Flutter Entertainment plc: a global compounder navigating a contained shift

Who they are. Flutter Entertainment is the world’s largest listed online sports betting and iGaming operator, formed through the 2020 merger of Paddy Power Betfair and The Stars Group. Its US business, FanDuel, is the market-leading sportsbook and iGaming platform following the state-by-state legalization of sports betting since 2018. Its International division spans established and newly acquired local brands across the UK & Ireland, continental Europe, Brazil, and Asia-Pacific — including Sisal and SNAI in Italy, Betnacional in Brazil, and Sky Bet in the UK. The group delisted from the London Stock Exchange in August 2026 and now trades solely on the NYSE.

We are long, and here is the shape of the case

Spider Eyes Research is long Flutter Entertainment. This note sets out why, states plainly what would have to go wrong for that to be a mistake, and builds a price range to FY2029 rather than a single target — because the honest range of outcomes here is wide, and stating that range clearly is more useful to a reader than false precision.

The case in brief: a still-dominant US sportsbook business (FanDuel) working through a deliberately elevated investment year, sitting alongside a larger, structurally insulated International collection of businesses — built substantially through debt-funded acquisition over the past 18 months — that generates roughly three-quarters of group profit and faces a materially different competitive landscape than the one dominating US headlines. We believe management’s turnaround narrative for the US business is broadly credible on the evidence, and — this is the more contrarian of our two central calls — we believe the threat from prediction-market upstarts is more contained than current fear reflects, for reasons specific to scale, geography, and regulatory structure that we lay out in Section 5, not because we are dismissing it.

The rest of this note builds the case in full before returning to those caveats in weight: the two-engine business (Section 3), what the 2025 debt-funded acquisition spree means for the balance sheet and the range (Section 4), the competitive and regulatory landscape (Section 5), and the explicit bear case, uncertainties, and catalysts that would tell us if we’re wrong (Sections 7–9). Source data and full model detail are in the appendices.

Valuation methodology and range

Why not a peer-multiple approach. A single-year multiple applied to a currently depressed earnings base understates a genuine multi-year recovery thesis — and pins the entire valuation to the reliability of one or two peer comparables (DraftKings, Entain), each of which carries its own distortions (DraftKings’ multiple partly reflects hoped-for rather than realised profitability; Entain’s enterprise value includes a 50%-owned joint venture whose earnings aren’t fully reflected in its own multiple). Both problems compound in a single-year snapshot. This note instead combines two methods that don’t share that weakness.

Method 1 — Discounted cash flow

Segment-level revenue and EBITDA are built up from disclosed operating drivers (US sportsbook and iGaming modelled separately, given materially different growth and margin dynamics — see Section 3) through FY2029, converted to unlevered free cash flow, and discounted at a scenario-specific WACC (7.8–9.5%, reflecting genuinely different financial risk across scenarios — see Section 4 on leverage trajectory).

Method 2 — EV/Revenue

Revenue is the most stable line in Flutter’s financials — it has grown every year since at least 2021, including through periods when EBITDA fell by double digits — making it a useful counterweight to a discounted cash flow’s sensitivity to terminal-value assumptions. A real historical EV/Revenue series (Q1 2023–present, built from market price data and primary net debt/revenue figures) shows the multiple ranged 2.9x–3.9x through 2023–2025 and compressed sharply — in a single quarter, not gradually — to its current ~1.6x during Q1 2026. That compression is now behind the business, not still unfolding.

The exit multiple used in this note’s range is held flat at 1.6x through FY2029, deliberately, so the range reflects an operating case rather than a re-rating layered on top of one. Worth being explicit about what that gives up: if growth genuinely returns across the US, International, and prediction-market participation simultaneously — the bull case, not the base case — there is a real, historically precedented possibility the multiple itself drifts back toward the 2.9x–3.9x band seen as recently as 2023–2025, rather than staying pinned near current levels. This is not built into any figure in this section and should not be read as a second, hidden target — it is flagged here as a further, genuinely uncertain source of upside the methodology intentionally excludes, not something to rely on.

Combining them

Rather than averaging two independent numbers, EV/Revenue is used as the terminal-value method within the DCF — replacing a perpetuity-growth assumption (which proved extremely sensitive to small changes in discount rate and long-run growth) with an exit multiple anchored to Flutter’s own trading history. A single, current-level multiple (1.6x) is held constant across all three scenarios, so that scenario differences reflect only the operating question this note is actually about — not a stacked assumption about sentiment also re-rating in the same direction as the operating case. An earlier version of this exercise let the multiple expand with the bull case and compress with the bear case simultaneously with WACC and revenue growth; the result was a bull case worth more than double the current price purely from methodology, which was discarded as not credible.

Annualized to the position’s current cost basis, the range implies roughly −2.6% per year (bear), +5.6% (base), and +10.6% (bull) to the FY2029 outcome — the bull case in particular is the shape of return a multi-year holder should be looking for, if the operating case actually lands. This is a price-return calculation to the stated target, not a formal return-on-capital-employed metric, and it excludes the further compounding discussed in Section 6.

What this range deliberately excludes. Net debt is held at its current level ($11,305m, including Fox Option and minorities) across all three scenarios — this is the technically correct DCF convention (enterprise value already reflects the present value of the cash that would pay debt down; deducting a scenario-specific future net debt figure would double-count that value). What does vary correctly by scenario is the discount rate, which reflects a genuinely different financial-risk profile — quantified directly in Section 4.

The two engines: US and International

United States — sportsbook and iGaming are different businesses

US segment Adjusted EBITDA has been volatile enough that treating it as one number obscures more than it reveals.

FY2026’s $760m guide would be the first year-over-year decline in US Adjusted EBITDA since the segment turned profitable in FY2023 — worth stating plainly, since it’s a materially different claim than “growth decelerated.” The disclosed guidance-cut bridge attributes the reduction mostly to elective spend ($270m of the $210m full-year cut, netted against underlying improvement and an NFL timing shift) rather than product deterioration — corroborated by structural revenue margin rising 40bps to 14.0% even as reported margin fell. Management’s own near-term guidance, from the Q2 2026 call: Q3 2026 roughly break-even, Q4 2026 approximately $500m (down from an original ~$700m Q4 guide) — a precise, dateable checkpoint.

Sportsbook and iGaming have diverged: iGaming revenue has grown every single quarter since 2021 with no down quarter, while sportsbook revenue and margin swing with sports results (Q2 2025’s 22.3% EBITDA margin — the best quarter on record — coincided with the highest sportsbook net revenue margin and highest revenue-per-player of any recent quarter, consistent with a favourable-variance quarter rather than a new structural level). Revenue per player fell 13.9% year-on-year in Q2 2026 even as player counts grew 9.2% — a genuine, quantified version of “growth is being bought at a lower price per customer,” which is the central open question behind the US recovery case.

International — larger than US profit, facing a two-wave tax story not fully priced by the market

The step-down from 23.1% to 18.0% has a clean, dateable cause: the UK Remote Gaming Duty increase (21%→40%) took effect April 1, 2026 — a one-time, already-absorbed shock, in the market’s likely framing. It isn’t the whole story. Flutter’s own November 2025 disclosure lays out a second wave: UK sports betting duty (ex-horseracing) rises 15%→25% effective April 2027 — a further ~$540m gross impact, netting to ~$339m after first-order mitigation. This is not yet reflected in most market commentary on the “UK tax headwind,” which tends to treat it as a 2026 event. It is carried explicitly through FY2027 in every scenario in this note.

Regional revenue growth, FY2023→FY2025: UKI $3,047m→$3,547m (steady, low-single-digit organic); Southern Europe & Africa $1,430m→$2,746m (the real growth engine — the Sisal and SNAI Italy acquisitions, plus Turkey); Brazil $64m→$227m (rapid initial ramp on market regulation, now decelerating — organic revenue turned negative in H1 2026, which management attributes to “government socioeconomic measures and regulatory uncertainty,” not execution); APAC $1,599m→$1,428m (shrinking, driven by the India/Junglee exit); CEE $286m→$604m (steady organic gainer).

Capital structure: what debt-funded acquisition means for the range

Net debt roughly doubled in 2025 — from $5,160m (FY2024) to $10,591m (FY2025) — funding a concentrated few quarters of acquisition activity (Sisal/SNAI in Italy, Betnacional in Brazil, and the buyout of FanDuel’s remaining minority stake), almost entirely debt-funded rather than cash-funded (a single quarter, Q2 2025, saw $6,004m of new debt issuance against $2,688m of acquisition spend). Leverage has risen further since — to 4.3x against a 2.0–2.5x target — but net debt in absolute dollar terms has actually fallen slightly since (down $111m over H1 2026): the leverage problem is currently an earnings problem more than a debt-growth problem.

Rolling forward each scenario’s own free cash flow (after a ~5.2% blended after-tax cost of debt, with interest expense declining as balances fall) against 100% directed to paydown — the assumption consistent with the company’s own stated buyback policy, which conditions further capital return on leverage being at or returning toward target:

This is the derivation behind the scenario-specific WACCs used in Section 2: the bear case genuinely carries elevated financial risk through the entire horizon; the bull case genuinely de-risks fast. It also produces a clean, falsifiable near-term signal: a resumed share buyback (the fifth $250m tranche completed as planned in May 2026; no sixth has been announced as of this note, consistent with — not proof of — the stated policy no longer being met) would be a real-time indicator of which path is unfolding, well before a full year’s results confirm it. The firm’s largest disclosed shareholder, Candle Lake (Kenneth Dart, ~28.9%), has a documented pattern — a swap-heavy structure and Schedule 13G passivity certification consistent with a long-horizon economic position rather than an active trading stake — plausibly more aligned with debt paydown and market-share retention than with near-term capital return, though this is a mild supporting read, not evidence of intent.

Competitive landscape

Market leadership, quantified. FanDuel and DraftKings together control 73% of US online handle — the highest concentration since legalization began, and rising. Compliance and technology infrastructure costs are becoming a structural moat: industry research explicitly notes “the costs associated with competing with the technology infrastructure of FanDuel and DraftKings is unsustainable for many” mid-tier operators, with the top five operators forecast to control 55%+ of regulated global GGR by 2028.

Prediction-market newcomers are well-capitalized — this cuts against an earlier assumption in this research, and is stated as found, not softened. Kalshi’s valuation moved from $2B to $22B (a $1B Series F in May 2026, explicitly earmarked to scale toward “100 million consumers”); Polymarket moved to $15B backed by a $600m investment from ICE, the NYSE’s parent. Robinhood — already public and profitable ($573m net income, Q2 2026) — generated $156m of event-contract revenue in Q2 2026, up 50% sequentially, while its CFO stated directly that the company is compressing the spread it collects specifically to win share, via its own newly launched exchange (Rothera). These are not cash-constrained entrants likely to exit on their own; the risk is real and well-funded.

Two things temper that threat materially. First, scale: Robinhood’s entire prediction-markets quarter is about 9% the size of Flutter’s US segment revenue in the same period. Second, and more structurally important: the prediction-market threat is substantially a US phenomenon, resting on a specific American legal mechanism (CFTC federal preemption of state gaming law) with no equivalent abroad. Where Kalshi and Polymarket have tried international expansion, they have done so without local licenses — and both the UK and Brazil, two of Flutter’s most important International markets, have already moved to block them by explicit government action. International generates roughly 74% of segment EBITDA and is, on current evidence, largely insulated from this specific competitive thread.

Bull case

• The US guidance cut is substantially elective, not structural — corroborated by structural revenue margin gains even as reported margin fell.
• International is larger, cash-generative, and mostly insulated from the single biggest competitive narrative currently pressuring US-focused sentiment.
• Debt-funded 2025 acquisitions are a genuine, high-quality, already-owned collection of assets (SNAI, Betnacional, full FanDuel ownership) — this is a digestion problem, not a strategy problem.
• Net debt is already modestly declining in dollar terms; a disciplined capital-allocation posture (paused buybacks, no further large M&A) plausibly returns leverage to target by FY2027–28.
• Prediction-market newcomers, despite deep funding, remain small in absolute scale and structurally confined to a US-specific legal mechanism.
• Rising compliance and technology costs are pushing sector consolidation toward scaled incumbents — a tailwind Flutter is positioned to benefit from as a bystander, separate from any further M&A it might itself pursue.
• The largest disclosed shareholder’s documented positioning is consistent with a patient, long-horizon view aligned with deleveraging over near-term capital return.

Bear case

• FY2026 would already be the first down year for US Adjusted EBITDA since profitability — the “trough year” framing requires something the segment has never actually delivered before.
• Revenue per US player fell 13.9% year-on-year in Q2 2026 even as player counts grew — a real, quantified risk that growth is being bought at a lower price, not evidence the elective-spend thesis is working yet.
• International faces a second, larger UK tax wave in April 2027 that is not yet fully reflected in most market commentary on the “UK tax problem.”
• Leverage does not return to target within this note’s horizon under the bear case (3.58x by FY2029) — a genuinely worse balance-sheet position than the original thesis assumed, and one that constrains optionality precisely when the business may need to invest through a downturn.
• Prediction-market newcomers are demonstrably well-funded and actively compressing fees to win share — a real, ongoing margin-pressure signal, not a resolved risk.
• The combined valuation range is highly sensitive to WACC and terminal-multiple assumptions; a modestly worse operating outcome combined with a modestly higher discount rate compounds rather than adds.
• Group leadership has changed twice in three months, with an International operator inheriting a US-centric problem at the exact moment the US determines the thesis.

Key uncertainties

Catalyst calendar

Conclusion

Conviction here is moderate: the range is wide by design, and the bear case is a genuine, not a token, scenario — leverage that doesn’t normalize within the note’s horizon, a second UK tax wave not yet priced by consensus commentary, and well-funded competitors actively compressing margin are all live, not resolved. What keeps this a contrarian long rather than a pass is that the market’s own current pricing sits closer to this note’s base case than its bear case, International’s scale and insulation are larger and more durable than a US-headline-driven read of this stock currently credits, and the capital-structure story — while genuinely worse than a prior version of this thesis assumed — is a digestion problem with visible, checkable milestones rather than an open-ended one.

Watching next: Q3 2026 US Adjusted EBITDA against the break-even guide; any sixth buyback tranche; International margin trajectory into Q4; and the first FY2027 guidance under new leadership, which this note treats as the single most important scheduled test of the entire thesis.

Appendices