DraftKings triples investment in Predictions and publishes downside model

DraftKings increased its investment in Predictions for 2026 to $700m and published a model showing how that investment would be recovered if event sports contracts ended before July 31, 2028. The company said that under its downside assumption a Predictions customer would generate 20% of current gross profit without sports contracts, extending payback from 18 to 41 months.

What DraftKings announced

In its third-quarter shareholder letter and in the earnings presentation published on Thursday, November 6, 2026, DraftKings reported it increased adjusted EBITDA investment for Predictions in 2026 by $50m, moving from $650m–$700m to $700m. The company also modelled a scenario in which sports contracts end on July 31, 2028 and detailed the implications for investment recovery.

"Even if sports contracts were to end in July 2028, we estimate our 2026 cohort would be fully paid back before that date," wrote Jason Robins, CEO of DraftKings, in the letter.

Robins added that Predictions customers acquired in 2027 would be more than 50% paid back by that date and should be fully paid "within four years without sports contracts." The company held its earnings call on Friday morning after publishing the materials.

How the downside model works

The presentation lays out the central assumption of the no-sports scenario: from August 2028 DraftKings projects adjusted gross profit per customer would fall to 20% of the level achieved with active sports contracts. The company attributes that recovery to cross-sell into non-sports markets, Pick6 and Lottery.

Under that assumption, the 2027 cohort shifts from payback at month 18 to month 41. DraftKings indicated that each customer's sports contracts would end between month 8 and month 19, depending on signup date, and at month 13 for the average customer. The company did not explain why it chose July 31, 2028 as the cut-off.

Size of the bet and financial position

DraftKings changed its adjusted EBITDA guidance for 2026. In August the company maintained a range of $700m–$900m and described the investment in Predictions as a gap of $200m–$300m versus a core business of roughly $1bn. In Thursday's update the total guidance was $175m–$275m, with Predictions investment now $650m–$700m, and core business EBITDA at $875m–$925m.

Sales and marketing expenses for the third quarter reached $605.7m, an increase of 68% year-on-year and nearly double the $322.5m spent in the second quarter. Adjusted EBITDA for the quarter was a loss of $316.9m, a year-on-year decline of 23.8%.

DraftKings tied the marketing increase to the returns it expects to generate: more than 1.6 million customers have interacted with Predictions year-to-date, versus more than 600,000 reported in August.

Composition of volume and take rate in Predictions

The company broke out the economics of Predictions for the first time. In September and October, trading on DKeX produced a take rate of 11.1% on consumer volume, which includes exchange fees, broker commissions and market-making returns. DraftKings retained 7.3% of consumer volume acting as exchange, broker and market maker; third-party market makers captured the remaining 3.8%.

DraftKings noted that figure approximates its sportsbook hold, which "eclipsed 11%" in the quarter. Total annualised traded volume in October was $56.8bn, of which market makers represented $46.7bn. Annualised consumer volume, the money customers actually traded, was $10.1bn on a linear basis.

The company also reported that nearly 20% of Predictions customers have traded non-sports markets. DraftKings said per-customer volume in non-sports is at sports levels and that crypto leads that category, approaching a double-digit share of total consumer volume in recent weeks. DraftKings intends to move crypto markets to DKeX, its own exchange, over the next month.

iGaming, core and 2027 guidance

Growth in iGaming also carried a cost: customer acquisition in iGaming rose 76% year-on-year in the third quarter. Gross gaming revenue increased 7%, but net revenue grew only 2.1% to $460.9m, because the company invested in promotions for those new customers. DraftKings operates iGaming in five states covering roughly 11% of the U.S. population, so the expansion reflects deeper penetration in active markets, not new launches.

The core business continued to fund experiments: sportsbook net revenue grew 36% on a handle that rose 12%. Robins used those numbers to assert that "prediction market operators are having no discernible impact on our Sportsbook business." DraftKings said it maintained cash around $1.1bn during the investment cycle.

For 2027 the company guided revenue of $7.35bn–$7.75bn and adjusted EBITDA of $800m–$1bn, including $650m–$750m in Predictions revenue and $250m–$350m in Predictions investment.

Regulatory context and litigation

The July 31, 2028 date closely follows the next Supreme Court term; New Jersey filed in September a petition for the Supreme Court to review the Third Circuit Court of Appeals decision in the Kalshi case. The Ninth Circuit denied Kalshi's rehearing petition in the Nevada case on Wednesday. A Supreme Court ruling this term would reduce the cushion DraftKings has modelled.

DraftKings also excludes upside in its model: it does not assume future legalisation of online sports betting or iGaming in the scenario, even though it reports spending to expand gambling markets. Adjusted EBITDA excludes $59.1m in defence and legal activity to date, which the company attributes mainly to efforts to legalise iGaming, the failed Nebraska initiative and other lobbying activities.