Developing DraftKings Put a Number on Losing at the Supreme Court. Wall Street Took the Stock Down 29% Anyway, Its Worst Day Ever.

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

DraftKings held its third-quarter call at 8:30 a.m. ET on Friday, November 6, and the stock, already down about 22% after Thursday evening's release, kept falling while Jason Robins was still talking. InGame had the low at $16.60, down 29%, which it reports as the largest one-day fall in the company's history, worse than March 12, 2020, the day the leagues shut down. By InGame's count the shares have lost 65.6% since the end of August 2025 and the company is now worth about $8.4 billion. Flutter was down 8% and Penn 4% in early trading, per 24/7 Wall St. We wrote up the release on Thursday: $650-$700 million of prediction-market investment this year, a 2026 adjusted EBITDA guide cut from $700-$900 million to $175-$275 million, and a core guide that slipped too.

Markets reopen Monday with the weekend's reading done. The call added three things the release did not have, and they are what the weekend's arguments have been about. Robins put a payback number on the scenario in which the Supreme Court takes sports contracts away. He disclosed what the exchange keeps on each trade, and how much of its volume is market makers trading with each other. And the analysts, for the first time in this cycle, argued with the premise rather than the spend.

The downside case, in the company's own words

DraftKings' base case, per Robins, is a Supreme Court decision in the first half of 2028. Asked by Jefferies' David Katz about outcomes between the extremes, Robins said the company had modelled "the most draconian likely scenario": every sports contract ends in July 2028. Covers' Ryan Butler puts the date at July 31, 2028. In that case, Robins said, the 2026 customer cohort is fully paid back by then and the 2027 cohort is more than half paid back. "We still think that cohort should pay back in four years or less." Butler's report has that at 41 months, and the 2026 cohort recovering its marketing cost by May 2028.

Citi's James Hardiman asked about 2028 directly. Robins: "probably somewhere north of $1 billion in 2028, were we to see sports predictions stick around for the whole year," and if not, "we would significantly dial back the marketing." Either way, "it should be positive." To Citizens' Jordan Bender, who asked whether next year's $250-$350 million was a ceiling, Robins said that "if there is an early negative signal, we might actually pull back on some of that investment." Bender's own note, per InGame, called this year's spend "beyond even the worst-case scenario."

We have not heard a licensed operator put a Supreme Court loss into a payback model on a public call before, and it is to DraftKings' credit that it did. What the model lacks is inputs. The 10-Q says prediction revenue is "included in Sports revenue" and does not quantify it. No cohort cost, margin or churn figure was disclosed. The 20-month payback is the company's claim about its own arithmetic, and nobody outside the company can check it.

The court calendar is also looser than "mid-2028" suggests. We read the Supreme Court docket in Flaherty v. KalshiEX on Sunday. Kalshi's response to New Jersey's petition was due October 8, then November 9, and on October 30 the Clerk moved it again, to December 7. Eleven amicus briefs sit on the petitioners' side, from a bloc of states led by Ohio to the NFL to Gary Gensler, and none yet on Kalshi's. Kalshi's own clock to petition in the Nevada case runs to about February 2, by our count from Wednesday's Ninth Circuit denial. A decision in the first half of 2028 requires the Court to take a case this term or early next. If it declines both, the circuits stay split and the July 2028 scenario becomes a patchwork rather than a cliff, which is neither the base case nor the draconian one.

What the exchange actually keeps

Robins gave the economics of DKeX for the first time. The take rate on trades was "approximately 11.1%" in September and October, and "We captured approximately 2 thirds of that gross revenue, third party market makers capturing the rest." By our arithmetic that leaves DraftKings about 7.4% of consumer volume before promotions, against a Sports Net Revenue Margin of 5.6% on the quarter's combined sportsbook and prediction volume.

The volume figures are the more revealing pair. Annualised consumer volume in October was $10.1 billion, nearly three times July. Annualised total volume was $56.8 billion, more than five times July. The difference is market makers, which means roughly 82% of what trades on DKeX is not a customer. Robins said DraftKings has "become the clear number 3 predictions exchanged by total volume," which we read as "exchange" and which, by that definition, counts the market makers. Combos were 27% of consumer volume in October and more than 70% of customers have traded one. About 20% of customers have traded a non-sports market, crypto is approaching a double-digit share, and crypto markets move onto DKeX within about a month.

Then the net. Robins said this year's prediction net revenue is "pretty close to zero" and that "almost entirely 90% of that is due to customer new customer promotions." CFO Alan Ellingson used the release's word, "immaterial," and guided fourth-quarter investment at "$175 million to $225 million."

The 2027 number, and what has to happen for it

Next year's guide is $650-$750 million of prediction revenue on $250-$350 million of investment. Set the $700 million midpoint against October's $10.1 billion annualised consumer volume and it comes to 6.9% of volume, close to the 7.4% net take. So the guide is, roughly, October's run rate with the promotions switched off and no growth, or growth with the promotions kept on. It cannot be both. The 2026 cohort is "run-rating" about $350 million of annualised gross profit, per the call, and 1.6 million customers have engaged with the product this year.

Promotional spend therefore has to fall from about $675 million to about $300 million while volume at least holds. That is the bet investors sold on Friday, and it is not an unreasonable thing to doubt. Before the report, analysts tracked by TIKR expected about $1.13 billion of 2027 EBITDA; 24/7 Wall St. has the Bloomberg consensus at $1.12 billion. The guide is $800 million to $1 billion.

The pushback

UBS's Robin Farley asked the question that cuts closest. If sports contracts survive, DraftKings gets sports volume through its sportsbook anyway in the states where it is licensed, and the non-sports paybacks look long. Robins' answer: "I do not think we can count on every state suddenly legalizing sports betting." He went further with Covers' framing of Tuesday's elections: "It could also be states just take a while or don't do anything at all or decide to tax prediction." On non-sports: "we put basically no effort into it, and it is already approaching 10% of our sports volume," and it "could be you know, 25%, 30%, maybe even more of the size that sports is today." He added, "it was not what I was thinking when we got into this."

That is an honest answer, and it narrows the thesis. DraftKings told Covers that 99% of its prediction volume comes from roughly two dozen states where it has no sportsbook. The exchange's value is therefore concentrated in the large states that still have no legal online book, a list Nebraska's voters added themselves to on Tuesday, plus whatever non-sports becomes. In licensed states it is cannibalisation with a promotion attached. Goldman's Eric Sheridan asked what remains if the product goes away; Robins said the brand halo was "likely minimal" but cross-sell and lifetime value would stay. Morgan Stanley's Stephen Grambling asked whether marketing per active customer was eroding the core; Robins tied the rise to early-season NFL state launches.

On the states, Robins said "most states are waiting to see what the courts decide before taking more severe action, but you never know," and that DraftKings has "managed to stay completely out of the litigation everywhere." Asked about Thursday's CFTC order on operating state by state, he said, "I would not make anything of that." We argued on Saturday that the order's explanation clause will be answered with letters, and Robins' answer is the first such letter in draft.

The strongest case for DraftKings

The bull case is straightforward and Robins made it. The core business is guided to $1.15-$1.25 billion of 2027 EBITDA, which by itself exceeds what analysts expected from the whole company. The 2026 cohort pays back in 20 months against a two-year target. Cash is $1.1 billion, Ellingson expects "significant excess free cash flows in 2027 even after funding our prediction investments," and a stock at its lowest since early 2023 already prices in a loss at the Court. Robins' line was that "we believe we are set up for an exceptional 2027."

Most of that is defensible. What we would answer is that the two unauditable claims carry the valuation. The payback is undisclosed arithmetic. The $700 million rests on a promotion that has to end without the customers leaving. iGaming, the other growth story, grew net revenue 2% in the quarter while customer acquisition rose 76%, which is the same shape. And the comparison the market is drawing is not with DraftKings' own core but with Flutter, whose US business earned $3 million of adjusted EBITDA last quarter and whose CFO, Rob Coldrake, said prediction spending would continue "at similar or higher levels." Penn's Aaron LaBerge took the other side of the whole trade on Thursday, saying his company is "not getting sucked into the noise of prediction markets," and Penn fell 4% anyway. Two of the three largest US operators are now spending to defend against a product that neither says it needs in its licensed states.

What would change our mind is a fourth quarter in which prediction net revenue turns positive while investment lands inside the $175-$225 million range. That would show promotions coming off without volume following them out. Until then, "exceptional 2027" is a forecast about a court and a promo budget, and on Friday the market priced it as one.

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