Developing DraftKings Nearly Tripled Its Prediction-Market Bill. The More Aggressive Number Is the One It Gave for 2027.

Editor's note: TrueEdge builds odds tools that include prediction-market prices and earns affiliate commissions from licensed sportsbooks. Our standing position is that states should not criminalise federally registered exchanges. Weigh the conflict.

DraftKings told investors on Thursday that it will spend $650 million to $700 million on its prediction-market business this year. In August the plan was an "incremental $200 million-$300 million." Adjusted EBITDA guidance for 2026 fell from $700 million to $900 million down to $175 million to $275 million, and revenue guidance fell from $6.5 billion to $6.9 billion down to $6.35 billion to $6.55 billion. The stock dropped hard after the close, though the coverage caught it at different moments: 18.4% to $19.20 in Benzinga's report, 20% at Investing.com and 23% when Casino.org filed. The 52-week low is $18.36.

The spending is the headline, and DraftKings wants it read as a choice. We think that is the less interesting part of the release. The company has now asked investors to accept two harder things: that its core business slipped this year, and that next year a product with almost no reported revenue supplies most of its growth.

The cut, split into its parts

Take the midpoints. EBITDA guidance fell from $800 million to $225 million, a drop of $575 million. Predictions investment rose from a midpoint of $250 million to $675 million, an increase of $425 million. That leaves about $150 million the prediction-market story does not obviously cover.

DraftKings' own words narrow it. In August, its release said the core business "remains on track to generate approximately $1 billion of Adjusted EBITDA this year." Thursday's figure is $875 million to $925 million. So the core guide came down by about $100 million, and "approximately" and the old ranges do not reconcile to the dollar. A company describing a pure investment decision would not also be shaving the business that funds it.

The third quarter shows the shape of it. Revenue rose 16% to $1.329 billion, under a consensus that outlets listed at $1.398 billion and $1.41 billion. Sports revenue rose 31.5% to $783.6 million on $13.9 billion of Sports Consumer Volume, and Sports Net Revenue Margin improved to 5.6% from 5.2%. That line was healthy. iGaming grew only 2.1%, to $460.9 million, and adjusted EBITDA came in at negative $316.9 million against negative $126.5 million a year earlier. The income statement shows where the money went: sales and marketing of $605.7 million against $360.4 million a year ago, up 68%, or about 46 cents of every revenue dollar.

Subtraction gives a rough sense of the pace. With $650-$700 million for the year and $175-$225 million planned for the fourth quarter, about $475 million went out in the first nine months, already well above the top of the original $200-$300 million plan. Guggenheim's pre-report assumption of about $400 million, already above the company's guide, turned out to be light.

Where 2027's growth is supposed to come from

The new 2027 guide is the piece investors should read twice. Revenue of $7.35 billion to $7.75 billion is up 17% at the midpoint and sits only slightly under the $7.59 billion consensus Casino.org cites. Broken apart, it looks like this: core revenue of $6.7 billion to $7.0 billion, and Predictions revenue of $650 million to $750 million.

From the 2026 midpoint of $6.45 billion to the 2027 midpoint of $7.55 billion is $1.1 billion of growth. Core supplies about $400 million of it. Predictions supplies about $700 million, or roughly 64%, by our arithmetic. This year, DraftKings says, Predictions revenue will be "immaterial" because new-customer promotions offset most of the gross revenue.

So the company is forecasting a swing from approximately nothing to about $700 million in one year, in a product category whose legal footing lost ground twice this week, in the Second and Ninth Circuits. Needham's estimate, cited in TIKR's preview, had Kalshi at 76% of sports prediction-market volume in the NFL's opening week and DraftKings' exchange at 3%. Those are one week's numbers from September, and DraftKings has spent heavily since. They still describe the size of the climb.

The second assumption is on the profit side. Core adjusted EBITDA is guided to $1.15 billion to $1.25 billion next year, up about a third at the midpoint on about 6% core revenue growth. That takes the core margin, on our division, from about 14% to about 17.5%. A day earlier, Flutter told investors the opposite about its own world: category media spending up more than 50% in October, and marketing and promotion intensity continuing through 2027. We do not know yet how DraftKings squares those two pictures. It may have an answer on Friday's call, and that is the first question we would ask.

The case that the market has this wrong

The strongest argument for DraftKings is the one Jason Robins made in writing. He said the 2026 cohort should "pay back in approximately 20 months, well below our two-year payback target," and that more than 1.6 million customers have engaged with Predictions this year, up from more than 600,000 in August. "The third quarter marked a pivotal moment for DraftKings," he said, adding: "we believe we are set up for an exceptional 2027." CFO Alan Ellingson said the core business "continues to generate significant free cash flow, which we are deploying to fund growth in our Predictions business at attractive payback periods."

Take that seriously. If 20 months is right, $675 million spent in 2026 is an acquisition cost with a defined return, spent during the one land grab this industry may get. Predictions investment falls to $250 million to $350 million next year and total EBITDA rises to $800 million to $1 billion. The stock was already more than a third below its January high before the report, and Needham argued days ago that the selloff had gone too far. A buyer at $19 is paying for fear about a line item that the company says shrinks by more than half.

Our answer is that the 20-month figure cannot yet be checked. A 2026 cohort has, by definition, less than a year of behaviour behind it, so every month beyond that is a projection. The figure also comes after customers have been paid to arrive. "Engaged" is the company's word, and nothing we read defines it. Divide the nine-month spend by 1.6 million and you get roughly $300 a head, which is not a true acquisition cost and is only a scale check. What matters is how many of those customers remain active, and at what margin, once the promotions taper. That is the number DraftKings has not shown, and it is the number 2027 depends on.

What would change our mind

Three things, and two of them could arrive quickly. Friday's call could disclose retention by cohort, which would turn the payback claim from an assertion into evidence. Next quarter's Predictions gross revenue, net of promotions, will show whether the immaterial line is starting to move toward a run rate that makes $650 million plausible. And if the core margin expands while Flutter's does not, we will have been wrong about the media bill being an industry-wide tax, a test we set in our Flutter piece on Wednesday. The 68% jump in sales and marketing points our way, but the release does not split that line between core and Predictions or discuss media rates, so the test stays open until the call.

For now, the plain reading of Thursday is that DraftKings cut this year's EBITDA guidance by more than two-thirds and asked investors to wait for next year's exchange revenue to pay it back.

Gambling involves risk. If you or someone you know has a gambling problem, call 1-800-GAMBLER.