DraftKings Inc. (NASDAQ: DKNG) — Profitable At Last, Priced for the Prediction-Market Disruption It Is Racing to Own

Independent fundamental research note. Report date: 2026-06-20. Price referenced ~$26.39 (close 2026-06-18).

⚡ Claude’s Take

This block is the author’s own independent opinion and general information, not investment advice. The analysis that follows is deliberately position-free and carries no price target; this section is the single exception where a view is expressed.

Verdict: HOLD / accumulate-on-weakness / not-a-short. Low-to-medium conviction. Constructive accumulation zone ~$20–24 (≈1.7–1.9x EV/forward-sales, ~10–12x “core” adjusted EBITDA — roughly the March-2026 panic low and the bottom of the stock’s own ten-year sales-multiple band). The risk/reward inverts above ~$40 (≈3x EV/sales), where the price again demands the prediction-market bull case to underwrite it. At $26.39 the name sits in the middle: cheap on sales, un-cheap on the questions that actually matter.

DraftKings is the rarest thing in this market — a hyper-growth story that has actually crossed into GAAP profitability and free cash flow (FY2025: first profitable year, +$508M FCF; revenue $615M→$6.05B in five years) — and the market has simultaneously de-rated it to the 4.8th percentile of its own ten-year price-to-sales history, the cheapest it has ever been on sales. That gap is the trade. The reason the gap exists is not silly: the betting profit pool is being taxed away by states with no operator ability to pass it through (the Aug-2024 surcharge that collapsed in 24 hours is the cleanest proof DKNG has no real pricing power), and prediction markets (Kalshi, Polymarket, Robinhood) are attacking the licensed, taxed channel from outside under federal CFTC rules that — as of the June-10-2026 proposal — look like they will entrench the arbitrage, not close it. The market is pricing a structurally-threatened, hold-volatile, politically-contestable rent. It is not wrong to worry. But at ~2.1x EV/sales it is now paying ~nothing for either the profitability inflection that is already in the financials or for DKNG’s own ~$250M bet (Railbird) to win the very disruption that scares everyone — and DKNG, uniquely, has the brand, the funded customer base, the pricing models, and now a federally-regulated exchange to do it. This is a high-beta, de-rated growth name that found a floor on profitability — a falling knife that has, for now, landed on a real balance sheet. I would not chase it, but I would own a starter and add into tax-panic or hold-driven air-pockets.

Framing: de-rated growth/quality-at-a-price with embedded regulatory optionality — not deep value (negative tangible book; medium-quality narrow moat; governance and the insider tape are genuinely bad) and not a momentum name (down ~63% from its 2021 peak, ~31% trailing-12-month relative strength, negative alpha). The factor tape agrees: DKNG trades like a high-beta, high-volatility, small-cap-flavored, credit-sensitive risk-on vehicle whose nearest neighbors are software/cloud ETFs and Block — not Caesars or MGM. Conviction: low-to-medium (the central question — whether prediction markets entrench DKNG’s margins or commoditize its core — is genuinely unresolved, and ~75% of the stock’s variance is idiosyncratic). Flips bullish on: evidence that Predictions scales profitably and the final CFTC framework lets licensed operators co-opt the channel (margin-up, all-50-states). Flips bearish on: a prediction-market take-rate race-to-the-bottom that commoditizes OSB, a New-York-style tax shock in a big state, or Predictions spend ballooning past $300M with no payback. Tag: “The house finally turned a profit — just as someone proposed changing the rules.”

📈 Stock Price Action — Five-Year Event Map

Factual price history. Price moves are FACT; attributed drivers are INTERPRETATION. No recommendation, no target. Source: five-year daily price history; earnings/filings/news cross-reference.

DraftKings has round-tripped a generation of sentiment in six years: from a ~$10 SPAC vehicle to a $71.98 all-time high (Mar-19-2021) at the peak of the pandemic-retail-SPAC mania, down ~84% to a $11.39 trough (end-2022) in the rate shock, back up more than 4x to $53.49 (Feb-14-2025) as profitability came into view, and then a long ~50% slide to a 52-week low of $20.72 (Mar-27-2026) on tax and prediction-market fears, before a partial recovery to $26.39 today. The stock now sits ~63% below its all-time high, in the lower third of its 52-week range ($20.72–$48.23), and on the cheapest price-to-sales multiple in its public life despite revenue having grown ~10x since the peak.

Cycle narrative. (1–2) DraftKings came public via the Diamond Eagle (DEAC) SPAC in April 2020 and rode the pandemic-retail-SPAC wave to a $71.98 peak — a pre-revenue-scale “story stock.” (3) The 2021–2022 rate shock punished every cash-burning growth name; DKNG, then losing ~$1.5B of EBITDA a year in the promotional land-grab, fell 84%, with a June-2021 Hindenburg short report on SBTech’s black-market exposure adding fuel. (4) From 2023 the narrative flipped to discipline: marketing efficiency improved, the first adjusted-EBITDA-positive quarters arrived, revenue more than doubled again, and the stock 4x’d to $53.49 by February 2025. (5) The 2025–2026 de-rate is the heart of this memo: a brutal Q3-2025 (−$257M GAAP net loss on customer-friendly sports outcomes, illustrating hold volatility), the Illinois per-wager tax and New Jersey/Illinois rate hikes signalling the states will keep taking more, and the rise of CFTC-regulated prediction markets threatening the licensed channel — together cut the stock 61% to a $20.72 low. (6) Since late March, traction in DraftKings’ own Predictions product, the 2026 FIFA World Cup catalyst, and Street upgrades (UBS to $49) drove a bounce, partially given back on a mid-June Fed-hawkish risk-off move that hit all high-beta names. The stock enters mid-2026 cheap on sales, profitable for the first time, and squarely in the cross-hairs of an unresolved regulatory contest.

1. Executive Summary

DraftKings is one of two dominant operators in US online sports betting (OSB) and a top-two force in online casino (iGaming) — a business that scaled revenue from $615M (2020) to $6.05B (2025), a ~58% five-year CAGR, and in FY2025 crossed into its first GAAP-profitable year (+$3.7M net income, $260M EBITDA, +$508M free cash flow). FY2026 guidance is $6.5–6.9B revenue / $700–900M adjusted EBITDA, and Q1-2026 grew revenue 17% with adjusted EBITDA up 64%. After a decade of cash burn, the operating model has visibly inflected: gross margin is rising (parlay/same-game-parlay mix lifts structural hold), marketing efficiency is improving, and the company is generating real cash.

The business is a FanDuel (Flutter)/DraftKings duopoly — the two control roughly 68% of US handle — and the competitive cycle has turned in the incumbents’ favor: ESPN Bet (Penn) shut down after a ~$2B failure, Fanatics’ promo-bought share is melting, and BetMGM is restructuring. Well-funded challengers have proven they cannot buy their way into the top two. That is a real economies-of-scale barrier — and it is financially load-bearing (it is why DKNG reached profitability while sub-scale peers bleed).

But the moat is narrow and partial. Customer captivity is weak: multi-homing is rampant, switching costs are ~one app download, and the Aug-2024 “gaming tax surcharge” — DKNG’s attempt to pass a tax through to customers — collapsed within 24 hours once FanDuel declined to follow. A business with real pricing power can pass costs through; DraftKings demonstrably cannot. Meanwhile the profit pool itself is politically determined and shrinking at the margin: states are raising OSB taxes (New York 51%, Illinois’s new graduated rate plus a per-wager fee, New Jersey 14.25%→21%), and the federal One Big Beautiful Bill Act capped gambling-loss deductions at 90%.

The defining issue — and the reason the stock is down ~63% from its 2021 peak and trades at the 4.8th percentile of its own ten-year price-to-sales history — is prediction markets. Kalshi, Polymarket, and Robinhood offer CFTC-regulated sports “event contracts” that replicate betting while bypassing state licensing and the 20–51% state tax, in all 50 states (including California, Texas, and Florida, where OSB is illegal). The courts have leaned toward federal preemption, and the CFTC’s June-10-2026 proposed rules would broadly permit sports event contracts — entrenching the arbitrage. DraftKings’ response is to become a prediction-market operator itself: it bought the Railbird exchange (a CFTC-registered designated contract market), is investing $200–300M in 2026, and is folding Predictions and Sportsbook into a single “Super App.” This is simultaneously the largest threat to DKNG’s taxed core and its only credible path to a genuine network-effect moat (an exchange has liquidity network effects that house-banked betting never had).

Capital allocation and governance are the weak links. Growth was bought largely with an over-valued equity currency (Golden Nugget Online ~$1.56B all-stock; Jackpocket ~$750M), leaving negative tangible book equity (~−$1.86B). The $1B→$2B buyback was executed at a ~$35.8 average — underwater at $26.39 — and is best read as a debt-financed offset to ~$339M/yr of stock-based comp rather than surplus-cash return. Founder/CEO Jason Robins controls 88.3% of the vote on ~2.1% of the economics, executive incentives gate on Revenue + Adjusted EBITDA with no ROIC or TSR hurdle, and the insider tape is ~110:1 net selling by dollar (~$362M sold vs. ~$3.25M bought since mid-2024, the latter entirely from two outside directors).

The investment question is not whether DraftKings is a good business — it is a medium-quality, narrow-moat duopolist that finally makes money — but whether the market’s de-rating has overshot a real but contested set of threats. At ~2.1x EV/sales the price embeds a lot of bad news and almost no credit for the profitability inflection that is already happening or for the prediction-market optionality DKNG is itself funding. The memo lays out the case on both sides; the body takes no position.

2. Business Overview

What DraftKings does. DraftKings is a US-centric digital sports-entertainment and gambling operator, headquartered in Boston and founded in 2012 (public since the April-2020 Diamond Eagle/SBTech SPAC merger). It makes money by taking the other side of bets (house-banked sportsbook and online casino) and, increasingly, by operating an exchange for event contracts. It reports as a single reportable segment but discloses revenue by product line (FY2025 10-K):

• Sportsbook (OSB) — the largest line, ~two-thirds-plus of revenue, live in ~25+ states. Revenue ≈ handle × hold, where handle is dollars wagered and hold (also “net revenue margin”) is the share the house keeps after paying winners and promotions. National hold runs ~10–11%; DKNG’s structural hold is rising as the mix shifts toward parlays and same-game parlays (SGPs), which carry far higher theoretical hold than single-game bets. Q1-2026 net revenue margin rose 140 bps to 7.8%, with parlay handle mix up ~300 bps.
• iGaming / online casino — higher-margin and more recurring than OSB (no sport-outcome volatility; slots/table games have fixed mathematical hold), but live in only ~7–8 states (NJ, PA, MI, WV, CT, DE, RI and recent additions). Management openly concedes DKNG has lagged the market here and is re-investing (new “Flex Spins” promo, slots-first acquisition).
• Daily Fantasy Sports (DFS) — the legacy 2012 product; mature, small, slow-growing, but the historical brand and customer-acquisition seed.
• Media (the Vegas Stats & Information Network, VSiN) and Lottery courier (Jackpocket, acquired 2024) — adjacencies that feed the funnel.
• DraftKings Predictions / Railbird (NEW) — CFTC-regulated event contracts (prediction markets), launched late 2025 and being scaled aggressively in 2026, combined with Sportsbook into a unified “Super App” and to be reported as a single “Sports revenue” line from Q2-2026.

How the money is made — and why the model matters for the moat. OSB and iGaming are house-banked: the operator is the counterparty and earns the vig/hold. This is not a two-sided exchange, and the distinction is central to — house-banked betting has no network effect (more users do not improve the odds or liquidity any individual user sees), whereas prediction markets (an exchange model) do. A meaningful part of the strategic pivot is an attempt to bolt a network-effect business onto a core that structurally lacks one.

Recurring vs. non-recurring. Revenue is recurring in the sense that customers re-bet continuously, but it is not contracted and is highly seasonal (Q4 NFL-driven; Q1 March Madness/NBA) and volatile quarter-to-quarter on sport outcomes. The clearest illustration is FY2025’s own quarters: Q1 −$34M net income, Q2 +$158M, Q3 −$257M (a quarter of customer-friendly outcomes — favorites won, the house paid out — that flipped a profitable run to a large loss), Q4 +$136M. The full year netted to +$3.7M. Revenue is real and growing, but the earnings stream is lumpy and partly exogenous to management’s control.

The 10-K saturation signal. The FY2025 10-K mentions “prediction markets” 83 times, “event contracts” 24 times, and “Railbird” 58 times — versus essentially zero a year earlier. In a single filing cycle, prediction markets moved from absent to both a central strategic theme and a top-tier risk factor. That is the tell that the entire forward thesis now pivots on this adjacency.

Verdict (Business Overview): A genuine category leader with a real, scaled, now-profitable revenue engine — but a single-country, regulation-gated, structurally house-banked business whose earnings are seasonal and sport-outcome-volatile, and which is betting heavily on a new, legally-contested adjacency to defend its core.

3. Industry Dynamics

3.1 Market size and the penetration runway

US regulated sportsbooks processed ~$165.6B in handle in 2025, producing ~$16.8B in gross gaming revenue (GGR) at a ~10.15% national hold, and paying ~$3.66B in state taxes. The combined North American online gambling market (OSB + iGaming) was ~$16.6B in 2024, growing roughly 12% annually toward ~$33B by 2030 on third-party estimates.

The penetration runway is the core bull pillar and it is genuine but back-end-loaded. Only ~30 states have legal online sports betting; iGaming is legal in only ~7–8. Roughly half the US population — including California and Texas (~65M people, ~20% of the country) — has no legal OSB, and ~90% has no legal online casino. The problem is that the highest-value unbet states are precisely the ones structurally hardest to legalize: California ballot measures drew <20% support in 2022 amid tribal opposition, and Texas faces constitutional and legislative blocks. The TAM is real, but the next leg of it is not in management’s control and may take years. Management’s own framing at the March-2026 Investor Day put the 2030 gross-revenue opportunity (the addressable pool, not DKNG revenue) at $55–80B, with a long-term adjusted-EBITDA margin target of ≥30%.

3.2 The duopoly and competitive intensity — Greenwald’s market-share-stability test

US OSB is a FanDuel (Flutter) / DraftKings duopoly. By handle as of early-2026, DraftKings runs ~36% and FanDuel ~32% (FanDuel typically leads on GGR/revenue), for a combined ~68% of dollars wagered — down from ~75% historically. The decisive evidence for a moat against new entrants is the failure of the trailing pack:

• ESPN Bet (Penn Entertainment) — Penn exercised its opt-out in November 2025, shut ESPN Bet and relaunched as theScore Bet; share collapsed to ~1.2%. The ESPN brand plus Disney distribution plus ~$2B of marketing could not break into the top tier. This is the single most important capital-cycle datapoint in the sector.
• Fanatics Sportsbook — spiked to ~13% share on promotional spend in 2025, then decayed to ~7% by early 2026 — share bought with promo dollars evaporates the moment the spending stops.
• BetMGM — a distant #3 at ~10–12%, restructuring.
• Caesars, bet365, Hard Rock — niche/regional.

Greenwald read. Greenwald’s primary moat diagnostic is market-share stability. The reading here is split: the top-two share is stable-to-consolidating (a scale barrier protects #1 and #2 — you cannot buy in), but within the duopoly, share oscillates with promotions and product, the signature of weak customer captivity. This is a partial economies-of-scale advantage with weak demand-side captivity — a real but incomplete moat (developed below).

3.3 The rising-tax structural threat — the core bear pillar

State OSB taxation is rising and idiosyncratic, directly compressing the profit pool:

• New York: 51% of GGR (the highest; ~36% of all US sports-betting tax, ~$1.3B in 2025).
• Illinois: moved from a flat 15% to a graduated 20–40% (Jan-2025) and added a per-wager tax (up to $0.50 per bet). The per-wager structure is uniquely damaging because it taxes volume regardless of margin and hits small-stake recreational bettors hardest.
• New Jersey: raised OSB tax from 14.25% to 21% (2025), with further hikes floated.
• Pennsylvania ~36%, Vermont ~31%.

The defining datapoint on pricing power: DKNG’s August-2024 “gaming tax surcharge” — a small fee on winning bets in high-tax states, explicitly designed to pass the tax burden to customers — was withdrawn within roughly 24 hours after FanDuel publicly declined to follow. In a true oligopoly with pricing power, the leaders can pass costs through. DraftKings tried and could not, because its only peer undercut it. That is direct, dated evidence that captivity is too weak to support pricing power, and it caps how attractive this industry can be: the profit pool is contestable both by the states (which keep raising the rate) and by the duopolists against each other.

Layered on top, the federal One Big Beautiful Bill Act (Jul-2025) capped gambling-loss deductions at 90% (from 100%), effective tax-year 2026, creating “phantom income” for high-volume and professional bettors — a modest but real headwind at the top of the customer pyramid, where a disproportionate share of handle originates.

3.4 Prediction markets — the existential threat and the chosen defense

This is the swing factor for the entire thesis. Kalshi, Polymarket, and Robinhood (which routes through Kalshi) offer CFTC-regulated event contracts — moneylines, spreads, and totals reframed as binary/derivative contracts. Because they are regulated federally by the Commodity Futures Trading Commission as swaps/derivatives, they argue they bypass state sports-betting licensing and state GGR taxes entirely, and they are offered in all 50 states (including CA/TX/FL) at a thin exchange fee instead of a 20–51% state tax. It is regulatory arbitrage on the tax base: the same economic product, a fraction of the tax, a national footprint.

The legal battle is live and, so far, tilting toward the prediction markets:

• State regulators in 10+ states issued cease-and-desist letters or sued; Arizona pursued criminal charges.
• Courts have leaned federal-preemption (pro-Kalshi): a February-2026 preliminary injunction held sports event contracts are likely “swaps,” and an April-2026 Third Circuit ruling strengthened preemption.
• The CFTC sued Arizona, Connecticut, and Illinois (April-2026) to block state enforcement — the federal regulator is actively defending the prediction-market operators against the states.
• CFTC proposed rules (June-10-2026) would generally permit sports event contracts (final scores, spreads, win/loss, season-long stats) while prohibiting micro-bets (a specific play/pitch, injuries, officiating). The comment period runs to July-27-2026.

The proposed framework is, on balance, bullish for the legality of the channel — it signals the CFTC intends to legitimize and federalize most sports event contracts, entrenching the tax-arbitrage rather than closing it.

DraftKings’ response is to join, not fight. It acquired the Railbird exchange (a CFTC-registered designated contract market) for ~$85M booked (up to ~$250M with earnouts), giving it a federally-regulated venue to list contracts, run market-making, and capture exchange fees. It is investing $200–300M in 2026 in DraftKings Predictions, launching a proprietary exchange and “combos,” and folding everything into a Super App. Early traction is real: Q1-2026 annualized Predictions trading volume topped $2.3B, customer-acquisition cost fell >80%, and market-making was profitable within months of launch.

The bull/bear on the pivot:

• Bull: Predictions is an adjacency DKNG can win. It brings the brand, the funded customer base, the pricing/risk tech, and now a DCM. An exchange has a real network effect (liquidity attracts traders attracts liquidity) — DKNG’s first genuine path to demand-side captivity. If the CFTC federalizes the channel, DKNG can offer betting in all 50 states at a lower effective tax, structurally raising margins and bypassing the state-tax bear case. Management says Predictions carries 10–30% higher gross margin than sportsbook.
• Bear: The pivot is defensive, dilutive, and may cannibalize the high-tax-but-licensed core while inviting financially-native exchange competitors (Kalshi, Polymarket, CME, Robinhood, Interactive Brokers) into DKNG’s lane. In an exchange, liquidity is the moat, and DKNG is late. Worse: if prediction markets win broadly, they commoditize the entire OSB model — why pay a 20–51% state tax when a federal exchange charges a thin fee? — a take-rate race to the bottom that could destroy the very profit pool DKNG just became profitable on. The $200–300M spend is a tax on 2026 EBITDA for a contested, uncertain return.

Verdict (Industry): Structurally mixed, tilting bad at the margin. Large, under-penetrated, and consolidating into a disciplined duopoly — but the profit pool is politically determined and being taxed away with no operator ability to pass it through, the best growth states are structurally locked, and a regulatory-arbitrage disruptor is attacking the licensed channel from outside. This is not a structurally attractive industry like a tollroad or a payment network; its rent is contestable from three directions at once (the other duopolist, the states, and prediction markets).

4. Competitive Position — The Moat Assessment

Running the candidates against Greenwald’s three genuine advantage types (supply/cost, demand/captivity, economies-of-scale+captivity):

(a) Economies of scale — REAL but partial. Fixed costs — the proprietary trading/risk platform (built in-house post-SBTech), national brand marketing, product R&D (the parlay/SGP and live in-game pricing engines), and promotional spend — are spread over a far larger GGR base at #1/#2 than at any sub-scale rival. The empirical proof is the tail’s failure (ESPN Bet → ~1%, Fanatics 13%→7%): you cannot buy into the top two. The advantage is financially load-bearing — strip scale out and the marketing/promo line alone makes the model unprofitable; with it, DKNG reached its first profitable year in 2025 while sub-scale peers bleed.

(b) Demand-side captivity / switching costs — WEAK (the moat’s fatal gap). Multi-homing is rampant — bettors line-shop across 3–4 apps for the best price or promo, and switching costs amount to one app download plus a deposit match. House-banked OSB has no network effect (the house is the counterparty). The August-2024 surcharge fiasco is definitive proof: DKNG could not raise effective price ~3.2% without instantly losing share. A business with real captivity passes costs through; DraftKings cannot.

© Brand / intangibles — moderate, asymmetric. “DraftKings” and “FanDuel” are genuine top-of-funnel brands that lower CAC versus challengers — a real intangible advantage at the duopoly level (why challengers can’t break in cheaply) but not a within-duopoly moat (it doesn’t stop a DKNG customer also using FanDuel).

(d) Data / pricing models — a product edge, not a structural wall. Better parlay/SGP pricing, faster in-game markets, and proprietary risk management let the top-two run higher structural hold (parlay mix rising; hold climbing year-over-year). This is margin-accretive and self-reinforcing with scale (more handle → more data → better pricing → more hold) — the most attractive part of the story — but it is replicable with capital and time, a gradient advantage rather than a wall.

(e) Network effects — ABSENT in OSB, POTENTIALLY REAL in prediction markets. The strategic crux: the pivot to an exchange (Railbird) is, in moat terms, an attempt to manufacture the network effect the core business lacks. Win exchange liquidity and DKNG gains its first true demand-side captivity; lose to Kalshi/Polymarket/CME and it has spent $200–300M/yr fighting on the disruptor’s home turf with the disruptor’s better DNA.

Greenwald synthesis. DKNG has a real but incomplete economies-of-scale advantage at the duopoly level, paired with weak customer captivity — below a true scale+captivity moat (the gold standard). The scale barrier protects the franchise from new entrants (ESPN/Fanatics proved it) but does not protect margins from the other duopolist (the surcharge proved it) or the profit pool from the states and prediction-market arbitrage.

Verdict (Competitive Position): A narrow, partial moat — real enough to keep DKNG one of two survivors, not deep enough to guarantee pricing power or a protected profit pool. It is not a crowded, undifferentiated market (the duopoly and the failure of well-funded challengers prove differentiation exists), but it is not a wide-moat compounder either. The durable advantage is “hardest to dislodge from the top-two,” not “earns excess returns at will.” The single best chance to upgrade the moat — adding a genuine network effect via prediction-market exchange liquidity — is exactly the contested, late-entry bet that could also commoditize the core.

5. Growth History and Forward Opportunities

Historical growth has been extraordinary and largely organic-plus-bolt-on. Revenue: $615M (2020) → $1.30B (2021) → $2.24B (2022) → $3.67B (2023) → $4.77B (2024) → $6.05B (2025) — a ~58% five-year CAGR. Growth has decelerated as the base scaled (FY2025 +27%; Q1-2026 +16.8%; FY2026 guide midpoint ~+11%), which is natural and healthy: the hyper-growth land-grab phase is maturing into a profitable-growth phase. The drivers have been (i) new-state launches (each legalization adds a market), (ii) handle growth per state as betting normalizes, (iii) rising hold from parlay/SGP mix, and (iv) acquisitions (Golden Nugget Online for iGaming customers; Jackpocket for lottery cross-sell).

The quality of the growth is improving. Earlier growth was bought with enormous promotional spend (the 2021–2022 −$1.5B-EBITDA promo war). The 2024–2026 vintage is different: revenue is still growing double-digits while marketing intensity falls and margins rise — Q1-2026 adjusted gross margin up ~200 bps, adjusted operating expense roughly flat ex-Predictions/Arkansas, and “15 consecutive weeks of year-over-year net-revenue growth.” This is the signature of a business transitioning from buying growth to compounding it. Management even claims AI-driven productivity (some teams at “2–3x last year’s output”).

Forward opportunities, ranked by credibility:

1. Rising hold / parlay mix (high credibility) — the most reliable near-term margin lever; parlay handle mix is still climbing with “a ton of runway left,” and structural hold rises mechanically as it does.
2. iGaming re-acceleration (medium) — DKNG admits under-indexing here; online casino is the higher-margin, less-volatile business, and a focused slots-first push (Flex Spins, new marketing) is upside not in the guide.
3. Prediction markets (high variance) — the largest TAM expansion (all 50 states) and the largest risk; could add meaningful incremental adjusted EBITDA or commoditize the core (above).
4. New-state legalization (medium, back-loaded, not management-controlled) — half the country still lacks legal OSB and ~90% lacks iGaming; DKNG argues, plausibly, that the threat of untaxed prediction markets is now a lobbying lever to push states toward legalizing (and taxing) operators instead.
5. 2026 FIFA World Cup (catalyst, not structural) — management expects a “tremendous” customer-acquisition event, especially in prediction-market-only states (CA/TX/FL) and via a new Spanish-language app, though it cautions the revenue impact will be smaller than the acquisition impact.

Verdict (Growth): High-quality and improving — decelerating but now profitable growth, with a credible near-term margin lever (hold) and real optionality (iGaming, Predictions). The caveat is that the biggest TAM unlocks (CA/TX, prediction markets) are the least controllable and the most contested.

6. Financial Quality

The inflection is real. The operating-loss arc tells the story: −$843M (2020), −$1.56B (2021), −$1.51B (2022), −$789M (2023), −$609M (2024), and −$16M (2025) — essentially breakeven — with +$3.7M GAAP net income, the first profitable year. EBITDA was +$260M, and crucially free cash flow turned firmly positive: −$729M (2022) → −$115M (2023) → +$297M (2024) → +$508M (2025), on $663M of operating cash flow. Q1-2026 extended the run (second consecutive positive-net-income quarter; adjusted EBITDA +64% to $168M).

Gross margin is structurally rising — 33.8% (2022) → 37.5% (2023) → 38.1% (2024) → 41.3% (2025) — as parlay mix lifts hold and per-customer promotional intensity falls. This is the single best evidence that economics improve with scale.

The quality-of-earnings caveats are material, however:

• Stock-based compensation remains enormous. SBC was $339M in FY2025 — down from a $683M peak (2021) but still ~91x GAAP net income and ~5.6% of revenue. The entire FY2025 GAAP profit ($3.7M) is a rounding error against SBC; on a cash basis the business is healthier than GAAP shows (SBC is non-cash), but on a true-economic-cost basis SBC is a real expense to shareholders via dilution, and “adjusted EBITDA” (which adds it back) flatters the picture. The honest read: cash generation is real and improving; per-share economics are still being diluted.
• Hold volatility makes quarterly earnings partly exogenous. Q3-2025’s −$257M net loss on customer-friendly outcomes shows a single quarter of bad sports luck can erase a year of profit. Over time hold normalizes, but the variance is real and un-hedgeable in the core (one of the appeals of the prediction-market/market-making model is that it can hedge).
• ROIC is still negative on a strict basis. Despite the GAAP profit, return on invested capital remains negative (ROIC ~−24% on a standard invested-capital computation, distorted by goodwill/intangibles and a tiny numerator); return on assets is barely positive (~+0.08). The business does not yet earn its cost of capital — it has reached accounting profitability, not yet economic profitability. This is the gap between “makes money” and “creates value,” and it is the crux of the valuation debate.

Balance sheet. Q1-2026: cash ~$999M, total debt ~$1.92B (predominantly the $1.265B zero-coupon convertibles due 2028 plus a $600M Term B Loan), net debt ~$836M — modest against a ~$13B+ enterprise and improving FCF. But tangible book equity is negative (~−$1.86B): goodwill ($1.60B) plus intangibles ($0.89B) exceed total equity ($605M), the residue of stock-funded M&A. The accumulated deficit is −$6.42B against $8.5B of paid-in capital — i.e., the enterprise has consumed ~$6.4B of the ~$8.5B ever raised. This is improving (the deficit shrinks as the company earns), but it frames why book-value multiples are meaningless here (P/B ~23x on a tiny, near-negative-tangible base).

Verdict (Financial Quality): Economics genuinely improve with scale — rising gross margin, positive and growing FCF, the first GAAP profit — and that is the bull case in one line. But the business is at the start of profitability, not the middle: SBC is still huge, ROIC is still negative, tangible book is negative, and earnings are hold-volatile. This is a real inflection, not yet a proven compounder.

7. Capital Allocation

M&A — scale-buying, not value-buying. DKNG grew substantially by acquisition, mostly funded with its own (high-priced) equity: Golden Nugget Online Gaming (~$1.56B all-stock, closed 2022, for iGaming customers and brand), Jackpocket (~$750M, ~55/45 cash/stock, 2024, for the lottery cross-sell funnel), and a string of bolt-ons (SBTech via the SPAC, Simplebet, Sports IQ, Mustard, and now Railbird). The residue is ~$2.49B of goodwill plus intangibles against $605M of equity → negative tangible book. None of the deals has an isolable post-acquisition ROIC (they are folded into a single reporting unit, by design), and the improving consolidated economics are driven far more by the organic core (hold, marketing efficiency, state expansion) than by acquired assets. This is the Marathon anti-pattern: asset growth funded by an over-valued currency, where per-share value creation is far less obvious than headline scale.

The buyback — optics and SBC-offset, not accretion. The board authorized $1.0B (Jul-2024), then doubled it to $2.0B (Nov-2025). FY2025 program spend was ~$523.5M at a ~$35.8 average — underwater at today’s $26.39 — plus another ~$154.9M of treasury purchases just to cover tax withholding on vesting RSUs (a pure SBC-offset). DKNG simultaneously issued ~$339M of new SBC and drew a $600M Term B Loan, so the buyback is best read as a debt-financed mop-up of dilution, not a return of surplus cash. The tell: weighted-average diluted share count still rose (~482M FY24 basic → ~496M FY25, higher diluted). Repurchasing 16M shares while SBC, option exercises, and acquisition stock issue more than that is treading water — and doing it at $35 was poor timing.

Capital structure. The $1.265B convertibles are 0% coupon (free money while it lasts) but struck at a $94.85 conversion price — ~3.6x above today’s stock, so no near-term dilution but also no equity take-out: absent a huge rally, they become a cash refinancing wall in 2028. Manageable given FCF, but a watch-item. The position is net debt (~$836M), not net cash — a correction to any “fortress balance sheet” framing.

Governance — extreme founder control. Class B carries 10 votes per share; Jason Robins controls 100% of Class B plus ~2.1% of Class A → ~88.3% of total voting power on ~2.1% of the economics. Public holders have, in practice, no governance leverage. Layered on top: Robins has pledged ~1.0M shares via prepaid variable forward contracts despite a stated anti-pledging policy, and personal-security/aircraft perks run ~$1.3M/yr.

Incentive alignment — rewards scale, not returns. The founders take $1 salaries (genuine skin-in-the-game optics), but the substance is in the equity: CEO total comp was ~$22.6M (FY2025), almost entirely RSUs/PSUs, and the PSUs gate on Revenue + Adjusted EBITDA only — no ROIC, no ROE, no relative-TSR. That is precisely the metric set that rewards asset growth and acquisition-fueled scale (Adjusted EBITDA can be grown by buying revenue and adding back SBC). Management is paid to get bigger and more “adjusted-profitable,” not to maximize per-share value.

Insider tape — a non-endorsement. Across all Form 4s since mid-2024: ~$362.5M of open-market sales vs. ~$3.25M of purchases — a ~110:1 sell ratio by dollar. Every founder and named officer was a net seller (Liberman ~$152M, Kalish ~$79M, Dodge ~$56M, Robins ~$51M, CFO Ellingson ~$10M). The only open-market buyers were two outside directors (Sloan, Wendt) putting up modest sums near $22–30. Even allowing that much founder selling is 10b5-1-planned diversification, the complete absence of any officer buy through a 50%+ drawdown is a meaningful non-endorsement. This is not a buy-the-dip insider tape.

Verdict (Capital Allocation): Mixed, tilting negative on the per-share test. The bull facts are real (>$508M FCF, $1 salaries, modest leverage, free converts), but growth was bought with an over-valued currency into negative tangible book, the buyback is an underwater SBC-offset, incentives reward scale over returns, and 88% founder control plus a 110:1 net-sell insider tape removes external accountability. The improving economics are an operating/TAM story, not a capital-allocation story.

8. Changes and Headwinds — Last Two Years

Strategic. The dominant change is the prediction-markets pivot (Railbird closed Oct-2025; $200–300M 2026 investment; Super App; combined “Sports revenue” reporting from Q2-2026) — a response to the Kalshi/Polymarket/Robinhood disruption (above). Secondary: Jackpocket (lottery, closed May-2024) and a renewed iGaming push.

Profitability inflection. FY2025 was the first GAAP-profitable year and FCF turned firmly positive — the most important fundamental change, and the one the de-rating has largely ignored.

Regulatory/tax headwinds. Illinois’s per-wager tax and graduated rate (2025), New Jersey 14.25%→21%, the federal 90% loss-deduction cap (2025), and the broader signal that states will keep raising rates — the structural threat to the profit pool. Notably, management claims no OSB tax increases have landed in 2026 so far, crediting the prediction-market threat as a lobbying lever (“states would be crazy to raise taxes on licensed operators while untaxed prediction markets exist”).

Capital actions. Buyback authorized (Jul-2024) and doubled to $2B (Nov-2025); $600M Term B Loan (Mar-2025); ~$523M of FY2025 repurchases at ~$35.8.

Headwinds into 2026. (i) Customer-friendly sport outcomes (the Q3-2025 −$257M reminder); (ii) prediction-market competition and the unresolved CFTC/state legal contest; (iii) the $200–300M Predictions investment drag on 2026 EBITDA; (iv) decelerating revenue growth; (v) a high-beta multiple in a risk-off tape (the mid-June Fed-hawkish pullback).

Verdict (Changes): Net neutral-to-slightly-negative for the multiple, net positive for the fundamentals. The profitability inflection strengthens the long-term thesis; the prediction-market disruption and tax trajectory weaken the visibility of the rent — which is exactly why the stock de-rated even as the financials improved.

9. Risk Analysis (Risk Matrix)

Catastrophic-loss assessment. The risk of a total loss is low: DKNG is FCF-positive, modestly levered, one of two scaled survivors, and sits on a saleable customer franchise. The realistic downside is a de-rating and earnings-power impairment (prediction markets erode the take-rate; taxes keep rising) rather than insolvency. The realistic upside is a re-rating if profitability compounds and Predictions is won. The distribution is wide but not bimodal-to-zero.

10. Valuation Discussion (Embedded Expectations)

No price target; no recommendation. Valuation is discussed only as embedded expectations and scenarios.

Where the multiple sits. At ~$26.39, market cap is ~$13.1B and enterprise value ~$13.9B (net debt ~$836M). On the FY2026 guide midpoint (~$6.7B revenue), that is ~2.1x EV/forward-sales and ~17x EV/adjusted-EBITDA on the $800M midpoint (or ~13–14x on the implied ~$1B+ “core” adjusted EBITDA before the $200–300M Predictions drag). The own-history valuation-percentile ranks are the cleanest tell: P/S at the 4.8th percentile of DKNG’s ten-year history (the cheapest it has ever been on sales) and EV/sales (~2.1–2.25x) sitting just above the 2022 trough of ~2.2x, versus 7.9x (2021), 4.5x (2023), 3.9x (2024). The P/E (241x) and P/B (89th percentile) ranks are not meaningful — GAAP EPS just crossed zero, and book value is tiny with negative tangible equity. The honest valuation anchor is sales and EBITDA, and on both the stock is near the low end of its own range.

What the price embeds. Reverse-engineering the ~2.1x EV/sales / ~17x EV/EBITDA: the market is underwriting roughly continued low-double-digit revenue growth and adjusted EBITDA of ~$0.8–1.0B near-term, at a compressing multiple — i.e., it is paying for the current profitable operator and giving little-to-no credit for (a) the structural margin ramp toward management’s ≥30% long-term target, (b) the all-50-states TAM that prediction markets could unlock, or © the iGaming re-acceleration. Equivalently, the market is pricing the bear case as the base case: that taxes and prediction-market arbitrage cap DKNG’s earnings power near current levels and the multiple should stay near trough.

Scenario analysis (illustrative, adjusted-EBITDA-based — the metric the market uses):

• Bear (~$15–18/share): Prediction-market take-rate competition + rising state taxes cap adjusted EBITDA near $0.8–1.0B; Predictions spend doesn’t pay back; multiple compresses to ~10x EV/EBITDA → EV ~$8–10B, equity ~$15–18.
• Base (~$32–45/share, 2–3yr): Revenue compounds low-double-digits, hold/parlay mix and operating leverage lift adjusted EBITDA toward ~$1.3–1.6B by 2027–28; multiple holds ~15–17x → EV ~$20–27B, equity ~$32–45 (≈the recent trading range).
• Bull (~$60–75+/share): Predictions is won and federalized (all-50-states, higher margin), iGaming re-accelerates, margins approach the ≥30% target; adjusted EBITDA ~$2B+ by 2028–29 at a re-rated ~18–20x → EV ~$36–42B, equity ~$60–75+.

At $26.39 the market sits between bear and base — closer to “this is roughly a $1B-EBITDA business at a trough multiple” than to either tail. The asymmetry depends entirely on which way the prediction-market question resolves; the current price does not require the bull case, but it does require the bear case not to fully play out.

Comp context. Pure-play OSB comps are thin (FanDuel is inside Flutter; PENN/CZR/MGM are land-based hybrids), so the sales/EBITDA own-history bands and the factor-peer set (high-beta growth, not casinos) are more informative than a casino comp table. On EV/sales DKNG screens cheaper than its own history and than high-growth consumer-internet peers, but that discount is the market’s verdict on profit-pool durability, not an oversight.

Verdict (Valuation): The stock is demonstrably cheap on its own ten-year sales/EBITDA history and is pricing the bear case as base. Whether that is opportunity or a value trap turns on the single unresolved question — prediction markets as margin-accretive adjacency vs. core-commoditizing disruption — which no multiple can settle.

11. Variant Perception

Consensus. The sell-side is moderately constructive-but-divided: price targets span ~$30 (TD Cowen) to ~$49 (UBS), clustering around “good operator, real profitability, but the prediction-market and tax overhangs justify a discounted multiple.” The market’s revealed view (the 4.8th-percentile sales multiple) is more bearish than the sell-side: it is pricing structural impairment of the rent.

Strongest bull case. A profitable, FCF-generative duopolist at the cheapest sales multiple in its history, with (i) a reliable near-term margin lever (rising hold), (ii) a credible path to ≥30% margins, and (iii) free optionality on prediction markets — where DKNG, far from being only a victim, is one of the few players with the brand, customers, pricing tech, and a CFTC exchange to win the disruption and unlock all 50 states at a lower effective tax. If the CFTC framework entrenches the channel for licensed operators, the state-tax bear case inverts into a margin tailwind. Buy the inflection the market is ignoring.

Strongest bear case. A narrow-moat, weak-captivity operator with no pricing power (the surcharge proved it), whose politically-determined profit pool is being taxed away from above and arbitraged away from outside, run by a founder with 88% control and no return-on-capital incentive, whose insiders are net sellers ~110:1, whose tangible book is negative, and whose own $200–300M defensive pivot may commoditize its core. The cheap sales multiple is correct: the market is right to discount a contestable rent. The profitability is early, SBC-flattered, and ROIC-negative.

The 3–5 assumptions that matter most:

1. Do prediction markets entrench DKNG’s margins (all-50-states, higher-margin exchange) or commoditize the OSB take-rate? (The single swing factor.)
2. Can rising hold + operating leverage carry adjusted EBITDA toward ~$1.5–2B despite tax drag?
3. Does the state-tax trajectory accelerate (a big-state NY-style shock) or stall (as 2026 has so far)?
4. Can a late DKNG win exchange liquidity vs. Kalshi/Polymarket/CME/Robinhood?
5. Does iGaming re-accelerate (free upside) or keep lagging?

Falsification evidence. Bull is falsified by: a prediction-market take-rate race-to-the-bottom visible in compressing hold/fees, a major-state tax shock, or Predictions spend ballooning past $300M without acquisition payback. Bear is falsified by: Predictions scaling to positive contribution with a final CFTC rule favoring licensed operators, hold/parlay mix carrying margins higher, and a stabilizing or falling state-tax trajectory.

Factor-positioning read (the tape’s verdict). DraftKings trades like a high-beta (1.18–1.38), high-volatility (negative LowVol loading −0.38), small-cap-flavored (+0.42), credit-sensitive (+0.13) risk-on growth vehicle — its nearest factor neighbors are software/cloud growth ETFs and Block (XYZ), not Caesars or MGM. It carries negative alpha (−0.27), ~31% negative trailing-12-month relative strength, and ~75% idiosyncratic variance (R² ~0.22–0.28) — i.e., the market prices DKNG on stock-specific regulatory/hold/competitive risk far more than on any style factor, and currently prices it as an abandoned high-beta growth name (down ~63% from the peak), not a momentum trade. The recent three-month bounce (Predictions/World Cup) is a counter-trend move within a year-long downtrend. The factor evidence supports the “de-rated, contrarian, idiosyncratic-risk” framing in Claude’s Take, not a “crowded long” reading — consensus is bearish/agnostic, which is where contrarian setups live, with the caveat that the bearishness is grounded in a genuine, unresolved structural threat.

12. Fact vs. Interpretation Table

13. Open Questions

1. Final CFTC framework + appellate/Supreme outcome on preemption — does it entrench prediction markets for licensed operators (DKNG margin-up, all-50-states) or trigger a take-rate race-to-the-bottom that commoditizes OSB?
2. Can a late DKNG win the exchange-liquidity network effect vs. Kalshi/Polymarket/CME/Robinhood, or does it subsidize a structurally-losing position?
3. Does prediction-market revenue cannibalize high-tax licensed OSB (margin-positive mix shift) or merely cannibalize it while inviting competition (rent destruction)?
4. What is the actual standalone ROIC of GNOG and Jackpocket? (Undisclosed; single reporting unit by design.)
5. Can the $1.265B 2028 converts be refinanced/repaid without a dilutive raise if the stock stays well below the $94.85 strike?
6. Does the state-tax trajectory accelerate (a big-state NY-style shock) or stall (as 2026 has so far)?
7. Does iGaming re-accelerate toward market growth, providing the free upside management hints at?

14. What Must Be True (Bull and Bear, with Falsification Tests)

Bull thesis — what must be true:

• Rising hold/parlay mix + operating leverage carry adjusted EBITDA toward ~$1.5–2B over 2–3 years despite tax drag.
• Prediction markets become a net positive — DKNG captures a profitable share of a federalized, all-50-states channel that raises blended margins, rather than commoditizing the core.
• The state-tax trajectory stabilizes (the prediction-market threat acts as the lobbying brake management claims).
• Falsification test: if, over the next 2–4 quarters, Predictions spend exceeds $300M without visible acquisition/EBITDA payback, OR a major state enacts an NY-style tax hike, OR blended hold/take-rate compresses as prediction markets scale — the bull thesis breaks.

Bear thesis — what must be true:

• Prediction-market arbitrage and rising state taxes cap DKNG’s earnings power near $0.8–1.0B adjusted EBITDA and justify a permanently low (~10–12x) multiple.
• The narrow moat (weak captivity) means any margin gain is competed away by FanDuel; pricing power never materializes.
• ROIC stays negative; the buyback and SBC keep per-share value flat.
• Falsification test: if Predictions scales to positive contribution under a final CFTC rule favoring licensed operators, AND hold/parlay mix lifts margins toward the ≥30% target, AND the state-tax trajectory stalls — the bear thesis breaks and the trough sales multiple re-rates.

The two falsification tests pivot on the same events (prediction-market economics, tax trajectory, hold trend) — which is why this is a genuinely two-sided, catalyst-dependent situation rather than a settled call.

15. Source Appendix

Primary sources: DraftKings FY2025 Form 10-K (filed 2026-02-13); Q1-2026 Form 10-Q and earnings call (2026-05-08); DEF 14A proxy (2026-03-26); Form 4 corpus (EDGAR CIK 0001883685, since 2024-06-01); FY2020–2025 financial statements (public financial databases, reconciled to filings); valuation-percentile and daily price-history data; a quantitative factor model (FactorsToday public API). Industry/regulatory: state gaming-revenue reports (2026), CFTC June-2026 proposed event-contract rules, court rulings on federal preemption (Feb–Apr 2026), One Big Beautiful Bill Act (Jul-2025). See Appendix B for the full citation list.

APPENDIX A — Standard Diligence Questionnaire

Supplemental to the research memo. Report date 2026-06-20. Fact/Interpretation/Assumption labels applied where it matters.

General

What thoughtful questions have other investors asked about this company? The recurring institutional questions cluster on five axes: (1) Are prediction markets (Kalshi/Polymarket/Robinhood) an existential disintermediation of the taxed OSB channel, or an adjacency DKNG can co-opt? — the dominant debate. (2) Is the FY2025 GAAP-profitability inflection durable, or SBC-flattered and hold-lucky? (3) How much further will states raise OSB taxes, and can operators ever pass it through? (the surcharge-failure question). (4) Is the duopoly with FanDuel a stable, disciplined oligopoly or one promo cycle away from re-escalation? (5) Does founder super-voting control (~88%) plus a scale-not-returns comp plan mean capital allocation will always favor growth over per-share value?

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? Neither cleanly — DKNG is at an early-cycle profitability inflection (first GAAP profit FY2025), not a mid-cycle peak or trough. (Interpretation.) Driven by external environment or internal actions? Both: revenue growth is partly exogenous (state legalization, sport outcomes/hold) and partly internal (marketing efficiency, parlay-mix engineering, product). A single quarter of customer-friendly outcomes (Q3-2025, −$257M) shows the exogenous swing. How stable are revenues? Growing but lumpy/seasonal (Q4 NFL-heavy) and hold-volatile quarter-to-quarter; not contracted. Outlook for products/services? Core OSB maturing to profitable double-digit growth; iGaming under-penetrated upside; Predictions high-variance optionality. How big will this market be? US online gambling ~$16.6B (2024) → ~$33B (2030E) at ~12% CAGR; management’s 2030 gross-revenue opportunity framing is $55–80B (TAM, not DKNG revenue); domestic-only today.

Business Quality & Competitive Moat

Is the industry getting more or less competitive? Less at the top (duopoly consolidating; ESPN Bet/Fanatics failing) but the profit pool is contested from outside (prediction markets) and above (state taxes). How profitable is the business (ROIC, ROE)? GAAP-profitable as of FY2025 but ROIC still negative (~−24% strict, distorted by goodwill; ROA ~+0.08) — accounting profit, not yet economic profit. (Fact.) How profitable is the industry? ~$16.8B GGR on $165.6B handle (2025), but ~$3.66B taxed away; high-tax states (NY 51%) earn little. Barriers to entry? Real scale/brand barrier against #3+ (proven by ESPN/Fanatics failure); weak captivity between #1 and #2. Can it be easily understood? Yes — handle × hold, minus promo/tax/marketing. Undermined by foreign low-cost labor? No (regulated, domestic, license-gated). Do brands matter? Yes at the top of the funnel (lower CAC) but not for within-duopoly retention. Nature of competition? Price/promo and product (parlay variety, live markets), plus now exchange liquidity. Switching costs? ~Nil — multi-homing is rampant (one app download + deposit match).

Financial Condition & Balance Sheet

Assets not fully recognized on the balance sheet? The customer database, brand, and market-access licenses are partly intangible-capitalized; the DFS/OSB brand value is understated. Off-balance-sheet liabilities? Operating leases (capitalized under ASC 842); customer-deposit liabilities; the $1.265B converts and $600M Term B are on-balance-sheet. How conservative is the accounting? Mixed — GAAP is conservative (SBC fully expensed, large accumulated deficit) but the company emphasizes “Adjusted EBITDA” which adds back ~$339M SBC; treat adjusted figures with that caveat. How CapEx-hungry? Low — capex ~$154M FY2025 (~2.5% of revenue), mostly capitalized software; this is an asset-light, working-capital-favorable model (negative cash-conversion cycle — customers pre-fund).

Capital Allocation & Management

How much FCF, and how is it used? +$508M FY2025 FCF; used for buybacks (~$523M, underwater at ~$35.8 avg) and M&A; no dividend. Philosophy? Growth/scale-first; buybacks framed as confidence but functionally an SBC-offset. Significant acquisitions? GNOG (~$1.56B all-stock, 2022), Jackpocket (~$750M, 2024), Railbird (prediction-market exchange, ~$85M booked / up to ~$250M, 2025). Buying back shares? Yes ($2B authorization), but net diluted share count still rising under ~$339M/yr SBC. Issuing shares to insiders? Yes — large RSU/PSU programs; CEO comp ~$22.6M FY2025 on a $1 salary. Compensation policy? PSUs gate on Revenue + Adjusted EBITDA only — no ROIC/TSR (rewards scale, not per-share returns). Motivations of management? Founder Robins controls ~88% of the vote on ~2.1% economics; aligned with growth and control, less with minority per-share value. Insider tape ~110:1 net selling.

Valuation & Market Data

ADR / MLP / K-1? No — US C-corp, NASDAQ-listed Class A common (1099, not K-1). Dividend policy? None (no dividend; capital returned via buyback). How profitable? Just GAAP-breakeven (FY2025 +$3.7M NI; thin Q1-2026 +$21M); adjusted EBITDA $700–900M guided FY2026. Is net income diverging from cash from operations? Yes, favorably — OCF ($663M FY2025) vastly exceeds GAAP NI ($3.7M), driven by ~$339M non-cash SBC, D&A, and favorable working capital (customer pre-funding). This helps the cash story but means GAAP NI understates cash generation while overstating per-share value (dilution).

Risks & Downside

What would cause the stock to decline? A prediction-market take-rate race-to-the-bottom; a big-state NY-style tax hike; a bad-hold quarter; Predictions spend ballooning without payback; high-beta risk-off macro. Risk of catastrophic loss? Low — FCF-positive, modestly levered, duopoly survivor. Chance of total loss? Low absent a regulatory black-swan (e.g., a federal crackdown that simultaneously bans the OSB and prediction-market channels). The realistic downside is earnings-power impairment + de-rating, not insolvency.

Recent News & Events

Has the business environment changed recently? Yes, materially — the rise of CFTC-regulated prediction markets (Kalshi/Polymarket/Robinhood) and the CFTC’s June-2026 proposed rules are reshaping the competitive and regulatory landscape; DKNG launched its own Predictions product and Super App. Significant acquisitions? Railbird exchange (Oct-2025). Change in accounting policies? None material; note the shift to reporting combined “Sports revenue” (Sportsbook + Predictions) from Q2-2026 — a presentation change that will reduce segment transparency. Recent changes — new markets/facilities/management? New-state launches (Arkansas, Alberta 2026); $600M Term B Loan (Mar-2025); buyback doubled to $2B (Nov-2025); a Spanish-language app and FIFA World Cup 2026 acquisition push.

APPENDIX B — Source Appendix

Report date 2026-06-20. Primary sources first. Accessed 2026-06-20 unless noted.

Primary — SEC filings (EDGAR, CIK 0001883685)

• Form 10-K (FY2025) — filed 2026-02-13. Segment/product-line revenue, risk factors (prediction markets, taxes), balance sheet (goodwill $1.60B, intangibles $0.89B, equity $605M), convertible notes ($1.265B, 0%, due 2028, $94.85 strike), Term B Loan ($600M), buyback authorization.
• Form 10-Q (Q1-2026) — filed ~2026-05; Q1-2026 income statement (rev $1.646B, NI +$21M), balance sheet (cash $999M, net debt $836M).
• DEF 14A proxy — filed 2026-03-26. Dual-class structure (Class B 10 votes; Robins 88.3% vote / 2.1% economics), executive comp (CEO ~$22.6M FY2025, $1 salary), PSU metrics (Revenue + Adjusted EBITDA, no ROIC/TSR), founder share-pledging (forward contracts).
• Form 4 corpus — 249 filings since 2024-06-01. Insider sales ~$362.5M (Liberman $152M, Kalish $79M, Dodge $56M, Robins $51M, Ellingson $10M) vs. purchases ~$3.25M (directors Sloan, Wendt only). 67 Form 144s corroborate 10b5-1 selling.
• 8-K timeline — buyback authorization (2024-07-30) and doubling to $2B (2025-11-06); Jackpocket close (2024-05-22); Term B Loan (2025-03); Railbird close (2025-10-21); quarterly earnings.

Primary — Earnings calls / company

• Q1-2026 earnings call transcript — 2026-05-08 (public financial databases). FY2026 guide ($6.5–6.9B rev / $700–900M adj EBITDA); $200–300M Predictions investment; “$1B+ core adj EBITDA”; Super App; Investor Day $55–80B 2030 gross-revenue opportunity, ≥30% LT margin; April-2026 metrics (rev +22%, >$100M monthly adj EBITDA, handle +6%); Predictions Q1 annualized volume >$2.3B, CAC −80%.
• Investor Day (March 2026) — strategic framing; prediction-market opportunity; long-term margin target.

Quantitative data

• Public financial databases — FY2020–2025 + quarterly income statements, balance sheets, cash flows, profitability ratios, enterprise value, valuation multiples (reconciled to filings). EV ~$13.9B; EV/sales ~2.1–2.25x; ROIC ~−24%.
• Valuation-percentile data — own-history percentile ranks (P/S 4.8th pctile / P/B 89th / P/E 25.9th; composite 39.9th); latest price/book/sales-per-share.
• Five-year daily price history — daily OHLCV, EMAs, beta (1.38), alpha (−0.27). Price arc ($71.98 ATH Mar-2021; $11.39 trough end-2022; $53.49 Feb-2025; $20.72 low Mar-2026; $26.39 2026-06-18).
• Quantitative factor model (FactorsToday, public API) — loadings (Market +1.12–1.17, SmallSize +0.42, LowVol −0.38, CreditRisk +0.13; R² ~0.22–0.28); stock-info (rs_12m −31%, rs_peak −63%); leaderboard (y5 −12%/yr, y5 maxDD −84%, m3 bounce); related stocks (software/cloud ETFs, Block).
• Financial news aggregators — recent-events timeline (prediction markets, CFTC rules, World Cup, analyst actions, Fed-driven pullback).

Industry / regulatory (public secondary)

• US 2025 handle/GGR/tax figures (~$165.6B / ~$16.8B / ~$3.66B) — state gaming-revenue trackers (RG.org / CasinoReports, 2026).
• OSB market shares (DKNG ~36%, FanDuel ~32%; ESPN Bet ~1%, Fanatics ~7%) — CasinoReports/RG.org, early-2026.
• State tax rates (NY 51%, IL graduated 20–40% + per-wager fee, NJ 14.25%→21%, PA ~36%, VT ~31%) — state statutes / gaming-commission reports, 2025–2026.
• CFTC proposed event-contract rules (2026-06-10); federal-preemption rulings (Feb-2026 preliminary injunction; Third Circuit Apr-2026); CFTC suits vs. AZ/CT/IL (Apr-2026).
• One Big Beautiful Bill Act (Jul-2025) — 90% gambling-loss-deduction cap.
• Deal terms: GNOG ~$1.56B all-stock (2021-08 announced, 2022 close); Jackpocket ~$750M (2024); Railbird (~$85M booked, up to ~$250M, 2025) — company releases / trade press.

Note: management commentary (guidance, Predictions traction, margin targets) is treated as hypothesis and validated against filings, financials, and external data . Third-party aggregated data is reconciled to primary filings; where they disagree, the filing governs.