Robinhood research
Robinhood: Losing Crypto Traders, Gaining Everyone Else
- 1. The Thesis
- 2. The Dual Mandate: Growth And Discipline, Together
- 3. Relentless Asset Accumulation
- 4. New Customers Are Arriving Richer Than Old Ones
- 5. Gold: The Subscription That's Quietly Carrying The Company
- 6. Trump Accounts: A Government-Seeded Customer Pipeline
- 7. The Crypto Number, In Context
- 8. Rothera: Owning The House, Not Just Working In It
- 9. Eleven Different Ways To Make Money
- 10. Where The Money Actually Goes
- 11. Spending More On Growth While Guiding Expenses Down
- 12. The War Chest And Shareholder Returns
- 13. The Rule of 112
- 14. The Price: A Pullback That Looks Overdone
- 15. What Could Go Wrong
- 16. Muffett's Take
1 The Thesis
The megatrend we keep coming back to across our technology and fintech coverage is platformisation — the way a company that starts out doing one thing well eventually builds enough trust and enough infrastructure that it can sell its customers everything else too. Amazon did it with retail into cloud computing. DigitalOcean, which we covered recently, is doing it with cloud hosting into AI. Robinhood is doing it with stock trading into, functionally, a full financial superapp: brokerage, options, crypto, prediction markets, retirement accounts, a credit card, banking, and now a government-mandated investment account for every American child. Management isn't hiding this ambition. One recent headline put it plainly: Robinhood isn't trying to build a better brokerage, it's trying to build the financial superapp. The second-quarter numbers are the clearest evidence yet that this strategy is working, even in a quarter where its original signature product — crypto trading — went through a real, ugly slowdown.
2 The Dual Mandate: Growth And Discipline, Together
Robinhood's own framing for the quarter was a "dual mandate," and the numbers back up the label rather than just sounding good on a slide.
Total net revenues of $1.3 billion, up 32% year over year. Total platform assets of $369 billion, also up 32%. Net deposits of $21.7 billion for the quarter, a 28% annualized growth rate. That's the growth side. On the discipline side: a record 57% adjusted EBITDA margin, net income of $573 million, up 48% year over year — meaningfully faster than revenue growth, which tells you costs aren't scaling anywhere near as fast as revenue is — and trailing-twelve-month net income of $2.1 billion. Most companies growing this fast are still burning cash to do it. Robinhood is growing this fast while getting more profitable at the same time, and that combination is rare enough that it's worth sitting with for a moment before moving on to the details.
3 Relentless Asset Accumulation
Revenue and profit are the scoreboard. Funded customers and platform assets are the engine underneath the scoreboard, and that engine is running hotter than it has at almost any point in the company's history as a public company.
28.4 million funded customers, up 32% year over year — that's not a company coasting on an existing base, that's real net new account growth at meaningful scale for a company this size. $369 billion in total platform assets. And $76 billion of net deposits over the trailing twelve months, 27% growth, with the growth rate itself accelerating quarter over quarter based on the shape of that chart. People aren't just opening Robinhood accounts. They're moving real, growing amounts of money into them.
4 New Customers Are Arriving Richer Than Old Ones
The data point in this whole deck that we think gets the least attention relative to how much it actually matters is the cohort chart, because it answers the single most important question about any growth story: is the newest growth as good as the growth that came before it, or is the company scraping the bottom of the barrel to keep the headline numbers up?
The answer here is unambiguous: newer cohorts are starting with dramatically larger deposits and compounding their platform assets faster than any prior generation of Robinhood customers. The 2021 cohort — signed up during the meme-stock and free-trading boom, often stereotyped as small, speculative accounts — took sixteen quarters to reach a modest deposit base. The 2026 cohort is already starting well above where the 2024 and 2025 cohorts were at the same point, and climbing at a steeper angle. That's the opposite of a company running out of growth. It's a company whose newest customers are, on average, wealthier and more serious than the ones who came before — which is exactly what you'd want to see if the "financial superapp" positioning is actually landing with a broader, more affluent audience rather than just the original day-trading crowd.
5 Gold: The Subscription That's Quietly Carrying The Company
If there's one single product that explains how Robinhood absorbed a 38% crypto revenue decline and still posted record results, it's Robinhood Gold, the company's premium subscription tier.
Gold subscribers reached a record 4.84 million, about 17% of the total funded customer base, each paying a recurring subscription fee for meaningfully better economics: 3.35% yield on uninvested cash instead of a token 0.01%, a 3% match on retirement contributions instead of 1%, lower fees on futures and index options, and access to a growing bundle of premium features including the Gold Card, Robinhood Banking, and the more advanced version of the company's Cortex AI tools. The behavioral data on these subscribers is the real story: Gold subscribers carry 4.2 times the assets under custody of standard users, adopt retirement products at 3.1 times the rate, and grow their net deposits 1.1 times faster. This is a subscription product that doesn't just generate its own fee revenue — it measurably changes customer behavior in every direction that benefits Robinhood's broader platform. That's a powerful, high-margin offset sitting directly next to whatever crypto trading happens to be doing in a given quarter.
6 Trump Accounts: A Government-Seeded Customer Pipeline
This is the part of the current Robinhood story that doesn't show up anywhere in the company's own investor deck yet, because it's simply too new — and we think it deserves more attention than it's gotten. Under new federal legislation, every American child born between January 1, 2025 and December 31, 2028 is eligible for a "Trump Account" — officially a 530A account — seeded with a $1,000 government contribution, invested by default in an S&P 500 index fund, with parents, relatives, employers, and charities able to contribute up to $5,000 more per year. Robinhood, alongside custodian Bank of New York Mellon, was tapped as the brokerage and initial trustee for this program. The accounts officially began accepting contributions on July 4, 2026, and Bloomberg has estimated the program could ultimately touch as many as 14 million children.
Think about what that actually means for a customer acquisition strategy. Robinhood isn't paying for these accounts to be opened — the federal government is funding the seed deposit and doing the marketing simply by making this a nationwide, headline-generating policy rollout. Every one of these accounts is a brand-new, funded Robinhood relationship with a family that will, in most cases, still be a Robinhood customer when that account converts automatically into a traditional IRA when the child turns eighteen. This is a multi-decade customer acquisition funnel, essentially underwritten by the federal government, landing directly inside Robinhood's existing infrastructure. We'd be surprised if this doesn't show up as a distinct, quantified growth driver in the company's investor materials within the next two or three quarters.
7 The Crypto Number, In Context
We're not going to bury this, because the user asking for this note asked us to look at it directly, and because pretending it away would make for a dishonest piece of research. Crypto trading revenue fell 38% year over year in Q2 2026, to $100 million, as crypto notional trading volume dropped 39% quarter over quarter to $40.4 billion and activity on Robinhood's own app specifically fell 35% year over year. That's a real, meaningful slowdown in what was, not long ago, one of the company's most talked-about growth engines, and it's the specific reason the stock initially sold off on an otherwise record earnings day.
Here's the context that we think the initial market reaction missed. Options revenue was $342 million, up 29%, still comfortably the largest single transaction revenue line. Equities revenue was $129 million, up 95%. And event contracts — prediction markets — generated $156 million on 13.6 billion contracts traded, both records, with revenue up roughly tenfold year over year by the company's own description, making prediction markets the fastest-growing business in Robinhood's history. Put simply: crypto went from being one of two or three pillars holding the roof up to being one of eleven, and the roof didn't come down, because the other pillars got dramatically stronger at almost exactly the same time crypto weakened. That's not luck. That's what a genuinely diversified platform is supposed to do when one line of business hits a rough patch.
8 Rothera: Owning The House, Not Just Working In It
The prediction markets number above is impressive on its own, but the more important development happened earlier in the year and changes the economics of that entire business line going forward. In January 2026, Robinhood and Susquehanna acquired 90% control of MIAXdx, rebranding it as Rothera Exchange with its own clearing arm, Rothera Clearing — giving Robinhood full Designated Contract Market and Derivatives Clearing Organization status under CFTC regulation.
Before Rothera, Robinhood earned ordinary retail brokerage economics on prediction market volume, while a third-party exchange — Kalshi — captured the underlying exchange and clearing economics on every single contract traded. That's the equivalent of running a busy casino floor while a different company owns the building and takes a cut of every table. Rothera flips that arrangement: Robinhood now owns the exchange and the clearinghouse, not just the storefront selling access to it. By June 2026 the platform was fully operational, handling flow including a slate of World Cup event contracts that Robinhood shifted away from Kalshi onto its own exchange. As prediction market volume keeps compounding — and a tenfold year-over-year revenue jump suggests it's compounding fast — an increasing share of that volume now clears through infrastructure Robinhood itself owns, at economics Robinhood itself sets. That's a structurally better business than the one that existed a year ago, on a product line the company itself calls its fastest-growing ever.
9 Eleven Different Ways To Make Money
Zoom out from any single product and look at the full shape of the business, and the diversification argument becomes hard to argue with.
Eleven separate business lines, each individually generating more than $100 million in revenue, with the two newest additions — prediction markets and the credit card — already large enough to earn a spot on that list. Robinhood Retirement alone now holds $34.5 billion in assets under custody. The WonderFi acquisition is completed and folded into the international push, alongside a new Indonesia brokerage launch, in-principle regulatory approval in Singapore, and a base of more than 1 million non-US customers already. This is what a genuine platform business looks like from the inside: no single product line, including crypto, is large enough on its own to meaningfully derail the whole company when it has a bad quarter.
10 Where The Money Actually Goes
Diversification only matters if it also shows up in the margin structure, and here the trend is about as clean as we've seen anywhere in our fintech coverage.
Average revenue per user climbed to $187, and as it rose, the adjusted EBITDA margin rose right alongside it to a record 57%, because — as management put it plainly — nearly all incremental revenue now falls directly to the bottom line rather than requiring proportional new spending to support it. Annualized revenue per employee surged 21% year over year to $1.73 million, which is an extraordinary figure for a financial services company of this scale and a strong signal that the technology and automation underneath this platform is doing real work, not just supporting a larger headcount doing the same tasks as before.
11 Spending More On Growth While Guiding Expenses Down
Perhaps the single most impressive item in the entire deck, from a capital discipline standpoint, is what management did to its own expense guidance in the same quarter it was actively spending on acquisitions.
The prior 2026 total expense outlook was $2.7 billion to $2.825 billion. Into that base, management layered $50 million of new costs from the Rothera prediction markets venture and $20 million from the completed WonderFi acquisition — real, incremental spending on two of the growth initiatives detailed above. And the resulting, current expense outlook still came out lower, at $2.675 billion to $2.775 billion, because the core business found $110 million of efficiencies to more than offset the new spending. That's a company funding its own expansion out of its existing operating leverage, rather than asking shareholders to underwrite growth investments with a wider loss. Very few management teams manage to lower guidance and increase strategic investment in the same breath. This one just did.
12 The War Chest And Shareholder Returns
None of this diversification is happening on a stretched balance sheet, which matters given how much new spending we've just walked through.
$5.6 billion of corporate cash, investments, and stablecoin sitting on the balance sheet, topped up opportunistically with a further $2.2 billion raised through zero-coupon convertible notes due 2029 — effectively an interest-free loan from the market, taken out because management judged the opportunity to deploy that capital was worth having in reserve. All three capital allocation levers are switched on at once: organic growth, M&A, and shareholder returns. $414 million was returned to shareholders in the second quarter alone through the repurchase of 4.4 million shares, and diluted share count has been held roughly flat through 2025 and 2026 despite all of this activity — meaning growth investors aren't quietly being diluted to fund the expansion story.
13 The Rule of 112
There's a well-known heuristic in software and fintech investing called the Rule of 40: add a company's revenue growth rate to its profit margin, and if the sum clears 40, you're looking at a genuinely efficient, well-run growth business. Robinhood's own materials make a simple, almost blunt case using that exact framework.
56% trailing-twelve-month revenue growth, plus a 56% trailing-twelve-month adjusted EBITDA margin, equals 112 — nearly three times the standard 40 that's usually held up as the mark of a well-run growth company. We don't love leaning on a single flashy composite metric as the entire basis for an investment case, and we won't do that here. But taken alongside everything else in this note — the diversification, the cohort acceleration, the expense discipline, the balance sheet — it's a fair, honest summary of a business firing on essentially every available cylinder at once, with one specific, well-understood exception in crypto trading that the rest of the business has already absorbed.
14 The Price: A Pullback That Looks Overdone
Now to the part that actually matters for whether this is a stock worth owning today rather than just a well-run company worth admiring. HOOD trades at $95.82, down 3.6% today and down sharply — roughly 38% — from its 52-week high of $153.86, against a 52-week low of $63.52. That's an enormous round trip even by the standards of a high-beta fintech name, and a meaningful chunk of the drop from the highs happened specifically around the Q2 earnings release, when the stock initially fell despite beating on nearly every headline metric, purely on the crypto revenue softness we've walked through above.
The trailing P/E sits at roughly 42 times, with a forward P/E near 40 — not a cheap stock by traditional value screens, but also not obviously expensive for a company compounding revenue at 32%-plus with expanding margins and a genuine, multi-pronged growth story still in its early innings. The analyst community's reaction to the print has been telling: the majority of banks trimmed price targets in the days following earnings — Goldman Sachs cut to $118 from $137, Barclays to $105 from $122, Deutsche Bank to $114 from $120 — largely citing the crypto miss specifically, while almost every one of those same banks kept a Buy or Overweight rating rather than downgrading the stock outright. BofA moved the other direction entirely, raising its target to $140 from $132 after the print. The consensus across 28 analysts remains a solid Buy, with an average target of $119.82, implying roughly 25% upside from today's price. One piece of coverage put it about as plainly as we would: Robinhood's record quarter makes the pullback look overdone.
15 What Could Go Wrong
- Crypto could stay soft, or get softer: a 38% year-over-year decline is a real trend, not a one-quarter blip, and if crypto trading volumes continue to shrink as a share of retail activity, it will keep pressuring transaction revenue even as other lines pick up the slack.
- This is a high-beta, sentiment-driven stock: a 52-week range spanning $63.52 to $153.86 tells you everything about how violently this name can move in both directions on sentiment shifts around crypto, meme-stock activity, and broader risk appetite, independent of the underlying business trend.
- Regulatory risk touches nearly every new business line: prediction markets, crypto, the Trump Accounts program, and international expansion into markets like Indonesia and Singapore all carry real, evolving regulatory exposure that could slow or reshape any one of these growth initiatives.
- Trump Accounts economics are still unproven: the program is brand new, the accounts start small, and it will likely be years before this shows up as a clearly quantified, material contributor to revenue rather than a promising but early customer-acquisition story.
- Valuation is not undemanding: a forward P/E near 40 leaves real room for disappointment if growth decelerates meaningfully from here, particularly if the market's patience for "story stocks" more broadly cools off.
- Competitive intensity in prediction markets is real: Kalshi, DraftKings, and Coinbase are all pushing hard into the same space Rothera is built to capture, and owning exchange infrastructure doesn't guarantee Robinhood wins the volume war outright.
16 Muffett's Take
Robinhood spent its first several years as a public company being written about almost entirely through the lens of two products: commission-free stock trading and crypto speculation. This quarter is the clearest evidence yet that management has methodically built past that reputation. Gold subscriptions with a documented 4.2 times asset multiplier. A government-seeded pipeline of new accounts arriving through Trump Accounts. A fully owned prediction markets exchange in Rothera capturing economics it used to hand to a competitor. Eleven separate hundred-million-dollar business lines. A 57% EBITDA margin and a Rule of 112. And all of that happened in the same quarter crypto revenue fell 38% — which tells you the diversification isn't a slide-deck aspiration anymore. It's already load-bearing.
The stock got marked down anyway, on the one number that fit the old, narrower story about what Robinhood is. We think that's exactly the kind of mispricing worth acting on: a genuinely diversifying platform business, growing revenue at 32% with margins still expanding, trading nearly 40% below its own 52-week high, with the broad analyst community still pointing toward meaningful upside even after trimming targets.
