Robinhood booked $156 million of event-contract revenue in the quarter ended 30 June 2026, on 13.6 billion contracts. Annualised, that is $624 million from a line that was $3 million six quarters ago.
Run our Robinhood forward model forward and the line crosses $1 billion of trailing twelve-month revenue in 2028 Q2 — $1,050 million — and finishes 2028 at $1,243 million. Bernstein's press-reported estimate for the same year is about $1.7 billion, which is 37% above our full-year 2028 number and 62% above the quarter we cross the billion in.
That gap is worth understanding, because it is not where anyone thinks it is.
The points
- $1 billion arrives in 2028 Q2 on a trailing-twelve-month basis, and $2.25 billion is where the line peaks, in 2031 Q1. Both are our projection, not guidance.
- Quarterly revenue peaks at $569 million in 2030 Q2 and then declines — $565M, $561M, $558M, $554M — ending the horizon 2.6% below its own peak.
- The reason is a ceiling we wrote. Our model caps event-contract volume at 60 billion a quarter. It binds in 2030 Q2. After that, volume stops while revenue per contract keeps falling 2% a quarter, so the line turns down.
- The line's revenue share peaks at 18.8% of total company revenue in 2030 Q2 and falls back to 15.7% by 2031 Q2. It never becomes the company.
- Doubling the assumed growth rate is worth 52 cents. Moving volume growth from 10% a quarter to 20% takes fair value from $94.42 to $94.94.
- Stalling costs $11.19. Moving the same assumption from 10% to 0% takes fair value to $83.22.
- The whole vertical is worth $14.80 a share — 15.7% of our $94.42 base. Delete it entirely and the model prints $79.62.
The conversion: this line is violently asymmetric
The sensitivity, run at the published basis:
| Volume growth per quarter | Fair value | Delta |
|---|---|---|
| 0% | $83.22 | −$11.19 |
| 5% | $88.01 | −$6.41 |
| 10% (published) | $94.42 | — |
| 15% | $94.75 | +$0.34 |
| 20% | $94.94 | +$0.52 |
| 30% | $95.12 | +$0.70 |
Downside is eleven dollars. Upside is fifty-two cents. Tripling the growth rate to 30% a quarter — a rate nothing sustains — adds seventy cents.
The ceiling table says the same thing from the other direction:
| Volume ceiling per quarter | Fair value | Delta |
|---|---|---|
| 30bn | $87.38 | −$7.04 |
| 45bn | $90.96 | −$3.46 |
| 60bn (published) | $94.42 | — |
| 90bn | $98.79 | +$4.37 |
| 120bn | $98.87 | +$4.45 |
| 400bn | $98.87 | +$4.45 |
Above 90 billion the ceiling stops mattering at all. A 400-billion cap and a 120-billion cap give the identical answer, because 10% a quarter compounded from 13.6 billion never gets near either one inside the horizon. We wrote the cap to stop a compound quietly becoming the company. It also caps how bullish this model is structurally capable of being, and that is the finding.
The denominator: what the gap to Bernstein is actually made of
Bernstein's press-reported target on Robinhood is $160. Our base is $94.42, against a share price of $91.53 as of 19 August 2026. The distance is $65.58 a share.
Here is the arithmetic that changes the story. The entire prediction-market vertical — every contract, at every growth rate, for five years — is worth $14.80, or 22.6% of that gap. And no achievable assumption about its growth moves it by more than about $4.45, which is 6.8% of the gap.
So under our structure, someone who is merely very bullish on event contracts cannot get from $94.42 to $160. They would need to be right about the ceiling and about growth and still find another fifty dollars somewhere else.
We have said otherwise. Replying on X last week about the same $160 target, we attributed the distance to Bernstein being "more optimistic on the speed and scale of prediction markets." This model run says that explanation cannot carry the gap. The honest version is the reverse: the difference to Bernstein is mostly crypto, the exit multiple and the discount rate. Prediction markets are the part of the disagreement that matters least.
What this does to the model
Nothing, and the size of the nothing is the point.
Our event-contract assumptions were already deliberately conservative on volume — 10% a quarter is a sharp deceleration from twenty-fold — and deliberately pessimistic on price, at a 2% quarterly fee decline, because this is a fee-per-contract business in an open fight with Kalshi, Polymarket and the sportsbooks. The sensitivity says those choices barely matter to fair value in the upward direction. They matter a great deal downward.
Two things this model does not contain, both worth stating plainly:
Regulatory risk is priced nowhere. State gaming regulators are challenging sports event contracts, a Ninth Circuit case is pending and CFTC rulemaking is unresolved. An adverse outcome does not slow this line, it removes a large share of it, because sports is where the volume is. The $14.80 brackets that neatly: it is roughly what a total loss of the vertical costs.
Legal sportsbook handle is not the same market. Americans wagered a press-reported $166.94 billion with legal sportsbooks in 2025. A binary contract cannot settle above $1, so Robinhood's 13.6 billion quarterly contracts carry at most $13.6 billion of settlement value, $54.4 billion annualised — at the arithmetic ceiling, about a third of the sportsbook handle. The two overlap and compete, but equating them is the most common error in this argument, and we are not going to make it to inflate a number.
The fuller history of the product, the venues and the regulatory fight is in today's deep dive.
What to watch
- August and September contract volume. July 2026 was 6.1 billion contracts, down 5% on June — the first sequential decline this line has had. If the next two months confirm a stall, the relevant row is the 0% one, worth −$11.19, not the ceiling.
- Revenue per contract. 1.15 cents in the June quarter. Robinhood publishes contracts monthly, so a fee falling faster than 2% a quarter shows up long before a print.
- The Ninth Circuit ruling and CFTC rulemaking. The only inputs here that can take the line to zero rather than slow it.
- A category split. Robinhood discloses total contracts, not a sports-versus-everything-else breakdown. Until it does, nobody outside the company can size the regulatory exposure — including us.
- Whether 60 billion turns out to be the wrong number. If quarterly volume passes 30 billion before 2029, our ceiling is doing more work than the evidence supports and the assumption should move.
Sources and provenance. Event-contract revenue of $156M on 13.6 billion contracts, total net revenues of $1,308M and the six-quarter history from $3M are as reported by Robinhood for the quarter ended 30 June 2026; July 2026's 6.1 billion contracts is from the company's monthly operating data. Derived by us: the $624M annualisation, 1.15 cents of revenue per contract, and the $13.6 billion quarterly and $54.4 billion annualised bounds on settlement value, which follow from a binary contract being unable to settle above $1. Ours, and assumptions rather than facts: the 60-billion-contract quarterly ceiling, 10% quarterly volume growth, the 2% quarterly decline in revenue per contract, and every fair value quoted here — $94.42 base, $79.62 without the vertical, and each row of both sensitivity tables — which are runs of our own model at its published 2026 Q2 basis. The $1,050M trailing crossing in 2028 Q2, the $569M quarterly peak in 2030 Q2, the $2.25B trailing peak and the 18.8% revenue share all fall out of those assumptions and inherit their uncertainty. Bernstein's approximately $1.7 billion 2028 estimate and $160 price target are press-reported and are not series this site stores or verifies. The $166.94 billion of 2025 US legal sportsbook handle is press-reported from state gaming regulators; handle and settlement value are different quantities and the comparison above is stated as a bound, not an equivalence. Regulatory risk is priced nowhere in the model. The share price of $91.53 is as of 19 August 2026 and will differ from the live quote.
