DraftKings predicts its predictions unit must turn profitable by 2027

DraftKings 2027 EBITDA guidance implies predictions unit must turn profitable

Jason Robins called it a pivotal moment, which is the kind of language CEOs reach for when the numbers are good enough to celebrate but complicated enough to explain. The Q3 report gave him both: $1.329 billion in revenue, up sixteen percent on the same quarter last year, 4.1 million monthly unique payers, and a prediction market platform that has drawn 1.6 million users since January. By most measures, a strong quarter.

The 2027 guidance is where it gets interesting. DraftKings is projecting revenue of $7.35 to $7.75 billion and EBITDA of $800 million to $1 billion. That EBITDA figure is the one worth holding up to the light. The 2026 EBITDA guidance sits at $175 to $275 million — a range that already reflects, as the company acknowledged, increased investment in Predictions and iGaming. The implied jump to $800 million-plus in a single year is not a rounding error. It is a structural claim: that the prediction market segment, currently consuming capital at a rate that cut $625 million from this year's earnings forecast, will contribute meaningfully to margin within twelve months.

I have seen this construction before in growth narratives — the year of investment followed by the year of harvest, with the harvest modeled more precisely than the conditions that would produce it. The 1.6 million users figure Robins cited is real engagement, but engagement and profitability are different animals, and the prediction market space is currently operating inside a legal environment that could reprice the entire business overnight.

The regulatory picture is worth pricing into that guidance, and I am not convinced DraftKings has done so fully. The company's predictions platform operates across a landscape where CFTC preemption authority is being contested in multiple federal courts, where state attorneys general are pursuing criminal charges against one of its main competitors, and where the question of which regulator actually governs event contracts remains genuinely unresolved. The 2027 harvest assumes stable ground. The ground is not stable.

What Robins has done, and what deserves credit, is commit publicly to a number. The $300 million investment in the predictions business is now on record with specific return expectations attached. If the EBITDA target is missed, analysts will know exactly which line to examine. That accountability is useful — it forces the company to make the case for the segment's economics rather than treating it as a speculative option.

My read is that the 2026 guidance range is honest and the 2027 range is aspirational in a way that underweights legal risk. The midterm results shifted the Senate, and Kalshi's federal preemption argument has now lost two appeals court rulings. A CFTC chair who issued the geofencing order is operating in a more hostile political environment than the one that existed when DraftKings built its 2027 model. Prediction market firms that are currently treated as federally regulated DCMs could find themselves reclassified by a commission that changes posture — or by courts that beat the commission to the conclusion.