Key Points
- Locke’s public letter argues Genius earns from prediction markets regardless of regulatory outcome, backed by Q2 revenue of $195.5m against $185m guidance.
- Kalshi’s $18.6m Michigan settlement failure on 5 September became the most visible real-world proof of Locke’s core claim that authoritative settlement data cannot be improvised.
- Connecticut’s 10 September enforcement sweep subpoenaed Genius Sports Media and Genius Tech International directly, a detail buried in the official state release that sits in direct tension with the letter’s confident framing.
Genius Sports co-founder and CEO Mark Locke published a public letter on 14 September built around a line designed to land with nervous investors: “Americans are going to keep wagering on sports, and Genius will be paid however they do it.” Most coverage read this as corporate reassurance. Read against what happened in the nine days before it, the letter is considerably more complicated than that.
The Argument, and What Most Coverage Glossed Over
Locke frames Genius as a layer beneath the regulatory conflict, supplying official data and integrity services simultaneously to DraftKings, FanDuel, Kalshi and Polymarket. He adds a second revenue channel that has been largely glossed over: the institutional market makers providing liquidity on prediction exchanges are separate buyers of Genius data and pricing, distinct from the platforms themselves. The Legend media platform sits on top of both, earning from customer acquisition for operators competing on all sides of the market.
On the economics, Locke is direct. He writes that “under our current commercial arrangements, we believe the economics to Genius from a comparable dollar of prediction market activity match or exceed those generated through a sportsbook.” The structure behind that claim matters. A sportsbook’s margin sits with the house. On an exchange, the spread goes to the market maker, and Genius earns from that market maker separately from whatever the platform retains. The margin has moved, not disappeared.
The Eilers & Krejcik data Locke cites gives this positioning some weight, though it carries caveats. As of July 2026, 69% of Kalshi’s retail demand came from states with no legal online sportsbook, while its combined handle in mature betting states sat at roughly 2%. These are modelled estimates, and Locke’s own letter acknowledges some of those users were previously betting offshore. The argument is not that prediction markets have created entirely new bettors, but that the demand underneath them is durable enough to survive whatever regulatory form eventually contains it. If prediction market customers are displaced, they either migrate to sportsbooks Genius already supplies or they pressure holdout states to legalise betting and open new markets. If they stay, Genius keeps serving the exchanges.
The Michigan Incident Made the Argument Concrete
On 5 September, Kalshi settled its Western Michigan versus Michigan market prematurely, paying the wrong side after treating the broadcast clock hitting zero as the official final state. Michigan quarterback Bryce Underwood then threw a 47-yard Hail Mary touchdown to JJ Buchanan; officials had restored one second after a review ruling, and Michigan won 13-12. Kalshi reversed the payouts on an $18.6m market and issued $50 credits as compensation.
Coinbase, which settles off Kalshi’s infrastructure, handled the error differently: it chose not to reverse payouts already made under the incorrect result, leaving traders on its platform paid under the wrong outcome while Kalshi clawed back and re-settled. The divergence was not a procedural footnote. It illustrated precisely what Locke had argued to investors for months: pricing tolerates approximation because a market can correct it in real time, but settlement is final and pays. Once Kalshi treated the zeroed broadcast clock as authoritative, with no official data relationship to arbitrate the disputed second, the error was already locked in. The Michigan incident put that argument on sports pages across the country, nine days before Locke’s letter appeared.
Connecticut Named Two Genius Entities Before the Letter Went Out
On 10 September, Connecticut issued nine cease-and-desist orders to prediction market platforms and nearly 30 subpoenas to surrounding infrastructure, covering 9 licensed gaming service providers, 15 media organisations and 5 app store or payment companies. Among the nine licensed service provider recipients are Genius Sports Media and Genius Tech International.
The state’s release is explicit that subpoena recipients are not under investigation; the documents seek information to assist Connecticut’s evaluation of the sector. Still, two Genius entities appear on a state enforcement list four days before Locke published a letter positioning Genius as a neutral infrastructure provider benefiting regardless of regulatory direction. Connecticut Consumer Protection Commissioner Bryan Cafferelli stated: “Our laws are clear: sports betting may only be offered by legal, licensed sportsbooks that adhere to our regulations and technical standards.” The enforcement followed the state’s earlier lawsuit against Kalshi and a federal court ruling that sports event contracts constitute illegal unlicensed gambling under Connecticut law. Underdog, one of the nine operators served cease-and-desist orders, sued Connecticut in federal court this week arguing CFTC jurisdiction supersedes state enforcement. The legal architecture is far from settled.
The Financials That Made the Letter Necessary
Genius Sports released its Q2 2026 results on 6 August, meeting expectations across all key metrics. The Group generated $195.5 million in revenues, above estimates of $185 million, representing a 65% year-over-year increase. EBITDA was $52.6 million against expectations of $45 million. The group updated its full-year guidance for 2026, with a midpoint of $1.02 billion for revenue and $290 million for EBITDA.
There was another story about the stock. GENI closed at $7.31 on 8 September and at $6.80 on 14 September, marking a 7% decrease within five trading days due to regulatory uncertainty surrounding the prediction markets that continued to impact confidence in traditional sportsbook operators used by Genius too. The Q2 GAAP net loss was $76.7 million compared to $53.9 million the prior year, mainly due to transaction costs of $28.9 million related to the $1.2 billion Legend acquisition and increased interest expense post-acquisition. While the revenue was up, GAAP loss was up. There was a need for shareholders to hear more than just numbers. The letter is Locke’s response.
The NFL Holdout Is the Most Significant Unlocked Variable
Genius holds exclusive official data rights to the NFL through the end of the 2029 season, a deal renewed in June 2025. No other current Genius league partnership generates comparable data licensing revenue or carries the same commercial weight in the US market. Locke said on the Q2 earnings call: “Clearly it would be very significant. There’s a financial significance that comes directly with the sale of the data for the most important league. There’s obviously the value of the affiliation that they get, which has a real monetary value as well… if I were you, I wouldn’t be expecting that to come through this season.”
The NFL’s SVP of Public Policy, Brendon Plack, submitted a comment to the CFTC in late July arguing the regulator’s prediction market proposal falls significantly short of the standards required to address integrity risks. Locke reframed that as public confirmation that the category depends on exactly the infrastructure Genius has spent two decades building. “The largest league in America has put on record that this category runs on infrastructure,” he said on the earnings call. That reading is clever, but the NFL’s caution also means the single largest upside for Genius in prediction markets remains entirely conditional.
Expert Analysis: Confident Letter, Specific Blind Spot
We find the supply-side logic credible and the timing worth examining. Locke’s structural argument holds. A company positioned between competing operators, market makers and leagues carries genuine resilience to individual platform outcomes. The Michigan incident handed him the clearest live evidence yet that settlement cannot improvise its way to credibility, and the CFTC’s own framework proposals, placing weight on objective settlement and cooperation with governing bodies, push in exactly the direction that benefits official data holders.
What we find harder to absorb without question is the framing of Genius as a passive beneficiary of the regulatory conflict rather than an entity actively caught inside it. Two of its licensed subsidiaries were subpoenaed by Connecticut in the same enforcement sweep targeting the platforms whose legal survival Locke describes as commercially irrelevant to his business. Those subpoenas are informational, not accusatory. But a company whose entities appear on a state enforcement document four days before the CEO publishes a letter arguing the company sits above the fray is making a stronger claim than the circumstances fully support.
The NFL holdout is the variable the letter leaves unresolved. Every league partnership Locke cites as evidence of the sector’s direction, including Serie A, Liga MX and the UFC through Polymarket, sits well below the NFL in US commercial scale. Until that changes, the thesis that Genius wins regardless of regulatory direction is real but structurally incomplete. Locke describes three ways the company benefits from prediction markets. The fourth way, the one that would make the NFL paragraph genuinely transformative, is the one outcome Locke himself said not to expect this season.
Explore more:
Share
20 views
