Robinhood prediction markets revenue reaches crypto parity

Eighteen months ago, Robinhood did not have a prediction markets business. In Q2 2026, that business generated $156 million in revenue — the same weight class as the company's crypto trading operation, a segment Robinhood spent years building into a flagship.

That number deserves a sentence before the next one. A product line that did not exist in March 2025 reached revenue parity with crypto trading by mid-2026. The growth curve implied is not gradual. It is the kind of curve that makes a CFO reclassify a line item.

The mechanism behind it is worth examining carefully. Robinhood's Prediction Markets Hub runs on CFTC-regulated binary event contracts — each one settles at one dollar or zero, depending on whether a stated outcome occurs. The contracts cover Federal Reserve decisions, election results, sports outcomes, economic indicators. Between launch and mid-2026, Robinhood facilitated between eleven and sixteen billion of them. The volume range is wide enough to suggest the internal accounting is still catching up to the scale.

What made this move structurally different from other retail entries into prediction markets is that Robinhood did not rent infrastructure — it built vertically. The Rothera Exchange, established in early 2026 through a joint venture with Susquehanna International Group, gave Robinhood a CFTC-licensed exchange it controls at roughly ninety percent ownership. In a previous project I built on borrowed infrastructure, and the dependency cost more than the convenience was worth. Vertical control over matching, clearing, and settlement is not an operational detail. It is the difference between a product and a business.

The Crypto.com talks reported for mid-2026 follow a recognizable logic. Robinhood has distribution — millions of retail accounts, a mobile interface that reduced friction until friction was no longer the constraint. Crypto.com has existing prediction market contract inventory. A deal of this kind would not primarily expand Robinhood's user base. It would expand the contract menu, which is where engagement compounds. More resolution events mean more reasons to return, more positions to manage, more Robinhood Cortex prompts to act on.

The consensus read on this is that Robinhood has executed a distribution play — take prediction markets mainstream by embedding them in a platform that already has the users. I think that framing understates what has actually happened. Robinhood has built the first vertically integrated retail prediction market exchange at scale, with its own licensed venue, its own clearing, and AI tooling designed to increase position frequency. The distribution was always there. What Rothera added was the structural independence to set terms — on contract design, on settlement timing, on what categories get listed — without negotiating with an upstream exchange on every decision.

That independence becomes significant when the regulatory environment moves, as it currently is. Maryland users are excluded. Other states carry variable availability. A platform that owns its exchange can respond to a favorable ruling or a new exemption faster than one licensing access from a third party. The Crypto.com partnership, if it closes, adds inventory. But the Rothera structure is what makes the inventory matter long-term.

The $156 million figure also reframes the broader market size debate. Bernstein's $240 billion annual volume projection for 2026 was reported elsewhere on this desk. Robinhood's revenue alone, annualized from Q2, implies a take rate and a volume that anchors those projections in something concrete rather than extrapolation.