DraftKings reported its second quarter on Thursday after the close, and it was ugly. Revenue of $1.443 billion fell 4.6% from $1.513 billion a year earlier, against a Street looking for roughly $1.51 billion. The company swung to a net loss of $67.6 million from net income of $157.9 million. Adjusted EBITDA fell 62%, to $114.6 million. Shares dropped as low as $21.40 overnight.
Then management held the call on Friday morning, and the stock closed up 8.4% at $24.03 on 36.1 million shares — roughly 173% above its three-month average volume.
What flipped it was not in the earnings release. It was a sentence from chief financial officer Alan Ellingson: the core business “remains on track to generate approximately $1 billion of Adjusted EBITDA this year.” The company’s actual guidance, reaffirmed the same day, is $700 million to $900 million. Subtract one from the other and DraftKings has just told investors it intends to spend somewhere between $100 million and $300 million on prediction markets in 2026 — without ever printing that figure.
The quarter was outcomes, not demand
Sports Consumer Volume — the total settled on sportsbook and prediction markets — rose 14.5% to $13.14 billion. Sports revenue fell 10.6% to $891.9 million anyway, because the net revenue margin on that volume dropped to 6.8% from 8.7%. Management put the outcome headwind at about $80 million of revenue, driven by the Knicks’ championship run (New York is its largest sportsbook state) and bettor-friendly World Cup group-stage results.
The engagement lines held. Sportsbook handle grew 11%, monthly unique payers rose about 9% to 3.6 million, and iGaming revenue climbed 7.5% to $461.9 million. Revenue per payer fell 13%, or $19, to $132 — what happens when hold collapses and promotional spend rises. Sales and marketing hit $322.5 million, up 38%.
Our take: A sportsbook’s revenue is volume times hold, and hold is the half it does not control in any given quarter. Volume grew 14.5%; that is the durable number. Margin moved because one basketball team in one state won. The line that deserves real scrutiny is the extra $89 million of sales and marketing — that one was a decision, not a bounce.
The number that isn’t in the guidance
Ellingson framed the $1 billion core figure as “financial flexibility to invest behind the significant opportunity that we are seeing in Predictions.” DraftKings launched the product in December 2025, launched its DKeX exchange in June, and won approval in July to operate as a futures commission merchant — which lets it move most major sports prediction volume in-house and keep more of the economics.
The scale so far is modest. Citizens analyst Jordan Bender estimated gross prediction-market revenue of about $7 million in the quarter, with July volume implying roughly a $65 million annual run rate excluding exchange and market-making. Chief executive Jason Robins said volume grew nearly fivefold from April to July. The spend is being sized at up to $300 million against a business running in the tens of millions.
Robins also argued the category is not cannibalising the sportsbook: roughly 1% customer overlap with the largest prediction-market operator in states with legal online betting, where he estimates 80–90% of volume comes from professional syndicates. The customers DraftKings wants are in California and Texas, where it cannot take a bet.
Guide low, hand the market a bigger number verbally, let it choose — Atlassian did a version of this last quarter. It works until it doesn’t: AppLovin hit its guidance and lost a fifth of its value. And the split here is the one DoorDash is managing too — growth and monetisation are no longer the same story.
What to watch
- Fourth-quarter leverage. Half-year adjusted EBITDA is $282.5 million against a full-year floor of $700 million. Most of the year has to land in one NFL season.
- Whether “core adjusted EBITDA” gets promoted. If it turns up as a reported metric in the third-quarter release rather than a call remark, the yardstick changed mid-year.
- Handle in September and October. July handle was up 20%. That is the cleanest read on whether the World Cup cohort stuck.
- Advocacy spend. $19.9 million in the quarter, $46.2 million in the half, zero a year ago — and all of it added back in adjusted EBITDA.
Bull case and bear case are the same sentence: DraftKings is using a profitable sportsbook to fund a land grab in a category it does not yet lead. Whether that is discipline or a subsidy depends on what Predictions looks like in February.