DraftKings: Volume Grows, Margins Crack — Hold Until Predictions Pays

Rating: Hold

DraftKings (DKNG) swung from a $158 million profit to a $68 million loss in the second quarter. Revenue fell 5% to $1.44 billion while sports consumer volume (handle) grew 15% to $13.1 billion. The disconnect between engagement and profitability is the problem investors need to parse before deciding whether the 45% decline from the 52-week high is an entry point or a warning sign.

The quarter broke down into two separate problems: one temporary, one structural. Sports outcomes worked against DraftKings in the second quarter — the Knicks' NBA title run and U.S. World Cup results favored bettors. Management has previously estimated that a one-standard-deviation swing in sports outcomes can move revenue roughly $150 million in either direction. Last year's second quarter benefited from favorable results; this year didn't. That portion of the miss is noise.

The promotional spending surge is not noise. Sales and marketing expense jumped 38% year-over-year to $322.5 million. Average revenue per monthly unique payer (ARPMUP — the revenue each paying customer generates in a given month) fell 13% to $132. Sports net revenue margin (the percentage of handle that converts to revenue after payouts) collapsed from 8.7% to 6.8%. Adjusted EBITDA — earnings before interest, taxes, depreciation, and amortization — tumbled 62% to $114.6 million from $300.6 million.

Management attributes the margin compression to "increased promotional reinvestment" tied to customer acquisition in both Sportsbook and its new Predictions product. CEO Jason Robins called the quarter one of "real momentum" in handle, users, and engagement. The engagement metrics do support that claim: monthly unique payers grew 9% to 3.6 million, and iGaming revenue climbed 7.5% to $461.9 million.

But user growth funded by 38% higher promotional spend is growth at a price. The question isn't whether DraftKings can attract bettors. The question is whether it can do so without permanently eroding the margin profile that first turned the business profitable.

The Predictions pivot and its cost

DraftKings launched its federally regulated prediction-market product, DraftKings Predictions, in December 2025 and is planning $200 million to $300 million in investment this year. The product operates under CFTC oversight and is available in 38 states — including California and Texas, where traditional sports betting isn't yet live. Management views prediction markets as a potential $10 billion annual revenue opportunity.

The rationale isn't just growth. It's defense. Competitors like Kalshi and Polymarket saw combined trading volumes reach $63.5 billion in 2025, a fourfold jump. Kalshi has overtaken both DraftKings and FanDuel in the app download race. Management says it hasn't seen "material cannibalization" of its core sportsbook yet, noting that early Predictions impact has been limited to lower-margin customers.

But the second quarter shows what heavy investment in an unproven vertical does to the P&L. Full-year 2026 guidance of $6.5 billion to $6.9 billion in revenue and $700 million to $900 million in adjusted EBITDA deliberately excludes any Predictions revenue contribution. Management has signaled the core business should generate approximately $1 billion of adjusted EBITDA this year, with the $200 million to $300 million Predictions investment eating into the top of that range.

That math requires the core business to perform strongly through the fourth quarter — the traditionally largest quarter, which delivered $1.99 billion in revenue and $343 million in adjusted EBITDA in 2025.

What the valuation says

At a $12 billion enterprise value and 1.9 times trailing sales, DraftKings isn't priced at hypergrowth levels anymore. The stock has been cut roughly in half from its $48.78 high. Free cash flow over the trailing twelve months sits at $497 million, or about a 4% yield on enterprise value. That provides a floor.

But the forward picture is murkier. Forward P/E sits at 47.5 times, which assumes earnings reaccelerate after this quarter's loss. The trailing twelve-month EV/EBITDA multiple is 52 times — inflated because second-quarter EBITDA collapsed. Gross margin holds at 41.8%, and free cash flow margin at 9.0%, but operating margin has been squeezed to 0.6%. The company carries $3.7 billion in total debt against roughly $1 billion in cash, leaving net debt of $836 million.

Compared to traditional gaming operators like MGM Resorts (7.4 times EV/EBITDA) and Las Vegas Sands (11.2 times EV/EBITDA), DraftKings still commands a massive growth premium. That premium is defensible only if user growth monetizes without sustained promotional intensity and the Predictions investment converts to revenue within the next couple of years.

The case to buy and the case to wait

The bullish case is straightforward: sports outcomes are mean-reverting, the platform's engagement metrics are growing, the geographic footprint covers 53% of the U.S. population for sports betting, and the valuation has reset sharply from its highs. If Predictions converts its user base into a second profitable revenue stream, the multiple re-expands.

The caution case is equally direct: ARPMUP is falling while promotional spend accelerates. Predictions will cost $200 million to $300 million this year with no revenue contribution counted in guidance. The core EBITDA target of approximately $1 billion is an internal management estimate, not a commitment. If sports outcomes continue to swing against the house, or if promotional intensity becomes permanent to retain users in a competitive market, the margin trajectory tilts lower.

What changes the call

I'm sitting at Hold. The volume growth and user engagement are real, and the valuation reset has narrowed the gap between price and intrinsic value. But the margin compression in this quarter — driven by promotional spend, not just sports outcomes — and the material investment in an unproven product line create enough uncertainty that the current price doesn't offer a clear enough risk/reward to buy.

An upgrade to Buy would require two things: (1) a quarter where ARPMUP stabilizes or grows while promotional spend moderates, proving the margin profile is recoverable, and (2) early evidence that Predictions is generating meaningful revenue rather than burning through its allocation. The third quarter earnings in November will be the first test of whether Q2 was a sports-outcome blip or the beginning of a structural margin shift.

Until then, the $12 billion enterprise value reflects a business that's still earning its growth premium.