In the quarter ended 2026-06-30, Robinhood's event-contract revenue was $156M and its cryptocurrency revenue was $100M. The line the company only began breaking out this cycle is now 1.56x the crypto line that carried its last one.
The crossover is the smaller fact. Crypto fell 38% year over year, from about $160M. So event contracts today are roughly 97% of what crypto was a year ago. The new line did not just pass the old one — it landed at almost exactly the size the old one used to be.
Both halves of that sentence have been on this site since the print. They had never been in the same sentence until this one.
The four lines, and the $49M nobody breaks out
From the Q2 2026 release capture — Exhibit 99.1 to an 8-K, accession 0001783879-26-000113, quarter ended 2026-06-30:
| Transaction-based revenue | Q2 2026 | YoY |
|---|---|---|
| Options | $342M | +29% |
| Equities | $129M | +95% |
| Event contracts | $156M | >10x |
| Cryptocurrencies | $100M | −38% |
| Total transaction-based revenues | $776M | +44% |
One caution before any of the percentages below. Those four lines add to $727M, not $776M — the release leaves $49M, 6.3% of the total, unbroken-out. Every share in this piece is taken against the $776M total, never against the sum of the four. A reader adding the column should know the column does not close.
Crypto is the only one of the four that fell.
The symmetry, and which $160M
The release discloses crypto's percentage change, not its prior-year dollar figure. Our July analysis printed $160 million; the disclosed −38% back-solves to $161M. We are keeping $160M, both because it is what this site has already published and because the argument does not depend on the choice:
- $156M is 97.5% of the published $160M.
- $156M is 96.7% of the back-solved $161M.
About 97% either way. That is the whole reconciliation, and it is why the sentence survives the rounding — a derived figure is doing work here, and it is worth saying so rather than presenting ~$160M as a reported number. It is not one.
The same back-solve, applied to the mix, gives the cleanest single line available. Both ends are derived: crypto was roughly 30% of transaction-based revenue a year ago ($160M against the reported Q2'25 base of $539M) and is 12.9% of it now ($100M against $776M). Against total net revenues of $1,308M it is 7.6%. A business does not usually cut a revenue line's share of the mix by more than half and post a record quarter at the same time; Robinhood did, because three other lines grew into the space.
The arithmetic behind the prior-year column is sound rather than assumed: the release's own disclosed +44% on transaction revenue takes the $539M Q2'25 base to $776.2M, which is the $776M reported. The base and the growth rate reproduce the absolute.
Bitstamp, which had never appeared on this site
The outcome is a revenue line falling 38%. The mechanism is one row of the capture, and it is the reason this is more than a chart:
Crypto Notional Trading Volume: $40B ($18B app + $22B Bitstamp) — Robinhood App volume −35%
Fifty-five percent of Robinhood's crypto notional no longer runs through the Robinhood app. The majority now runs through the acquired institutional exchange. The shrinking half — down 35% — is the retail app, the part that made the franchise.
That reframes the line entirely. It is not "crypto is down." It is that the retail crypto business Robinhood built its last cycle on is now a minority of Robinhood's own crypto volume. The word Bitstamp appears nowhere else on this site; the split has been sitting in the capture since the print.
One derived figure, and then a limit. Derived: the app's disclosed −35% back-solves to roughly $27.7B of app notional a year ago, against $18B now. The limit: we are not computing a year-over-year on total crypto notional, and neither should anyone quoting this. The release discloses a change for the app line only. Bitstamp is an acquired exchange and nothing in the capture establishes whether it sat in the year-ago base, so a total-notional comparison would be measuring two different perimeters and calling the difference a trend.
Why this sentence went unwritten
It would be easy, and wrong, to say nobody noticed. The observation was written down. The capture's own list of angles, filed the day of the print, says the mix shift away from crypto "is a notable reversal worth its own piece."
So it was recorded, in this repository, and still never reached a reader. That is the more uncomfortable version of the story, and the cause is visible in the article it did not make it into. Our July piece put the crypto decline under Why the Stock Fell on a Genuine Beat and the event-contract number under What's Actually Working — sixteen lines apart, one writer, one print. Its framing was "Strip out the crypto line and the macro noise, and the underlying business is diversifying…", and that framing is what hid this: crypto went in the bad-news column, event contracts in the good-news column, and the possibility that they were the same line in sequence never came up. Good-news and bad-news columns are a fine way to organise a print. They are a poor way to notice a substitution.
The predecessor to this piece has the same shape from the other side. Monday's article asked what event contracts became — $156M, 20.1% of transaction revenue, against a prediction market that had priced Robinhood's own launch at 24% in February. This one asks what they displaced. Neither earlier piece is wrong and nothing here corrects them; both printed accurate figures from the same capture. What was missing was the line between them.
What this piece does not claim
Three things, stated because each is a way this argument could be overextended:
- No quarterly streak. The capture's Q1'26 column is a dash for every transaction sub-line, so there is no Q1 2026 crypto figure anywhere in this repository. The year-over-year story stands on its own and needs no sequence of quarters behind it.
- No total-notional trend, for the perimeter reason above.
- No claim about substitution in the customer. Event contracts arriving at the size crypto used to be is an accounting fact about one company's revenue mix. Nothing in the release says the same customers moved from one product to the other, and this piece is not saying it.
What is left after those three deletions is still the thing worth knowing. In four quarters, Robinhood replaced the dollar value of a marquee revenue line with a product this site described in July as "a rounding error a year ago" — and the line being replaced did not so much collapse as move off the app it was built on.
Every reported figure — the four transaction lines and their YoY changes, the $776M and $1,308M totals, the $539M Q2'25 transaction base, and the $40B crypto notional with its $18B/$22B split and the −35% app change — is from this site's capture of Robinhood's Q2 2026 release for the quarter ended 2026-06-30, filed as Exhibit 99.1 to an 8-K (Item 2.02) under accession 0001783879-26-000113 and reported 2026-07-29. Derived figures, marked as such above: the 1.56x ratio and $56M gap, the $49M unbroken-out residual, crypto's 12.9% and 7.6% shares, the ~$160M prior-year crypto revenue and the ~30% prior-year share back-solved from the disclosed −38%, the 97% comparison, the 55% Bitstamp share of notional, and the ~$27.7B implied prior-year app notional back-solved from the disclosed −35%. The $160M prior-year crypto figure is the one already published in our July analysis; the release itself discloses only the percentage. Two figures here come from our own earlier pieces rather than from this capture and are attributed in the text where they appear: the $160M, as above, and the February prediction-market pricing of 24%, which Monday's article derives from the prediction-market data behind Robinhood's ticker page — neither is a release figure. No share price, price move or market-capitalisation figure appears here — this repository stores no previous close for any company — and no comparison to another crypto business is drawn, because no second capture exists to support one.