecoins where USDC/Tether share barely moved despite dozens of launches. Claims: Base is the largest Ethereum L2, #1 in stablecoin volume with ~$32T in stablecoin transfer volume over the last 12 months, leader in agentic finance, with a "2-year head start" and a stated path to decentralization. He floated blockchain M&A as a future competency.
- Crypto-native re-engagement: Cobie taking over the Base app and Armstrong's podcast circuit are deliberate outreach to the crypto-native constituency — a tone shift worth noting given the Everything Exchange's mainstream pivot.
- Leadership turnover dismissed as non-strategic: multiple senior departures (people, legal, institutional functions implied) were framed as normal succession with a "deep bench" — Dominique (people), Molly (legal), Liz named as internal successors "groomed by the outgoing leaders."
Key numbers and quarter mechanics- No Q2 financials in the transcript. No revenue, transaction/S&S split, EBITDA, opex, guidance, or restructuring-charge update. The prior quarter's $50–60M Q2 restructuring charge, S&S guide of $565–645M, and tech/dev+G&A guide of $820–870M are unaddressed here — verify against the shareholder letter.
- Bitcoin ~12% of revenue (questioner's figure, affirmed by Armstrong's "thanks for noticing that") — a striking diversification datapoint if accurate; the denominator and period were not specified.
- Coinbase One: all-time-high paid subscribers in Q2, growing despite falling crypto volumes. No absolute number given (prior quarter disclosed >1M); no ARPU, churn, or revenue mix.
- Base: ~$32T stablecoin transfer volume, trailing 12 months (Armstrong, stated imprecisely — "like, 32 bill 32 trillion"); claimed #1 in stablecoin volume and largest Ethereum L2.
- Growth marketing: typical 1-year payback, recently outperforming that benchmark; early signal that prediction-market users drive incremental spot volume — "we are not seeing cannibalization."
- Circle contract: renewal conditions already met; renews on the same terms. Haas was unambiguous — this removes the renewal overhang entirely and answers Owen Lau's leverage question directly.
- Hyperliquid partnership economics: management confirmed sharing USDC economics with Hyperliquid to embed USDC in a major perps ecosystem — a deliberate trade of near-term economics for network effect. Worthington's framing (third parties leveraging USDC holdings to "command the majority of USDC economics") was not rebutted on the economics-protection point; Haas answered with strategy, not safeguards.
- Coinbase One unit economics: Haas gave an admittedly "unsatisfying answer" — on average subscribers trade more and have higher unit economics, revenue is earned across staking/card/subscription, and the relationship is "accretive," but she conceded revenue will shift within the P&L (away from transaction revenue) as adoption grows, and declined to answer whether revenue per dollar traded is higher or lower for members.
Product and launch scorecard- Pre-IPO perpetual futures (SpaceX first): "early traction is definitely encouraging," strong customer demand, more private companies implied in the pipeline, US access "on the road map" with no timeline. No volume or revenue figures.
- Prediction markets: cited as a growth-marketing success and a cross-sell entry point driving incremental spot trading — qualitative only; no update to last quarter's $100M annualized figure.
- Coinbase One: ATH subscribers in a down market — the strongest product datapoint of the call; positioned as accretive across the stack with better retention, but the take-rate-dilution question was explicitly not answered numerically.
- Base: $32T TTM stablecoin transfer volume, #1 stablecoin-volume chain claim, agentic-finance leadership, decentralization "progress." Cobie's appointment to lead the Base app signals a consumer/crypto-native push. Base token not mentioned.
- x402/agentic: still "very early days" — Armstrong explicitly said there are no specific numbers or forecasts to share, a second consecutive call with no agentic monetization data. Majority of agentic transactions on USDC/Base/x402/CDP remains the claim.
- Stock trading / stock options: mentioned only in passing as launched/horizon items — still zero equities metrics, now two-plus quarters of silence on the rollout.
- Onyx USD: joined as consortium member; no launch date, economics, or terms disclosed.
- Not mentioned at all: Deribit integration/US options, tokenized equities, the RIF/AI-native transition, the restructuring charge, buybacks, the June 1 convertible retirement, BTC purchases, and any Q2 financial results.
Sell-side read-through- Format drift continues and worsens: last quarter was curated written questions; this quarter is a live-but-brief X Spaces call with a handpicked mix of institutional analysts (Lau/Clear Street, Worthington/JPMorgan, Markgraff/KeyBanc) and independent/retail-adjacent questioners (Eric Pan, Brian Jung, Austin Hankwitz). No prepared results section at all. The call ended early on a mic failure rather than extending to other analysts — accountability time was minimal.
- Eric Pan (independent) — CLARITY downside scenario: the most strategically important exchange. Armstrong's "business as usual" fallback (agency rulemaking by Atkins/Gensler regardless) is new and meaningfully de-risks the failure case — but also quietly walks back last quarter's signed-by-summer framing. The 30% passage odds from prediction markets went unchallenged.
- Owen Lau (Clear Street) — Onyx USD and Circle leverage: produced the call's hardest disclosure — renewal conditions met, same terms, ambiguity removed. Lau's negotiating-leverage angle was deflected into the multi-stablecoin strategy answer.
- Brian Jung (independent) — crypto-native pivot and Robinhood L2: soft questions, but Armstrong's fragmentation-to-consolidation answer and the $32T Base volume stat were the most substantive competitive framing of the call.
- Ken Worthington (JPMorgan) — Hyperliquid/USDC economics protection: the sharpest institutional question. Management confirmed the strategy of sharing economics for network effect but gave no mechanism for protecting long-term economics as large holders gain leverage — an unanswered structural question on the USDC P&L.
- Austin Hankwitz (independent) — agentic price competition and Coinbase One take rate: two genuinely good questions. On agents, Armstrong argued trust/reliability/compliance transfer to machine customers (AWS analogy) alongside Base's sub-cent, sub-second settlement — a thesis, not evidence. On Coinbase One, Haas self-described her answer as "unsatisfying" and conceded P&L mix shift without quantifying per-dollar revenue — the take-rate question that disappeared from last quarter's curated Q&A resurfaced and remains unanswered.
- Alexander Markgraff (KeyBanc) — cross-product adoption: Haas's answer (storage-first funnel, 1-year marketing payback, prediction-markets incrementality) was the most operationally informative of the call. His second question was lost to audio failure.
- Notably absent: no questions on Q2 results, guidance, the RIF, restructuring charge, Deribit, tokenized equities, buybacks, or the convertible — the format ensured the quarter's hard numbers went undiscussed.
Management credibility- Circle renewal disclosure is a clear positive: specific, unambiguous, and ahead of the event — "we have already met the conditions... it will renew on the same terms." This resolves a multi-quarter analyst concern definitively.
- CLARITY messaging has quietly de-escalated: last quarter Grewal gave markup/floor-vote/signature dates with confidence; this quarter Armstrong is "pretty optimistic" about a floor vote only, acknowledges sub-50% external odds, and pre-builds the failure narrative. The fallback framing is prudent, but the prior quarter's timeline now looks overconfident — check whether "signed by end of summer" is still claimed anywhere.
- Coinbase One ATH claim is unverifiable as stated: no subscriber number given, so "all-time high" cannot be checked against last quarter's >1M. Volunteering a superlative without the number fits the established pattern of flattering-but-fuzzy disclosure.
- Haas's candor on Coinbase One economics ("I'm gonna give an unsatisfying answer") is honest but confirms management will not disclose member vs. non-member revenue per dollar — a real analytical gap as mix shifts.
- Leadership-turnover answer was thin: "nothing to read into it" plus succession framing, with no explanation of why multiple senior departures clustered in one quarter. Not evasive in tone, but low-information.
- The absence of results discussion is the dominant credibility fact: a quarter containing a restructuring charge, a down market, and (per prior guidance) key financial tests was discussed with zero numbers. Whatever the filing shows, the call format minimized management's on-record accountability for it.
- Base's $32T figure was delivered imprecisely (the transcript shows Armstrong stumbling over billion/trillion) — treat the number as approximate until confirmed in print.
What changed versus the prior quarter- Circle contract: from "auto-renews into perpetuity, cannot be terminated" (Q1 disclosure) to renewal conditions already met on the same terms — the overhang is now fully resolved, not just structurally described.
- CLARITY: from dated confidence to probabilistic hedging — Q1 had markup-this-month/signed-by-summer; Q2 has ~30% external odds acknowledged and an agency-rulemaking fallback. The legislative timeline from Q1 appears to have slipped (markup did not produce a passed bill by this call).
- Stablecoin strategy broadened: from "we are the platform that powers stablecoins" with USDC at the center to an explicit multi-stablecoin posture including joining a competing consortium (Onyx USD) — a real strategic evolution, partly answering the concentration question while raising the economics-dilution question Worthington pressed.
- Coinbase One: from first disclosure (>1M) to claimed ATH with growth through a downturn — but the absolute number was withheld this quarter, a partial disclosure step back.
- Revenue mix milestone: Bitcoin at ~12% of revenue (per questioner, affirmed) — the diversification narrative's most concrete single datapoint to date.
- New product category opened: pre-IPO perps (SpaceX) live for non-US users — a genuinely novel product with early traction claims; US timeline absent.
- Senior leadership turnover emerged as a new topic — multiple departures across people/legal/institutional functions, unexplained beyond succession.
- Disclosure format deteriorated further: from curated written Q&A with results discussion to a short X Spaces Q&A with no results section at all.
- RIF/AI-native narrative vanished: last quarter's defining topic received zero mentions — no savings-realization update, no AI-productivity metrics.
Bull case- Circle renewal risk is gone — conditions met, same terms, removing the single largest contractual uncertainty over the S&S revenue base.
- Coinbase One growing to an ATH through a down market supports the subscription-stabilization thesis: engaged, retained, multi-product users accruing revenue across staking, card, and fees even as trading volumes fall.
- Diversification is measurable: Bitcoin ~12% of revenue, prediction markets driving incremental (non-cannibalizing) spot volume, pre-IPO perps opening a new asset category with visible demand.
- Base's network-effect moat is holding against well-funded new L2s (Robinhood, Stripe): $32T TTM stablecoin volume, #1 stablecoin chain, agentic leadership, and a credible consolidation-endgame framing.
- Regulatory downside is now explicitly de-risked: even CLARITY failure yields agency rulemaking from publicly committed chairs — the existential-regulatory scenario is off the table in management's telling, and CLARITY upside remains.
- Multi-stablecoin posture converts a competitive threat (Onyx USD) into incremental economics and FX/trading opportunities while USDC remains #1 in transaction volume.
Bear case- The call contained zero financial results in a quarter with a restructuring charge and a down market — investors must rely entirely on the filing, and management chose a format with no accountability for the numbers. The disclosure trend is now a pattern, not an event.
- CLARITY's timeline slipped versus last quarter's confident dates, and management's own fallback (agency rules) implicitly concedes the legislative path may fail — the "signed by end of summer" credibility test appears to be failing.
- Hyperliquid precedent is unaddressed: management confirmed sharing USDC economics for distribution but offered no answer on protecting long-term economics as large holders gain leverage — the USDC margin structure may be structurally negotiable downward.
- Coinbase One's accretion claim is unquantified: Haas conceded revenue shifts away from transaction revenue as membership grows and declined to say whether revenue per dollar traded is higher or lower — take-rate dilution remains a live, unanswered risk.
- Agentic commerce remains narrative-only: two consecutive calls with explicitly no numbers or forecasts, while the trust-transfers-to-agents argument is asserted, not evidenced.
- Clustered senior leadership departures were waved away without explanation — succession depth is claimed, but simultaneous exits across people, legal, and institutional functions in a restructuring year warrant scrutiny.
- Key Q1 promises went unmentioned: RIF savings realization, Deribit/US options, tokenized equities, equities-trading metrics — the "more news next quarter" deferral pattern continues.
Next-quarter watchlist- The Q2 shareholder letter/filing (immediate): revenue, S&S vs. the $565–645M guide, tech/dev+G&A vs. $820–870M, the $50–60M restructuring charge, GAAP marks, buyback pace, and the June 1 convertible retirement — none of this is in the transcript.
- CLARITY Act: Senate floor vote before/after August recess; final rewards language; if it fails, whether SEC/CFTC rulemaking actually proceeds as Armstrong claims — and whether management acknowledges the Q1 timeline miss.
- Circle renewal execution: formal renewal on the disclosed same terms; any Onyx USD launch details, economics, or consortium commitments.
- Coinbase One: an actual subscriber number (the ATH claim needs a figure), and any quantification of member vs. non-member revenue per dollar traded.
- USDC economics protection: whether the Hyperliquid-style economics-sharing model expands, and any disclosure of aggregate rewards/economics-sharing costs as a share of stablecoin revenue.
- Pre-IPO perps: pipeline additions beyond SpaceX, volume/open-interest data, and any US regulatory pathway.
- RIF follow-through: first update on $500M savings realization and restructuring execution — entirely absent this quarter.
- Leadership stability: whether departures continue or the named successors (Dominique, Molly, Liz) are confirmed in role.
- Base metrics in print: verification of the $32T stablecoin-volume claim and any Base-app relaunch metrics under Cobie.
- Call format: whether results discussion returns to the call or the X Spaces Q&A-only format persists — and whether Deribit, tokenized equities, and take-rate questions ever resurface.
|
| May 7, 2026 | +4.25% | Q1 FY2026 | Read transcript briefingQuarter in one view- Q1 FY2026: total revenue $1.4B (-21% QoQ vs. $1.8B), transaction revenue $756M, S&S $584M (-16%, within the $550–630M guide), GAAP net loss $394M, adjusted EBITDA $303M — the 13th consecutive adjusted-EBITDA-positive quarter. Management claims delivery "within or better than every range" set in February.
- The quarter was macro-driven: total crypto market cap and total crypto trading volume both fell >20% QoQ, and long-tail asset volatility hit historic lows. Management's framing: "price headwinds outpaced strong growth," but share, native units, and new-product revenue all hit records.
- The Everything Exchange finally got economics: retail derivatives >$200M annualized revenue (on track to be the next $250M-tier product), prediction markets $100M annualized in March — two months after launch, on track to be the 13th $100M product. This was the data deferred from the Q4 call, and it arrived.
- The quarter's biggest structural event was not in the P&L: a headcount reduction announced two days before the call, framed as an "AI-native" transition, removing ~$500M of costs vs. the Q4 2025 run rate, with a $50–60M Q2 restructuring charge and Coinbase's first-ever annual adjusted-expense guide ($4.3–4.6B).
- Q2 guide: S&S $565–645M (management flags "opportunity for quarter-over-quarter growth"); tech/dev + G&A $820–870M (-4% to -9% QoQ). Notably, no intra-quarter transaction revenue partial was given — a disclosure step back from last quarter's ~$420M-through-Feb-10 data point.
- Balance sheet tightening: >$10B cash, $12B total available resources (vs. $11.3B/$14.1B last quarter); $1.3B 2026 convertible due June 1 will be retired in cash unless the conversion price is reached; $1.1B of buybacks (~6M shares) executed in Q1.
What management is focused on- The RIF and AI-native transition as the quarter's defining action: announced Tuesday, two days before the call. Haas framed it as two inseparable forces — market headwinds and AI leverage — and explicitly declined to weight them. Supporting metrics: pull requests per engineer +78–80% YoY, integration test coverage up 3x in six months. The $500M cost removal (vs. Q4 2025 run rate) and the first annual expense guide convert last quarter's one-quarter-deep discipline commitment into a full-year framework.
- Everything Exchange validation: Armstrong claimed Coinbase has transformed "from a primarily spot-focused crypto platform into a place where you can now trade any asset class" — stocks, 24/7 equity perps, prediction markets, non-crypto futures (silver/gold/oil volumes +4x QoQ). The $200M/$100M annualized figures are the proof points; Choi emphasized this revenue is cross-sold into an already-acquired customer base.
- Stablecoin full-stack positioning: Choi rejected the "network participant" framing outright — "we are the platform that powers stablecoins" — citing USDC distribution (>25% of all USDC held in Coinbase products; ~50% of all USDC economics captured), Base as settlement layer (62% share of stablecoin transactions), payments APIs, and x402. Stablecoin transaction volume doubled QoQ; USDC and partner stablecoins drove >80% of it.
- Agentic commerce as the new frontier narrative: 99% of x402 transactions settled in USDC, >90% of agentic stablecoin volume on Base, x402 now governed under the Linux Foundation with Cloudflare, AWS, Stripe, Shopify, and Google contributing; agentic.market directory launched. Armstrong positioned this as the 1–3 year excitement thesis alongside tokenization ($30B today, $16T by 2030 cited) and stablecoin payments.
- CLARITY Act as the regulatory unlock: Grewal expressed confidence in markup this month, floor vote early summer, signed law by end of summer. The Tillis/Alsobrooks compromise — preserving activity-based rewards while prohibiting passive deposit-style yield — was framed as workable and protective of "key elements of our current program."
- Share gains as the downturn KPI: all-time-high crypto trading market share, gains in both spot and derivatives globally, ~5x share growth since Q1 2023, 12th consecutive quarter of native-unit inflows. Choi's claim: share captured in down markets is sticky.
Key numbers and quarter mechanics- Revenue bridge: total $1.4B (-21%). Transaction $756M: consumer $567M (-23% vs. a 35% decline in consumer spot volumes — outperformance attributed to mix shift toward core trading from advanced, plus derivatives/prediction-market revenue that sits outside the spot-only volume metric); institutional $136M (-27%, in line with institutional volume trends).
- S&S $584M (-16%): stablecoin revenue $305M with average USDC in Coinbase products at an ATH of $19B; blockchain rewards $101M (down on price and protocol reward rates, but native-unit growth in staked balances); interest and finance fee income $68M (+13% QoQ) on ATH average daily loan balances of $1.4B and double-digit active-customer growth. Armstrong stated S&S is now 44% of net revenue.
- Reporting change: $18M of corporate stablecoin revenue reclassified to other revenue (cash/USDC treated as fungible in corporate operations, consistent with reporting payment stablecoins as cash equivalents); historicals recast. Directionally small but reduces stablecoin-revenue comparability.
- GAAP net loss $394M: no mark-to-market breakdown (crypto portfolio, strategic investments/Circle) was provided on the call — a gap versus last quarter's explicit $718M/$395M bridge. Check the filing for the marks.
- Opex $1.4B (-5% QoQ): tech/dev $526M, up modestly on one-time Q4-acquisition costs; G&A -17% QoQ on lower deal-related legal, customer support, and policy expenses — management says it "got a head start" on cuts before the RIF.
- RIF economics: ~$500M of total costs removed vs. Q4 2025 run rate; $50–60M restructuring charge as a standalone Q2 line item; 2026 adjusted expenses guided to $4.3–4.6B (~$500M below the Q4 2025 annualized exit rate at midpoint; flat YoY vs. 2025 excluding USDC rewards growth).
- Capital return and balance sheet: $1.1B buyback of ~6M shares (~$183/share implied); cumulative buybacks have offset ~90% of SBC-issued shares since Q4 2024. Cash >$10B; total resources $12B (down ~$2.1B QoQ); $1.3B convertible retirement intended June 1.
- Coinbase One: over 1 million paid subscribers — the first subscriber disclosure after five-plus quarters of refusal. Members described as higher-volume, higher-revenue, most-engaged, with "strong unit economics." No ARPU or churn given.
- USDC contract specificity (new): auto-renews every three years into perpetuity, cannot be terminated; revenue share tied to overall USDC supply and adoption and "unaffected by any rewards language" per Haas — a direct answer to the Circle-mechanics question that last quarter was only addressed directionally.
- Derivatives mechanics: both US and international derivatives exchanges hit ATH revenue contribution (in institutional transaction revenue); institutional derivatives revenue more than offset a Deribit options-activity decline driven by low volatility; Deribit open-interest share held steady. No Deribit revenue figure — second consecutive quarter without one.
Product and launch scorecard- Prediction markets: $100M annualized revenue in March, two months after launch — the fastest-scaling new product disclosed and on track to be the 13th $100M product. This is the first hard monetization data for any December-event launch. No margin, take-rate, or Kalshi revenue-split detail.
- Retail derivatives: >$200M annualized revenue, ATH quarter, on track for the $250M tier. Real revenue, though still ~3% of total revenue annualized.
- Non-crypto futures (silver, gold, oil): volumes +4x QoQ — growth off an undisclosed base; no revenue figure.
- Equities/stock trading: mentioned as launched ("we've added stock trading, 24/7 equity perps") but zero metrics — no volumes, users, or revenue. The ~10,000-ticker rollout target from last quarter was not updated.
- Tokenized equities: not mentioned at all on the call — the largest undelivered Everything Exchange promise has gone silent again.
- Deribit/options: integration "progressing nicely," full unification of spot/perps/futures/options expected in 2026 with incremental milestones through the year; US options launch has no timeline ("actively working on it, very optimistic"). Options activity declined QoQ on low volatility after Q4's ATH.
- Base: 62% share of all stablecoin transactions; 10x YoY growth in stablecoin transactions on Base; >90% of agentic stablecoin volume; DEX volumes 2x QoQ; DeFi borrow/lend balances >$1B (grown over the last year). Base token not mentioned.
- x402/agentic: real adoption signals (99% USDC settlement, blue-chip governance contributors, agentic.market) but zero monetization data — Kumar's question on incremental USDC growth and facilitator fee potential was answered with distribution stats, not economics.
- Coinbase One: >1M subscribers — the quarter's disclosure win; Armstrong explicitly positioned it as the zero-fee answer to TradFi price competition.
- Institutional pipeline: 45 major financial institutions "moved tokenization from concept to production" in Q1; ETFs (including staking), DEX, and prime custody activations cited as new TAM. No revenue attached.
Sell-side read-through- Format change is the story: Q&A moved to curated written questions (analyst + X submissions) read by the new IR head — no live follow-ups, no ability to press evasions. This structurally reduces accountability versus prior calls and coincided with a quarter featuring a RIF, a GAAP loss, and a -21% revenue print.
- James Yaro (Goldman) — CLARITY status: Grewal gave the most specific legislative timeline yet (markup this month, floor vote early summer, signature by end of summer) and endorsed the Tillis/Alsobrooks rewards compromise. Confident but hedged: "not declaring victory," "details matter."
- Ken Worthington (JPMorgan) — post-CLARITY ecosystem: Armstrong gave a GENIUS-Act analogy (hundreds of companies integrating stablecoins afterward) and positioned CDP as the integration layer. Vision answer, no quantification.
- Peter Christiansen (Citi) — Circle revenue share under rewards policy change: the most economically informative exchange. Grewal: contract is set and auto-renews. Haas: revenue share tied to USDC supply/adoption, "unaffected by any rewards language." This directly addresses last quarter's open question — and partially defuses the rewards-ban risk to Coinbase's own economics.
- Ramsey El-Assal (Cantor) — competitive environment: Choi delivered the share stats (ATH, ~5x since Q1 2023, spot + derivatives gains) and the stickiness claim. No denominator or third-party source, as before.
- Andrew Jeffrey (William Blair) — stablecoin infrastructure ambitions: Choi's "we are the platform that powers stablecoins" answer was strategically clear but contained no new product, revenue, or CPN-specific detail.
- Rayna Kumar (Oppenheimer) — x402 monetization: the key unanswered question of the call. Armstrong described distribution and governance but gave no USDC-growth sensitivity, no Base fee potential, no facilitator economics.
- Patrick Moley (Piper Sandler) — Everything Exchange monetization timeline: Haas gave the $200M/$100M figures but explicitly refused a per-product outlook ("we're not going to give an outlook on a per product basis") and deferred again — "more news to share next quarter."
- Owen Lau (Clear Street) — US options timing: no timeline; integration milestones through 2026 promised. Second consecutive quarter of deferral on US options specifics.
- Devin Ryan (Citizens) — speculative vs. utility transition: Armstrong rejected the premise that utility is "waiting," citing stablecoins, prediction markets, tokenization, agentic, DeFi. Philosophically confident; the 44% S&S mix stat was the only hard support.
- Alex Markgraff (KBCM) — RIF sizing: Haas gave the $500M figure and the two-forces framing but declined to split market-driven vs. AI-driven — the question everyone wanted answered, answered deliberately vaguely.
- John Todaro (Needham) — institutional weakness: Choi attributed it to low volatility hitting Deribit hedging demand, noted most of the decline occurred in January with engagement recovering into quarter-end, and pointed to loan balances and the tokenization pipeline. A reasonably direct answer.
- Notably absent from Q&A: no question on the Q4 outage follow-up, take-rate trajectory (breaking a five-plus-quarter streak of asking), BTC purchase cadence, tokenized equities, or the GAAP loss composition. The curated format made these absences costless.
Management credibility- Guidance delivery claim is checkable and holds: S&S $584M landed within $550–630M; opex came in under guidance; management's "within or better than every range" claim is consistent with the February outlook. The 2025 guidance streak extends.
- The Q4 monetization promise was kept: Haas deferred Everything Exchange economics to this call and delivered specific annualized revenue figures for derivatives and prediction markets. A real credibility deposit.
- Coinbase One disclosure after five-plus quarters of refusal (>1M subscribers) partially repairs the selective-disclosure pattern — though it arrives when the number is flattering.
- The RIF timing and framing cut both ways: cutting $500M of costs into a downturn is the discipline management promised, and the first annual expense guide is a genuine step-up in commitment. But announcing it two days before the call, attributing it to an inseparable mix of macro and AI, and front-running it with a G&A "head start" suggests the revenue environment drove more of the decision than the AI narrative implies. The AI-productivity metrics (PRs +78%, test coverage 3x) are real-sounding but unaudited and self-selected.
- Disclosure tightened where it hurts: no Q2 transaction revenue partial (given last quarter), no GAAP loss bridge on the call, no Deribit revenue (second quarter), no equities metrics, no tokenized-equities update, no outage follow-up, and a curated Q&A format that prevents follow-up pressure. The information ratio fell again even as one long-withheld number (Coinbase One) was released.
- Circle contract specificity is a credibility positive: the auto-renew-in-perpetuity, cannot-be-terminated, supply-linked framing is the most concrete contractual disclosure to date and directly addresses a standing analyst concern.
- CLARITY confidence is now on the record with dates: markup this month, signature by end of summer. This is falsifiable within one quarter — a credibility test management chose to set.
- Armstrong's AI-code clarification was handled cleanly: the X question on nontechnical staff pushing code to production was answered with a direct correction (human engineers review all production code) plus a quality-investment data point from Haas. Responsive, not defensive.
What changed versus the prior quarter- Revenue decline steepened sharply: -21% QoQ vs. -5% in Q4; adjusted EBITDA $303M vs. $566M; GAAP loss narrowed to $394M from $667M (Q4's was mark-driven; Q1's composition undisclosed on the call).
- Cost posture shifted from "flat guide" to structural action: a RIF, $500M cost removal, a Q2 restructuring charge, and the first annual expense guide — versus last quarter's one-quarter-flat commitment. Discipline moved from promise to restructuring.
- Everything Exchange crossed from shipped-product to disclosed-economics: $200M derivatives and $100M prediction-markets annualized revenue versus "zero economics yet" last quarter. Equities, however, went from rollout metrics promised to complete silence.
- Regulatory posture flipped from defense to offense: last quarter the Senate rewards amendments were a live threat; this quarter management endorses a compromise preserving activity-based rewards and gives a signed-by-summer timeline. The Circle contract was disclosed as effectively unbreakable.
- Stablecoin narrative shifted from "flatlining market cap" to volume and share: transaction volume doubled QoQ, Base at 62% share, USDC-in-products ATH $19B — the velocity-over-supply story Haas previewed last quarter is now the official framing.
- Q&A format changed to curated written questions including X users — a structural governance/disclosure change that reduces real-time accountability.
- Balance sheet began shrinking: resources down ~$2.1B QoQ with a $1.3B convertible retirement flagged for June 1 — capital deployment (buybacks, BTC, M&A) is now visibly drawing down the pile.
- New disclosure wins and losses: Coinbase One subscribers and Circle contract terms disclosed for the first time; transaction-revenue partial, GAAP mark bridge, Deribit revenue, and tokenized-equities status all withheld or absent.
- Agentic commerce elevated to a top-tier narrative (x402, Linux Foundation governance, agentic.market) — largely new emphasis versus last quarter's "Base as onchain home for AI" framing.
Bull case- The diversification thesis now has revenue behind it: derivatives >$200M and prediction markets $100M annualized within months of launch, 12 products >$100M, S&S at 44% of net revenue — and consumer revenue fell 23% against a 35% spot-volume decline, showing mix resilience.
- Cost structure is being reset proactively: $500M removed, 2026 adjusted expenses guided ~$500M below the Q4 exit rate, flat YoY ex-USDC rewards — operating leverage is now a guided, falsifiable framework rather than a one-quarter promise.
- Share gains in a down market (ATH share, ~5x since Q1 2023, 12 straight quarters of native-unit inflows) support the consolidation-to-trusted-platform thesis, with management arguing the share is sticky into recovery.
- Regulatory path is the clearest it has ever been: a rewards compromise management can live with, a dated legislative timeline, an effectively permanent Circle contract, and a post-CLARITY institutional onboarding wave (45 institutions already moving tokenization to production).
- The stablecoin/agentic stack is genuinely differentiated: >25% of USDC held on-platform, ~50% of USDC economics captured, 62% Base share of stablecoin transactions, 99% USDC settlement on x402 with blue-chip governance — if agentic commerce materializes, Coinbase owns the full stack.
- Capital return continued through the drawdown ($1.1B in Q1, ~90% of SBC since Q4 2024 offset) with $10B+ cash remaining even after the convertible retirement.
Bear case- Revenue is still deeply macro-cyclical: -21% QoQ with both market cap and volumes down >20%, and the S&S "growth engine" has now declined two consecutive quarters on price and rate sensitivity. Q2's guided S&S "opportunity for growth" depends on rates and prices, not execution.
- The RIF is an admission: you do not cut $500M two days before an earnings call purely because AI made you efficient. Management's refusal to split macro vs. AI drivers, plus the pre-announced G&A "head start," implies the revenue environment forced the move — and restructuring into a downturn risks execution capacity if the market turns.
- Disclosure quality is deteriorating at the margin: curated Q&A with no follow-ups, no transaction-revenue partial, no GAAP mark breakdown on the call, no Deribit revenue, no equities metrics, no tokenized-equities update, no outage follow-up. The pattern of volunteering ATHs while withholding unflattering granularity intensified.
- New-product revenue is still small: $300M combined annualized from derivatives and prediction markets is ~5% of total revenue — promising slope, immaterial base, and no margin or take-rate data to assess quality.
- Institutional revenue fell faster than retail (-27%), Deribit options activity declined on low volatility, and US options remain undated — the named 2026 growth driver is itself vol-dependent.
- Balance sheet resources fell ~$2.1B in one quarter with a $1.3B convertible due June 1; continued buybacks plus BTC purchases plus M&A into a revenue decline compresses the cushion if the downturn extends.
- CLARITY confidence is management's own read of a live process; the rewards compromise is not final text, and Grewal himself cautioned "details matter" and rules remain unwritten.
Next-quarter watchlist- Q2 vs. guide: S&S within $565–645M and whether the "opportunity for QoQ growth" materializes; tech/dev + G&A within $820–870M; the $50–60M restructuring charge landing as a clean standalone line.
- RIF execution and savings realization: evidence the $500M cost-out is flowing through without product-velocity or reliability damage; whether the AI-productivity metrics (PRs, test coverage) are updated with the same specificity.
- CLARITY Act against management's own timeline: markup this month, floor vote early summer, signature by end of summer — and the final rewards text versus the Tillis/Alsobrooks compromise. Any slippage directly tests Grewal's credibility.
- June 1 convertible: retirement of the $1.3B obligation in cash and the resulting balance-sheet position.
- Everything Exchange per-product update: management again promised "more news next quarter" — equities trading metrics (volumes, users, revenue) are now the most overdue disclosure; tokenized equities need any status at all.
- Derivatives detail: Deribit integration milestones toward 2026 unification, any US options timeline, and whether Deribit revenue disclosure returns after a vol-driven activity decline.
- GAAP loss composition in the filing: crypto-portfolio and strategic-investment marks not broken out on the call; also whether BTC purchase cadence continued.
- Coinbase One trajectory: subscriber growth from the 1M base and any unit-economics detail now that the number is public.
- x402/agentic monetization: any first evidence of facilitator fees, Base fee revenue, or measurable USDC growth attributable to agentic volume — the question Kumar asked and management didn't answer.
- Q&A format: whether the curated written-question format persists, and whether absent topics (take rate, outage remediation, tokenized equities) resurface.
- Buyback pace: deployment against the $2B authorization and whether the discount-to-SBC framing from Q4 returns.
|
| Feb 12, 2026 | +16.46% | Q4 FY2025 | Read transcript briefingQuarter in one view- Q4 FY2025: total revenue $1.8B (-5% QoQ vs. $1.9B), transaction revenue $983M (-6%), S&S $727M (-3%), adjusted EBITDA $566M (vs. $801M), adjusted net income $178M (vs. $421M) — a softer quarter against an 11% QoQ decline in crypto market cap, but with claimed volume/share outperformance driven by derivatives.
- GAAP net loss of $667M, driven by a $718M unrealized loss on the crypto investment portfolio and a $395M loss on strategic investments (including Circle) — the pro-cyclical marks now dominate GAAP in both directions, exactly the pattern flagged in prior quarters.
- FY2025: revenue $7.2B (+9% YoY), S&S $2.8B (+23%, 5.5x the 2021 cycle peak), 12 products >$100M annualized revenue (6 of them >$250M), 12th consecutive adjusted-EBITDA-positive quarter, 9th consecutive quarter of native-unit inflows.
- Capital return became real: $1.7B of buybacks (8.2M shares, Q4 through Feb 10), fully offsetting 2025 SBC dilution at a claimed $815M notional discount to SBC issuance price; new $2B authorization approved in January. This ends four-plus quarters of buyback silence.
- Q1 FY2026 guide: ~$420M transaction revenue through Feb 10 (with an explicit volatility caveat); S&S $550–630M (down sharply on lower crypto prices, rates, and staking rewards); tech/dev + G&A flat at $925–975M; S&M flat-to-down at $215–315M.
- The Everything Exchange actually shipped: prediction markets (via Kalshi) rolled out to 100% of customers, equities rolling out with ~10,000 tickers targeted this month, gold/silver futures drove record notional during the recent crypto selloff. The December 17 event promises converted into product, though with no revenue data yet.
What management is focused on- 2026 priorities, explicitly ranked: (1) grow the Everything Exchange (derivatives, prediction markets, equities, tokenized equities, international expansion); (2) scale stablecoins and payments (USDC utility, CDP, Coinbase Business, defending stablecoin rewards in legislation); (3) bring the world onchain (DeFi integrations in the main app, Base app with trading focus, Base chain volume).
- Everything Exchange as the monetization thesis: Armstrong's "asset accumulation flywheel" — trust drives asset custody (claims 12% of all crypto, more than the next four competitors combined; assets on platform ~3x over 3 years), products connect to assets, stickiness and monetization follow. New asset classes are framed first as stickiness tools for existing users, second as a front door for traditional investors.
- Derivatives as the 2026 growth driver: Haas named derivatives the big 2026 driver, citing US and international momentum plus Deribit options integration; Q4 saw ATH derivatives volume and revenue, and Deribit had another ATH quarter.
- Stablecoin rewards as a policy fight: management is actively lobbying to preserve rewards on regulated US stablecoins (CLARITY Act Senate amendments would restrict them); Armstrong argued rewards are needed for US stablecoins to compete with offshore/CBDC alternatives, and made the striking point that a rewards ban would *increase* Coinbase profitability since it passes the majority of Circle economics to customers.
- Capital allocation as a downturn playbook: buy the dip on Bitcoin weekly (modestly increased purchase size; doubled BTC native units in the investment portfolio in 2025), buy back stock on dislocations, continue opportunistic M&A (10 deals in 2025; 2026 themes: Everything Exchange, onchain infrastructure, stablecoin/payments rails).
- Cost discipline entering 2026: Q1 opex guided flat to Q4's guide; 2025 framed as "an investment year" with USDC rewards the single largest YoY expense driver and 16% of the increase from M&A (mostly deal costs). The "absorb and slow" commitment from last quarter is holding, at least for one quarter.
Key numbers and quarter mechanics- Revenue bridge: total $1.8B (-5%); transaction $983M (-6%); S&S $727M (-3%). Management attributes resilience to derivatives growth offsetting softer spot conditions; no spot/derivatives split given on the call.
- GAAP loss mechanics: -$667M net loss = $718M unrealized crypto-portfolio loss + $395M strategic-investment loss (Circle included). Adjusted net income $178M — the GAAP/adjusted gap is now entirely mark-driven, and the marks are getting bigger as the portfolio grows.
- Opex: total $1.5B (+9% QoQ), in line with outlook; tech/dev + G&A + S&M +14% QoQ on Deribit/Echo costs and higher USDC rewards (record USDC balances); +11% ex-deal costs. Headcount 4,951 (+3% QoQ) — hiring has clearly slowed from the +12% pace of Q3.
- Balance sheet: $11.3B cash; ~$14.1B total available resources including crypto investments and collateral (down from $11.9B cash in Q3, reflecting buybacks, BTC purchases, and M&A).
- Buyback math: $1.7B for 8.2M shares (~$207/share average implied), claimed $815M notional discount to the average SBC issuance price — a new, specific framing of buyback ROI.
- Q1 S&S guide of $550–630M is a material step-down (midpoint -19% vs. Q4's $727M): lower average crypto prices, lower interest rates, lower staking protocol reward rates. This is the most important forward number on the call — the S&S growth streak is guided to break.
- Q1 transaction revenue ~$420M through Feb 10 (~40% of the quarter elapsed) with heightened volatility; management explicitly cautioned against extrapolation — a more hedged version of the monthly-revenue disclosure than prior quarters.
- USDC: all-time-high USDC stored in Coinbase products in Q4; USDC market cap ATH ~$75B (vs. $74B last quarter) — but stablecoin market cap has flatlined for "the last couple of months" per Haas, who attributed it to reduced long-tail speculation/leverage, partly offset by higher payment velocity.
- Take rate: mix shift to advanced trading and Coinbase One continued; Haas confirmed Coinbase One trades carry up to zero fees but still generate spread recorded as retail transaction revenue. No quantification — fifth-plus consecutive quarter declining take-rate specifics.
Product and launch scorecard- Everything Exchange (launched Q4): the quarter's defining delivery. Prediction markets live to 100% of customers via Kalshi (non-exclusive; Coinbase retains the option to launch its own venue — "nothing to announce"); equities rolling out toward ~10,000 tickers this month; commodities (gold/silver futures) drove record notional during the selloff; highest 24-hour volume in over a year. Evidence so far is anecdotal ("early encouraging signals," Super Bowl engagement); Haas explicitly deferred monetization data to the Q1 call. Real launch, zero economics yet.
- Derivatives/Deribit: ATH derivatives volume and revenue in Q4; Deribit another ATH quarter; options integration cited as 2026 momentum. No updated revenue contribution figure this quarter (Q3 was $52M partial).
- Tokenized equities: "working on shipping," with a claimed path via a "crypto-forward SEC" — still no date, but now framed as regulatory-path-dependent rather than purely conceptual.
- Base: new transaction ATH; positioned as "the onchain home for AI" (agent wallets, stablecoin payments gaining traction); #1 L2 on Ethereum per Armstrong; privacy features in development; Base token still "exploring" with no new detail. Haas corrected the record that sequencer fees sit in *other transaction revenue*, not S&S — a useful disclosure clarification.
- Stablecoins/payments: USDC balances in Coinbase products at ATH; payments vertical (Base + USDC + APIs) described as "early in our product journey" with Q4 product/API build-out and go-to-market now starting. No updated Coinbase Business count (was 1,000+ last quarter).
- Coinbase One: central to the take-rate strategy — more volume migrating under the membership umbrella; subscriber count still undisclosed.
- Platform reliability: a same-day outage (buy/sell/transfer interruptions on retail and Prime; derivatives and equities unaffected) disclosed in response to a Baird question — attributed to a technical issue unrelated to volume or market conditions, now resolved. Notable given last quarter's resilience claims.
Sell-side read-through- Andrew Jeffrey (William Blair): Everything Exchange monetization timing — Haas pointed to derivatives as the 2026 driver and deferred prediction-market/equities data to the Q1 call ("weeks and days of data"). A disciplined but real deferral.
- Ken Worthington (JPMorgan): would CLARITY eliminate Circle reserve-income sharing? — Armstrong: no; and the counterintuitive disclosure that a rewards ban would make Coinbase *more* profitable since it passes the majority of Circle economics to customers. The most economically informative answer of the call.
- Owen Lau (Clear Street): buybacks and M&A in a down market — Haas gave the $1.7B/8.2M-share detail, 10 acquisitions in 2025, doubled BTC holdings. Direct, quantified answer.
- Patrick Moley (Piper Sandler): prediction-market adoption and own-venue plans — Armstrong: early, positive, Kalshi non-exclusive, own-market optionality open. No metrics.
- James Yaro (Goldman): crypto winter and cycle KPIs — Armstrong declined to predict, dismissed fundamental explanations, cited net buying on-platform. Philosophical rather than analytical.
- Ben Budish (Barclays): 2026 spending flex — Haas: Q1 flat to Q4 guide; USDC rewards were the largest 2025 expense driver; 16% of YoY growth was M&A (mostly deal costs); "nimble" thereafter. No full-year framework — the 2026 opex commitment remains one quarter deep.
- Robbie Bamberger (Baird): the day-of outage and ecosystem leverage (BlockFill context) — Haas confirmed a brief buy/sell/transfer interruption, technical in nature, resolved; derivatives/equities unaffected. Answered directly but minimized; no duration, scope, or customer-impact quantification.
- Alex Markgraff (KeyBanc): will equities/prediction markets bring net-new users? — Armstrong: flywheel first, new-user acquisition second. Strategy answer, no data.
- Ramsey El-Assal (Cantor): M&A strategy — Choi: selective but aggressive; themes are Everything Exchange, onchain infrastructure, stablecoin/payments rails. Consistent with prior messaging.
- Pete Christiansen (Citi): Base L2 value prop as L1 improves — Armstrong engaged substantively (Vitalik's post, Base as #1 L2, privacy features, L1/L2 blurring). No DeFi-regulation answer despite it being half the question.
- Devin Ryan (Citizens): stablecoin market-cap flatlining — Haas gave the most analytically honest answer of the call: reduced long-tail speculation and leverage, offset by rising payment velocity; monetization shifting to new vectors.
- John Todaro (Needham): January/February USDC balances — declined; pointed to the shareholder letter. A small but notable tightening of intra-quarter disclosure.
- Bo Pei (US Tiger): take-rate compression and normalization — Haas ("my quarterly take rate question") confirmed mix shift and Coinbase One spread economics but again declined any quantified trajectory. Fifth-plus quarter of refusal.
- Gus Gala (Monness): B2B payments S-curve timing and x402 — Haas: early, building go-to-market, "you'll see more through the year." No timeline, no metrics — the payments S-curve question remains unanswered.
- Dan Dolev (Mizuho): casual-trader behavior in a downturn — Haas: HODL pattern intact, active users net buying the dip, diversification is the hedge.
Management credibility- Guidance streak claimed and partially checkable: Haas stated Coinbase "delivered or outperformed revenue and expense guidance every quarter" in 2025; Q4 opex came in line with the $925–975M/$215–315M framework. S&S of $727M landed within the $710–790M guide.
- Buyback promise kept with specifics: after four-plus silent quarters, $1.7B deployed, share count, dilution offset, and a novel $815M discount-to-SBC-issuance metric. This is the quarter's biggest credibility repair.
- Everything Exchange delivered on schedule: tokenized equities remain undated, but prediction markets, equities, and commodities all shipped within weeks of the December event — a real conversion from narrative to product.
- Cost discipline holding so far: headcount +3% (vs. +12% in Q3), Q1 opex guided flat — consistent with last quarter's "absorb and slow" commitment, though only one quarter of evidence and no full-year framework.
- Outage disclosure under pressure was direct but thin: confirmed promptly when asked, but no duration, affected-user count, or remediation detail — and it cuts against last quarter's infrastructure-resilience narrative.
- Disclosure tightening in places: refused intra-quarter USDC balances (Todaro), no Deribit revenue update, no Coinbase One subscribers, no perp/DEX economics, no payments metrics. The pattern of volunteering favorable data (ATHs, share gains) while withholding unfavorable or unflattering granularity persists.
- The rewards-ban admission is double-edged: Armstrong's candor that a stablecoin-rewards ban would raise Coinbase profitability is credible and useful — but it also quantifies, in direction if not magnitude, how much of the USDC economics currently flow out to customers, and frames the lobbying fight as partly self-interested.
- "Share doubled year-over-year" claims (global volume and market share) are asserted without a denominator or third-party reference; treat as management claim pending volume disclosures.
What changed versus the prior quarter- Profitability stepped down: adjusted EBITDA $566M vs. $801M; adjusted net income $178M vs. $421M; GAAP swung to -$667M on $1.1B of combined crypto/Circle marks (vs. near-cancellation in Q3).
- Capital allocation flipped from silence to action: $1.7B buyback executed, $2B new authorization, weekly BTC purchases increased — the biggest behavioral change this quarter.
- Everything Exchange went from event-tease to shipped product: prediction markets (Kalshi), equities rollout, commodities record volumes — versus "tune in December 17" last quarter.
- S&S trajectory broke: $727M (-3%) and guided to $550–630M — the first guided double-digit sequential decline in the periods covered, ending the beat-and-raise S&S pattern.
- Opex growth decelerated as promised: headcount +3% vs. +12%; Q1 guided flat — the 2026 discipline commitment is being honored early.
- Stablecoin narrative got more nuanced: market-cap flatlining acknowledged and explained (speculation down, velocity up) versus unqualified ATH framing last quarter.
- New risks surfaced: a day-of platform outage; active Senate-amendment threat to stablecoin rewards (with management conceding a ban would be perversely profitable); Deribit revenue contribution no longer disclosed.
- Base token still exploratory — no progress on specifics despite being raised again.
Bull case- Diversification is showing up in the P&L mix: 12 products >$100M annualized revenue (6 >$250M), S&S at $2.8B for the year (5.5x the 2021 peak), and record derivatives volume/revenue in a down-crypto quarter — revenue is demonstrably less spot-correlated than prior cycles.
- The Everything Exchange is real and already differentiating: gold/silver futures drove record notional during a crypto selloff; equities and prediction markets create cross-asset stickiness and a net-new-user funnel; tokenized equities have a claimed regulatory path.
- Capital return is now a credible lever: $1.7B bought back at a claimed $815M discount to SBC issuance, $2B more authorized, $11.3B cash / $14.1B resources, and management explicitly frames downturns as deployment opportunities (buybacks + BTC + M&A).
- Cost discipline is arriving: flat Q1 opex guide, +3% headcount, and a stated need-vs-want spending framework — operating leverage is the explicit 2026 setup if revenue stabilizes.
- Derivatives momentum (Deribit ATH quarter, options integration, US/international growth) is the named 2026 driver with actual revenue behind it, unlike the newer launches.
- Retail behavior is resilient: ninth straight quarter of native-unit inflows, net buying through the Q1 dip — the HODL base keeps assets on platform through drawdowns.
Bear case- The Q1 S&S guide ($550–630M, -19% at midpoint) exposes how rate- and price-sensitive the "diversified" revenue base still is; the S&S growth story is paused, and staking-reward-rate declines add a new structural headwind.
- GAAP losses will keep whipsawing: $718M crypto + $395M Circle marks this quarter, and management is *increasing* weekly BTC purchases — deliberately adding pro-cyclical GAAP volatility into a downturn.
- New-launch economics are entirely undisclosed: prediction markets, equities, and commodities shipped with zero adoption, revenue, or margin data; the Q1 call is now the third deadline for Everything Exchange monetization evidence.
- The outage — on the day of the call, during a volatile tape — undercuts the trust-and-resilience pitch, and management's answer lacked any scope or remediation specifics.
- Take-rate compression is structural and unquantified: advanced mix plus Coinbase One migration (zero-fee, spread-only) keeps pressuring consumer monetization, and management has now declined to quantify it for five-plus quarters.
- Regulatory risk is live on the core USDC economics: Senate amendments threatening stablecoin rewards could reshape the Circle relationship; management's own admission that a ban would raise Coinbase profit invites scrutiny of whose interest the lobbying serves.
- Disclosure is tightening where it matters: no Deribit revenue update, no intra-quarter USDC data, no Coinbase One subscribers, no payments metrics — the information ratio is falling as the product count rises.
Next-quarter watchlist- Q1 vs. guide: does ~$420M transaction revenue (through Feb 10) hold given volatility; S&S within $550–630M; opex flat as promised — the first test of the 2026 discipline commitment.
- Everything Exchange monetization data: management committed to prediction-market and equities updates "at the end of Q1" — adoption, crossover trading, revenue contribution, and any take-rate disclosure are now owed.
- Derivatives detail: Deribit revenue contribution (absent this quarter), options integration milestones, US options licensing progress, and whether ATH volumes convert to visible revenue.
- Stablecoin legislation: CLARITY Act progress "in the next few months" per Armstrong; any rewards-restriction language and its quantified P&L impact in either direction.
- USDC trajectory: whether market-cap flatlining persists; payment-velocity monetization evidence; Coinbase Business/payments go-to-market metrics (none given this quarter).
- Outage follow-through: root cause, scope, remediation, and whether reliability investment claims survive another high-volume day.
- Buyback pace: deployment of the new $2B authorization; whether the $815M discount framing recurs; BTC purchase cadence and portfolio marks at quarter-end.
- Coinbase One: subscriber disclosure (still absent) and the spread-economics trajectory as more volume migrates under membership.
- Base: token exploration specifics (still none), privacy-feature launch, AI-agent wallet traction converting to measurable sequencer/transaction revenue.
- Tokenized equities: any dated milestone or SEC-path detail — currently the largest undelivered Everything Exchange promise.
|
| Oct 30, 2025 | +4.65% | Q3 FY2025 | Read transcript briefingQuarter in one view- Q3 FY2025: total revenue $1.9B (+27% QoQ vs. $1.5B), adjusted EBITDA $801M (vs. $512M), GAAP net income $433M, adjusted net income $421M — a dramatic normalization from Q2's $33M adjusted print, and this quarter GAAP and adjusted nearly converge.
- GAAP vs. adjusted bridge: +$424M crypto-portfolio remeasurement gain, −$381M other expense driven by unrealized losses on the Circle stake (stock down QoQ) — the Circle mark now whipsaws both directions, as flagged last quarter.
- Consumer spot volume +37% to $59B vs. US/global spot markets +29%/+38% — Coinbase outgrew the US market this quarter (reversing Q2's -40% vs. -31/32% gap); consumer transaction revenue +30% to $844M, lagging volume on higher advanced-trading mix (lower fee rate).
- Institutional transaction revenue +122% to $135M, driven by derivatives; Deribit closed August 14 and contributed $52M revenue and $30M opex (incl. $16M deal amortization, mostly in S&M).
- S&S +14% to $747M, above the $665–745M guide; native-unit inflows in USDC balances, institutional financing loans, custody. Assets on platform $516B. Average on-platform USDC $15B; USDC market cap ATH $74B.
- Opex -9% to $1.4B (Q2 included the $307M breach expense); tech/dev + G&A + S&M +14% to $1.1B on headcount and USDC rewards. Headcount 4,795 (+12%).
- Balance sheet: $11.9B USD resources (up from $9.3B), $2.6B long-term crypto investments (up from $1.8B).
- Q4 guide: October transaction revenue ~$385M (vs. July ~$360M); S&S $710–790M; tech/dev + G&A $925–975M (+~$100M at midpoint, ~half Deribit/Echo, ~half headcount); S&M $215–315M; ~$70M D&A embedded (up on acquisition intangibles).
What management is focused on- Everything Exchange execution: DEX integration live — tradable assets expanded from ~300 to 40,000+ in the US; US perps driving ATH US derivatives volumes and share; Deribit closed; next asset classes (prediction markets, tokenized stocks) teased for the December 17 H2 product event rather than detailed on the call.
- Derivatives integration: Deribit + Coinbase did $840B+ Q3 derivatives volume; near-term plan is integrating spot, perps, futures, and options under one roof with cross-margining (spot/derivatives cross-margining already live for US customers). US options are a "multi-quarter road map" requiring licenses and product.
- Payments/stablecoins: vertically integrated stack (Base + USDC + payment APIs + apps); Coinbase Business has 1,000+ businesses onboarded with another 1,000 on the waitlist; Citi partnership announced last week; x402 protocol (stablecoin payment attached to web requests) attracting Cloudflare, Vercel, Google; AgentKit for AI-agent wallets.
- M&A acceleration: Echo acquired (capital formation / token issuance via Sonar) — framed as vertical integration "where coins are issued before they graduate to the exchange"; Choi explicitly cites regulatory clarity as enabling a faster M&A pace, all in service of trading and payments.
- Base token exploration: Armstrong confirmed they are "early on exploring a Base network token" — no specifics on governance, distribution, or timing; Haas detailed Base monetization (sequencer fees + indirect ecosystem monetization; Base app monetizing via trading fees and advertising, with a revenue profile potentially "similar to the Coinbase main app").
- Cost absorption narrative for 2026: after a year of heavy hiring, management plans to "absorb the employees" and slow sequential opex growth versus the Q4 rate — a soft commitment to operating leverage next year.
Key numbers and quarter mechanics- Volume vs. market: consumer volume +37% vs. US spot +29% — outperformance this quarter; Haas attributed the revenue/volume growth gap (30% vs. 37%) to higher advanced-trading mix at lower fee rates, plus a "white glove" concierge service for high-value advanced traders (part service, part sales function per Armstrong).
- Institutional +122% to $135M: $52M from Deribit (partial quarter, closed Aug 14); remainder from exchange and Prime growth. Notably, Haas said derivatives rebates/incentives were scaled back with "more profitable growth" in international derivatives on stickier organic open interest — a shift from last quarter's buy-share posture. She declined to quantify take-rate impact, citing Deribit mix and unreported derivatives volume.
- S&S $747M vs. $665–745M guide: beat the top end; drivers were USDC balances, loan balances, custody assets. Q4 guide of $710–790M implies flat-to-up on higher crypto prices and Coinbase One subscriber growth.
- GAAP/adjusted convergence: net income $433M vs. adjusted $421M — the Circle mark flipped to a $381M loss, offset by a $424M crypto gain. The two noisy lines nearly canceled; adjusted net income of $421M on $1.9B revenue is a far healthier recurring-profit read than Q2's $33M.
- Opex: total -9% to $1.4B (breach expense lapped); core three lines +14% to $1.1B on headcount (+12% to 4,795) and USDC rewards growth. Deribit added $30M opex including $16M amortization.
- Q4 opex step-up: tech/dev + G&A $925–975M, +~$100M at midpoint — ~half acquisitions (Deribit, Echo), ~half headcount, with hiring slowing in Q4 vs. Q3. D&A rising to ~$70M/quarter on acquisition intangibles.
- October ~$385M transaction revenue vs. July ~$360M — continued sequential momentum, though one analyst (Gala) flagged that October spot share appeared to reverse September's outperformance; Haas gave a non-specific "we've always faced competition" answer.
Product and launch scorecard- Deribit (closed Aug 14): $52M revenue in ~6 weeks, record August volume, 100 employees onboarded in September; 75%+ global options share, all non-US. Early cross-sell signal: existing clients trading more and holding more assets on the combined balance sheet. US options path is multi-quarter and license-dependent. Real revenue, real integration risk ahead.
- DEX integration (launched): 300 → 40,000+ tradable US assets; day-1 token access. No adoption, take-rate, or cannibalization metrics disclosed — the Q2 questions remain unanswered.
- US perps: "early traction is strong," driving ATH US derivatives volumes and share — but no volume, OI, or revenue figures this quarter, and no monetization update (deferred since launch).
- Echo (acquired): capital-formation platform (Sonar); strategic logic is issuance-to-listing vertical integration against Coinbase's $500B+ asset base. No price, revenue, or deal terms discussed on the call.
- Coinbase Business / payments: 1,000+ businesses onboarded, 1,000+ waitlist; businesses earn rewards on idle USDC. First concrete adoption metric for the payments push.
- x402 protocol + AgentKit: open payment standard for web requests and AI agents; Cloudflare, Vercel, Google "started working with" it; driving CDP sign-ups. Early, no revenue.
- CaaS/CDP: 264 institutions (vs. 240+ last quarter); Citi added to JPMorgan, BlackRock, PNC, Stripe, PayPal, Revolut, Webull. Still no distinct P&L line.
- Base app: monetizing via trading fees and advertising; Haas floated a Coinbase-main-app-like revenue profile over time. No user metrics beyond last quarter's 700K waitlist.
- Base network token: confirmed under exploration; no distribution/governance/timing details — a new potential value-accretion (and regulatory) question for shareholders.
- Coinbase One: new basic tier plus Coinbase Card "showing a lot of traction"; subscriber count still not disclosed. Q4 S&S guide leans on subscriber growth.
- Tokenized equities / prediction markets: deferred to the December 17 event — still no product, timeline, or milestone.
Sell-side read-through- Craig Siegenthaler (BofA): Echo's role in capital formation — got the vertical-integration thesis (issuance → exchange graduation) but no economics.
- Ken Worthington (JPMorgan): M&A pace and themes — Choi: regulatory clarity enables more M&A; buy/build/partner/invest framework; all in service of trading and payments. No pipeline specifics.
- Pete Christiansen (Citi): infrastructure resilience — Armstrong conceded AWS-outage exposure and no full multi-cloud redundancy (cost/benefit trade-off), but cited clean operation through the October 10 record-activity day when "several major exchanges" had extended outages. Choi: 65% of support interactions fully automated; LLM agents to automate most compliance investigations in 2026.
- Ben Budish (Barclays): white-glove service and retail competition — Haas: concierge for high-value advanced traders; claimed volume growth above US spot market. Armstrong: whales drive disproportionate volume; service is partly sales.
- Owen Lau (Clear Street): Coinbase Business and Citi — Haas gave the 1,000-businesses metric and idle-USDC rewards; Armstrong gave the CDP/AWS framing.
- Devin Ryan (Citizens): Deribit integration — cross-margining live for US spot/derivatives; options integration over "the next few quarters."
- Patrick Moley (Piper Sandler): Everything Exchange milestones — answer was essentially "tune in December 17." An evasion on specifics.
- James Yaro (Goldman): October 10 liquidations — Haas: no significant liquidations on-platform, no systemic fallout; attributed ecosystem risk to opaque, non-public venues. Confident, checkable answer.
- Andrew Jeffrey (William Blair): stablecoin adoption timing and economics — Armstrong: "not seeing a change in economics yet"; reiterated $40T cross-border / 75% B2B / ~$100B stablecoin volume framing.
- Bo Pei (US Tiger): derivatives incentive cuts and take rate — Haas confirmed scaled-back rebates and "more profitable growth" but declined to quantify take-rate impact. The key margin question got a directional, not numerical, answer.
- Alex Markgraff (KeyBanc): contribution margins by product — Haas: no per-product margin targets; manage to total adjusted EBITDA and whole-customer economics (3+ products per institutional client). Deliberately non-committal.
- Dan Dolev (Mizuho): take-rate outlook — declined; pointed to Coinbase One/basic tier diversification away from trading fees.
- Ed Engel (Compass Point): is Q4 opex the full reinvestment run-rate? — Haas: ~half M&A, ~half headcount; hiring slows in Q4; 2026 opex growth to slow. The clearest guidance-quality answer of the call.
- Gus Gala (Monness): September share gains vs. October reversal — Haas gave a generic competition answer; Armstrong deflected with a prediction-market joke. The most evident dodge of the call.
Management credibility- Guidance delivered and beaten: S&S $747M vs. $665–745M guide; July's ~$360M transaction revenue extrapolation held and improved (October ~$385M). The Q2 rebound thesis was real.
- Deribit closed as promised (year-end commitment, closed Aug 14) with immediate revenue contribution and opex disclosure — a clean delivery on last quarter's biggest open item.
- Incentive-reduction claim is checkable and consistent: institutional revenue +122% alongside scaled-back rebates supports the "organic open interest" claim, though the refusal to quantify take rate leaves it partially unverifiable.
- October share question dodged: Gala's specific September-vs-October share observation got no data — notable because management volunteered share-gain claims when the trend was favorable.
- Everything Exchange specifics deferred to a marketing event: pushing tokenized equities/prediction markets to December 17 is showmanship over disclosure; two quarters in, these remain undated.
- Opex transparency improved: the Q4 step-up was decomposed (~half M&A, ~half headcount) with a forward commitment to slower 2026 growth — a testable promise.
- Adjusted-metric stability held this quarter: no redefinition; GAAP and adjusted nearly converged, reducing (for now) the metric-gaming concern from Q2.
- Armstrong's closing joke (stuffing keywords to move a prediction market on call content) is trivial financially but a poor look on market-integrity optics for a company selling trust.
- Still undisclosed: Coinbase One subscriber count, buyback activity (fourth quarter of silence), perp monetization, DEX economics, Echo/Deribit deal terms on the call.
What changed versus the prior quarter- Profitability normalized: adjusted net income $421M vs. $33M; GAAP $433M vs. $1.4B — the Circle mark flipped from +$1.5B gain to −$381M loss, exactly the whipsaw risk flagged last quarter.
- Volume/share dynamic reversed: Q2 volume -40% vs. market -31/32% (stablepair repricing); Q3 consumer volume +37% vs. US market +29% — outperformance, though October may have partially reversed (unanswered).
- Deribit went from "no update" to closed and contributing ($52M revenue, $30M opex, $16M amortization) — the biggest structural change to the institutional P&L.
- Derivatives posture shifted from land-grab to monetization-adjacent: incentives scaled back, "profitable growth" language — though take-rate impact undisclosed.
- M&A cadence stepped up: Echo added; ~half the Q4 opex step-up is acquisition-driven; D&A structurally higher (~$70M/quarter).
- New disclosures: average on-platform USDC ($15B), USDC market cap ($74B ATH), assets on platform ($516B), CDP count (264), Coinbase Business traction (1,000+), support automation (65%).
- New strategic optionality: Base network token confirmed under exploration — not present in prior quarters.
- Breach fully lapped: no residual breach costs mentioned; remediation story shifted to automation and infrastructure resilience.
- Cost trajectory: Q3 core opex +14%; Q4 guided +~$100M more — but with the first explicit commitment to slower 2026 growth.
Bull case- The earnings engine works at scale: $1.9B revenue, $801M adjusted EBITDA, $421M adjusted net income with GAAP/adjusted convergence — the cleanest profitability print in the periods covered, with S&S beating guide and October transaction revenue accelerating to ~$385M.
- Derivatives thesis is converting: Deribit closed early, $52M revenue in weeks, record August volume, $840B+ combined quarterly volume, cross-margining live, incentives being cut while volumes grow — the share-buying phase may be ending into a profitable-growth phase.
- Share momentum: consumer volume outgrew the US market (+37% vs. +29%), aided by 40,000-asset DEX coverage and white-glove retention of whales.
- Payments has its first hard adoption metrics (1,000+ businesses, 1,000 waitlist), a new bank flagship (Citi), and a credible protocol wedge (x402 with Cloudflare/Vercel/Google) into AI-agent payments.
- Balance sheet expanded to $11.9B USD + $2.6B crypto, funding an accelerating M&A flywheel (Deribit, Echo) under a favorable regulatory regime.
- Management committed to slower 2026 opex growth — operating leverage on a growing revenue base is the explicit setup.
Bear case- Q4 opex guide steps up ~$100M at midpoint with D&A structurally higher; the "absorb and slow in 2026" promise is unproven against a multi-year pattern of reinvestment.
- Take-rate opacity is worsening: advanced-trading mix diluted consumer revenue growth, institutional take rate is unquantifiable post-Deribit, and management declined every margin-by-product and forward take-rate question.
- October spot share may have reversed September's gains — the one direct question on it was dodged, and competition (newly listed exchanges, legacy players expanding) was acknowledged only generically.
- Circle exposure is a recurring P&L swing factor in both directions (−$381M this quarter); the $2.6B crypto portfolio adds more pro-cyclical GAAP noise.
- Tokenized equities and prediction markets remain undated narratives deferred to a product livestream; the Base token introduces regulatory and governance uncertainty with zero detail.
- AWS-dependency conceded with no full redundancy investment; the trust franchise still carries post-breach remediation execution risk.
- Persistent non-disclosures: Coinbase One subscribers, buyback (fourth silent quarter), perp monetization, DEX economics, Echo/Deribit terms.
Next-quarter watchlist- December 17 product event: concrete launches/timelines for tokenized equities and prediction markets vs. more vision; any Base token specifics (distribution, governance, regulatory posture).
- Q4 vs. guide: October ~$385M holding through November/December; S&S in $710–790M; tech/G&A in $925–975M; whether the ~$100M step-up decomposes as stated.
- Deribit full-quarter contribution: revenue run-rate vs. the $52M partial, integration milestones (options cross-margining), US options licensing progress, and any purchase-accounting detail.
- Derivatives economics: whether incentive cuts show up in institutional revenue quality; any take-rate quantification; perp monetization commentary (now multiple quarters deferred).
- October share question resolved: full-Q4 volume vs. US/global spot markets — did the late-quarter softness Gala flagged persist?
- 2026 opex framework: whether management formalizes the "slower sequential growth" commitment into guidance; headcount trajectory from 4,795.
- Coinbase One: subscriber disclosure (still absent); basic-tier and Card traction as S&S drivers.
- Payments metrics: Coinbase Business growth beyond 1,000; Citi partnership scope and launch timing; x402 adoption converting to CDP revenue; Shopify volume data (still none).
- Circle stake and crypto portfolio: quarter-end marks and whether disclosure/hedging policy evolves given the demonstrated ±$400M swings.
- Buyback and capital allocation: a fifth silent quarter would cement the pattern; Echo deal terms; further M&A signals.
- Breach aftermath: any residual legal/regulatory costs; automation milestones (65% → higher; LLM compliance agents in 2026).
|
| Jul 31, 2025 | -16.70% | Q2 FY2025 | Read transcript briefingQuarter in one view- Q2 FY2025: total revenue $1.5B (-25% QoQ); adjusted EBITDA $512M (vs. $930M in Q1); GAAP net income $1.4B — but driven by a $1.5B unrealized gain on strategic investments (Circle stake, now public) and a $362M crypto-portfolio gain; adjusted net income was just $33M (vs. $527M in Q1), with the metric redefined to exclude both crypto and strategic-investment gains/losses.
- Trading volume -40% QoQ vs. US/global spot markets -32%/-31%; management attributes the gap to a deliberate March repricing of previously ~zero-fee stablepairs in Advanced ("revenue over volume"). Consumer volume $43B (-45%), consumer transaction revenue $650M (-41%); institutional volume $194B (-38%), institutional transaction revenue $61M (-38%).
- S&S $656M, within the $600–680M guide; native-unit growth (USDC balances, staking units, custody assets, ATH Prime financing loans) offset by ETH/SOL price and protocol reward-rate headwinds — exactly the guided mechanic.
- Opex $1.5B, inflated by a $307M data-theft incident expense (May breach); ex-incident, opex would have declined 9% QoQ. Headcount ~4,300 (+8%).
- Balance sheet: $9.3B total USD resources; crypto investment portfolio $1.8B, "fueled by weekly Bitcoin purchases."
- Q3 guide: July transaction revenue ~$360M (strong rebound vs. April's ~$240M); S&S $665–745M (+8% QoQ at midpoint); tech/dev + G&A $800–850M (up, on deliberate headcount growth); S&M $190–290M.
- Strategic headlines: "everything exchange" vision (300+ assets listed, DEX integration into the app "next week," tokenized equities in development); US perpetual-style futures launched to retail with volume "doubling week over week" and an ATH this week; $1T+ Q2 derivatives volume and $1B open-interest ATH; Shopify live on stablecoin payment APIs; PNC and JPMorgan CaaS partnerships; GENIUS Act passed; SEC "Project Crypto" single-license proposal.
What management is focused on- The "everything exchange": every asset class on crypto rails — millions of tokens via in-app DEX integration (rolling out next week), full derivatives suite (24/7 BTC/ETH contracts, US perps, Deribit options pending), and tokenized equities next, possibly via traditional-broker integration as a stepping stone. Armstrong quantified the TAM hook: capturing 3% of equities trading would double the crypto market. No timeline given for tokenized equities — "coming quarters."
- Payments as the second growth engine: vertically integrated stack (USDC + Base + apps/wallets + payment APIs). B2B cross-border framed as a $40T opportunity, 75% B2B; ~$100B annual stablecoin payment volume "up from about zero 2 years ago." Shopify live; merchants sharing saved card fees as 1% USDC cashback. Armstrong explicitly said Coinbase is *not* competing with Visa/Mastercard — decentralized protocols are.
- CaaS as AWS-for-crypto: 240+ institutions (up from 200+ last quarter), now including PNC and JPMorgan alongside BlackRock, Stripe, PayPal; 80%+ of crypto-ETF custody; 150+ government agencies. Fully white-labeled; monetization flows through existing lines (institutional volume, custody assets), not a separate P&L item — per Haas.
- US perps land-grab: launched ~2 weeks ago, volumes doubling week over week; Haas explicitly deferred monetization — "not focused in the near term on monetization and margin," focused on liquidity, open interest, share. Derivatives are 75% of the crypto market, 90%+ offshore — the US capture thesis.
- Base app as consumer super-app: beta with 700K waitlist; creator-monetization and ENS-based identity ("Base names") as the differentiator vs. ad-based Web2; 1-billion-people-onchain goal.
- Post-breach remediation: bringing BPO customer-service functions in-house, AI/automation push, new Charlotte NC onshore support office, $25M bounty for threat-actor information.
Key numbers and quarter mechanics- Volume vs. market gap explained: total volume -40% vs. spot markets -31/-32%; the delta is the intentional stablepair repricing (historically ~0-fee). Ex-stablepairs, volume tracked the market. This is a mix/price decision, not share loss — but it also means prior-period volumes were flattered by zero-revenue flow.
- Consumer: volume -45% to $43B, revenue -41% to $650M — revenue fell less than volume, consistent with the stablepair fee change and a shift toward simple (higher-fee) trading in low volatility.
- Institutional: volume and revenue both -38% ($194B / $61M). Critically, the guided $30–40M derivatives incentive drag did not hit transaction revenue as contra revenue — it was recorded as a transaction expense due to program design and client mix. The P&L line moved, the spend happened; watch where it sits going forward.
- S&S $656M: in the guided $600–680M range; unit growth across USDC, staking, custody, and ATH Prime financing loans was offset by ETH/SOL prices and reward rates — the guided headwind materialized.
- GAAP vs. adjusted divergence, inverted: Q1 had GAAP $66M vs. adjusted $527M (crypto losses); Q2 has GAAP $1.4B vs. adjusted $33M (Circle mark +$1.5B, crypto +$362M, breach -$307M). Adjusted net income of $33M on $1.5B revenue is the cleanest read on recurring profitability — and it is thin.
- Metric redefinition: adjusted net income now excludes strategic-investment gains/losses too — conveniently stripping the Circle gain, but also committing to stripping future Circle marks in both directions.
- Opex: $1.5B including $307M breach expense; ex-breach -9% QoQ. Q3 tech/G&A guided up to $800–850M with explicitly higher headcount growth — cost discipline is reversing into investment mode.
- July ~$360M transaction revenue vs. April ~$240M — a sharp rebound on higher prices/volatility; annualizes to ~$1.1B/quarter, still below Q1's $1.3B.
Product and launch scorecard- US perpetual-style futures (launched, ~2 weeks): first-of-kind CFTC-regulated retail perps in the US; volumes doubling week over week; ATH volume this week. Real but early; monetization explicitly deferred — no revenue contribution disclosed.
- Derivatives overall: $1T+ Q2 volume (vs. $800B+ in Q1), $1B open-interest ATH; 24/7 BTC/ETH futures live. Deribit still not closed — no update on timing, purchase accounting, or integration this call.
- DEX integration: rolling out "next week," eventually millions of tokens; monetized at the brokerage layer with fees "similar, if not even a little higher" than CEX, plus sequencer fees if on Base; Armstrong holds investments in some DEXs. Cannibalization question answered with a routing-agnostic framing — unproven.
- Stablecoin payment APIs: Shopify live in market; merchant savings (200–300bps card fees) partially recycled as 1% USDC cashback. Concrete first deployment; no volume or revenue figures.
- Base app: beta, 700K waitlist; creator economy + ENS identity. Pre-revenue; no usage metrics beyond the waitlist.
- CaaS: 240+ institutions (200+ last quarter); PNC and JPMorgan named as new; white-label, subscription-style, monetized through existing lines. Haas explicitly warned there will be no distinct P&L line — verification will be indirect.
- Tokenized equities: in development, broker integration possible as interim step; no timeline, no product. Armstrong: "Nobody has really launched tokenized securities yet."
- Coinbase Card / Coinbase Business: cited as "positive reception" — no metrics.
- USDC: network-effect and distribution-partner strategy reiterated; rewards will continue post-GENIUS Act (Coinbase is not the issuer; rewards ≠ interest). No updated on-platform balance or revenue split this call.
Sell-side read-through- Ken Worthington (JPMorgan, via email): payments ambition — alternative network vs. Coinbase-customer use cases, and monetization (fees vs. USDC economics). Got: "not competing with Visa/Mastercard," both first-party and infrastructure, monetization via payment fees + USDC balances + Base sequencer fees.
- Owen Lau (Oppenheimer): SEC Chair's Project Crypto single-license — Grewal gave a policy answer (lower compliance cost, rulemaking follow-through) but no commitment on whether/when Coinbase would obtain or need the license.
- Devin Ryan (Citizens JMP): everything-exchange revenue streams and tokenized-equities timing — Armstrong listed brokerage front-end, CEX/DEX routing, and primary capital-formation services (noting heavy inbound interest); timing answer was "coming quarters," no date.
- John Todaro (Needham): on-platform USDC upside from banks/neobanks/remittance integrations — Haas confirmed partners grow both on- and off-platform balances (50% off-platform share) and that rewards are the partner-acquisition tool.
- Ben Budish (Barclays): payments monetization — Armstrong: direct payment fees (vastly below incumbent rates), sequencer fees, USDC balances; costs should fall "an order of magnitude or two."
- Alex Markgraff (KeyBanc): PNC/JPMorgan economics and UX control — fully white-labeled, partner controls UX; Haas: no unique P&L line, shows up in institutional volume/custody. Ramp timing not answered.
- James Yaro (Goldman): GENIUS Act yield prohibition vs. rewards and tokenized money funds — Haas/Armstrong: rewards are loyalty programs, not issuer interest, and will continue; utility/network effect will determine winner vs. tokenized funds. The key regulatory-risk question got a confident, checkable answer.
- Patrick Moley (Piper Sandler): DEX economics and cannibalization — brokerage-layer fees similar or higher; cannibalization risk waved off via order-routing framing.
- Bo Pei (US Tiger): US perps traction and materiality — doubling weekly volumes, but Haas explicitly declined any monetization timeline: "in coming quarters, as we gain traction, we'll speak more about this."
- Pete Christiansen (Citi): customer service post-breach — Choi conceded BPO model weaknesses, in-housing, automation; Armstrong cited Charlotte office and $25M bounty. No quantified CSAT or cost impact beyond the $307M.
- Note: technical difficulties forced several questions to be read by IR; the call took fewer live analyst questions than typical.
Management credibility- Guidance delivered: S&S $656M landed inside $600–680M; the ETH/SOL headwind thesis played out as stated; the $30–40M derivatives incentive spend happened — though it surfaced as transaction expense rather than contra revenue, a classification shift from what was guided.
- Adjusted net income redefined again: Q1 introduced it (excluding crypto gains/losses, flattering a weak GAAP quarter); Q2 expanded it to exclude strategic investments, cutting a $1.4B GAAP profit to $33M. Symmetric treatment is defensible, but two redefinitions in two quarters, each conveniently timed, warrants skepticism about metric stability.
- Stablepair repricing disclosed after the fact: the March change explains the volume shortfall vs. market, but it was not flagged on the Q1 call — investors learned the volume-quality caveat only when volumes dropped 40%.
- Breach handling: $307M expense disclosed with ex-incident opex math provided — transparent on cost; Choi's BPO admission was candid. But the incident itself is a security-culture negative for a company whose core pitch is trust.
- Perp monetization deferral is now explicit and repeated — consistent with the multi-quarter "share first, monetize later" stance, but the unbooked-revenue narrative keeps extending.
- No Deribit closing update, no buyback disclosure, no Coinbase One subscriber count — three straight quarters of silence on the buyback; Deribit timing ("year-end") unaddressed on this call.
- "Weekly Bitcoin purchases" and a $1.8B crypto portfolio (up from $1.3B) — the pro-cyclical GAAP exposure is growing, not shrinking, despite the adjusted-metric reframing.
What changed versus the prior quarter- Cycle troughed and rebounded within the quarter: April ~$240M → July ~$360M monthly transaction revenue; Q2 volume -40% but Q3 guided up on higher prices/volatility.
- GAAP/adjusted relationship inverted: Q1 GAAP $66M / adjusted $527M → Q2 GAAP $1.4B / adjusted $33M, driven by the Circle IPO mark ($1.5B) — a new, permanently volatile P&L line tied to Circle's stock price.
- Derivatives incentives migrated from contra revenue to transaction expense — same spend, different line; institutional revenue comparability across quarters is now muddier.
- US perps launched — the Q1 "interim step" (BTC/ETH futures) became full perpetual-style futures for US retail, with early volume doubling weekly.
- GENIUS Act passed (Q1: "vote expected next week") and SEC Project Crypto announced — the regulatory catalysts from last quarter converted; rewards model explicitly defended as compliant.
- $307M data-breach expense — a new risk category realized; BPO customer-service model being unwound.
- Cost posture flipped: Q2 ex-breach opex -9%, but Q3 tech/G&A guided up to $800–850M with accelerating headcount — from discipline back to investment.
- CaaS momentum: 200+ → 240+ institutions, with PNC and JPMorgan added — bank-partnership thesis from Q1 now has named logos.
- Tokenized equities elevated to a stated product goal ("everything exchange"), with broker integration floated — a new TAM narrative not present in Q1.
Bull case- The trough is in: July transaction revenue ~$360M (+50% vs. April pace), Q3 S&S guided +8% at midpoint, and volatility/prices rising — Q2 looks like the cyclical bottom with operating leverage intact (ex-breach opex -9%).
- US perps are the largest untapped pool in crypto (75% of volume, 90%+ offshore) and Coinbase is first to market with a CFTC-regulated retail product already doubling weekly — plus Deribit (75% global options share) still to close.
- Regulatory catalysts delivered: GENIUS Act law, SEC Project Crypto single-license path, dismissed lawsuit — and Coinbase's rewards model was explicitly defended as compliant, de-risking the USDC economics.
- CaaS is compounding: 240+ institutions, PNC/JPMorgan logos, 80%+ ETF custody — white-label infrastructure revenue embedded in existing lines with an AWS-like expansion motion.
- Payments has a live flagship (Shopify) with a demonstrated value split (merchant savings → 1% USDC cashback) and a $40T B2B framing; the full stack (USDC + Base + APIs + apps) is genuinely differentiated.
- $9.3B USD resources, $1.8B crypto portfolio with a $1.5B Circle mark already realized on paper — balance-sheet optionality for the promised "lot more" M&A.
Bear case- Recurring profitability is razor-thin: adjusted net income of $33M on $1.5B revenue; the $1.4B GAAP headline is a Circle stock mark that will now whipsaw the P&L quarterly.
- Core trading shrank faster than the market (-40% vs. -31/32%), and the explanation (stablepair repricing) reveals prior volumes were inflated by zero-fee flow — revenue quality of past quarters is retrospectively weaker.
- Derivatives growth is still being bought, the cost now sits in transaction expense, and management again declined any monetization timeline — the unbooked-revenue promise is entering its fifth quarter.
- The $307M breach expense and the BPO failure cut against the core trust proposition; remediation (in-housing, automation) will cost money and CSAT risk is unresolved.
- Q3 opex guidance jumps (tech/G&A $800–850M vs. ~$1.2B actual ex-breach run-rate components) with accelerating headcount — cost growth into a revenue base that just printed $33M adjusted.
- Tokenized equities, Base app, and payments are all pre-revenue narratives with no timelines or metrics; the "everything exchange" is currently a vision statement.
- No Deribit closing update, no buyback execution for a third quarter, no Coinbase One count — capital-allocation and integration transparency remain gaps.
Next-quarter watchlist- Deribit: closing vs. year-end, purchase accounting, pro-forma contribution, US options path — no update this call makes Q3 disclosure overdue.
- US perps trajectory: whether weekly doubling persists off the tiny base; any first monetization commentary; open interest vs. the $1B ATH.
- Q3 S&S vs. $665–745M guide: stablecoin revenue must drive the +8% midpoint; watch USDC balances, rewards cost, and Circle-stock-driven noise in GAAP lines.
- July extrapolation: does ~$360M hold through August/September, and does volume again lag or lead the market ex-stablepairs?
- DEX integration rollout ("next week"): adoption, brokerage-layer take rate vs. CEX, and any evidence of cannibalization or incremental volume.
- Transaction-expense line: whether derivatives incentives recur there and at what magnitude — the new home for the share-buying spend.
- Opex and headcount: tech/G&A landing in $800–850M; pace of hiring; any residual breach costs or litigation beyond the $307M.
- Breach fallout: regulatory/legal developments, customer attrition signals, CSAT metrics, and whether the $25M bounty produces arrests.
- CaaS monetization evidence: since no P&L line will appear, watch institutional volume, custody assets, and any named-partner launches (PNC/JPMorgan live dates).
- Tokenized equities: any concrete milestone — broker partnership, regulatory filing, or launch window — vs. continued "coming quarters."
- Shopify payments: first volume or merchant-count data; additional payment-API partners.
- Buyback: a fourth quarter of silence would confirm the pattern; also watch RSU-withholding disclosure and weekly-BTC-purchase pace.
- Coinbase One and Base app: subscriber re-quantification; Base app public launch timing and waitlist conversion.
|
| May 8, 2025 | -3.48% | Q1 FY2025 | Read transcript briefingQuarter in one view- Q1 FY2025: total revenue ~$2.0B; adjusted EBITDA $930M; GAAP net income $66M; new metric "adjusted net income" $527M (GAAP net income excluding tax-adjusted crypto investment-portfolio gains/losses).
- Transaction revenue $1.3B, -19% QoQ. Consumer volume $78B (-17%), consumer transaction revenue -19% with similar mix; institutional volume $315B (-9%) but institutional transaction revenue -30% — the gap driven by derivatives rebates/incentives (contra revenue) and a spot mix shift toward market makers/liquidity providers.
- S&S hit an all-time high $698M (+9% QoQ), led by stablecoin revenue +32% QoQ to $298M — the Q4 decline reversed as guided. USDC market cap hit ATH $60B; average USDC in Coinbase products +49% QoQ to $12B; Base stablecoin balances $4B (+12%).
- Opex $1.3B (+7% QoQ) on higher variable expenses (market-maker activity) and losses on crypto assets held for operations.
- Headline event: Deribit acquisition announced same day — ~$2.9B ($700M cash + 11M Class A shares), the largest crypto acquisition ever; Deribit has 75% global crypto options share, $30B open interest, $1T 2024 volume, positive adjusted EBITDA; close expected by year-end.
- Q2 guide: April transaction revenue ~$240M; S&S $600–680M (down sequentially on ETH -36% / SOL -25% vs. Q1 averages); tech/dev + G&A $700–750M; S&M $215–315M; derivatives incentives to cost an incremental $30–40M QoQ in Q2.
What management is focused on- Deribit as the defining move: makes Coinbase the #1 crypto derivatives platform globally by open interest; spot + futures + options under one roof; cross-sell/hedging on one platform; accelerates the international exchange. Choi walked through the M&A track record (Xapo 2019, Tagomi 2020, FairX 2022, One River 2023) and promised "a lot more" — balance sheet deliberately held in cash for bigger bets.
- Derivatives share-buying strategy, now explicit and quantified: $800B+ Q1 derivatives volume; international perps market share +60%; 39 new perp books internationally, 4 in the US; rebates/incentives are contra revenue depressing institutional transaction revenue, with a guided $30–40M incremental Q2 drag. Monetization remains deferred.
- USDC economics transparency: new shareholder-letter table; 100% of reserve income on on-platform USDC, 50% of remaining off-platform economics; Binance added as a distribution partner (economics shared by agreement with Circle) to grow market cap. $298M revenue vs. ~$100M rewards paid; ~$26M net margin on-platform plus $171M full-margin off-platform revenue.
- Regulatory wins converting: SEC lawsuit dismissed; strategic Bitcoin reserve executive order; stablecoin bill vote expected "next week"; CFTC engagement on US perps (BTC/ETH futures trading launching "tomorrow" as an interim step).
- Emerging utility: business accounts with B2B stablecoin pay-ins/payouts (first pilot businesses onboarding in Q2); BTC-backed USDC lending on Base ($100M loans in first 100 days, now $160M); Base acquisitions Spindle (on-chain ads) and Iron Fish (private transactions).
- Competition framing: Armstrong's three-part moat answer (100% crypto focus, hard-to-replicate infrastructure, crypto-as-a-service powering new entrants — 200+ institutions including BlackRock, Stripe, PayPal). No bank charter plans — fully reserved model preferred, charter would slow product velocity.
Key numbers and quarter mechanics- Institutional revenue -30% vs. volume -9%: two stated drivers — derivatives incentives (contra revenue) and spot mix toward lower-fee market makers. This is the quarter's key negative mechanic and is deliberate spend, not pricing loss.
- Consumer take rate stable: volume -17%, revenue -19%, "similar mix" — no repeat of prior stable-pair distortions.
- Stablecoin revenue $298M (+32% QoQ) vs. Q4's -9% decline: the guided recovery delivered. New disclosure enables margin math: ~$100M rewards against on-platform revenue, $171M off-platform at full margin.
- USDC unit economics: MTUs holding USDC doubled over 2 years; average balance per holder tripled; on-platform balances +49% QoQ to $12B.
- Opex $1.3B (+7%): variable expenses from elevated market-maker activity early in the quarter plus operating crypto-asset losses — note the opex line now absorbs crypto losses, a new wrinkle.
- GAAP net income $66M vs. adjusted net income $527M — the ~$461M gap is crypto investment-portfolio losses, the pro-cyclical GAAP pattern swinging negative again. Portfolio: ~$150M new crypto bought in Q1 (mostly BTC); fair value ~$1.3B = 25% of net cash.
- Capital allocation: $700M cash to Deribit; ~$100M spent withholding ~390K RSUs (transcript says "390 million RSUs," likely a transcription artifact) to pay employee taxes — framed as dilution reduction; $1B buyback remains "opportunistic," no execution disclosed.
- April: ~$240M transaction revenue; Coinbase spot volume -12% MoM vs. global -13% — slight outperformance continuing.
Product and launch scorecard- Deribit (announced, not closed): 75% options share, $30B OI, $1T 2024 volume, consistently positive adjusted EBITDA; Haas says accretive to adjusted EBITDA but caveats purchase accounting. US availability requires CFTC work — Deribit is ex-US only; no timeline for US options.
- Derivatives (organic): $800B+ volume, perps share +60% internationally — first hard derivatives volume disclosure after four quarters of refusal, though still no take-rate or revenue split. Incentive spend is now quantified ($30–40M incremental Q2 drag).
- USDC: best quarter of disclosure ever — revenue split, rewards cost, margin structure all tabled. Binance partnership is concrete evidence of the "share economics to grow market cap" strategy; MiCA compliance cited as European edge; "fastest-growing major US-backed stablecoin" claimed.
- BTC-backed USDC lending: $100M loans in 100 days → $160M — real early traction, small base.
- Business accounts / B2B stablecoin payments: pilot onboarding in Q2 — pre-revenue, timing now dated.
- Base: $4B stablecoin balances (+12%); Spindle and Iron Fish acquisitions closed; still no sequencer revenue disclosure.
- Coinbase One: "continued to add subscribers" — no updated count beyond Q4's 600K+; contribution folded into S&S growth narrative.
- US perps: BTC/ETH futures launch "tomorrow" as an interim step; full perps "going to take a little bit longer" with the CFTC.
Sell-side read-through- Ken Worthington (JPMorgan): Deribit cross-sell and accretion — got "adjusted EBITDA accretive" with a purchase-accounting caveat; Armstrong added the single-platform hedging logic.
- James Yaro (Goldman): Binance joining the USDC partnership and the revenue split — extracted the full waterfall (Circle issuer fee → 100% on-platform to Coinbase → 50% of remaining off-platform) and confirmation the Circle agreement is filed with the 10-K. Most economically useful answer on USDC to date.
- Devin Ryan (Citizens): banks as partners vs. competitors — Armstrong: "every major bank will integrate crypto," some deals signed; crypto-as-a-service framing; Haas cited 200+ institutions on Coinbase rails.
- Owen Lau (Oppenheimer): 5–10 year identity — "number one financial services app in the world"; explicitly rejected looking backward to TradFi equities/bonds — tokenization of those assets onchain instead.
- Brett Knoblauch (Cantor): Circle change-of-control — Haas: contract persists through any acquisition, no change to economics. Clean, important answer given Circle M&A rumors.
- Pete Christiansen (Citi): bank charter speculation — Armstrong firmly shut it down: no current plans, fully reserved model preferred, charter would slow product velocity.
- Ben Budish (Barclays): why one options venue dominates — nascent product, Deribit's focus/execution; users are prosumers/advanced traders, mixed institutional and retail, all ex-US.
- Patrick Moley (Piper Sandler): US derivatives go-to-market post-Deribit — CFTC path is the gating item; near-term US options not offered; go-to-market is direct outreach to large-volume traders plus incentives.
- Joseph Vafi (Canaccord): USDC P&L bottom line — produced the $298M / ~$100M rewards / $26M + $171M margin breakdown.
- Dan Dolev (Mizuho): USDC rewards opportunity — flywheel answer (rewards → balances → engagement → trading), no incremental quantification.
- Alex Markgraff (KeyBanc): more large M&A — Choi: "expect a lot more from us," cash position deliberate, regulatory clarity enables larger swings. Payments follow-up was cut off by time.
- Bo Pei (US Tiger): bear-market flexibility — Haas: scenario-planned, expense growth set against bear cases; would take a bear market "plus something else really bad" to revisit spending.
Management credibility- Delivered on the Q4 guide's key claim: stablecoin revenue resumed growth (+32% vs. guided "higher stablecoin revenue"), and S&S hit the guided range's upper half ($698M vs. $685–765M). The offset thesis is now 2-for-4 but the latest print supports it.
- Disclosure improved materially: first derivatives volume number ($800B), full USDC economics table, rewards cost, adjusted net income metric. This addresses multiple quarters of withheld data — though the new metric also conveniently reframes a $66M GAAP quarter as $527M.
- Incentive-spend candor: management proactively quantified the derivatives revenue drag and guided the Q2 increment ($30–40M) — consistent with Q4's "fees will evolve" roadmap, but monetization is still deferred, now with a bigger stated cost.
- Buyback question answered but not resolved: retail question produced a capital-allocation philosophy answer (Deribit cash, RSU withholding) — no disclosure of actual buyback execution for the second straight quarter.
- Circle contract durability stated unambiguously under M&A-rumor pressure — checkable.
- Adjusted net income introduction in a quarter where GAAP was depressed by crypto losses is optically convenient; the metric is legitimate but the timing favors management's narrative.
- "No meaningful impact" February claim from last call is consistent with Q1's $1.3B transaction revenue print — validated.
What changed versus the prior quarter- Cycle turned down: volume, transaction revenue, and GAAP net income all fell QoQ; crypto portfolio swung from +$476M pretax gains to losses driving $66M GAAP vs. $527M adjusted. Q2 guided softer still (April $240M annualizes well below Q1's monthly pace).
- M&A went from "pipeline" talk to the largest crypto deal ever — Deribit ($2.9B) announced on call day, plus Spindle and Iron Fish closed in Q1.
- Stablecoin mechanics flipped again: Q4's -9% became +32% with full economics disclosed for the first time, including the Binance distribution partnership — the "new ecosystem participants" drag from Q4 is now formalized as strategy.
- Derivatives moved from narrative to numbers: first volume disclosure ($800B), perps share +60%, and an explicit quantified incentive drag — the monetization deferral now has a visible P&L cost.
- SEC lawsuit dismissed — the biggest outstanding legal overhang from prior quarters is gone.
- New profitability metric (adjusted net income) introduced; opex now includes crypto operating-asset losses.
- Bank charter speculation killed — a strategic option explicitly taken off the table.
- S&M range narrowed ($215–315M vs. $235–375M) and tech/G&A guided down ($700–750M vs. $750–800M) — modest cost discipline signal into a softer tape.
Bull case- S&S at an all-time high ($698M) in a down-volume quarter, with stablecoin revenue +32% and fully disclosed positive margins — the diversification engine held up exactly when transaction revenue fell 19%.
- Deribit instantly makes Coinbase the global #1 derivatives platform by open interest with 75% options share, positive EBITDA, and accretion expected — a durable, cycle-tested revenue stream added to the mix, with cross-sell into Coinbase's institutional base.
- Derivatives share-buying is working: perps share +60%, $800B volume — and management has a stated path to monetize later, an unbooked margin lever.
- USDC flywheel has real numbers now: balances +49% QoQ, holders doubled, balances per holder tripled, $171M of full-margin off-platform revenue, Binance distribution, MiCA compliance moat in Europe.
- Regulatory overhang cleared: SEC suit dismissed, stablecoin bill advancing, CFTC engaged on perps — each unlock (US perps, options, tokenized assets) is additive TAM.
- Balance sheet deployed productively: $2.9B deal funded with $700M cash + stock, $150M added to the crypto portfolio, RSU withholding reducing dilution — and Choi signals more deals coming.
Bear case- The core trading business is shrinking again: transaction revenue -19% QoQ with April at ~$240M (a ~$720M quarterly run-rate vs. Q1's $1.3B) and Q2 S&S guided down $18–98M sequentially on ETH/SOL price declines.
- Institutional revenue quality deteriorated: -30% on -9% volume, with management committing to *more* incentive spend ($30–40M incremental) — derivatives growth is being bought, and the monetization date remains unspecified.
- GAAP earnings remain hostage to crypto marks: $66M net income, and the response was to introduce a new adjusted metric rather than change the exposure; the portfolio is now 25% of net cash and still growing.
- Deribit is ex-US only with no US regulatory path timeline; $2.9B (including 11M new shares) is being paid before purchase accounting is complete, and integration of an offshore options venue carries execution risk.
- USDC growth is increasingly paid for: ~$100M quarterly rewards, economics shared with Binance and future partners — on-platform net margin is thin ($26M) relative to the headline revenue.
- Opex still rose 7% QoQ in a down quarter, partly on crypto operating losses — cost discipline is guided for Q2 but not yet demonstrated.
Next-quarter watchlist- Deribit: closing progress vs. year-end target, purchase accounting, any pro-forma revenue/EBITDA disclosure, integration plan, and CFTC path for US options/perps.
- Derivatives incentive drag: whether the $30–40M Q2 impact lands as guided and whether perps share gains persist; any take-rate evolution commentary.
- Q2 S&S vs. $600–680M guide: stablecoin growth must offset ETH/SOL-driven blockchain-rewards declines — the stated mechanics are testable.
- USDC economics: rewards vs. balance growth, Binance partnership's effect on off-platform revenue share, market-cap trajectory vs. the $60B ATH.
- April extrapolation: whether ~$240M monthly transaction revenue was a trough or the trend; May/June volume vs. global spot.
- Stablecoin legislation: the "vote next week" Armstrong cited — passage, terms (state pathway, non-bank issuers), and any impact on USDC economics.
- Business-account pilot: first B2B payments customers onboarded in Q2 — count, volume, any revenue signal.
- BTC-backed lending: growth from $160M; any credit-loss or risk disclosure.
- Buyback: third quarter of silence would be a pattern — watch for any repurchase disclosure in the Q2 letter.
- Coinbase One: subscriber count update beyond 600K was absent this quarter — watch for re-quantification.
- Opex discipline: tech/G&A landing in $700–750M and S&M in $215–315M as guided, and whether crypto operating losses recur in the opex line.
|
| Feb 13, 2025 | -7.98% | Q4 FY2024 | Read transcript briefingQuarter in one view- Q4 FY2024: total trading volume $439B (+137% QoQ from $185B), consumer volume $94B (+176%), institutional volume $345B (+128%). Consumer transaction revenue $1.3B (+179%); institutional transaction revenue $141M (+156%). MTUs +24% to 9.7M; nearly half of Q4 trading customers were new or resurrected (dormant 1+ year).
- S&S $641M, +15% QoQ — the Q3 decline reversed, driven by higher asset prices, USDC market cap, and native-unit inflows. But stablecoin revenue declined $21M (-9% QoQ): lower rates plus "new USDC ecosystem participants" more than offset market-cap and on-platform balance growth. Other S&S +$33M (+56%), largely Coinbase One.
- Adjusted EBITDA $1.3B; net income $1.3B, which included $476M pretax ($357M after-tax) crypto investment-portfolio gains, mostly unrealized — the pro-cyclical GAAP pattern now swings positive.
- Opex $1.2B (+19% QoQ), driven by transaction expenses on higher volume; tech/dev + G&A + S&M up $84M (+10%) on performance marketing, USDC rewards, and policy spend. USD resources $9.3B (from $8.2B).
- FY2024: revenue $6.6B (more than doubled), adjusted EBITDA $3.3B, S&S $2.3B (+64% YoY) — the $2B S&S pace claim from Q3 was delivered. International revenue share reached 19% in Q4.
- Q1 2025 guide: ~$750M transaction revenue already generated YTD; S&S $685–765M (sequential growth); tech/dev + G&A $750–800M; S&M $235–375M — an unusually wide range.
What management is focused on- Post-election "golden age" framing: Armstrong calls it "the dawn of a new era" — pro-crypto president and Congress, stablecoin and market-structure legislation in motion, SAB 121 rescinded (SAB 122 early-adopted), SEC crypto task force under Commissioner Peirce. Regulatory clarity is repeatedly equated with TAM expansion.
- 2025 three-prong plan: (1) grow revenue in core businesses (derivatives, accelerated asset listings, USDC, international, Coinbase One, custody); (2) drive utility (stablecoin payments integrated across the product suite, Base as #1 chain for builders, CDP); (3) scale foundations (policy — Stand With Crypto to 4M advocates by 2026 midterms, further Fairshake donations; back-end re-architecture for higher volume orders of magnitude).
- USDC #1 stretch goal: explicitly framed as multi-year ("over the next few years" per Haas) — a deliberate softening of an aggressive-sounding headline.
- New opportunity set unlocked by regulation: US perpetual futures (bring offshore volume onshore), tokenized securities/equities (24/7 trading, fractional shares, real-time settlement), prediction markets (interested, nothing to announce), and a revived M&A pipeline ("M&A is legal again," said jokingly).
- Platform-of-platforms ambition: "primary financial account" aspiration spanning brokerage, banking, and payments; Coinbase as infrastructure provider (CDP, Prime APIs) for banks, fintechs, and Fortune 500s entering crypto.
Key numbers and quarter mechanics- Volume +137% QoQ vs. consumer transaction revenue +179% — the Q3 mechanic (fee-less stable-pair mix deflating take rate) reversed: Haas says consumer volume mix was "very similar" to Q3 and simple/advanced grew proportionally, so the retail fee rate ticked up slightly sequentially. Bo Pei's question confirms the rate rose despite volume nearly tripling — no pricing change claimed, just mix normalization.
- Macro backdrop: average crypto market cap +33%, volatility +27% in Q4; US spot market +126% — Coinbase outperformed it on both consumer and institutional volume, hence the all-time-high US spot and global derivatives share claims.
- Stablecoin revenue -9% QoQ despite USDC market cap and on-platform balance growth — the Q3 offset thesis (native units outgrowing rate drag) failed this quarter; Haas attributes it to lower rates plus new USDC ecosystem participants (i.e., revenue sharing). This is the quarter's most important negative mechanic.
- Assets on platform reinstated as key metric post-SAB 122: $404B, ~12% of total crypto market cap. Disclosure location change only; no operational/legal change.
- Prime financing hit all-time-high loan balances; elevated trading among financing clients.
- 13 new assets listed including meme coins; Armstrong cites ~1M tokens created per week industry-wide as rationale for DEX integration into the core product.
- Marketing payback discipline restated: one-year CAC payback target; Q1 S&M range width reflects volatile week-to-week performance-marketing opportunities post-election.
Product and launch scorecard- Derivatives: all-time-high global derivatives market share claimed (unquantified). Haas gave the clearest monetization roadmap yet: currently subsidizing liquidity via incentives, fees "will evolve and become more mature" as scale builds — an explicit commitment that take rates rise from today's levels. US perps now framed as a realistic regulatory unlock. Still no notional/volume disclosure — fourth straight quarter.
- Coinbase One: finally quantified — exceeded 600,000 paid members (announced early December), with "strong growth since." Drove most of the +56% other-S&S growth. International expansion is the growth engine; users described as monthly-active traders with deeper multi-product engagement. Two quarters of withheld subscriber counts resolved.
- USDC: market cap and on-platform balances "grew substantially" but revenue fell 9% — growth is being bought via rewards and shared with new ecosystem partners. #1 stablecoin goal is now a stated multi-year stretch target with partnership levers (Stripe, Yellowcard named).
- Base: still "number one chain" narrative; Q4-specific metrics absent from this call (no transaction growth or fee figures given). New strategic role: settlement layer for stablecoin payments and tokenized assets; continued fee reductions planned.
- Listings/DEX integration: Armstrong's X post operationalized — plan to deeply integrate DEXs so customers trade long-tail tokens without knowing the venue, balanced with disclosures/consumer protection. A material product-direction signal given meme-coin-driven volume.
- Prediction markets: elevated from passing mention to repeated strategic interest ("better source of truth than traditional media"), but explicitly nothing to announce; Choi frames fast-follower posture as deliberate.
- CDP/Prime APIs: Fortune 500 adoption path sketched (treasury BTC allocation → B2B/cross-border payments → on-chain rewards/governance); still no CDP revenue or usage metrics.
Sell-side read-through- Owen Lau (Oppenheimer): share gains vs. low-cost competitors — Armstrong/Haas attribute to trust, asset additions, UX, stability, marketing; Haas concedes week-to-month share wobble from missing pairs/pricing but claims long-term share gains. No structural moat evidence beyond brand asserted.
- Devin Ryan (Citizens JMP): derivatives take-rate trajectory — extracted the call's most economically significant answer: current fees are deliberately below market to build liquidity; monetization will rise with scale. A future margin lever is now on the record.
- Ben Budish (Barclays): new/resurrected cohort quality — Haas declined cohort detail; described two waves (new listings vs. first-time BTC/ETH buyers) and reiterated one-year CAC payback. Coinbase One users trade monthly. Cohort economics remain undisclosed.
- Ken Worthington (JPMorgan): legislative priorities — Armstrong's checklist: market structure/token classification, stablecoin bill (state pathway, non-bank issuers, 100% reserve audits), strategic Bitcoin reserve, fair banking access, self-custody rights. Grewal cites the executive order and Peirce task force as foundations.
- Pete Christiansen (Citi): crypto-native vs. TradFi value proposition — Choi's "digital native vs. digital feature" (Amazon analogy) is the clearest articulation of the competitive moat claim; still assertion, not data.
- Patrick Moley (Piper Sandler): listing-process rethink — produced the DEX-integration disclosure; consumer-protection balancing acknowledged but mechanics unspecified.
- John Todaro (Needham): stablecoins + tokenized RWAs on Base — Armstrong fully endorsed the vision; named partnership levers for USDC growth.
- Dan Dolev (Mizuho): prediction markets — interested, watching, nothing to announce; Choi's fast-follower framing covers the gap.
- Bo Pei (US Tiger): retail fee-rate uptick — answered cleanly (mix stable, broad-based growth), a reversal of prior quarters' evasions; competition question (Robinhood's 400% crypto volume growth implied) deflected with TAM-expansion framing, no direct share-defense answer.
- Alexander Markgraff (KBCM): wide S&M range — explained as market-condition-driven flexibility, not a forecasting change; CDP non-crypto-native adoption — roadmap narrative only.
Management credibility- Delivered on the big dated claims: S&S exceeded $2B in 2024 ($2.3B, +64%); Q3's guided headwinds played out and reversed; all-time-high share claims coincide with volume outperformance vs. the US spot market (+126% market vs. +176% consumer volume).
- The stablecoin offset thesis broke: Q3's "native units outgrow rate cuts" mechanism failed in Q4 (stablecoin revenue -9% despite balance growth) — and management disclosed it plainly with a new driver (ecosystem participant sharing). Candor is intact, but the thesis the bull case rested on is now empirically weaker.
- Disclosure pattern improved at the margin: Coinbase One subscribers finally numbered (600K+); derivatives monetization path articulated. Still withheld: derivatives notional (four quarters), Base sequencer dollars, stable-pair volume share, cohort economics, CDP metrics.
- Take-rate consistency: third straight quarter of "no pricing change, it's mix" — this quarter the mix explanation was simpler and survived direct questioning; the Q3 stable-pair distortion did normalize as Haas predicted ("not structural" call validated).
- Guidance style shift: the $235–375M S&M range is unusually wide; management explained it transparently, but it reduces near-term forecastability and embeds an option to spend heavily into strength.
- USDC #1 goal immediately hedged: Armstrong's "stretch goal" was qualified by Haas to "over the next few years" within the same call — ambition walked back to plausible timeline in real time.
- February liquidation question dismissed: Haas laughed off the early-February liquidation event as ordinary volatility with "no meaningful impact" — a checkable claim against Q1 results, given ~$750M transaction revenue was already booked YTD at call date.
What changed versus the prior quarter- Regime change is the story: Q3 was pre-election optionality (Grewal's hypothetical unlock list); Q4 is post-election execution — SAB 121 rescinded, SEC task force active, perps/tokenized securities now plausible US products, legislation in motion.
- S&S reversed: $556M → $641M (+15%), but the internal composition flipped — Q3's stablecoin growth (+3%) became a -9% decline while Coinbase One became the growth driver. The diversification engine rotated.
- Take-rate mechanic inverted: Q3's fee-less stable-pair drag disappeared; retail fee rate rose slightly on tripled volume with stable mix. Haas's "not structural" call from Q3 proved correct.
- Coinbase One went from unquantified "all-time high" to 600K+ paid members — resolving a two-quarter disclosure gap.
- Derivatives narrative advanced from adoption metrics to monetization roadmap: Q3 gave 100K CFM users; Q4 gives all-time-high global share and an explicit future take-rate increase commitment. US perps moved from "blocked" to "hopeful path."
- Balance sheet and GAAP swung with crypto: USD resources $8.2B → $9.3B; portfolio marks flipped from -$121M pretax to +$476M pretax. No buyback execution discussed on this call despite the Q3 $1B authorization.
- Policy spend continued post-election as promised: Fairshake donations extended, 4M-advocate goal reiterated — the "not going to slow down" commitment held.
- New disclosure architecture: SAB 122 adoption; assets on platform ($404B, ~12% of crypto market cap) reinstated as the headline custody metric.
Bull case- Volume +137% QoQ with outperformance vs. the market on both consumer and institutional sides, all-time-high US spot and global derivatives share, and 9.7M MTUs (+24%) with half of traders new or resurrected — the acquisition and re-engagement engine works in up markets.
- FY2024 proof points delivered: revenue doubled to $6.6B, $3.3B adjusted EBITDA, S&S $2.3B (+64%) — the diversification thesis now has a full-year record, and Q1 is off to a $750M-transaction-revenue start with S&S guided up again.
- Regulatory unlocks are converting from list to action: SAB 121 gone, perps and tokenized securities plausible in the US, stablecoin and market-structure legislation moving — each is a concrete TAM addition, and management's access to decision-makers is new.
- Derivatives has a stated margin lever: today's subsidized fees "will evolve" upward with scale, on top of all-time-high share — revenue upside not yet in the numbers.
- Coinbase One at 600K+ and growing, driving +56% other-S&S — a recurring-revenue line with international runway and quantified traction at last.
- $9.3B USD resources, $404B assets on platform (~12% of crypto market cap), buyback authorization in place, and a revived M&A pipeline — balance-sheet optionality at cycle highs.
Bear case- Stablecoin revenue fell 9% despite USDC market-cap and balance growth — rates plus ecosystem revenue-sharing are overwhelming unit growth, and Q1 S&M guidance embeds even higher USDC rewards. The USDC flywheel's economics are deteriorating at the margin even as the #1 goal implies more spending.
- Earnings quality: $1.3B net income includes $476M pretax mostly-unrealized crypto gains; the GAAP result is as pro-cyclical as ever, now with a larger balance sheet amplifying it.
- Opex +19% QoQ and Q1 tech/dev + G&A guided to $750–800M (vs. Q4's implied lower level within the $1.2B total) with a $235–375M S&M range — cost growth is re-accelerating into the upcycle, and the wide S&M range signals spend-to-volume pro-cyclicality.
- Share and fee-rate strength are cycle-conditioned: management concedes share wobbles when it lacks hot pairs, the DEX-integration plan is an admission that centralized listing can't keep up with ~1M tokens/week, and the Robinhood competition question was deflected rather than answered.
- Core disclosures still missing after four quarters: derivatives notional, Base sequencer revenue, stable-pair volume share, cohort economics — the biggest claimed growth engines remain unmodelable from company data.
- Prediction markets, tokenized securities, US perps are all pre-product; the "golden age" narrative is running well ahead of shipped functionality, and legislative timing is outside management's control.
Next-quarter watchlist- Q1 S&S vs. the $685–765M guide: specifically whether stablecoin revenue resumes growth as guided ("higher stablecoin revenue" is the stated driver) or the Q4 decline extends — the offset thesis is now 1-for-3 and needs a positive print.
- Derivatives monetization: any evidence of fee evolution off subsidized levels; US perp regulatory progress; whether global share holds; any notional disclosure after four quarters of refusal.
- Retail take rate: whether the slight sequential uptick holds or meme-coin/DEX-integrated long-tail mix re-dilutes it; watch for stable-pair share creep returning.
- S&M landing spot in the $235–375M range: where spend falls vs. transaction revenue, and whether the one-year CAC payback claim survives a heavy-spend quarter.
- USDC economics: rewards cost growth vs. balance growth; ecosystem-partner revenue-sharing terms; any market-cap share data vs. Tether supporting the #1 stretch goal.
- Legislative scorecard: stablecoin bill and market-structure/token-classification progress vs. Armstrong's checklist; SEC task force guidance; state staking restoration.
- DEX integration and listings: cadence of asset additions, consumer-protection mechanics, and any cannibalization of centralized-exchange fees.
- Buyback execution: whether the $1B authorization was used in Q4/Q1 — it went unmentioned on this call.
- February volatility impact: Haas's "no meaningful impact" claim is testable against full Q1 results given the early-February liquidation event.
- Coinbase One: subscriber trajectory beyond 600K and international mix; whether it keeps carrying other-S&S growth.
- M&A: Armstrong flagged an "incredible pipeline" and Choi's framework (international, Base, payments) — watch for a first post-election deal.
|
| Oct 30, 2024 | -15.34% | Q3 FY2024 | Read transcript briefingQuarter in one view- Q3 FY2024: total revenue $1.2B (down from $1.4B in Q2), adjusted EBITDA $449M (seventh consecutive positive quarter), net income $75M (fourth consecutive positive quarter). Trading volume $185B, -18% QoQ; transaction revenue $573M, -27% QoQ; S&S $556M, -7% QoQ — the first sequential S&S decline after the Q2 record.
- Net income absorbed $121M pretax crypto investment-portfolio losses (~$92M after tax, mostly unrealized) — a smaller mark than Q2's $319M but the same pro-cyclical GAAP pattern.
- Opex $1.0B, -6% QoQ, within guided ranges. USD resources $8.2B (+5% QoQ); Board authorized the first-ever buyback, up to $1B with no expiration.
- S&S is on pace to exceed $2B in 2024 (vs. $1.4B in 2023) — management's headline diversification proof point despite the QoQ decline.
- Q4 guide: S&S headwinds (ETH -10% in October vs. Q3 average, lower rates); tech/dev + G&A $690–730M; S&M $170–220M on higher USDC balances and brand spend.
What management is focused on- Election and post-election policy continuity: both candidates courting crypto voters; 350+ federal candidates with A/B Stand With Crypto grades; Stand With Crypto now 1.8M advocates (from 1.3M in Q2) with a 4M stretch goal by 2026 midterms; another $25M committed to Fairshake explicitly for the 2026 midterms, not this cycle. Armstrong: "we're not going to slow down post election."
- Payments as the underappreciated utility narrative: stablecoin transaction volume ~$10T in 2023, already 2x'd to $20T+ in 2024; Armstrong's long-term framing is crypto rails growing from ~0.5% of global GDP toward 20% (his own "tough to estimate" internal guess).
- USDC as the trusted-stablecoin wedge: market cap +45% YTD to $36B, claimed fastest-growing major USD stablecoin; EURC launched on Base and doubled its market cap in Q3 to become the largest euro-backed stablecoin.
- Capital allocation maturation: $8.2B USD resources + $1.3B crypto investment portfolio (~25% of net cash); buyback authorization framed as opportunistic; explicit rejection of the MicroStrategy playbook ("we are an operating company, not an investment company").
- M&A posture: Emilie Choi cited the MiFID license acquisition (closed August), Station Labs (June, smart wallet), and historical deals (Xapo, Tagomi, FairX); focus areas are international consolidation, Base tech/talent, and payments/stablecoin utility. Armstrong's "build, buy, invest" with Buffett-style selectivity.
Key numbers and quarter mechanics- Volume $185B (-18% QoQ) vs. transaction revenue $573M (-27% QoQ) — the gap is the quarter's key mechanic, and Haas pre-explained it: (1) US fiat-to-crypto share steady QoQ; (2) consumer stablecoin-pair volume grew significantly (aided by a product update easing stable-pair trading for advanced traders) and carries little-to-no fees; (3) nontrading consumer transaction revenue (DEX, minor fees) declined.
- Retail take rate: Haas stated no material fee-structure changes; blended decline is mix (stable pairs) plus lower nontrading revenue. Under Dolev-style pressure from Bo Pei, she added: excluding stables, advanced mix was slightly higher QoQ; stablecoin mix shift is not structural ("no guarantee ARPU opportunities exist every quarter") but refused to quantify stable-pair volume share.
- S&S $556M (-7% QoQ): native units grew in staking and custody but were offset by lower average crypto prices. Stablecoin revenue +3% QoQ — USDC market cap and on-platform balance growth exceeded the rate-cut drag, i.e., the native-unit offset thesis worked this quarter. Coinbase One hit all-time-high paid subscribers (still no number).
- Institutional: spot revenue declined, but prime broker and derivatives relatively outperformed; derivatives revenue growth has begun but is "not yet material."
- Opex $1.0B (-6% QoQ); Q4 guides: tech/dev + G&A $690–730M, S&M $170–220M (USDC rewards + brand spend).
- Tether reached 15% of platform trading volume; Haas attributed it to the stable-pair product change and denied any USDC-to-Tether shift.
Product and launch scorecard- Base: now claimed #1 L2 by transactions and TVL; transactions +55% QoQ (decelerating from Q2's +300% but off a larger base) with median fee held below $0.01. Sequencer fees acknowledged as the revenue mechanism but deliberately kept low; still no dollar figure. New launches: Base names (onchain identity) and cbBTC (wrapped BTC on Ethereum/Base for DeFi borrow/lend, monetization described only as "potential opportunities").
- Smart Wallet: first concrete adoption metric — time-to-first-transaction fell from 2.5 hours to ~8 minutes in integrated products; not yet integrated across the full product family. Station Labs acquisition (June) disclosed as an accelerator.
- USDC/EURC: $36B market cap (+45% YTD, all-time high post-Q1 2023 crisis); EURC doubled to largest euro stablecoin. Stablecoin revenue +3% QoQ despite rate cuts — the strongest evidence yet for the unit-growth offset.
- Derivatives: MiFID license closed (August), operationalization underway for 20+ EU markets; 100K+ retail advanced traders onboarded to Coinbase Financial Markets (first hard adoption number for US futures); referral program paying ~30% of trading-fee revenue driving volume; unified margin improvements; perps still blocked in the US. Armstrong: 2025 "will be a really pivotal year." Still no notional/volume disclosure — third straight quarter.
- Coinbase One: all-time-high subscribers claimed, unquantified for a second consecutive quarter.
- CDP: AWS-of-crypto ambition restated; new evidence is a surge of AI-agent developers embedding crypto wallets via CDP. No revenue or usage metrics.
Sell-side read-through- Pete Christiansen (Citi): long-tail altcoin share down ~10 points from Q4/Q1 — Haas attributed it to lower volatility reducing market-maker/hedge-fund long-tail activity and ETF-driven concentration in BTC/ETH. Framed as cyclical, not competitive loss.
- Owen Lau (Oppenheimer): retail fee-rate drivers — got the cleanest confirmation that core per-product fees are unchanged; decline is stable-pair mix plus nontrading revenue.
- Ken Worthington (JPMorgan): what does a friendlier federal environment unlock bottoms-up, and does it defuse state-level restrictions? Grewal gave the most concrete regulatory-unlock list to date: accelerated listings, crypto-securities trading under license, stablecoin payments at scale, restoring paused state staking, banking/payment on-ramp partners, and creator/social platform integrations; states "take their cue" from federal regulators.
- John Todaro (Needham): Tether at 15% of volume — is share shifting from USDC? Haas: no; it's the stable-pair product change, and USDC remains the fastest-growing major stablecoin. Notably, USDC volume is not broken out the same way, so the rebuttal rests on market-cap data, not platform flow data.
- Devin Ryan (Citizens JMP): 2025 expense framework — Haas declined any 2025 outlook; repeated "prudent," selective headcount, variable spend flexed to volumes. No expense-growth-vs-revenue-growth relationship offered.
- Ben Budish (Barclays): MicroStrategy-style balance-sheet play — Haas rejected it directly; crypto portfolio capped by operating cash needs at ~25% of net cash.
- Mike Colonnese (H.C. Wainwright): product gaps and build-vs-buy — Choi's answer signals international consolidation and Base/payments deals are the active hunting ground; Armstrong cautioned on large-M&A integration risk.
- Joseph Vafi (Canaccord): whether Base will use its capital advantage to outspend rival L2s — Armstrong pushed back on aggressive economics-sharing/token incentives ("right reasons" framing); Haas added the flywheel is Coinbase's full product-suite integration, not grants.
- Bo Pei (US Tiger): pressed for stable-pair volume quantification and whether the mix shift is structural — refused on the number; "not structural" on the mix. The most evaded quantitative question of the call.
- Mark McLaughlin (BofA): derivatives adoption path — produced the call's best derivatives disclosure (100K CFM users, referral economics, whale outreach, unified margin) and Armstrong's "pivotal 2025" commitment, which is now a checkable promise.
Management credibility- Guidance delivery held again: Q3 opex within the ranges given in Q2; the Q2-flagged S&S headwinds (ETH price, rate cuts, USDC costs, one-time reward roll-off) did produce the predicted S&S decline — management called the slowdown accurately rather than being surprised by it.
- The rate-cut offset thesis now has one data point: stablecoin revenue +3% QoQ despite lower rates, exactly the native-unit mechanism Haas described last quarter. One quarter is evidence, not proof.
- Consistent take-rate candor: for a second straight quarter Haas pre-explained a blended take-rate distortion before analysts could misread it, and the explanation (mix, not pricing) survived direct questioning from Lau and Pei. The corollary: management again confirmed no consumer pricing-power claim.
- Disclosure pattern persists: new voluntary metrics appeared where the story is good (100K CFM users, 8-minute onboarding, EURC doubling, $1.3B crypto portfolio) while requested quantifications were withheld (stable-pair volume share, Coinbase One subscribers, Base sequencer dollars, derivatives notional — third quarter running, ETF economics — second quarter).
- Buyback vs. policy spend optics: $1B repurchase authorization and a fresh $25M Fairshake commitment announced on the same call; management frames both as high-ROI, but the buyback is discretionary with no expiration and no execution commitment.
- Armstrong's "pivotal 2025" for derivatives and the $2B S&S pace claim are now explicit, dated commitments against which delivery can be scored.
What changed versus the prior quarter- S&S streak broke as guided: $599M record → $556M (-7%), the first sequential decline, driven by price effects despite native-unit growth — the Q2 headwind list materialized on schedule.
- Take-rate driver rotated: Q2's distortion was volume-less revenue streams (derivatives, wallet fees) inflating the rate; Q3's was fee-less stable-pair volume deflating it. Both times management held the "no pricing change" line.
- Capital return entered the picture: first-ever $1B buyback authorization; Q2's earmarked 2026 convert payoff is now described as completed opportunistic repurchases. Balance sheet $7.8B → $8.2B.
- Policy posture shifted from election-cycle to permanent: Fairshake commitment extended to 2026 midterms (+$25M), Stand With Crypto goal raised to 4M — policy spend is now a multi-cycle structural commitment, consistent with Q2's G&A reclassification signal.
- Derivatives got its first traction metrics: 100K+ US futures users and referral-program economics, versus Q2's "immaterial, no disclosure." MiFID moved from "close expected 2024" to closed and operationalizing.
- Base growth normalized: +55% QoQ transactions vs. +300% in Q2; narrative shifted from fee-cut-driven share grab to ecosystem hub (Optimism stack, interoperability) and new products (names, cbBTC).
- Smart Wallet went from launched to measured: first adoption KPI (time-to-first-transaction 2.5 hours → 8 minutes).
- Stablecoin debate changed direction: Q2's question was whether MiCA would migrate Tether share to USDC; Q3's was whether Tether's 15% platform volume share signals the reverse — management says no.
Bull case- The diversification thesis survived its second test: volume -18% and transaction revenue -27% QoQ, yet adjusted EBITDA stayed at $449M, net income stayed positive, and S&S is on pace for $2B+ in 2024 vs. $1.4B in 2023.
- The rate-cut offset worked in real time: stablecoin revenue grew 3% QoQ through falling rates on USDC market cap (+45% YTD to $36B) and on-platform balance growth — the core bear mechanism from prior quarters failed to bite this quarter.
- Derivatives finally has traction evidence (100K+ US futures users, MiFID closed for 20+ EU markets, referral program driving volume) with management committing to 2025 as the pivotal year — the 75%-of-volume opportunity now has a dated scorecard.
- Regulatory optionality is unusually concrete: Grewal enumerated specific unlocks (listings acceleration, crypto-securities, state staking restoration, payment partnerships) and both-candidate upside means the tailwind is not election-outcome-dependent.
- Capital return begins: $8.2B USD resources, $1B buyback authorized, debt repurchases done — balance-sheet strength converting into shareholder-relevant action.
- Smart Wallet's 8-minute onboarding and Base's #1 L2 position with sub-$0.01 fees give the utility narrative its first measurable consumer-friction evidence, plus an emerging AI-agent developer wedge on CDP.
Bear case- Transaction revenue fell faster than volume (-27% vs. -18%) because consumers migrated toward fee-less stable-pair trading — a mix headwind management enabled via its own product update and cannot quantify or call structural either way.
- S&S declined despite native-unit growth, and Q4's headwind list is steeper (ETH -10% in October vs. Q3 average, more rate cuts) — the offset mechanism worked at Q3's rate levels but is untested against a deeper cutting cycle.
- USDC rewards and brand spend keep scaling (Q4 S&M guide $170–220M), meaning the stablecoin growth engine carries a rising, balance-linked cost just as rates compress its revenue per dollar.
- Core monetization questions remain unanswered after repeated asks: no stable-pair volume share, no Base sequencer dollars, no derivatives notional (three quarters), no Coinbase One subscriber count (two quarters), no ETF economics (two quarters).
- Derivatives traction metrics, while new, are adoption counts, not revenue — Haas still says "not yet material," and the referral program pays away ~30% of trading-fee revenue to buy flow in a whale-driven market.
- GAAP earnings remain hostage to crypto marks ($121M pretax loss this quarter), and the $1.3B investment portfolio (~25% of net cash) amplifies drawdown correlation between results and the balance sheet.
- Policy ROI is asserted, not measurable: another $25M to Fairshake plus ongoing Stand With Crypto support with no framework for tying spend to legislative outcomes; FIT21 remains unpassed in the Senate.
Next-quarter watchlist- Q4 S&S vs. the steeper headwind list: does the native-unit offset hold against ETH -10% and further cuts, or does S&S decline accelerate? Watch stablecoin revenue specifically as the offset's proving ground.
- Derivatives delivery against "pivotal 2025": MiFID operationalization progress in 20+ EU markets, CFM user growth beyond 100K, any notional/volume disclosure resumption, and whether referral-driven flow persists without the ~30% fee giveaway.
- Retail take rate and stable-pair mix: whether the fee-less stable-pair share normalizes as Haas implied (non-structural) or persists; any quantification of stable-pair volume; nontrading consumer revenue (DEX/minor fees) recovery.
- Buyback execution: whether the $1B authorization is used opportunistically or sits idle; cadence and price discipline; interaction with the $25M Fairshake commitment and policy spend trajectory post-election.
- Election aftermath: FIT21 Senate movement, SEC case posture under new leadership, state staking restoration — Grewal's unlock list is now a checklist.
- Base monetization: any sequencer revenue figure; cbBTC adoption and the "potential" monetization path; whether +55% transaction growth sustains; EURC trajectory.
- S&M discipline: where Q4 S&M lands in the $170–220M range; whether USDC rewards growth tracks balance growth; payback guardrails on brand spend.
- M&A: Choi flagged international consolidation, Base tech/talent, and payments/stablecoin deals — watch for a first transaction under the stated framework.
- Coinbase One: whether the all-time-high subscriber claim ever gets a number.
- GAAP optics: Q4 crypto marks on the $1.3B portfolio and whether adjusted EBITDA remains the market's anchor.
|
Corbanu