$LNW: a slot machine oligopoly at half Aristocrat's multiple | Zack Buckley

Summary

• Zach Buckley’s core thesis: Light & Wonder ($LNW) is one of a three-player slot oligopoly with IGT and Aristocrat, has 70%+ recurring revenue, and trades at 7-8x EBITDA and “8 to 9 times my estimate of free cash flow” versus Aristocrat at ~14x EBITDA and 20-25x FCF. He believes the two “should trade very close to parity over time” — and since LNW is growing faster, arguably at a premium — making the discount “shocking.”
• The thesis flipped when Aristocrat’s brain trust migrated: former CEO Jamie Odell, former CFO Tony, current CEO Matt Wilson and roughly 50 executives moved to LNW, while the lottery-business sale cut leverage to ~3x net debt/EBITDA from a much higher prior level. Buckley had passed earlier on exactly those two problems — wrong strategy and too much debt — and got involved only after both changed, first buying in September 2022 in the low $40s USD and selling around $80-100.
• The Australia sole-listing created a forced-selling/reversal opportunity: Buckley re-entered around $73 USD in November when U.S. funds were forced to sell, then sold some at roughly $120 USD in January after the Dragon Train settlement. The stock has since retraced to the high-$70s USD (~AU$110), and “this is the largest I’ve ever sized it and the most excited I’ve ever been about it.”
• The 2026 weakness is launch timing, not necessarily share loss: Aristocrat front-loaded game launches into H1 while LNW’s slate lands in H2, so short-term Eilers data and a soft Q1 look bad. Buckley’s analogy: if PlayStation just launched a console and Nintendo launches later this year, “of course PlayStation sales from like a market-share standpoint are going to be better” — the wrong read is that Nintendo lost its edge.
• The 2028 math: management’s $2B EBITDA target implies roughly $1B of net income and, depending on buybacks and assumptions, roughly $13-15 USD of FCF per share; Walker calculates that at Aristocrat’s multiple, LNW could be a ~$280 USD stock, while at 15x it could be ~$210. Walker says that implies “several hundred percent upside.” Consensus sits clearly below the guide, but this same team set 2025 targets in 2022 and hit them “almost to a T” despite Dragon Train hurting 2025.
• On AI, Buckley thinks the SaaS-like drawdown is “completely unwarranted”: game development could get a productivity tailwind, and SciPlay is probably the only area he views as a genuine AI risk. SciPlay is ~20% of EBITDA, and he attributes its recent softness to sweepstakes rather than AI. He calls LNW “probably the best example of a company that really has very little, if any, AI risk over time… trading with a significant AI discount.” In payments, he cites Dave and Sezzle as meaningful investments made near the lows and later sold after both rose more than 100% off those lows.
• Andrew Walker’s main pushback — could Caesars build its own machines with AI — draws a layered rebuttal: nearly all top developers work for Aristocrat or LNW, poaching a team means waiting about two years before games ship, and a casino would lose access to Buffalo and Huff & Puff. Buckley’s line: “Marriott doesn’t build elevators. Delta doesn’t build airplanes.”
• Capital allocation leans toward the Australian shareholder base: management says repurchases have not paused, though Q1 buybacks were lower, and Buckley cited roughly $100M of Q2 buybacks; management also wants leverage toward 2.5x rather than running at 3.5x and, as Walker put it, “murdering the share count.” Buckley empathizes with the view that lower leverage could support a higher exit multiple, even if paying down 6-7% debt is less attractive than buying shares he believes could offer a roughly 30% IRR. The Grover charitable-gaming acquisition was around $750M, or roughly 7.5x EBITDA, and offers high recurring revenue and growth. Without multiple convergence, Buckley expects an IRR broadly in line with FCF-per-share growth.

Deep dive

1. The thesis was born when Aristocrat’s team defected — and the balance sheet got fixed

• Buckley’s setup: LNW is a slot-machine conglomerate in an oligopoly with IGT and Aristocrat, where Aristocrat was “the darling” and a multibagger from 2010-2020 while LNW “had the wrong strategy… over-leveraged… weren’t investing in game development.” Two things kept him out: strategy and leverage. Both changed.
• The talent migration is the crux: former Aristocrat CEO Jamie Odell (the strategist), former CFO Tony (the finance person), and Matt Wilson (the executor, now CEO) came over — “about 50 executives from Aristocrat in total… this mass migration of talent.” Selling the lottery business helped take leverage down to ~3x net debt/EBITDA, “a level I think is very comfortable, especially for a business that has 70% plus recurring revenue” — and under 10x earnings today.
• His trading history: first bought in September 2022 in the low $40s USD, sold about two years later at $80-100, then re-entered in the low $70s when the Australia sole-listing announcement forced U.S. funds to sell “with really no buying from Australia.” The stock then ran to roughly $120 USD (~AU$180) by mid-January after that flow reversed and the Dragon Train litigation was settled.

2. Why the lease-model oligopoly doesn’t race to the bottom

• The majority of revenue is a lease model — a cut of every pull — putting LNW over 70% recurring with “a high degree of predictability and consistency.” Walker’s framing of the financial-engineering appeal: install the box, take a percentage of revenue, inflation raises the company’s cut, and regional gaming held up much better through the GFC than people assume.
• Buckley says the comparison to video-game development is not perfect, but content power explains why Caesars, MGM and Boyd do not simply force prices down. “Would you want to be negotiating with Grand Theft Auto?” The top developers make games people demonstrably play more; casinos maximize revenue per square foot like retail, so the trade-off can be paying 2% more of revenue for a box that generates 20% more revenue. “There’s no need for a race to the bottom because there’s enough for both sides to be really happy.”

3. Walker’s stress test: what if Caesars builds its own AI-designed slots?

• Walker’s hypothetical: Caesars hires ten scientists and lets AI build and tune its own boxes, cutting out the revenue share on “the most profitable thing in our casino.” Buckley’s rebuttal starts with talent: “almost all of the most talented individuals” work for Aristocrat or LNW, so a competitor must poach a team and then wait — “it’s not like you just poach a team and you have a game tomorrow… you spend years developing games.”
• The full accounting of what a casino would sacrifice: “Marriott doesn’t build elevators. Delta doesn’t build airplanes.” A vertically integrating casino could lose access to Aristocrat’s Buffalo and LNW’s Huff & Puff, spread development costs over far fewer properties, and carry a roughly two-year risk that the games do not work.
• Walker adds the customer-behavior point from his mom and grandma’s slot ritual: players can be fiercely loyal to specific machines, and “you don’t have my Huff and Puff… I’m out” is how a casino could lose a lot of customers — easy to overlook in a spreadsheet. Buckley, an ex-professional poker player (“the only gambling I’ve ever done”), says that developing in-house games would be a major risk and could “honestly be a disaster for the casino.”

4. Australia listing: chasing Aristocrat’s halo, jury still out

• Why abandon a liquid U.S. listing for Australia while remaining Nevada-domiciled and Delaware-incorporated? Buckley: Australia is a heavily gaming-focused market, but “I really think it comes down to Aristocrat being listed there and getting a premium multiple.” LNW wanted the analysts, fund managers and shareholder base that already understand the model, to “help that gap converge.”
• The convergence hasn’t happened yet — some Australian investors choosing between the two pick Aristocrat because “they’ve seen the consistency of the execution over 10-plus years,” while Light & Wonder is “just a newer story.” Buckley’s hedge: “the jury’s out… we really have to judge them on a little bit longer period of time,” with back-half execution against the 2028 plan an important test.

5. Dragon Train and the guide the Street refuses to believe

• The litigation: an employee who came over from Aristocrat “brought information that was downloaded at Aristocrat,” some of that math went into Dragon Train, and LNW settled for roughly $125-130M in mid-January with “a small and partial admittance of guilt without a full acknowledgement.” The game was pulled, but Buckley expects minimal lingering drag on the $2B 2028 EBITDA target.
• His confidence anchor: this team set a 2025 guide back in 2022 — with Dragon Train hurting 2025 too — and “were able to essentially hit that guide almost to a T.” Consensus is clearly below the 2028 targets; his base case is $2B, “I’ll be modestly disappointed, maybe they do 1.9… I would be very surprised if consensus ends up being correct.”
• On why the Street stays skeptical, a sell-side analyst he spoke to last week said investors were focused on recent weakness and “what’s happened so far in 2026.” Buckley says that is the wrong focus. Aristocrat launched in H1, LNW launches in H2, so soft Q1 results and “modestly weak” Q2 Eilers data are, in his view, launch timing rather than evidence of lasting share loss. Walker’s gloss: nobody orders the Switch 1 three weeks before the Switch 2 ships.

6. The 2028 math: several hundred percent upside either way

• The numbers: $2B EBITDA, reasonable assumptions on capex, interest and taxes, gets to roughly $1B of net income in 2028 and — depending on the share count and buybacks — roughly $13-15 USD per share; Buckley’s own estimate was $13-14. Walker calculates that at Aristocrat’s multiple, “somewhere in the ballpark of a $280 stock” USD; at 15x, “which I still think is undervalued but more reasonable,” roughly $210. “It’s just a question of does this trade at 15 or 20 times and I’m not sure but I’m going to be happy either way.”
• Walker notes management cited a minimum AU$14.66 EPS for 2028 on the Q1 call — roughly seven times the current AU$110 price — and that the stocks look a lot like SaaS charts, with AI fears rising around September and peaking around March.

7. AI is a discount, not a risk — except possibly at SciPlay

• Buckley’s AI view: higher productivity per employee in game development could flow to margins, so “the fact that they’ve traded with SaaS is completely unwarranted.” He sees SciPlay, the social/free-slots business, as probably the only area with a meaningful AI risk — “closer to being commodity… less barriers to entry” — where a VC-funded competitor could plausibly appear in a way it could not readily do selling slot machines to Caesars.
• The sizing of that risk: SciPlay is ~20% of EBITDA, and Buckley thinks its current downtrend is driven by sweepstakes rather than AI. “It’s certainly not going away… maybe there’s a small incremental value difference, but nothing worth justifying the sell-off.” He extends the pattern to payments, where investors were “shooting first and asking questions later”: Dave and Sezzle were meaningful investments made near the lows, and he sold them after both rose more than 100% off those lows because the upside had become thinner and the risk more significant.

8. Buybacks, the deleveraging compromise, and the Grover acquisition

• Walker’s capital-allocation pushback: why guide leverage toward 2.5x instead of running at 3.5x and “just murdering the share count” when the stock is this cheap? Buckley’s honest split: paying down 6-7% debt versus buying “a stock that I think is like a 30% IRR — sure, I 100% agree” — but if lower leverage helps the multiple re-rate, for example from 15x to 20x earnings, “that five turns is a huge difference. You get the best exit.” Part of it is “playing to the Australian shareholder base who wants lower leverage and higher free cash flow conversion” — which Buckley empathizes with without fully agreeing. Management says buybacks have not paused; Q1 was lower, and roughly $100M was expected in Q2.
• The Grover charitable-gaming acquisition was around $750M, or roughly 7.5x EBITDA. Walker described it as officially licensed in five states, with nearly 15% year-over-year unit growth in only a handful of states and potential for broader deployment. Grover uses electronic pull tabs — predetermined outcomes, “10,000 tickets” rather than a true random-number generator — in places such as Elks Lodges and American Legions. Buckley sees high recurring revenue, rapid growth and a good fit with LNW’s portfolio.
• Buckley calls the deal arguably similar in quality to Aristocrat, which trades around 14-15x EBITDA, while Grover was acquired at roughly 7.5x. Management has done very few acquisitions, and he views this one as sensible.
• What keeps him up at night: almost nothing company-specific. “If key management team members left, that would certainly give us pause” — otherwise only a big macro shock. His closer: eight-to-nine times free cash flow, FCF per share growing “at 15 to 20% plus clip,” and either convergence to Aristocrat for a very high IRR or, at worst, an IRR in line with FCF-per-share growth itself.