Event Contracts Overtake Crypto and the Expense Guide Comes Down: Two Bear Legs Break, But July Halves the Deposit Run-Rate

Results vs. Consensus

Consensus figures carry a wider spread than usual this quarter. The revenue number most widely cited into the print was $1.29B, which is where the beat computes to about +1.4%; one major provider carried a materially lower base and framed the same result as a +3.5% beat. On EPS, three of the four compilations we cross-referenced sat at $0.43 and one at $0.44. We use the $1.29B and $0.43 pairing throughout.

Year-over-Year Comparison

Sequential Comparison

Revenue. A 32% year-over-year increase to $1,308M more than doubles the 15% growth rate printed in Q1, and it is the composition that matters more than the level. Transaction-based revenue grew 44% while net interest, the line that carried the franchise through the crypto downturn, grew only 9%. Two quarters ago the story was that net interest and subscriptions provided a floor under a declining transaction book. That has inverted: transaction revenue is now the growth engine again, and it is a different transaction book than the one that existed a year ago. Options at $342M and equities at $129M are volume stories, with 774M contracts traded (+50%) and $956B of equity notional (+85%), and the per-unit economics held roughly flat sequentially at about 44 cents per options contract and 1.3 basis points on equity notional. Nothing here is take-rate financial engineering.

Margins. The 57% Adjusted EBITDA margin is seven points above Q1 and a point above the year-ago quarter, reversing the compression that defined the March period. It was produced by two things working at once: 23% growth in Adjusted OpEx and SBC against 32% revenue growth, and a marketing line that grew only 5% year-over-year while the company added nearly a million funded customers. That second point deserves emphasis, because it is the opposite of what a customer-acquisition surge normally looks like. Management attributed the funnel reacceleration to product breadth rather than spend, and the marketing line supports that claim. The offsetting pressure is in the expense lines that scale with balance-sheet risk and headcount: operations grew 97%, general and administrative grew 51%, and provision for credit losses doubled.

EPS. The gap between the reported $0.62 and the underlying $0.48 is the most important single number in the release, and it is disclosed plainly rather than buried. Robinhood Ventures Fund I listed in Q1, and the Q2 deconsolidation produced a $106M gain plus $23M of realized and unrealized equity-securities gains. This is a legitimate outcome of a real strategic build, giving retail investors access to private markets, and Fund II is already on the roadshow. It is not, however, a run-rate earnings stream, and the consensus that this quarter is measured against was not modeling it. On the tax line, the 19.2% effective rate compares to 12.7% a year ago, which is the more durable drag on the year-over-year EPS bridge.

Revenue Line Performance

Event Contracts: The Line That Reframes the Story

Event contracts generated $156M of revenue on 13.6B contracts traded, both records, and passed cryptocurrencies to become the second-largest transaction line. A year ago this business was rounding error. Two quarters ago it was reported inside a residual "other transaction" line. It now carries nearly 12% of company revenue and is the reason total revenue accelerated to 32% growth while the crypto book fell by a quarter sequentially.

The volume figure implies roughly 1.15 cents of revenue per contract, which is worth watching as the mix shifts toward higher-frequency sports events. Management framed durability around event supply rather than any single catalyst.

"I think the great thing about prediction markets is there's events all the time. You mentioned football season. That's coming up. There's also the midterms, which I think are extremely important. They're going to be a topic of discussion, of course, customers are going to want to trade them and hedge their portfolios."
— Vlad Tenev, Chairman and CEO

The user base is scaling alongside the volume: management disclosed that cumulative prediction-market users moved from about 1.5 million to closer to two million. Against 28.4 million funded customers, penetration is still under 10%, which is the bull argument for durability. The bear argument is that the World Cup quarter is the easiest comparison this business will ever face, and the football-and-midterms bridge has to be built at a higher volume base.

Assessment: This is the most consequential development in the coverage since Trump Accounts, and it is more immediately monetizable. The concentration risk simply moved rather than disappeared: an $156M line exposed to a binary regulatory outcome is a larger exposure than the $147M version was in Q1, and the company's own risk factors continue to name event contracts specifically as a business that adverse regulatory or legislative action could prevent it from offering.

Cryptocurrencies: The Comp Improved, the Business Did Not

Crypto revenue of $100M was down 38% year-over-year, an apparent improvement on Q1's -47% and Q4 2025's -52%. That reading is wrong. The year-ago quarters get progressively easier through 2025, so the year-over-year line flatters. Sequentially, crypto revenue fell 25%, from $134M to $100M, and Robinhood-app notional volumes fell 25%, from $24B to $18B. Those two numbers matching is the tell: this is volume, not pricing.

Pricing, in fact, improved. Total crypto notional of $40B against $100M of revenue implies roughly 25 basis points, up from about 20 basis points in Q1. The improvement is a mix artifact: Bitstamp notional fell from $42B to $22B, so the lower-take-rate institutional book shrank faster than the retail app, lifting the blended rate. That is not a durable margin story.

In our Q1 note we set two consecutive quarters of sequential stabilization as the primary upgrade trigger. The first of those two quarters has now failed, and management's July commentary points the same direction.

Assessment: Crypto is now 7.6% of revenue, down from 12.6% one quarter ago. The right conclusion is not that the crypto problem is fixed; it is that the crypto problem has been outgrown. That materially reduces its power to determine the thesis, which is why the bear point is diluted rather than confirmed. But nothing in this print suggests a bottom, and the model should continue to assume flat-to-down.

Options and Equities: The Core Franchise at Record Engagement

Options revenue of $342M on 774M contracts and equities revenue of $129M on $956B of notional are both records, and both were achieved with stable unit economics. Equities revenue grew 95% against 85% notional growth, so the take rate improved slightly. Options revenue grew 29% against 50% contract growth, so per-contract revenue fell about 14% year-over-year, though it was essentially unchanged sequentially at roughly 44 cents.

June's elimination of the pattern-day-trading rule is a structural tailwind that is only partially in these numbers. Management characterized the benefit as industry-wide but disproportionate given the customer base.

"It was a nice tailwind in June when the rule went through. What we're seeing is it's broader across the whole industry, regardless of what brokerage. We probably had an outsized benefit relative to our peers, given our customers have smaller balances."
— Shiv Verma, CFO

Assessment: The active-trader pillar is performing better than at any point in the coverage. One month of a rule change and a strong tape are both in these figures, so treating Q2 volume as the new baseline would be a mistake. Management's own July guidance, that average daily volumes are running in a similar area to the Q2 average, is the more useful anchor and implies a flat rather than compounding volume line into Q3.

Net Interest: The Quality Compounder Loses a Step

Net interest revenue of $389M grew 9% year-over-year, down sharply from 24% in Q1. The underlying dynamics remain constructive on the asset side, with Margin Book up 127% to a record $21.6B and Cash and Deposits up 34% to a record $18.7B. The drag is on rates and on the funding mix: Cash Sweep fell 9% year-over-year to $29.7B, partly the deliberate February repositioning that moved over $6B of sweep balances onto the balance sheet, and short-term rates continue to fall.

Sequentially the line grew 8%, which is respectable. The concern is what sits alongside it. Provision for credit losses of $56M doubled year-over-year and grew 56% sequentially, reaching 4.3% of revenue against 2.8% a year ago. A margin book that has more than doubled will generate more provisioning arithmetically, but the provision is growing faster than the book on a sequential basis.

Assessment: We flagged credit losses as an early watch item in Q1 at $36M. At $56M it is no longer early. This does not change the thesis today, and Robinhood's margin lending is collateralized against liquid securities, but the combination of a rapidly growing on-balance-sheet lending book, a decelerating spread, and a provision line compounding faster than either deserves a specific place on the risk list rather than a footnote.

Key Operating Metrics

The funded-customer reacceleration is the metric the Street reacted to most positively, and it warrants a composition check. Management named four drivers: a strong market backdrop, new products including banking and the credit card, the SpaceX listing, and an acquisition that contributed a couple hundred thousand accounts. Two of those four are episodic. Backing out the acquisition alone puts organic adds closer to 800 thousand, still the best result in years, but the run-rate claim is weaker than the headline.

Key Topics & Management Commentary

Overall Management Tone: Markedly more confident than the April call, and the first since then where management was answering questions about how fast to grow rather than defending a decelerating line. Confidence was earned on the reported quarter and thinner on the forward setup: the expense-reduction answer was specific and quantified, while the question about July net deposits stepping down was answered by redirecting to the annual average rather than addressing the month. The venue itself has changed, with a live audience mixing sell-side analysts, retail shareholders, and content creators, which produced a wider range of question quality and gave management more room to spend airtime on product demonstrations than on the financial model.

1. Prediction Markets Pass Crypto and Rothera Goes Live

The strategic pivot articulated in Q1 as optionality became operating reality in Q2. Rothera, the CFTC-licensed exchange and clearinghouse held through a joint venture with Susquehanna International Group, launched in June and has cleared over 3.5 billion contracts. Management says it reached top-three status among US designated contract markets within a month, which converts Robinhood from pure distribution into a participant in the exchange economics of its own fastest-growing product.

"Obviously, we're top three after one month, the goal would be to keep growing that, right? We're not satisfied just with where we are after just a couple of months."
— Vlad Tenev, Chairman and CEO

The flow-migration plan was described in directional rather than numeric terms, and the venue strategy remains deliberately plural, with connections to multiple counterparties maintained alongside the captive venue.

"In terms of the flow, what we've shared is you should expect in the near to medium term more than the majority or a good portion will flow through Rothera. We're still making sure it can scale, and after one month, it's already doing tremendous volume."
— Shiv Verma, CFO

Assessment: Vertical integration in the fastest-growing line is a structural margin lever that consensus is unlikely to be modeling, because the economics still have not been quantified two quarters after the joint venture was announced. The B2B extension management floated, onboarding third-party futures commission merchants onto Rothera, is a second unmodeled option. Both remain qualitative, which is the recurring pattern with this asset.

2. The Expense Guide Comes Down, and How It Was Funded

Three months ago Robinhood raised its FY26 Adjusted OpEx and SBC outlook by $100M to build the Trump Accounts experience, and we treated that raise as one of the three legs of the downgrade to Hold. This quarter the range came down to $2.675B–$2.775B, and the new range absorbs Rothera and WonderFi costs that the prior range explicitly excluded. The midpoint fell roughly $38M while the scope expanded by two businesses, so the like-for-like reduction is larger than the headline.

"We're lowering and tightening our 2026 outlook for adjusted OpEx and SBC to a range of $2.675 billion-$2.775 billion. We believe it is a competitive advantage to not only be a growth company that can invest for the long term, but also leverage our lean and disciplined operating model to self-fund a meaningful amount of these new investments."
— Shiv Verma, CFO

Part of the funding is visible in the income statement rather than the narrative: a reduction in force announced in June produced $23M of restructuring charges plus $7M of associated share-based compensation. Neither the headcount reduction nor the affected functions were disclosed, and no question on the call addressed it.

Assessment: This is the cleanest reversal of a bear point in the coverage, and it is worth more than its dollar value because it re-establishes management's credibility on expense control after a quarter in which that credibility was the specific casualty. The qualifier is that a growth company printing record revenue does not usually need a reduction in force to fund its guide-down, and the absence of any detail on it is a gap.

3. July Net Deposits and the Metric Management Chose to Be Judged On

Management named three measures it wants investors to track, and put net deposits first. That makes the July disclosure the most important forward-looking statement on the call, and it is not a good one.

"July average daily volumes, compared to a record Q2, are in a similar area for equities, options, and event contracts. July net deposits are tracking towards the $4 billion area, and this does not yet include deposits into the Trump Accounts."
— Shiv Verma, CFO

Q2 delivered $21.7B across three months, a $7.2B monthly pace and a 28% annualized rate against prior-period platform assets. July at $4B annualizes to roughly 13% against the $369B closing base, which is below the 20% long-term target and roughly 45% below the pace just reported. Trump Account deposits are excluded and will help, though the disclosed $1.5B of cumulative contributions since the July 4 launch is not enough to close the gap on its own.

Pressed directly on the step-down, management declined to engage with the month.

"Our goal, as we've said, is 20% on an annualized basis. I don't look at any particular quarter and month. There will be things that change from time to time. If you look at year to date, including July, we're still well north of that 20%."
— Shiv Verma, CFO

Assessment: The answer is defensible and the seasonality point is real, but it is the only place on the call where management redirected rather than answered, and it happens to be the metric they asked to be measured against. Volumes and take rates are flat to Q2 and crypto is slower, so Q3 revenue rests on whether this deposit pace was a July artifact. This is now the binding constraint on the rating, replacing crypto.

4. Trump Accounts Launches and Starts Generating Revenue

The program went live on July 4 and has drawn over seven million account sign-ups with nearly $1.5B contributed, against 5.5 million pre-launch sign-ups disclosed in Q1. More consequentially for the model, the release attributes part of the 54% growth in other revenues to Trump Account service revenues, which means the line is monetizing in its first partial quarter rather than in 2027 as the Q1 framing implied.

"Seven million children have signed up. We've seen over $1.5 billion in, or nearly $1.5 billion, I should say, is the number of contributions into Trump Accounts already. Again, that's before a lot of the philanthropic contributions have started flowing in."
— Vlad Tenev, Chairman and CEO

The roadmap extends to employer funding, where a $2,500 annual tax-free employer limit exists, and to positioning the account as a charitable-giving vehicle. Robinhood built and runs the standalone app, provides support and educational content, and custodies assets alongside BNY.

Assessment: Faster than we modeled on both sign-ups and revenue recognition, and the strategic case is unchanged: this is customer acquisition at age zero with a federal brand association, and the value is in the 2030s rather than in 2026 estimates. The economics remain cost-plus at a thin margin by management's own prior description, so revenue growth here should not be read as margin growth.

5. Wallet Share: The Card Crosses a Million and Banking Crosses $3B

The Gold Card passed one million cardholders driving over $17B in annualized purchase volume, and the credit card business crossed $100M in annualized revenue. Banking deposits passed $3B from over 240 thousand funded customers with approximately 40% signed up for direct deposit, against $2B and 125 thousand customers a quarter ago. Gold subscribers reached 4.8 million at a 17% adoption rate, with roughly 40% of new funded customers attaching Gold in the quarter. Strategies passed 300 thousand funded customers and nearly $2B in assets, and TradePMR's newly launched advisor network reached $50B in assets under management.

"We get a customer in, they become a Gold subscriber. After they become a Gold subscriber, they look out at all the different products we offer. If those products are competitive with what's on the market or better, they'll adopt them."
— Vlad Tenev, Chairman and CEO

The cross-adoption evidence management offered was directional rather than cohort-level, though the specific data point is the useful kind.

"One thing we looked at, if you're a prediction market customer, for example, you're actually more likely to have a retirement account with Robinhood. Any vector you come in, whether it's banking, prediction markets, credit card, you tend to adopt others."
— Shiv Verma, CFO

Assessment: ARPU growing 24% against 7% funded-customer growth is the cleanest quantitative proof that the wallet-share thesis is working, and it is the metric to watch rather than any single product milestone. Banking direct-deposit attach holding near 40% as deposits scaled 50% sequentially clears the attach half of the test we set in Q1; deposits at $3B have not reached the $5B level that test also required. The pillar is intact and strengthening.

6. Top of Funnel Reaccelerates, With Caveats

Robinhood added nearly a million funded customers, the most in a quarter since the period around its 2021 listing, after several quarters in which top-of-funnel growth was the acknowledged weak point. Management described the drivers as diversified rather than singular.

"A really strong market backdrop, that definitely helped. We had a lot of new products that came out, banking, credit card, for example. Those contributed strong organic growth. We also had the SpaceX IPO, which helped. We had an acquisition in the quarter, which provided a couple hundred thousand accounts, and we continue to grow overseas."
— Shiv Verma, CFO

Marketing spend grew only 5% year-over-year while this happened, which is the strongest evidence that product breadth rather than paid acquisition drove the result. Management committed to funnel growth as a top KPI going forward.

Assessment: Genuinely good, and cheaper than expected. But an acquisition and a high-profile listing are inside the number, and at least one sell-side desk has flagged that new-account growth slowed meaningfully in July. Combined with the July deposit figure, the most likely read is that Q2 caught several tailwinds at once and Q3 catches fewer.

7. Robinhood Chain and the First Monetization Disclosure

Robinhood Chain's public mainnet launched at the company's London event with more than $12B in decentralized-exchange volume after launch, and management described it as the fastest chain to reach 100 million transactions. Stock Tokens are now available in more than 120 countries through the Robinhood Wallet, Robinhood Earn has taken over $200M of deposits at a 7% advertised yield, and perpetual futures went live in the EU.

The new information this quarter is economic rather than promotional. For the first time, management quantified how the chain makes money.

"On the monetization piece, per transaction, we make a few basis points. Not per volume, it's per transaction. Now it varies depending on the size of the transaction but just think about it as a few basis points on average. We do share approximately half of it with Arbitrum, who's the level two that we built on top of."
— Shiv Verma, CFO

Assessment: A few basis points per transaction, with half shared, on a volume base in the low tens of billions, is not a 2026 or 2027 revenue line, and management's commitment to break it out only "when it starts to get larger" is the correct posture. What matters is the optionality: tokenized real-world assets distributed to anyone with an internet connection is the largest addressable idea in the company, and it now has infrastructure, a wallet, developer traction, and a disclosed fee mechanism rather than a slide.

8. Agentic Trading: Early, Honest About the Friction

Agentic Trading launched in May and has drawn nearly 100 thousand accounts with over $100M in assets under custody, with equities and options tradable and crypto coming. The candid part was the friction discussion.

"In terms of friction, I think one piece of friction has been that not everyone loves, surprisingly, going to a Codex or a Claude code and kind of stitching together these two apps. You have to have quite a bit of sophistication in order to do that. In many cases, the models themselves aren't familiar with being used for trading, so sometimes they'll fight you and they'll say, 'Well, I don't know, I don't really want to trade,' and you kind of have to work hard to get it to do what you want to do."
— Vlad Tenev, Chairman and CEO

Assessment: $100M of assets across 100 thousand accounts is an average of $1,000 per account, which is exactly what a curiosity-stage product looks like. Being first matters more than being large here, and the honest description of why adoption is capped is more useful than a projection would have been. Not a 2026 revenue driver.

9. Capital Structure: A $2.2B Convertible and What It Costs

Robinhood raised $2.2B of convertible senior notes in June, and the balance sheet now carries $2,170M of long-term borrowings against none at year-end. Management characterized the terms as shareholder-friendly, and on coupon they are.

"In June, we opportunistically raised $2.2 billion of capital to give us even more flexibility to invest for future growth. We believe we have a massive opportunity ahead of us, and the capital gives us even more capacity to go after it. We raised this capital at attractive terms for shareholders, with both a 0% coupon and no net dilution until our share price exceeds $300."
— Shiv Verma, CFO

The company also purchased $123M of capped calls and repurchased $414M of stock at roughly $94 per share. Of that, $290M was executed in connection with the convertible offering and outside the existing authorization, so open-market repurchase against the standing program was closer to $124M. Year to date, 7.5 million shares have been retired for $664M. Diluted share count of 912 million is up 0.3% year-over-year and down 3 million sequentially.

Assessment: Zero coupon with a conversion reference roughly 3.5x the current share price is a good outcome, and the capped calls push the effective dilution point higher still. The nuance worth holding onto is that the quarter's buyback optics are stronger than the underlying capital return: most of it was the standard offsetting purchase that accompanies a convertible issue, not incremental capital return.

10. International: Licences, an Acquisition, and a Sequencing Change

International funded customers passed one million. Robinhood closed the WonderFi acquisition, entering Canada, and Robinhood Singapore received its capital markets services licence from the Monetary Authority of Singapore on July 1. Management flagged a deliberate change in how it launches new geographies, learned from the UK.

"I think what we've learned from the U.K. is it's better to get more at once rather than just launching equities, following up with options, then launching margin. I think what we're trying to do in subsequent launches is get more of the overall Robinhood ecosystem live at launch."
— Vlad Tenev, Chairman and CEO

Singapore is intended as the Southeast Asia and APAC hub with passporting into the region, and Indonesia is expected to get direct licensure.

Assessment: The sequencing change is the substantive disclosure and it is the right lesson, at the cost of slower time-to-first-revenue in each new market. International remains a multi-year build with roughly 3.5% of funded customers outside the US. Not a 2026 or 2027 model input of any size.

11. Framing the Long Term: Rule of 80 and the Trillion-Dollar Question

Management offered three yardsticks for the long-term story: net deposit growth above 20%, a Rule of 40 measure on which the company claims to have exceeded 80 for several years, and the count of business lines above $100M in annualized revenue, now 13 after Robinhood Legend and the credit card joined this quarter. Asked directly whether the company could reach a trillion-dollar market capitalization, the answer was unusually measured for the format.

"I think getting to a trillion will be very, very difficult. I don't think a financial company has ever hit a trillion market cap. I think it can be done."
— Vlad Tenev, Chairman and CEO

The CFO reframed it away from the share price and onto assets, arguing that adjacent markets are larger rather than smaller than the core.

"We don't comment on the stock price specifically, but what we've said before is we believe we can 10x the business over the next 10 years, that's through a lot of the vectors Vlad shared. When you look at assets are the greatest predictor of what's going to happen in financial performance. We're super excited. We're nearly $400 billion today."
— Shiv Verma, CFO

Assessment: The 13-business-lines metric is the most useful of the three because it is objective and it captures exactly what has changed about this company: diversification is no longer a promise. The Rule of 80 framing is accurate on the reported quarter (32% growth plus a 57% margin) but flatters, because it uses an adjusted margin that excludes credit losses and restructuring against a GAAP revenue line.

Guidance & Outlook

Robinhood does not guide revenue or EPS, so the expense range plus the intra-quarter operating color is the whole of the outlook. The expense revision is unambiguously positive and is the second revision to this range in two quarters, this time in the shareholder-friendly direction. One accounting change is worth noting for modelling continuity: beginning in Q3 2026, management intends to exclude interest expense classified as operating expenses from both Adjusted OpEx and Adjusted OpEx and SBC, which will make the next print's adjusted expense figures slightly flattering on a like-for-like basis.

Implied setup for Q3: flat volumes, flat take rates, slower crypto, and net deposits running near half the Q2 monthly pace. On those inputs, sequential revenue growth close to zero is the reasonable base case, with upside from a full quarter of Rothera flow migration, the first full quarter of Trump Account service revenue, and continued Gold and card scaling.

Street at: consensus sits near $0.49 for Q3 and $1.95 for FY26 revenue-adjusted earnings. Against first-half reported EPS of $1.00, of which $0.14 was the one-time gain, the FY26 number requires roughly $0.95 across the back half versus a clean $0.86 in the first half. That is achievable but is not the layup the headline beat implies, and it embeds a Q4 that is below Q3.

Guidance style: conservative on expenses and deliberately absent on revenue. Management's willingness to disclose an unflattering July deposit figure unprompted, in the same release that cut the expense range, is a point in favour of the disclosure culture even as it creates the quarter's main overhang.

Analyst Q&A Highlights

Durability of the funded-customer reacceleration

The most positively received datapoint in the print drew the first question, and the answer disaggregated it more honestly than the headline did. Management named four contributors, two of which are episodic rather than structural, and committed to treating top-of-funnel growth as a standing KPI rather than a recovered problem.

Q: "I think what surprised us the most on the positive side was the amazing growth in the funded accounts. Maybe you can shed some light on what part of that is structural, what are you doing, and maybe a quick update on the progress in Europe and how that is going."
— Dan Dolev, Mizuho

A: "We said on the last quarter, we were making a concerted effort to regrow top of funnel. As you mentioned, really great to see the progress. A million funded customers, the most we've added in a quarter in nearly five years since the IPO. There was a variety of different vectors, so it wasn't any one thing. A really strong market backdrop, that definitely helped. We had a lot of new products that came out, banking, credit card, for example. Those contributed strong organic growth. We also had the SpaceX IPO, which helped. We had an acquisition in the quarter, which provided a couple hundred thousand accounts, and we continue to grow overseas."
— Shiv Verma, CFO

Assessment: A candid answer that costs management something, since the cleanest version of the story would have left the acquisition and the listing out. Roughly 20% of the adds are inorganic and an unquantified further share is tied to a single high-profile listing. The organic core is still the best in years, but this is not yet a run-rate.

Composition of the reduced expense outlook

The most substantive exchange of the call. The question set up the original three-bucket construction of the 18% expense-growth plan and asked which buckets absorbed the reduction. The answer was that savings came from all three while investment in new initiatives actually increased, and that the two new businesses were self-funded rather than added to the range.

Q: "I wanted to ask on the updated expense guidance. At the start of the year, you guided to 18% expense growth, of which 10% or a bit more than half of the growth was really earmarked to support new product launches, 5% to support the core business, 3% for acquisitions. There have been a lot of moving pieces underpinning the new expense guidance, I just wanted to better understand how the buckets have evolved."
— Steven Chubak, Wolfe

A: "The short answer is the savings are coming from all of them. A little bit's coming from the M&A. We're being a little bit more efficient. The new M&A actually wasn't even included in those. We fully self-funded those. On the core business, the teams are really, really working hard to make sure they're delivering fast, but also doing it efficiently. We had some savings there."
— Shiv Verma, CFO

Assessment: The claim that spending on new initiatives rose while the total range fell is the part that matters, and it is consistent with the reported Adjusted OpEx and SBC growth of 23% against 32% revenue growth. What went unmentioned in the answer is the June reduction in force, which is visible in the $23M restructuring charge and is presumably part of how the core-business savings were achieved.

The July net-deposit step-down

The sharpest question of the call, and the only one management did not engage with on its own terms. The questioner identified that the disclosed monthly figure would be the lowest of the year and asked directly for commentary. The response reframed to the annual target and the year-to-date average.

Q: "If I could just ask a question about the quarter. You had a disclosure in the deck showing that you're trending toward roughly $4 billion in net deposits. That's down quite a bit from June. It's the lowest of the year. I was wondering if there's any commentary around that. I also want to ask about the health of investors considering the drawdown in semis and what's been happening in the market."
— Craig Maurer, FT Partners

A: "Our goal, as we've said, is 20% on an annualized basis. I don't look at any particular quarter and month. There will be things that change from time to time. If you look at year to date, including July, we're still well north of that 20%. As you mentioned, we had a really strong Q2. Great to see the engagement. It will fluctuate from time to time, but I don't read anything into it. Net deposits have a variety of factors. It's what's going on in the macro. It's when do we launch new products. There is some seasonality. Summer tends to be a little bit slower."
— Shiv Verma, CFO

Assessment: The seasonality argument is legitimate and the annual framing is the right way to run the business. It is still a redirection. Management chose net deposits as its first long-term yardstick earlier on the same call, which makes "I don't read anything into it" a difficult position to hold for a metric running 45% below the pace just reported. This exchange is the single best explanation for why a beat-and-raise quarter traded down.

July trajectory by product and the crypto omission

A follow-up that caught what the prepared remarks had left out. The July colour named equities, options, and event contracts but not crypto, and the question forced the missing item onto the record along with take-rate direction.

Q: "I was wondering if you could unpack the July commentary a little bit more. You said ADV was trending similar to Q2. I guess on a product-by-product basis, equities, options, should we assume that that's equities similar to Q2, options similar to Q2? Any color you could share on the take rate. I know particularly for options, that one bounces around a little bit. Crypto, you didn't include in that commentary."
— Ben Budish, Barclays

A: "First, the way you read it was correct. For July, the ADVs for equities, options, and prediction markets are all in the similar area of Q2. Those are seeing pretty healthy engagement, all of those. In terms of crypto, it's probably a little bit slower than what we saw in Q2, it's early in the month, that's what we're seeing to start. For take rates, for July, they're in a similar area to what we saw for the Q2 average. I would use that as kind of your starting point."
— Shiv Verma, CFO

Assessment: The most model-relevant answer on the call, and it confirms that the crypto omission was deliberate rather than accidental. Flat volumes and flat take rates across three of four asset classes with the fourth declining sets up a Q3 with little sequential revenue momentum from the trading book, which places the burden on Rothera migration, Trump Accounts, and subscription growth.

Why not launch perpetual futures in the US

A pointed challenge on competitive positioning, framed around the company's own prior advocacy and the fact that rivals are moving. The response declined the premise without supplying a timeline, and the reason given was product-roadmap crowding rather than regulatory caution.

Q: "I'm just a little confused why you all are not being more aggressive with launching perps in the U.S. I think a little over a year ago, you submitted a comment letter supporting CFTC allowing perps in the U.S. They've opened the door now on a number of products. You have a few of your competitors who are either live or they're moving in that direction, and what they're really trying to do is become more of a multi-asset class platform so that they can compete more closely with you. Could you just elaborate on what the source of hesitation is with perps in the U.S.? Are you worried about leverage?"
— Patrick Moley, Piper Sandler

A: "No, I don't think it's any of those things. I think we're, in some ways, in the fortunate position of having lots of great products that we see a huge opportunity in, right? The roadmap, just a couple of weeks ago, we launched 13 new products that we're in the process of rolling out. We have perps live overseas as well, both through our DeFi offerings and through centralized. I wouldn't take that to mean that we don't like perps. In fact, we like them quite a bit, and we intend to offer them to our customers. I didn't give you a timeline, but I wouldn't take the absence of a timeline at earnings to mean that we don't intend to be aggressive. I just don't like giving timelines and earnings. That's what our events are for."
— Vlad Tenev, Chairman and CEO

Assessment: A non-answer delivered with enough specificity to be credible. The substantive information is that regulatory conversations are described as constructive and the engineering work is characterized as small, which suggests the constraint is prioritization rather than capability. For a company whose crypto line is falling 25% sequentially, a higher-take-rate crypto derivative that is already live in Europe sitting behind other roadmap items is a genuine allocation question.

Robinhood Chain economics

The first quantification of how the blockchain earns money, offered in response to a question that paired customer-engagement strategy with monetization mechanics. The answer distinguished per-transaction from per-volume economics, which is the distinction that determines how the line scales.

Q: "Maybe one for both Vlad and Shiv on Robinhood Chain and the DeFi product expansion. I'm curious how you think, Vlad, about how the effort to engage with customers and the touchpoints change in DeFi products. Shiv, just on a related note, would be curious if there's anything you can share from a monetization standpoint on how we should be thinking about those relationships versus the sort of CeFi relationships that exist today."
— Alex Markgraff, KeyBanc

A: "On the monetization piece, per transaction, we make a few basis points. Not per volume, it's per transaction. Now it varies depending on the size of the transaction but just think about it as a few basis points on average. We do share approximately half of it with Arbitrum, who's the level two that we built on top of. When it starts to get larger and it goes through a couple of quarters, we'll break it down more specifically but think of it as a few basis points on transactions than a 50/50 rev share."
— Shiv Verma, CFO

Assessment: Useful and appropriately deflating. Per-transaction rather than per-volume economics means the line scales with transaction count, not with notional, so headline volume statistics overstate the revenue relevance. Half the economics going to the underlying layer-two provider halves it again. This is a strategic asset and an option on tokenization, not a near-term revenue line, and management said so.

Rothera's product roadmap and flow migration

A two-part question on whether the joint venture would pursue company-financial contracts in response to a competitor filing, and on how quickly existing Robinhood event-contract flow moves onto the captive venue. The answer separated the two: the product question was deferred on regulatory-regime grounds, the migration question was answered directionally.

Q: "My question's on Rothera, with specifically the company financial KPI contracts. With Cboe filing with the SEC to launch these potentially in the near term, what's the interest in Rothera launching these in the near term? Then also just in terms of migrating more of the event contracts at Hood now to Rothera, such as NFL, how are you thinking about that game plan for the rest of the year?"
— Brian Bedell, Deutsche Bank

A: "On the Cboe financial KPI contract, my understanding is those are under the securities-based regime. I don't know if Rothera would be launching those right now, but obviously, Robinhood is able to connect to a wide variety of counterparties. We have a brokerage business. We also have our FCM business, and of course, Rothera is a great joint venture that we have with SIG. That doesn't necessarily mean that we're not going to make all these products available to customers."
— Vlad Tenev, Chairman and CEO

Assessment: The distribution-agnostic posture is strategically sound and consistent with the multi-venue approach maintained since the first third-party connections. The unanswered question, for the second consecutive quarter, is what the economics actually are when flow routes through the captive venue versus a third party. Until that is quantified, the vertical-integration story cannot be modelled, only believed.

What They're NOT Saying

1. No Rothera economics, two quarters running: the joint venture is described as strategically important and is taking a majority of flow, but the take-rate differential between routing to Rothera and routing to a third-party venue has never been quantified. This is the largest unmodelled item in the story and the omission is now conspicuous.
2. No detail on the June reduction in force: a $23M restructuring charge plus $7M of related share-based compensation appears in the reconciliation, and the expense guide came down partly because of it. No headcount figure, no affected functions, and no question on the call.
3. No crypto bottom, and a deliberate omission from the July colour: the prepared July commentary named equities, options, and event contracts, and left crypto out until an analyst asked. The answer, "a little bit slower," is the third consecutive quarter without a stabilization signal.
4. No Trump Accounts revenue sizing: service revenue is now flowing and is cited as a driver of the 54% growth in other revenues, but the amount is not disclosed and neither is the contract structure beyond the earlier cost-plus description. The line is now large enough to matter and small enough to hide.
5. No event-contract per-unit economics: $156M on 13.6 billion contracts is disclosed, but the revenue mix between sports and non-sports events, and how per-contract economics differ across them, is not. Given this is now the second-largest transaction line, the disclosure gap is material.
6. No comment on the regulatory tail on event contracts: the risk factors continue to name potential enforcement or legislative action that could prevent Robinhood from offering event contracts at all. With the line at 11.9% of revenue, management offered no framing of that exposure on the call.
7. No update on the CFO-transition charges: $7M of transition-related share-based compensation ran through Q2 after $13M in Q1, and it remains a standing exclusion from the adjusted expense measures with no stated end date.

Market Reaction

• Pre-print setup: HOOD closed at $89.84 on July 29 entering the print, down 20.6% year to date against a +6.9% S&P 500, down 11.8% over the trailing 30 days, and down 13.0% over the trailing twelve months. The 52-week closing range was $65.16 to $152.46. The stock came into the print already de-rated and out of favour.
• After-hours: shares fell roughly 4% in extended trading on July 29 despite the beat, with coverage attributing the decline to the 38% year-over-year drop in crypto revenue.
• Next-day session: the stock opened higher at $90.91, traded as high as $92.57 intraday, then reversed to close at $86.60, down 3.6% or $3.24 on the day, against an S&P 500 up 1.7%.
• Volume: 29.1 million shares against a 27.0 million 30-day average, only 1.1x normal. This was a repricing on conviction, not a capitulation.

The intraday shape is more informative than the close. Opening up 1.2% and reaching +3.0% before finishing down 3.6% is a session in which the market read the release, liked it, then read the July disclosure and the earnings composition and changed its mind. A round trip from +3.0% to -3.6% on average volume is what happens when the headline and the fine print disagree.

Two things in the fine print explain it. The $0.62 that produced a 44% headline beat contains $0.14 that consensus was not modelling and will not recur, which turns a spectacular print into a good one. And the July net-deposit figure, disclosed voluntarily, tells the market that the metric management asked to be judged on is running at roughly half the pace of the quarter being celebrated. Neither is visible in the headline; both are unavoidable by the second read.

The sell-side response fits the pattern. Every published action we reviewed cut the price target while leaving the rating unchanged, with one desk explicitly citing a meaningful July slowdown in new accounts. Analysts trimming valuation anchors without touching ratings is the signature of a multiple reset rather than a thesis break, and it matches our own read. This is now the sixth negative day-of reaction in the last seven prints, on quarters that have been operationally strong throughout, which continues to suggest the stock's response function keys on forward deceleration signals rather than on reported growth.

Street Perspective

Debate: Does diversification finally break the cyclicality discount?

Bull view: Thirteen business lines above $100M in annualized revenue, event contracts overtaking crypto, and subscription revenue compounding at 54% mean the market can no longer value this as a levered bet on retail trading enthusiasm. The multiple should converge toward a diversified financial platform rather than a brokerage.

Bear view: Transaction revenue is still 59% of the total, and event contracts are arguably more cyclical than crypto, not less, since they depend on a sports and political calendar and carry an unresolved regulatory question. Diversification into a second volatile line is not the same as diversification into a stable one.

Our take: The bulls have the better of this now. The correlation structure matters more than the count: crypto fell 25% sequentially in a quarter when total revenue rose 23%, which is the first direct evidence that a single asset class can roll over without dragging the company. That is what diversification is supposed to buy, and it was just demonstrated rather than asserted. The regulatory point stands as a separate risk rather than a rebuttal.

Debate: Is the July deposit slowdown seasonal or structural?

Bull view: Summer is seasonally slow, Q2 was an exceptional comparison inflated by a high-profile listing and tax-season flows, Trump Account deposits are excluded from the July figure entirely, and the year-to-date rate remains comfortably above the 20% long-term target. One month is noise.

Bear view: The step-down is roughly 45% off the Q2 monthly pace, it lands alongside a reported slowdown in new account growth, and management declined to address the month when asked directly. Deposits lead assets, assets lead revenue, and this is the earliest point in the chain where deterioration would appear.

Our take: Unresolvable on one month of data, which is precisely why it caps the rating. The bull case is likely right about seasonality and the bear case is right that this is the wrong metric to be guessing about. The Q3 print resolves it, and until then a 44x multiple is being asked to price an open question on the company's self-declared primary KPI.

Debate: Is 44x defensible for a business growing 32%?

Bull view: A 57% Adjusted EBITDA margin at 32% revenue growth is an exceptional combination at over $5B of annualized revenue, the expense guide is coming down rather than up, and the second-year multiple falls into the mid-thirties on any reasonable 2027 estimate. The stock has de-rated 20% year to date while earnings compounded.

Bear view: On clean earnings excluding the one-time gain, the multiple is closer to 48x this year's number, consensus already embeds a fourth quarter below the third, and the forward indicators disclosed on this very call point to flat volumes and slower deposits. Paying a growth multiple requires the growth to be visible in the next quarter, not the next decade.

Our take: Both readings are defensible and that symmetry is the whole argument for Hold. The valuation is not obviously wrong for the franchise quality, but it leaves no compensation for the specific risk that Q3 is a flat quarter. We would want either a de-rating or a Q3 deposit recovery before paying it.

Model Update Needed

Valuation impact: at the $86.60 close on roughly 912 million diluted shares, the market capitalization is approximately $79B, or about 44x the $1.95 FY26 consensus and roughly 48x that number excluding the one-time gain. First-half clean earnings of about $0.86 require roughly $0.95 in the back half to meet consensus. Our fair-value range is $80 to $95, which brackets the current price and produces no actionable upside. The range widens meaningfully to the upside on a Q3 net-deposit recovery toward the 20% annualized target, and compresses toward $70 if the July pace persists through the quarter.

Thesis Scorecard Post-Earnings

Overall: The thesis is stronger than it was a quarter ago. Four of five bull pillars confirmed, two of the three bear points that drove the April downgrade have weakened materially, and the revenue base is demonstrably more diversified than at any prior point in the coverage. The offset is that the constraint moved rather than lifted: crypto has been replaced by net-deposit growth as the metric that determines whether the multiple is defensible, and the first month of data on it is poor.

Action: Hold. None of the four upgrade triggers set in April were met, and the one that was closest, quantifiable Trump Accounts contribution, is a partial. We are closer to an upgrade than we were three months ago and would move on a single quarter of confirmation. Upgrade to Outperform on any of: net deposits returning to a 20%-plus annualized pace in Q3; crypto printing flat or better sequentially; Rothera economics disclosed at a take-rate premium to third-party routing; or the stock trading below $75. Downgrade to Underperform on: net deposits sustaining a sub-15% annualized pace through Q3; an adverse regulatory or legislative development on event contracts; or credit-loss provisioning passing 6% of revenue.