MGM Resorts International (NYSE: MGM) — A Rent-Levered Cash Machine Shrinking Its Own Equity

Independent fundamental research · Report date: July 3, 2026 · Price: $47.10 (2026-07-02)

⚡ Claude’s Take

This block is the author’s own independent opinion and general information only — not investment advice. The analysis that follows (sections 1–15) takes no position, carries no price target, and is written to remain recommendation-free by design.

Verdict: HOLD / accumulate-on-weakness — not a short. “A great capital-allocation story trading at the wrong end of its own five-year range.” MGM is a legitimate, high-quality operator and capital allocator wrapped around a thin, rent-levered equity. Management sold its real estate at premium prices, turned BetMGM cash-positive, sells non-core casinos above the multiple its own stock trades at, and has retired ~47% of the share count in five years — mostly below today’s price. That is exactly the supply-side, per-share compounding behavior you want in a mature industry. But the equity you are buying is a leveraged residual: ~$2.26B of fixed, escalating triple-net rent and ~$4.2B of corporate net debt sit ahead of a cyclical, essentially flat ~$4.6B EBITDAR stream, and the stock carries a 1.28 market beta. At $47.10 — within ~7.5% of its five-year high after a ~+60% momentum rip off the October-2025 low — you are paying a full price for a cycle-late Las Vegas consumer bet and getting the China/BetMGM/Osaka optionality for a modest discount, not for free.

Framing and zone. This is a re-rating-with-room name, not a falling knife and not a blow-off top: the factor tape shows a stock that was dead money for five years (y5 +1.4%/yr, lifetime total return ~0) and has only just clawed back to the top of a range it has failed to break since 2021. On the correct operating lens, whole-company EV/EBITDAR of ~9x is fair-to-slightly-full versus the Caesars take-out (~7.5x) and below Wynn (~11x) — neither cheap nor distressed. The constructive accumulation zone maps to the low-to-mid $30s, where P/S returns toward its ~0.5x floor and the domestic stub stops embedding a full-cycle multiple; the high-$40s/$50 area already prices Vegas holding, BetMGM inflecting, and the buyback continuing — little margin of safety. Fair-value zone roughly $38–$48; today’s tape sits at the upper edge of it. Conviction: medium — the bull mechanics (accretive buyback, monetization arbitrage, under-credited optionality) and the bear mechanics (leverage-on-leverage, cycle-late Vegas, perennial-#3 BetMGM, China minority) roughly offset here, which is precisely why the honest call is HOLD rather than a directional bet.

Triggers. Flips bullish: two-plus consecutive quarters of accelerating Las Vegas Strip EBITDAR and a BetMGM print confirming the ~$500M-2027-EBITDA path — that removes the “cycle-rolling-over” and “digital-is-a-money-pit” bear legs at once. Flips bearish: declining Strip EBITDAR or a Macau/China shock that forces a buyback pause — kill the per-share compounding engine at a 1.28 beta and the thin equity de-rates fast. One honest discomfort: no insider has bought a share of the “cheapest-ever P/S” with personal cash, and directors sold modestly into the rally — there is no conviction-buy to corroborate management’s own valuation argument.

📈 Stock Price Action — Five-Year Event Map

Text-only. Price moves are FACT (AZI adjusted-close series); attributed drivers are INTERPRETATION.

MGM has spent five years going almost nowhere while doing a great deal — a full round-trip from a pandemic-crushed low near $26 to a post-recovery high of $50.90 (Jul 2023), back into the low-$30s, and now back near the top of the range at $47.10 (2026-07-02), just ~7.5% below the five-year high. The 52-week range is $30.72–$50.69, and year-end closes tell the flat-line story punctuated by violent swings: 2021 $44.86 → 2022 $33.53 → 2023 $44.68 → 2024 $34.65 → 2025 $36.49 → today $47.10. The defining feature is not direction but oscillation: this is a high-beta cyclical that has traded a wide ~$26–$51 band for half a decade, and it enters mid-2026 riding a sharp momentum rally off the October-2025 low.

Cycle narrative. (1) The 2020–21 leg was the reopening trade plus a genuine narrative shift — the BetMGM launch let the market re-cast a Vegas operator as a digital-gaming optionality story. (2) The 2021–22 re-rating was structural, not sentimental: MGM sold the real estate under its resorts to VICI and Blackstone, bought the Cosmopolitan’s operations, and reinvented itself as an asset-light OpCo funding one of the largest buyback programs in the S&P 500 — the share count began its ~47% multi-year collapse here. (3) The late-2022 drawdown was macro, not company-specific: a rate shock and recession scare punished high-beta discretionary names indiscriminately. (4) 2023’s rally to the adjusted all-time high was the demand peak — record Las Vegas room rates and convention pricing collided with Macau’s post-zero-COVID reopening. (5) The long 2023–2025 bleed to $25.79 was the mirror image: Vegas growth decelerated against impossible comps, Macau recovery doubts resurfaced, and April 2025’s tariff shock triggered a broad growth-scare repricing of cyclicals. (6) 2025 built a base near $30 on buyback support and a consumer that refused to roll over. (7) The current ~+60% rip since October 2025 is a real-fundamentals-plus-capital-allocation move: Vegas top line turned positive for the first time in six quarters, BetMGM crossed into profitability, and management sold Northfield Park at 6.6x trailing EBITDA — a multiple it pointedly noted is “significantly higher than what is implied by our current share price” — while reaccelerating repurchases. Across the full five years the through-line is a levered, high-beta claim on the Las Vegas consumer that the market keeps re-pricing between fear and greed inside a stubborn ~$26–$51 channel.

1. Executive Summary

MGM Resorts International is an integrated casino-resort operator that, since 2021, has re-engineered itself into an asset-light, rent-levered operating company and one of the most aggressive capital-returners in the S&P 500. It runs nine Las Vegas Strip resorts, a stable portfolio of US regional casinos, a controlling ~56% stake in Hong Kong-listed MGM China (two Macau resorts), a consolidated international online arm (MGM Digital / LeoVegas), and a 50/50 US online joint venture, BetMGM. In its 2021–22 transformation it sold essentially all of its owned real estate to VICI and Blackstone/BREIT and leased it back, and recycled the ~$25B of proceeds almost entirely into buybacks — retiring roughly 47% of the share count in five years (average diluted shares 487M in 2021 to 277M in 2025; ~258M today).

The central analytical fact is structural: MGM’s equity is a thin residual claim on a rent-encumbered EBITDA stream. FY2025 revenue was $17.5B and company Adjusted EBITDAR ~$4.6B, but ~$2.26B of fixed, escalating triple-net rent — nearly half of EBITDAR — is senior to equity, and ~$4.2B of corporate net debt sits alongside it. That leverage is why the stock’s own-history valuation percentiles split three ways and two of them are noise: GAAP P/E (~89th percentile) is broken because FY2025 net income to common was just $206M against ~$2.5B of operating cash flow — depressed by a $279M goodwill impairment, property-transaction swings, and, above all, the ~$315M of consolidated profit (60% of the total) that belongs to MGM China’s 44% public minority; P/B (~99th percentile) is meaningless because years of buybacks have hollowed equity to a near-negative, intangible-heavy residual. The one clean signal is P/S at the ~18.8th percentile — cheapest-ever territory on ~0.7x sales. On the correct operating lens, whole-company EV/EBITDAR is ~9x, fair-to-full versus peers (Caesars take-out ~7.5x; Wynn ~11x).

The investment case does not rest on a wide moat. MGM has one real, financially load-bearing moat — the Las Vegas Strip scale/convention/loyalty position (33.9% segment EBITDAR margin, 4M+ sq ft of meeting space) — one valuable-but-contested license position in Macau (share earned from 9.5% in 2019 to ~16–17% now, not structurally locked), and no moat in US online, where BetMGM is a distant #3 behind the FanDuel/DraftKings duopoly. Roughly half the EBITDAR those positions generate is claimed by rent before it reaches equity. What is genuinely high-quality is the per-share story: ~$1.2–1.6B of real owner free cash flow (not the $2.4B headline), almost no stock-based-compensation dilution, and disciplined asset monetization (Northfield at 6.6x, Mirage, Gold Strike) recycled into shrinking the equity at a bottom-decile multiple.

Consolidated EBITDAR has been essentially flat for three years (~$5.13B segment total in 2023, 2024, and 2025) with an unfavorable mix shift — the wholly-owned Strip cash cow eroding while the 44%-minority Macau business grows. This is a stable-to-mature cash harvester with embedded China/BetMGM/Osaka options, not a compounding machine. The equity is a leveraged bet that the cycle holds long enough for the buyback to keep compounding per-share value; at a 1.28 beta, the same leverage that lifts per-share cash flow on the way up amplifies the equity drawdown on the way down. This memo takes no position and sets no price target; it lays out the embedded expectations and the falsification tests on both sides.

2. Business Overview

What MGM does. MGM is an integrated casino-resort operator and, increasingly, an online-gaming company. It makes money four ways that map to its four reportable segments (FACT — 10-K FY2025, Item 1 and Note 17): Las Vegas Strip Resorts, Regional Operations, MGM China, and MGM Digital. Within each property, revenue splits into casino (table/slot win), rooms, food & beverage, and entertainment/retail/other — a diversified “resort” mix rather than a pure gaming take. On top of the consolidated segments sit two off-balance-sheet growth vehicles: the BetMGM US online JV (equity method) and the Osaka Japan integrated resort (a ~50% JV opening in 2030).

Revenue and profit segmentation (FY2025, FACT — 10-K Note 17):

The mix tells the story. Las Vegas is the profit engine — ~48% of segment revenue and the single largest EBITDAR pool at the fattest margin (33.9%). Regional is a stable ~$3.8B annuity at ~31% margins. MGM China is the growth: revenue compounded from $3.15B (2023) to $4.46B (2025), and its EBITDAR ($1.20B) now rivals Regional’s. MGM Digital (the LeoVegas international online book) is a ~$654M revenue, EBITDAR-negative (−$90M) investment bucket. Because the segment measure is EBITDAR (before rent), the positive segments overstate cash economics until ~$2.26B of triple-net rent is subtracted, taking company Adjusted EBITDAR of ~$4.6B down to post-rent Consolidated Adjusted EBITDA of $2.36B (FACT — Note 17).

The OpCo/PropCo transformation — MGM no longer owns its dirt. The defining structural fact about MGM is that it has sold essentially all of its owned real estate and now operates as an asset-light tenant. In April 2022 the MGM Growth Properties (MGP) REIT was merged into VICI Properties, and MGM’s owned real estate under a dozen domestic properties is now leased back from VICI under a single master lease: $775M annual cash rent (May-2025 lease year), a 25-year initial term with renewals and a fixed 2% escalator stepping to the greater of 2%/CPI capped at 3% (FACT — 10-K Note 11). Separately, the crown-jewel Strip trophies were sold to Blackstone/BREIT vehicles: Bellagio (leased back at $276M/year), Mandalay Bay + MGM Grand Las Vegas (a joint BREIT venture, ~$322M/year), and The Cosmopolitan real estate ($212M/year). Aggregate GAAP rent runs ~$2.26B; contractual cash rent ~$1.8B (FACT — 10-K Note 11 / risk factors). MGM also retains residual real-estate exposure through ~$6.0B of shortfall guarantees on the Bellagio and Mandalay Bay/MGM Grand landlord debt — the tail is not fully severed (INTERPRETATION).

The economic consequence is central to the whole thesis: MGM converted a hard-asset balance sheet into a stream of operating leases plus a war chest of cash, and recycled the proceeds almost entirely into buybacks. What remains on the balance sheet is a ~$25.3B capitalized lease liability, ~$6.3B of debt (of which ~$2.5B is non-recourse inside MGM China) against ~$2.1B cash, negative tangible book, and an operating position rather than a property position.

BetMGM — a 50/50 equity-method JV, off the consolidated top line. BetMGM is owned 50/50 with Entain plc; MGM contributed exclusive US access, Entain contributed technology, and all major decisions require both partners’ consent (FACT — 10-K Note 9). Because neither controls, it is equity-method: BetMGM’s revenue (JV total ~$2.8B) does not appear in MGM’s consolidated revenue; only MGM’s 50% share of the net result flows through “income/(loss) from unconsolidated affiliates.” That share was +$59.6M in 2025, a genuine inflection from −$110.1M (2024) and −$90.9M (2023) — BetMGM crossed into profitability, distributed $135M to MGM in 2025, and began paying a branding fee (FACT — 10-K Note 9; Q1-2026 call, 2026-04-29).

Recurring vs. cyclical. Convention/group room and F&B revenue is the most recurring and highest-margin stream — booked years ahead at premium ADRs, smoothing the leisure cycle. Casino win (especially Macau baccarat and Strip high-end play) is the most cyclical and, in Macau’s case, geopolitically sensitive. The lease rent is a fixed, escalating cost regardless of the cycle — operating leverage cuts both ways, and with ~$2.26B of rent ahead of equity, a downturn hits MGM’s residual cash flow harder than it would an unlevered owner-operator (INTERPRETATION).

Verdict. MGM is a good-quality collection of operating businesses wrapped around a financial-engineering strategy. The Strip and Macau assets are genuinely attractive; Regional is a stable annuity; BetMGM has turned the corner; MGM Digital is a cash-consuming option. But post-transformation this is an asset-light, rent-levered operator that owns brands, licenses, and customer relationships rather than land — the equity is a leveraged claim on resort EBITDAR after a large, fixed, escalating rent check. That is a legitimate model, but it means the moat, if any, has to live in the operating position, not the real estate.

3. Industry Dynamics

MGM competes in four structurally distinct arenas, and the honest answer to “good or bad industry?” differs by segment. Lumping them together is the most common analytical error here.

Las Vegas Strip — a genuine oligopoly on scarce land, but a mature, cyclically-exposed leisure market. The Strip is one of the world’s best examples of a supply-constrained oligopoly. Gaming-licensed, zoned Strip frontage is effectively fixed; no material new integrated resort has opened since Resorts World (2021). Three operators — MGM, Caesars, and Wynn — plus the Apollo-owned Venetian/Palazzo and a handful of independents control the bulk of rooms and gaming positions. In Marathon capital-cycle terms this is the favorable half of the cycle: supply is essentially frozen because land is scarce, construction costs are prohibitive, and licensing is a barrier, so demand growth accrues to incumbents rather than triggering a capacity flood (INTERPRETATION — Greenwald/Marathon lens).

The demand side, however, is cyclical and currently soft. Las Vegas drew ~38.5M visitors in 2025, down ~8% and the weakest since the 2021 rebound, hurt by softening consumer confidence, higher travel costs, and a slump in international visitation — notably Canada, down ~20–28% (FACT — LVCVA, Jan 2026; casino.org). Yet Strip gaming revenue still set a record (~$8.8B market-wide), and convention attendance held at ~6M, roughly flat YoY though still ~10% below the 2019 record of 6.6M (FACT — LVCVA). The convention/group flywheel is the structural attraction: it fills rooms midweek at premium rates, is booked years ahead, and is where MGM’s 4M+ sq ft of meeting space is decisive. The entertainment/sports flywheel reinforces it — Allegiant Stadium (Raiders), the F1 Las Vegas Grand Prix, recurring Super Bowls (2029), the Sphere, a College Football Playoff championship (2027), the Final Four (2028), and a targeted NBA franchise turn Vegas into a year-round events destination that lifts ADRs on event weekends. But MGM’s own 2025 print — Strip EBITDAR down ~10% over two years and margin down to 33.9% from 35.2% — shows the cyclical softness is real even as the structure remains sound (FACT — 10-K Note 17).

Macau — a licensed six-way oligopoly, recovering but not yet back to 2019, with a favorable secular mix shift. Macau is the world’s largest gaming market and structurally a government-created oligopoly: gaming is administered through concessions awarded to exactly six concessionaires (MGM China, Sands China, Galaxy, Wynn Macau, Melco, SJM), whose current 10-year concessions run January 2023 through December 2032 (FACT — 10-K Item 1). Full-year 2025 GGR reached ~MOP247.4B (~$30.8B), +9.1% YoY — the highest since the pandemic but still only ~85% of the 2019 level (FACT — GGRAsia/GGB, Jan 2026). The recovery’s character matters more than its level: the market has pivoted decisively from VIP junket play (gutted by the 2021–22 crackdown) toward premium-mass and mass, which carries lower credit risk and structurally higher margins. The concessions also obligate operators to invest heavily in non-gaming amenities — a regulatory tax on capital that suppresses returns but raises the barrier to entry (INTERPRETATION). This is a good industry structure — six licensed players, no new entrants possible before 2033 — attached to a demand base politically dependent on Beijing; June 2026 printed the first YoY GGR decline of the year, a reminder of the volatility (FACT — GGRAsia).

Regional US gaming — mature, competitive, low-growth. MGM’s regional footprint (MGM Grand Detroit, Borgata in Atlantic City, National Harbor, Beau Rivage, Empire City, Springfield) sits in a mature, largely saturated market where growth comes from occasional new-jurisdiction openings rather than same-store volume. Individual properties often enjoy quasi-monopoly local licenses (National Harbor near DC; MGM Grand Detroit as one of three Detroit licensees). US commercial gaming set a fifth straight record in 2025 — $78.7B GGR, +9.2% — but land-based casino/racino GGR grew only ~2% to ~$51B; the growth was online (FACT — AGA State of the States 2026). Regional is a stable-but-slow annuity, not a growth engine.

US online (OSB + iGaming) — a fast-growing but brutally competitive, politically-taxed profit pool. This is the industry’s growth vector and its most contested. US regulated sportsbooks processed ~$165.6B of handle in 2025 for ~$16.8B of GGR; iGaming GGR grew ~28% to $10.74B (FACT — AGA). But online sports betting has consolidated into a FanDuel/DraftKings duopoly controlling roughly two-thirds of handle, and the trailing pack (ESPN Bet, Fanatics, BetMGM) has demonstrably failed to buy its way in — evidence of a real economies-of-scale barrier favoring the top two (FACT — DKNG report; casino.org). Worse, the profit pool is politically determined and shrinking at the margin: New York taxes OSB at 51%, New Jersey raised its rate to 21% (2025), Illinois added a graduated rate plus per-wager fee, and the federal 2025 tax bill capped gambling-loss deductions at 90%. iGaming is legal in only a handful of states, so the biggest structural upside (broad iGaming legalization) remains an option, not a fact.

Barriers to entry. Across every segment the binding barrier is the same class of asset: government-issued gaming licenses plus irreplaceable land/position. Strip frontage and Macau concessions cannot be reproduced at any price; regional licenses are capped by statute; online market access requires state-by-state licensing. These are formidable, durable barriers — but they protect the industry structure, and in the online and regional segments they do not specifically privilege MGM over equally-licensed rivals.

Verdict. Structurally attractive by segment, unevenly. The Las Vegas Strip and Macau are genuinely good industries — licensed oligopolies on unreproducible assets with frozen or capped supply — though both carry real demand cyclicality. Regional is a mature, low-growth annuity — fine, not exciting. US online is a good market but a bad competitive position for the #3 player. On balance MGM operates in better industries than most consumer-cyclicals, but the quality is concentrated in the two segments where supply is physically constrained.

4. Competitive Position

The right question, post-transformation, is not “does MGM own great assets?” (it sold most of them) but “does MGM hold a durable operating position that produces returns a licensed competitor could not replicate?” The answer, in Greenwald’s taxonomy, is a real but narrow moat on the Strip, a licensed-but-contested position in Macau, and no moat at all in US online — and the consolidated returns confirm the advantages are modest and rent-diluted.

Las Vegas Strip — intangible + economies-of-scale moat, expressed through the operating position, not the dirt. MGM’s Strip moat is the strongest thing it owns. It operates nine resorts — Aria, Bellagio, The Cosmopolitan, MGM Grand, Mandalay Bay, Luxor, New York-New York, Excalibur, Park MGM — controlling roughly 35–40% of Strip room inventory, anchored by 4M+ sq ft of convention/meeting space (FACT — 10-K Item 1; INTERPRETATION on the aggregate share). This is a genuine economies-of-scale-plus-customer-captivity advantage: the convention footprint is a high-fixed-cost asset only the largest operators can offer at citywide scale; group customers are captive (multi-year contracts, the MGM Rewards loyalty database, cross-property play); and the density lets MGM flex room inventory across nine properties in a way a single-resort competitor cannot. Crucially, this moat survives the sale-leaseback — it lives in the brands, the meeting-space footprint, the loyalty program, and operating scale, not in owning the buildings. The financial signature is the 33.9% Strip EBITDAR margin and record gaming revenue in a down-visitation year. But it is narrow: it is a local scale advantage, it does not travel, and it is shared with Caesars in the same city. Remove the convention/scale advantage and Strip margins compress toward a single-property operator’s — so the moat is real and load-bearing, but it is a duopoly moat, not a monopoly (INTERPRETATION).

Macau — a license moat that is valuable but six-way contested, and MGM is the subscale share-gainer. MGM China’s advantage is the concession itself — one of only six, non-reproducible until 2033. Within that protected structure, MGM China has been the standout share-gainer: from 9.5% of Macau GGR in 2019 to 15.8% in 2024 and ~16–17% in 2026 (FACT — asgam; Q1-2026 call cited 15.4% Q1 exiting March at 17.3%), a ~600bp+ gain achieved with the fewest hotel rooms of the six concessionaires — i.e., the best revenue-per-room execution in the market, driven by premium-mass focus at MGM Cotai. By Greenwald’s share-stability test, a multi-point gain is the opposite of a stable-oligopoly signature: it signals share is earned by execution, not locked by structure, which cuts both ways — MGM took it and could give it back. The moat is the license; the outperformance is operational skill, which is emulable. Durable license, contestable share — a good position but not unassailable, and one exposed to Beijing policy and the ~$23M/year brand-fee increase (1.75%→3.5% of revenue) that now flows more cash to the parent (INTERPRETATION).

BetMGM — no moat; a distant, restructuring #3 in a duopoly market. Name the loser: this is MGM. In US online sports betting, BetMGM is a distant #3 at ~14% of revenue / ~10–12% of handle, in a market where FanDuel and DraftKings control roughly two-thirds and where well-funded challengers (ESPN Bet, Fanatics) have proven they cannot buy into the top two (FACT — DKNG report; casino.org). The network-effect/scale moat in OSB exists — but it belongs to the top two. In iGaming, BetMGM is relatively stronger — a solid #2–3 and still #1 in some states (Michigan, West Virginia) — but it ceded its former iGaming leadership to FanDuel starting mid-2023 (FanDuel ~28.5%, DraftKings ~23.8%, BetMGM behind; FACT — bonus.com, 2025 data). The turn to profitability (MGM’s +$59.6M 2025 share; JV ~+$220M EBITDA) is real and welcome, but it reflects the whole industry inflecting to profit as promo spend rationalizes, not a BetMGM-specific advantage; customer captivity is weak everywhere in online (multi-homing is one app download). BetMGM is a legitimate top-3 asset that has escaped cash-burn, but it is the weaker side of a duopoly-plus-one structure — there is no evidence of a durable competitive advantage, and it should be valued as a profitable-but-share-losing #3 (INTERPRETATION).

Returns reality — the tell that the moat is modest and rent-diluted. The cleanest disconfirming evidence is returns on capital. Pre-2020, MGM as an asset-heavy owner-operator earned mid-single-digit-to-low-double-digit ROIC — respectable but never the 15–25% sustained returns Greenwald associates with a strong moat. Post-transformation the picture is distorted by the ~$25.3B capitalized lease liability and negative tangible book, but the underlying cash economics are unambiguous: ~$4.6B of company Adjusted EBITDAR is reduced by ~$2.26B of rent to ~$2.4B of post-rent EBITDA, and net income to common was just $206M in 2025. A wide-moat business would not see roughly half its property EBITDAR consumed by a landlord’s fixed claim. The sale-leaseback created shareholder value through capital return — it did not create a competitive moat; if anything it thinned the residual economics a moat is supposed to protect (INTERPRETATION).

Direct peer comparison. Caesars (CZR) is the other asset-light Strip-plus-regional operator, more financially levered, with a weaker Vegas convention position but a similar #3-ish online book — a fair mirror of MGM’s model with more balance-sheet risk. Wynn (WYNN) is a Macau-heavy luxury near-pure play with a smaller, higher-end Strip footprint and superior per-property margins but far less scale and convention breadth — a narrower, deeper moat. Las Vegas Sands (LVS) exited Las Vegas entirely to become a pure Asia (Macau + Singapore) operator, with Singapore’s duopoly a genuinely better structure than Macau’s six-way. Relative to these, MGM is the most diversified and best-positioned on the Strip convention flywheel, but it is not the highest-return or highest-moat operator in any single arena — it is the scale generalist.

Verdict. MGM has one real, financially load-bearing moat (the Las Vegas Strip scale/convention/loyalty position), one valuable-but-contested license position (Macau), and one no-moat business (BetMGM). The moats that exist are narrow and local, and a large share of the EBITDAR they generate is now claimed by ~$2.26B of fixed rent before it reaches equity holders. This is a good operator with a narrow, segment-specific moat, not a wide-moat compounder. The investment case cannot rest on durable advantage protecting supernormal returns; it rests on the operating position being good enough to service the rent while management shrinks the equity through buybacks.

5. Growth History and Forward Opportunities

Historical growth — a recovery story layered on two acquisitions and one consolidation, not an organic compounding machine. Consolidated net revenue grew from $9.68B (2021) to $17.54B (2025) — a ~16% CAGR that badly overstates the underlying franchise. Three non-organic events explain most of it: the COVID recovery off a $5.16B 2020 trough; the May 2022 acquisition of The Cosmopolitan operations (~$1.7B); and the September 2022 consolidation of LeoVegas that created the MGM Digital segment (FACT — 10-K). Strip those out and the organic story is pedestrian. The most recent year makes the point: 2025 consolidated revenue rose just ~2%, with MGM China +11% and MGM Digital +19% offsetting a Las Vegas Strip that fell ~4% to $8.44B. And on the measure that matters — Consolidated Adjusted EBITDA — the company was essentially flat: $2.34B (2023) → $2.41B (2024) → $2.43B (2025) (FACT — 10-K). ~2% over two years while revenue rose ~9%. The operating engine is mature and margin-pressured, not growing (INTERPRETATION).

Forward drivers, ranked by materiality and probability:

(1) Buyback-driven per-share growth — the real near-term compounding engine (high probability, high materiality). MGM has repurchased ~$9.4B of stock since 2021, cutting average diluted shares from 487M (2021) to 277M (2025), with the balance-sheet count now ~258M. A fresh $2.0B authorization was approved in April 2025, and management explicitly frames post-Northfield proceeds as fuel to reaccelerate buybacks because asset-sale multiples imply a value above the share price. On roughly flat EBITDA, a ~5–8%/yr reduction in share count is where per-share growth comes from. This is financial engineering on a real free-cash-flow base — legitimate and accretive at a low P/S, but not business growth (INTERPRETATION).

(2) BetMGM path to profitability (high probability, moderate materiality). BetMGM inflected in 2025: MGM’s share swung to +$59.6M from −$110.1M (2024). At the JV level, ~$2.8B net revenue (+33%) and ~+$220M EBITDA (from −$224M), with iGaming (~$1.82B) the profit driver. But BetMGM is a clear #3 in OSB and ceded iGaming leadership, and FY26 JV guidance was cut to ~$2.9–3.1B revenue with a Q1 EBITDA miss — a warning shot (FACT — Gaming America, 2026). At MGM’s ~50% share, digital is a ~$100–150M EBITDA contributor, not a needle-mover against a $2.4B base (INTERPRETATION).

(3) MGM China premium-mass share gains + suite additions (moderate probability, moderate materiality). MGM China revenue grew to $4.46B (2025, +11%), EBITDAR $1.20B at 27.0%; share ~15.4% Q1’26, exiting March ~17.3%; the property is “under-suited,” with premium suites being added. Macau GGR grew ~10.9% YTD through May 2026, though June printed the first YoY decline of the year (FACT — GGRAsia). Share gains are real but the market is maturing and the 3.5% brand fee now skims more to the parent (INTERPRETATION).

(4) LV convention/group + sports/entertainment flywheel (moderate probability, low-moderate near-term materiality). Vegas grew top-line for the first time in six quarters in Q1’26 on record convention ADRs, the MGM Grand room remodel returning ~54k room-nights, and a Marriott distribution partnership. The event calendar (F1, targeted NBA franchise, CFP 2027, Final Four 2028, Super Bowl 2029) supports midweek demand. But 2025 visitor volume fell ~8% and Strip RevPAR slipped to $229 from $245. Group/convention is a stabilizer and a genuine edge, but today it offsets leisure/international softness rather than driving net growth (INTERPRETATION).

(5) Osaka Japan 2030 (low near-term materiality, back-end-loaded, capital-intensive). MGM is the sole licensed IR operator in Japan via a ~50%-owned JV (10-K states 50%; 2026 press reports MGM/Orix each ~44% after smaller investors joined — a discrepancy flagged, see Open Questions). Total project cost has risen to ~¥1.51T (~$10.2B); MGM is funding $200–225M of equity in FY26, pre-funded via a yen facility, with a remaining commitment of ~¥428B (~$2.8B) through the 2030 opening (FACT — 10-K commitments; GGRAsia 2026). Japan offers a 120M-population market and ~40M international visitors with no domestic casino competition. A well-located monopoly IR should generate attractive returns — but it is four-plus years out, subject to cost inflation and FX, and contributes zero EBITDA until 2030 (ASSUMPTION).

(6) UAE Dubai optionality (speculative). MGM is building a non-gaming resort in Dubai and holds ~0.25M sq ft for a potential casino license should the UAE authorize gaming (FACT — Q1’26 call). A free option with no cost basis in the current price; unquantifiable (INTERPRETATION).

Verdict — mixed quality, and the two “growth” stories point in opposite directions. Organic EBITDAR growth is low-quality and modest: the core Strip is flat-to-down and margin-compressing; Regional is flat; MGM China is growing but maturing and now taxed a higher brand fee. The genuinely high-return organic seeds — BetMGM’s iGaming position and the Osaka monopoly — are either small relative to the base or years away. What is high-quality and high-probability is the per-share story: a real ~$2.4–2.5B operating cash-flow base and an aggressive, sustained buyback shrinking the equity ~5–8%/yr. MGM is not compounding EBITDAR; it is compounding per-share claims on a flat EBITDAR stream by shrinking the count — a legitimate but distinct, and lower-multiple, kind of growth.

6. Financial Quality

The one-sentence frame. MGM is a genuine cash generator sitting on a thin, near-negative equity sliver, because two structural features — a ~$2.26B annual triple-net rent obligation and a 44%-owned Macau subsidiary consolidated at 100% — make every GAAP-equity, ROE, and P/B statistic meaningless. The economics are real; the reported equity is an accounting artifact of the buyback program. You underwrite MGM on cash and on rent coverage, not on book value.

Margin structure and the rent wedge. Consolidated Adjusted EBITDAR is ~$4.6B, a ~26% margin on revenue — a respectable resort/gaming figure. But between EBITDAR and operating income sits triple-net lease rent of $2,258.4M (FACT — Note 17), which converts a ~26% pre-rent margin into a ~13% post-rent EBITDA margin ($2,356M) and a ~5.7% operating margin. Rent consumes ~49% of consolidated Adjusted EBITDAR. This is the defining feature of the post-MGP/VICI, post-BREIT MGM: management sold the real estate for ~$25B and in exchange the business now carries a fixed, escalating, senior-to-equity cash claim that grows at the greater of 2% or CPI. Rent has been roughly flat at ~$2.26B for three years, but contractually it ratchets up every year regardless of how the properties perform. That is the equity’s biggest structural risk: it is a levered claim on a rent-encumbered EBITDA stream.

The real FCF bridge — this is the crux, and it is a long way below the “$2.4B” headline. True free cash flow = cash from operations minus capital expenditures, both from the 10-K cash-flow statement:

The “~$2.4B FCF” figure often quoted is not real free cash flow — it is roughly CFO before capex (the aggregator convention). Actual owner FCF is ~$1.46B on a simple basis and ~$1.26B once you net out cash that actually leaves to MGM China minorities and the equity calls into BetMGM/Osaka (FACT — 10-K cash flows). Even that overstates parent-level discretionary cash: capex of ~$1.07B is not purely maintenance (it embeds Macau room/suite reinvestment and Strip remodels), and the Osaka obligation flows through “investments in unconsolidated affiliates,” with ~$2.8B of equity still to fund through 2030. So the forward call on this cash stream is heavier than the trailing bridge suggests. Owner FCF is genuine and substantial, but it is closer to ~$1.2–1.6B than to $2.4B (INTERPRETATION).

Owner FCF yield. On the ~$12.2B market cap, simple FCF of ~$1.46B is a ~12% owner-FCF yield; the ~$1.26B economically-adjusted figure is ~10%. On an ex-lease enterprise value (~$16.4B), simple FCF is a ~8.9% yield. All three are attractive absolute numbers — this is not an expensive stock on cash — but the yield is a claim on the residual after a $2.26B senior rent check, carrying far more operating leverage to a Vegas downturn than the headline multiple implies (ASSUMPTION on the split of maintenance vs. growth capex).

Quality of earnings — reconciling $206M GAAP net income to owner cash. FY2025 net income attributable to MGM common was $206M, down from $747M (2024) and $1,142M (2023). The collapse is almost entirely non-cash and non-recurring distortion, not deteriorating cash economics:

• Minority interest: consolidated net income was $520.9M, of which $315.0M (60%) was attributable to noncontrolling interests — overwhelmingly the 44% of MGM China’s now-booming profit MGM does not own. Consolidation pulls 100% of MGM China’s revenue/EBITDAR into the top of the P&L, then strips 44% of the profit back out at the bottom. This is the single largest reason GAAP EPS is noise.
• Goodwill impairment: a $278.9M non-cash goodwill impairment (Empire City license withdrawal) hit 2025 and did not exist in 2024/2023.
• Property transactions & preopening: −$126.0M in 2025 (vs. a +$370.5M gain in 2023) swings reported operating income by ~half a billion year-to-year with zero cash content.
• SBC: stock-based compensation is only $90.5M (~0.5% of revenue, ~6% of FCF) — extraordinarily low, a sharp contrast to software/fintech names where SBC eats all of FCF. MGM’s reported earnings are not flattered by add-back accounting; owner FCF ≈ real FCF here.

Netting these out, MGM’s owner cash earnings (~$1.2–1.5B) are roughly 6–7x the $206M GAAP figure. The market looking at a ~64x trailing GAAP P/E is looking at a broken denominator; the P/S rank (18.8th percentile — cheapest-ever) and P/CFO (~4x) are the honest tells (INTERPRETATION).

ROIC/ROE/P/B are meaningless — use ROA and cash-on-cash. FY2025 total equity is $3,276M, of which $846M is minority, leaving ~$2,430M attributable. Against $4,902M goodwill and $1,357M other intangibles, tangible book equity is deeply negative — the mechanical result of buying back ~$9.4B of stock while selling the real estate. Consequently P/B (~5x, AZI 99.3rd percentile) is a nonsense signal, P/TBV is undefined, and ROE is uninterpretable and swings with each year’s impairment noise. On a cash basis the business earns a high-single-digit to low-double-digit unlevered cash return — decent, not exceptional (INTERPRETATION).

Balance sheet and real leverage — two entirely different pictures coexist.

• Ex-lease (OpCo/refinancing view): total debt principal $6,260M, of which ~$2,488M sits inside MGM China (non-recourse to the parent) and ~$3,772M is corporate. Against $2,063M cash, corporate net debt is ~$4.2B — versus ~$4.6B EBITDAR, <1.0x; versus post-rent EBITDA, ~1.8x. The revolver is undrawn, interest incurred is $421M, and near-term maturities are manageable but chunky ($1,150M due 2026, $1,425M due 2027).
• Lease-adjusted (rating-agency/economic view): capitalize the ~$25.3B of lease liabilities and total obligations balloon to ~$29.5B, or ~6.4x Adjusted EBITDAR. Moody’s/S&P treat the master leases as debt-like precisely because the rent is fixed, escalating, and senior. This is why MGM is a sub-investment-grade credit at the corporate level despite <1x “real” leverage — the rent is the leverage.

Verdict — a real cash machine, but a rent-encumbered one on thin equity; economics do not obviously improve with scale. MGM throws off ~$1.2–1.6B of genuine owner FCF against a ~$12.2B cap (~10–12% yield) with almost no SBC dilution, clearing the “is this a real business” bar decisively. But three quality flags keep it out of the top tier: (1) the ~49%-of-EBITDAR rent burden converts modest operating softness into large equity swings; (2) the growth is concentrated in the one 44%-minority-owned segment (Macau), so ~$315M of consolidated profit is not MGM’s while the wholly-owned Strip cash cow is shrinking; (3) reported equity, ROE, ROIC and P/B are structurally uninterpretable, forcing the entire thesis onto cash and rent-coverage. Scale has not produced expanding margins — post-rent margins are flat-to-down — so this is a stable-to-mature cash harvester with a Macau/Osaka call option, not a compounding machine.

7. Capital Allocation

The strategy in one line: for five years management has run a single, coherent, aggressive playbook — monetize the real estate and non-core casinos at high multiples, then shrink the equity relentlessly through buybacks — while funding two large growth options (BetMGM, then Osaka/UAE/Macau) on the side.

The buyback machine. MGM has repurchased approximately $9.4 billion of stock in five years (FACT — cash-flow statements): $1,753.5M (2021), $2,775.2M (2022), $2,291.9M (2023), $1,357.9M (2024), $1,228.3M (2025). Average diluted shares fell from ~487M (2021) to ~277M (2025), and balance-sheet shares are now 258.3M — a ~47% reduction in five years, one of the most aggressive programs in the S&P 500. Q1’26 continued it (~2.5M shares / ~$90M). Was it accretive? The stock traded a ~$26–51 band the entire period, and the bulk of the spend (2021–2023, ~$6.8B) was executed at a blended cost in the high-$30s to low-$40s — below today’s $47.10. So the program has been net accretive to continuing holders, and per-share cash flow has risen even as total EBITDAR stayed flat. The legitimate criticism is leverage, not price: funding buybacks partly by selling the real estate means the per-share cash flow that remains is a more levered, rent-senior residual. Buying back stock is accretive to EPS and defensible on valuation, but it has structurally lowered the quality of each remaining share by trading owned real estate for a fixed rent claim — a real trade-off, not a free lunch (INTERPRETATION).

Asset monetization — genuinely well-executed. This is the strongest part of the record. Over the cycle MGM merged MGP into VICI, sold Bellagio to BREIT, sold the Mirage operations ($1.075B, 2022), Gold Strike Tunica ($450M / $399M gain, 2023), and most recently Northfield Park to Clairvest for $546M at ~6.6x trailing EBITDA (closed early April 2026), with the VICI master lease amended to cut annual rent by $53M. Selling operating casinos at high-single-to-double-digit EBITDA multiples while the parent’s own equity trades near ~9x EV/EBITDAR and a bottom-decile P/S is textbook Marathon capital-return discipline: dispose where the private multiple exceeds the public multiple, recycle into the cheapest asset available — its own shares. The counter-argument (Greenwald lens) is that these sales have narrowed the moat: MGM sold its irreplaceable Strip real estate — the one genuinely scarce, un-reproducible asset it owned — and is now a tenant on the land that is its only durable advantage. You cannot buy back that optionality (INTERPRETATION).

M&A and growth capital. The offsetting deployments: The Cosmopolitan operations (2022, a high-quality Strip adjacency at a reasonable multiple — sound); LeoVegas (2022, now MGM Digital) and Tipico’s US technology (the question marks — MGM Digital still loses money, −$90.3M segment EBITDAR in 2025, a widening loss, and the “breakeven by 2027” path is management hypothesis, not evidence). The international-online build is a speculative growth option consuming ~$90M/yr with unproven unit economics (INTERPRETATION).

BetMGM and the growth options. BetMGM finally inflected — MGM’s share +$59.6M in 2025 (first meaningful profit), and MGM received $135M in distributions after years of contributing capital; the equity investment is carried below zero (a $160M liability for losses in excess of the investment). A clean win: MGM funded a multi-year cash-burning option, it turned cash-positive, and MGM now collects distributions and a rising branding fee. Against that, two large forward calls loom: Osaka — a remaining equity commitment of ~$2.8B for MGM’s ~43.5% share through the 2030 opening — and the UAE/Dubai project. These multi-billion-dollar, long-dated bets will compete with buybacks for cash over 2026–2030. The capital-allocation tension is explicit: keep shrinking the equity, or fund Osaka/UAE — MGM has committed to both, which will require either continued asset-sale proceeds or higher leverage (INTERPRETATION).

Dividend. Effectively none. MGM cut its dividend to a token $0.01/share in 2020 and suspended it entirely in February 2023 in favor of buybacks. Defensible — buying back a bottom-decile-P/S stock beats a sub-1% yield — and consistent with the shrink-the-equity thesis; not a red flag.

Incentive alignment. Per the 2026 DEF 14A: the annual bonus is weighted 75% to a Compensation Adjusted EBITDAR target, and long-term incentives are 50% relative-TSR PSUs (vs. the S&P 1500 Hotels/Restaurants/Leisure index) plus 50% time-vested RSUs. CEO Bill Hornbuckle’s FY2025 total compensation was $25.3M (up from $15.8M in 2024). Say-on-pay passed with 89.3% support. The alignment is reasonable but imperfect for this specific strategy: relative TSR and RSUs tie management to the share price (good, given the buyback focus), and EBITDAR is the right operating metric — but there is no per-share, FCF, ROIC, or leverage metric in the plan, a notable gap for a company whose entire equity thesis rests on per-share cash flow and rent coverage. Rewarding absolute Adjusted EBITDAR can, in principle, reward the empire-building (Osaka/UAE) the shrink-the-equity strategy is supposed to discipline. The jump in CEO pay to $25.3M in a year GAAP EPS collapsed also sits awkwardly (INTERPRETATION).

Insider read. Across every Form 4 filed in 2025–2026 there are zero open-market purchases (code P). All executive activity is routine: option/RSU exercises, tax-withholding dispositions, grants. There were discretionary open-market sales by directors (Daniel J. Taylor at $38.44; Keith Meister). Two strategic/activist holders sit on the board: Corvex Management (Keith Meister), ~5.35M shares / 2.09%, and IAC. CEO Hornbuckle owns 827,854 shares. No insider has stepped in to buy the “cheapest-ever P/S” with personal cash — there is no conviction-buy signal to corroborate management’s valuation argument, and the tape shows modest director selling into the rally (FACT — Form 4 corpus; INTERPRETATION on signal).

Verdict — a coherent, mostly well-executed program with one honest debate at its center. High marks for disposal discipline (selling casinos and real estate above its own equity multiple), for seeing BetMGM through to cash generation, and for relentless, largely-accretive share shrinkage (~47% of the count, ~$9.4B, mostly below today’s price). Applying the Marathon lens, this is exactly the supply-side, capital-returning behavior you want in a mature industry. The two reservations: (1) the buybacks were funded in part by permanently surrendering the Strip real estate — MGM sold its scarcest asset to shrink the equity, converting owners into levered residual claimants behind a $2.26B senior rent check; and (2) the plan now has to fund ~$2.8B of Osaka equity plus UAE against that same cash, with incentive metrics that don’t fully police the choice. Directionally, management has allocated capital intelligently given the hand it dealt itself — but the wisdom of the whole strategy is inseparable from whether the real estate was worth more owned than the buybacks were worth executed, and the negative insider signal offers no independent confirmation that today’s price is the bargain the CFO says it is.

8. Changes and Headwinds — Last Two Years

Asset-light monetization accelerated (thesis-strengthening). The defining strategic thread remains selling operations at high multiples to shrink the equity. Northfield Park ($546M, ~6.6x, closed April 2026) follows the Mirage ($1.075B, 2022) and Gold Strike Tunica ($450M, 2023). Management explicitly ties disposals to accelerated buybacks, arguing private-market multiples sit well above MGM’s equity multiple. Strengthens the thesis — it validates a higher private valuation of the OpCo and funds equity shrinkage — but progressively thins the asset base and adds rent-encumbrance with each sale (INTERPRETATION).

Empire City license withdrawal (thesis-neutral, a cleared overhang). In Q3’25 MGM withdrew its application for a NY downstate commercial gaming license for Empire City, taking a $256M goodwill impairment plus $93M of write-downs. It continues operating Empire City as a video-lottery facility. The GAAP charge is ugly and drove the 2025 net-income optics, but withdrawing avoided a large, low-return capital commitment in a saturated NYC-metro race — capital-discipline positive (INTERPRETATION).

MGM China brand-fee increase 1.75%→3.5% of revenue (mildly positive for parent, negative for minority). Effective 2026 the brand/services fee doubled, adding ~$23M in Q1’26 and routing more Macau cash to the parent. Net positive for MGM parent FCF; a transfer away from the ~44% MGM China public minority — a mild related-party flag, not a red one (INTERPRETATION).

Digital build-out — LeoVegas/Tipico US technology (thesis-supporting, still loss-making). MGM is migrating BetMGM and MGM Digital onto owned/controlled technology, reducing reliance on Entain’s stack and targeting ~breakeven for MGM Digital in 2027. Owning the tech is strategically correct for long-run margins; near-term it is a cash cost (INTERPRETATION).

Rising self-insurance and “frivolous, PE-backed” litigation cost pressure (thesis-weakening, structural). Management flagged a step-up in self-insurance reserves — Q1’26 Vegas EBITDA hit ~$37M and Regional ~$9M — attributed partly to rising premises-liability litigation. External data corroborate a “new normal” (federal tort filings up ~20% in 2023–24 vs 2021–22; litigation funding extending cases). This is a recurring margin headwind, not a one-time item, even though Q1’26 partly offset it with $31M/$10M of one-time business-interruption proceeds (INTERPRETATION).

Las Vegas demand softness at the low/mid end (thesis-weakening, cyclical). 2025 Strip visitor volume fell ~8%; ADR slipped to $249 (from $260) and RevPAR to $229 (from $245) at 92% occupancy. Canadian visitation is down ~20–28% on trade/geopolitical friction and reduced airline seat capacity; Q1’26 airport traffic still lagged ~5%. Operators are leaning into high-net-worth play (gaming budgets ~$850/visitor) to offset thinner low-end and midweek volumes. A genuine cyclical/discretionary headwind partly masked by mix-shift to whales; if the high end also softens, the Strip has real downside given its fixed rent (INTERPRETATION).

Osaka construction start (neutral-to-negative near-term). Ground-breaking (April 2025) converts Japan from an option into a multi-year cash outflow with cost-inflation risk before any return. Strategically positive, financially a drag until 2030.

Leadership continuity (neutral). Bill Hornbuckle remains CEO and Jonathan Halkyard CFO; the senior team is stable. A CISO search is underway after a departure — a modest governance gap given the 2023 cyberattack history.

Verdict — net thesis-strengthening on capital allocation, thesis-weakening on the operating core. The last two years confirm a disciplined, value-accretive capital-allocation program (high-multiple disposals funding equity shrinkage, avoidance of the low-return Empire City bet, more Macau cash to the parent) — the strongest part of the story. But the operating backdrop deteriorated at the margin: LV leisure/international demand softened, self-insurance/litigation costs became a structural headwind, and the digital and Osaka growth bets still consume cash. On balance the changes reinforce the per-share thesis while modestly weakening the organic-EBITDAR thesis.

9. Risk Analysis

Risk Matrix

1. The rent-encumbered, high-beta OpCo is the central risk — operating leverage cuts both ways. MGM’s asset-light model traded owned real estate for ~$2.26B/yr of fixed triple-net rent escalating ~2% annually regardless of the cycle (FACT — 10-K Note 11: VICI $775M, Mandalay Bay/MGM Grand $322M, Bellagio $276M, Aria/Vdara $233M, Cosmopolitan $212M, plus ground/China leases). Against that fixed obligation sits a cyclical, essentially flat ~$4.6B Adjusted EBITDAR and a 1.28 beta. In an up-cycle incremental revenue drops to the OpCo above a fixed rent line — wonderful. In a down-cycle it is punishing: rent does not fall, so a 10–15% EBITDAR decline (well within casino-cycle history) compresses post-rent free cash flow and the residual equity far more than 10–15%. The RENT structure is the single most important risk lens on this name, and it is why the ~$4.2B ex-lease net debt understates the true fixed-charge burden. The leases also carry covenants that, if breached, force MGM to post six-to-twenty-four months of rent as cash security — a liquidity accelerant in a severe downturn (FACT — Note 11).

2. Las Vegas concentration into softening discretionary demand. The Strip is the profit core and is geographically concentrated on a few square miles. 2025 already showed the vulnerability: visitors −8%, RevPAR down ~7%, Canadian visitation down ~20–28%, Q1’26 air traffic still ~5% light. The industry’s response — leaning on ~$850/visitor high-rollers — concentrates the base into the most cyclically fragile customer segment. If a US consumer slowdown or tariff-driven income shock reaches the high end, the Strip’s flat-to-declining EBITDAR could turn sharply negative — and the rent stays (INTERPRETATION).

3. Macau and BetMGM — the two growth flanks each carry structural fragility. Macau contributes ~$1.2B EBITDAR but sits under a 10-year concession the government can redeem from the eighth year with notice; it is exposed to Beijing policy, capital controls, and visitation swings. In digital, BetMGM is a durable #3 but structurally behind the duopoly; its defensible position is iGaming. Neither flank is a solvency risk, but both are “priced-for-growth” stories where disappointment removes the optionality embedded above the flat core (INTERPRETATION).

4. Catastrophic / total-loss assessment: LOW — but the equity is a levered claim. A permanent-capital-impairment scenario is unlikely. MGM generates real cash (~$2.5B CFO in 2025), carries modest corporate net debt of ~$4.2B ex-leases, has staggered maturities and an undrawn revolver, and holds monetizable assets (MGM China stake, Osaka JV, remaining owned real estate, BetMGM interest). The 2020 COVID stress — negative EBITDA and a ~−98% lifetime-scale drawdown — was survived without equity wipeout. The genuine tail risks are (a) the ~$6.0B of shortfall guarantees on the Bellagio/Mandalay Bay/MGM Grand landlord debt, which could crystallize only if collateral values collapsed and lenders exhausted remedies, and (b) the fixed-rent/covenant structure amplifying a severe, prolonged demand shock. Total loss is remote given the cash-flow base and asset coverage; but because the equity sits behind ~$2.26B/yr of fixed rent and ~$4.2B of net debt, a moderate downturn produces an outsized equity drawdown — the risk is severe volatility and impairment of the levered residual, not zero (INTERPRETATION).

10. Valuation Discussion

The central fact governing MGM’s valuation is structural, not numerical: MGM’s equity is a thin, highly-levered residual claim on a rent-encumbered operating company. The single most-quoted enterprise value — EV ~$42B at the current price (equity ~$12.2B + ~$25.3B capitalized leases + ~$4.2B corporate net debt + ~$0.85B minority) — is ~60% capitalized lease. Every valuation lens below is an attempt to see through that distortion. No price target and no recommendation appear in this section; this is embedded-expectations analysis only.

Why the headline GAAP multiples are false signals. MGM’s AZI own-history percentile ranks split three ways, and two are noise:

• P/E in the ~89th percentile — ignore. FY25 GAAP net income to common was just $206M against ~$2.5B of operating cash flow — the gap is impairment, minority-interest leakage to MGM China’s 44% float, and a distorted tax line. GAAP earnings are structurally suppressed; the P/E “looks expensive” only because the denominator is nearly fictional.
• P/B in the ~99th percentile — ignore. Years of buybacks (~$9.4B) have ground equity to ~$2.4B attributable, and tangible book is negative. P/B and ROE are meaningless when the book is a near-zero, buyback-hollowed residual.
• P/S in the ~18.8th percentile — the one real tell. On ~$17.5B revenue and a ~$12.2B cap, MGM trades at ~0.7x sales, cheapest-ever territory. Revenue is the only denominator untouched by leverage, buybacks, or accounting artifacts, and it says the market is paying a trough multiple on the top line.

The correct operating lens: EV/EBITDAR ~9x. Because rent is a financing-like fixed charge, the honest comparison capitalizes the leases into EV and adds rent back to EBITDA (EBITDAR). On EV ~$42B against company Adjusted EBITDAR of ~$4.6B, MGM trades at ~9x EV/EBITDAR (~8.9x on the FY-end mark). Against peers this is middling-to-full, not distressed: Caesars is being taken private at a ~6.6–7.5x forward EV/EBITDAR, and Wynn trades at ~11x EV/EBITDA (a premium Macau/luxury comp). (Sources: CDC Gaming, Caesars takeout scenarios; stockanalysis.com WYNN statistics; both accessed 2026-07-03.) On the right operating metric MGM is neither cheap nor dear — it sits between the CZR take-out floor and the Wynn premium. The “cheapest-ever” P/S and the “middling” EV/EBITDAR are reconciled by leverage: the same $4.6B EBITDAR supports a huge lease-plus-debt stack, so a low equity price still buys a fair enterprise multiple (INTERPRETATION).

Equity free-cash-flow yield — the bull’s engine. On the real FCF bridge (Section 6), owner FCF is ~$1.3–1.6B, a ~10–12% gross equity FCF yield on a ~$12.2B cap — though the look-through figure is lower because ~44% of MGM China’s cash flow belongs to the public minority. The per-share math is what matters: MGM repurchased ~$1.2–1.4B of stock annually, retiring ~5–8% of the float per year at trough multiples. At a ~0.7x sales / ~10–12% FCF yield, each dollar of buyback is powerfully accretive to per-share value — the mechanical core of the bull case.

Sum-of-the-parts — what the domestic stub implies. MGM is a holding company, and its cleanest valuation is a break-up:

Backing out MGM China (~$2.8B) and a conservative BetMGM mark (~$2.5B) from the $12.2B equity leaves roughly $6.9B of equity for the domestic OpCo, which — carrying ~$4.2B of net debt — implies a domestic enterprise value of ~$11–12B before leases. Against domestic post-rent EBITDA of roughly $1.25B (total $2.36B less MGM China’s ~$1.1B) — or ~$3.4B domestic EBITDAR once the ~$2.2B rent is added back against ~$25.3B capitalized leases — the domestic OpCo is valued at a full-to-rich ~10x post-lease EV/EBITDAR, richer than CZR’s ~7.5x. Read one of two ways: either the domestic Vegas/regional business is priced optimistically at cycle-late, or, equivalently, the market is handing you MGM China at market, BetMGM near cost, and Osaka/UAE for free, and the “expensive stub” is the price of the embedded options. The bull and bear divide precisely on which framing is right (INTERPRETATION).

Scenario analysis (embedded expectations). Explicit assumptions, no price target:

The scenarios illustrate the defining feature: because ~$29.5B of leases-plus-debt sits ahead of the equity, small swings in EBITDAR or the multiple produce enormous swings in the thin equity residual. A 1x multiple compression on $4.6B EBITDAR is ~$4.6B of enterprise value — roughly 38% of the entire equity.

What must be true at $47.10. The market is underwriting, in effect: (1) domestic Adjusted EBITDAR holding near ~$4.6B — the Las Vegas consumer and convention calendar do not roll over at cycle-late; (2) MGM China defending ~$1.2B EBITDAR against Macau competitive and macro risk; (3) BetMGM converting its ~$500M-EBITDA-by-2027 target rather than remaining a cash-burning #3; and (4) management continuing to retire ~5–8% of the equity annually at these multiples. What the market is correctly pricing: MGM’s real cash generation and buyback velocity (the low P/S). What it is arguably mis-pricing in either direction: the embedded China + BetMGM + Osaka optionality, which the stub either discounts to near-zero (bull) or treats as fairly valued against an over-rich domestic multiple (bear). The equity is a leveraged bet that the cycle holds long enough for the buyback to compound — not a value stock in any margin-of-safety sense.

11. Variant Perception

Consensus. The market treats MGM as a de-rated, high-beta cyclical whose best days are behind the current cycle: Las Vegas growth has decelerated from its 2023 peak, Macau carries perennial geopolitical uncertainty, and BetMGM is a sub-scale, distant #3 that has repeatedly disappointed (FY26 JV revenue guidance was cut and Q1 EBITDA missed by ~68% — Gaming America, 2026). That composite view is exactly why P/S sits at the ~18.8th percentile. Consensus is not that MGM is a bad business; it is that MGM is a rent-levered, cycle-late business whose equity is a leveraged coin-flip on the US consumer, deserving a trough multiple.

Strongest bull case. MGM is an asset-light cash machine using a trough multiple as a feature, not a bug. It generates ~$2.5B CFO and ~$1.3–1.6B real FCF, and has spent it retiring ~5–8% of its shares every year for four years (~47% cumulative) at ~0.7x sales — the single most accretive lever available at these prices. Management is simultaneously monetizing assets above the multiple the stub implies (Northfield 6.6x, Mirage and Gold Strike at higher multiples), with the CFO explicitly flagging the arbitrage. Layer in three embedded options the stub barely credits — a MGM China stake that could re-rate on Macau strength, a BetMGM that is now profitable and distributing cash (targeting ~$500M EBITDA by 2027), and a free Osaka call (sole Japan licensee, opens 2030) — and the bull argues you are buying a domestic OpCo at a fair multiple and getting the optionality for a modest discount while management shrinks the equity underneath you.

Strongest bear case. MGM equity is a thin, rent-encumbered residual with a 1.28 beta — a leveraged bet on the Las Vegas consumer at cycle-late, dressed up as a buyback story. The ~$25.3B of capitalized leases plus ~$4.2B net corporate debt sit ahead of a ~$12.2B equity, so a modest EBITDAR miss or 1x multiple compression can vaporize a third of the equity, and the buybacks that lever returns up lever losses down. Macau/China is a structural risk the parent cannot control and only owns 56% of. And BetMGM — the entire reason the stock ever re-rated as a digital-optionality name — is a perennial #3 in a FanDuel/DraftKings duopoly that has proven challengers cannot buy their way in; it keeps consuming the option value rather than realizing it (the Q1’26 guide-cut is the warning). The bear says you are paying a full domestic multiple for a rented-real-estate OpCo at the top of the cycle, with leverage and the China minority as accelerants when it turns.

The 3–5 assumptions that matter most, and what falsifies each:

1. Las Vegas EBITDAR holds (~$4.6B group run-rate). Falsified for the bull by two consecutive quarters of declining Strip EBITDAR or convention softening; for the bear by sustained ADR/occupancy growth through 2026–27.
2. The buyback keeps compounding at a trough multiple. Falsified by a leverage-driven pause or an FCF shortfall (Osaka/Macau capex crowding out repurchases).
3. BetMGM inflects rather than bleeds. Falsified for the bull by another guidance cut; for the bear if BetMGM sustainably distributes cash and iGaming share climbs.
4. MGM China defends ~$1.2B EBITDAR. Falsified for the bull by Macau GGR/share erosion or a policy shock; for the bear by continued share gains.
5. The embedded options (China re-rate, Osaka, UAE) get credited. Falsified for the bull if the stub stays discounted indefinitely; for the bear if a monetization event forces the market to mark them.

Factor-positioning read (input, not a call). MGM carries a market beta of ~1.28 (Base model, R² ~0.37) — a high-beta consumer cyclical, so much of the move is market-directional, not idiosyncratic (FACT — FactorsToday). The leaderboard shows a stock that was dead money for five years (y5 +1.4%/yr, lifetime total return ~0, lifetime max drawdown ~−98% at the COVID low) that has ripped on momentum in the last two-to-three quarters (m3 ≈ +28% actual quarter, Sharpe 3.68; m6 ≈ +29% actual; y1 +25%; m3 de-annualized per the model’s convention). Its factor-nearest peers are CZR (0.90) and RRR (0.88) — other levered domestic-gaming names — then hotel REITs, confirming the “levered Vegas-consumer” signature. The read: this is not a crowded, extended momentum trade in the classic sense — it is a dead-money cyclical that has only just recovered to the top of a five-year range it has failed to break for half a decade. The tape supports the bull’s “re-rating with room” framing more than the bear’s “blow-off top,” but the 1.28 beta and the round-trip history are the bear’s evidence that the near-term momentum is macro-borrowed and mean-reverting. Consensus may be offsides in treating a genuine buyback-plus-monetization compounder as a pure beta vehicle — but the factor read is equally a reminder that if the market turns, a 1.28-beta thin equity leads the decline.

12. Fact vs. Interpretation Table

13. Open Questions

1. What is MGM’s actual economic interest in the Osaka JV — 50% (per the 10-K) or ~43.5–44% (per 2026 trade press after smaller investors joined)? It changes the ~$2.8B remaining equity commitment and the eventual profit share.
2. How much of the ~$1.07B annual capex is true maintenance vs. Macau/Strip growth reinvestment? This determines whether real “maintenance FCF” is closer to $1.6B or $1.2B, and thus the true owner yield.
3. Will the self-insurance/litigation cost step-up prove structural or one-time? Management calls it unusual; external tort-filing data suggest a new normal. ~$46M/quarter matters against flat EBITDAR.
4. Does BetMGM’s Q1’26 guide-cut signal a stall or a timing issue? The $500M-2027-EBITDA target is the swing factor for the digital option’s value.
5. Will Osaka/UAE capital calls force a buyback pause before 2030? The plan funds both from the same cash; the tension is unresolved.
6. What is the realizable private-market value of the remaining owned real estate and the MGM China stake if management continued the monetization playbook to its conclusion?

14. What Must Be True

Bull case — what must be true, and its falsification test. The bull needs: (1) domestic Adjusted EBITDAR to hold or grow near ~$4.6B through the cycle — i.e., the Las Vegas consumer and convention calendar do not roll over; (2) the buyback to keep retiring ~5–8% of the float annually at a low multiple, funded by FCF plus continued high-multiple asset sales; (3) BetMGM to convert to a durable ~$500M-EBITDA contributor; and (4) the market to eventually credit the China/BetMGM/Osaka optionality the stub discounts. Falsification test: two consecutive quarters of declining Las Vegas Strip Adjusted EBITDAR, or a buyback pause driven by leverage/capital calls. Either breaks the per-share compounding engine that is the entire bull thesis.

Bear case — what must be true, and its falsification test. The bear needs: (1) the Las Vegas consumer to soften materially at the high end (not just the low/midweek end), turning flat Strip EBITDAR negative against fixed rent; (2) the leverage-on-leverage structure to convert that operating softness into an outsized equity drawdown at a 1.28 beta; and/or (3) a Macau/China shock or a BetMGM stall to remove the embedded optionality. Falsification test: sustained ADR/occupancy and convention-mix growth plus a BetMGM print confirming the $500M-2027-EBITDA path — which together would show the cycle is not rolling over and the digital option is real, dismantling the bear’s two load-bearing legs.

The two cases are unusually balanced because they rest on the same structural fact — the rent-levered, high-beta equity — read with opposite signs: the bull sees leverage amplifying an accretive buyback on a resilient cash stream; the bear sees leverage amplifying a cyclical downturn on a rent-encumbered residual. The evidence that resolves it is the trajectory of Las Vegas Strip EBITDAR and the buyback’s continuity — both observable quarter by quarter.

Section 15 (Source Appendix) is maintained as a separate deliverable and appended to the combined report.

APPENDIX A — Standard Diligence Questionnaire

MGM Resorts International (NYSE: MGM) · Report date: 2026-07-03 · Price: $47.10

Answers are grounded in the FY2025 10-K (filed 2026-02-11), the Q1’26 earnings call (2026-04-29), the 2026 DEF 14A, and third-party data cited in the memo. Labels: F = Fact, I = Interpretation, A = Assumption.

General

What thoughtful questions have other investors asked about this company? The recurring institutional debates: (1) Is the asset-light sale-leaseback a value-creating master-stroke or did MGM permanently surrender its scarcest asset for a fixed rent claim? (2) What is real free cash flow after growth capex and the MGM China minority — is the equity FCF yield ~10% or closer to ~6% on a look-through basis? (3) Is BetMGM a durable top-3 franchise or a perennial money-loser that will keep consuming the digital option? (4) Does the ~47% share-count reduction make MGM a per-share compounder, or does the leverage it required make the equity a leveraged beta vehicle? (5) How should one value a stub where GAAP earnings and book value are both broken? These map to Section 6, Section 7, Section 10, and Section 11.

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? (I) Roughly mid-cycle, tilted toward a high: Las Vegas visitation and RevPAR are off their 2023–24 peak (visitors −8% in 2025), while Macau is still ~85% of 2019 and BetMGM is early in its profit ramp. Consolidated Adjusted EBITDA has been flat (~$2.4B) for three years — a plateau, not a trough.

Driven by the external environment or internal actions? (I) Both. The flat EBITDAR is external (mature Vegas demand, Macau recovery ceiling); the per-share earnings growth is entirely internal (buybacks, asset sales, brand-fee capture).

How stable are revenues? (F/I) Convention/group and Regional revenue are stable and recurring; Strip leisure and Macau casino win are cyclical. The ~$2.26B rent is a fixed cost regardless — so post-rent earnings are less stable than revenue.

Outlook for products/services? (I) Modest organic growth: Vegas flat-to-up low-single-digits, Regional flat, MGM China mid-single-digits, digital growing off a small/negative base. Osaka (2030) is the one step-change, four-plus years out.

How big will this market be — growing, shrinking, domestic or international? (F) US commercial gaming set a record $78.7B in 2025 (+9.2%), but land-based grew only ~2%; the growth is online and international (Macau ~$30.8B, Japan a new market). MGM’s mix is ~70% domestic resort, ~25% Macau, growing international-digital.

Business Quality & Competitive Moat

Is the industry getting more or less competitive? (I) The Strip and Macau are stable licensed oligopolies (supply-constrained, favorable); US online is a brutal, consolidating duopoly-plus-one where MGM is on the weak side.

How profitable is the business (ROIC, ROE)? (I) ROE/ROIC/P/B are uninterpretable here — tangible book is negative and equity is a buyback-hollowed residual. The right lens is cash-on-cash: ~$1.3–1.6B owner FCF on a ~$12.2B cap (~10–12% yield), a high-single-digit unlevered cash return on the enterprise. Decent, not exceptional.

How profitable is the industry — how many competitors, what barriers to entry? (F) Barriers are formidable (gaming licenses, Strip land scarcity, Macau’s six-concession cap to 2032). Strip EBITDAR margins ~34%; Macau ~27%; Regional ~31%.

Can the business be easily understood? (I) The operations yes; the equity no — the lease capitalization, minority consolidation, and impairment noise make the reported financials a trap for the casual reader.

Can it be undermined by foreign low-cost labor? (F) No — physical, location-bound resort/gaming assets. Online is technology/marketing-driven, not labor-arbitrage-exposed.

Do brands matter? (F/I) Yes — Bellagio, Aria, MGM Grand, Mandalay Bay are genuine intangible assets and part of the Strip moat; the MGM Rewards loyalty database drives cross-property captivity.

What is the nature of competition? (I) Scale/convention footprint and loyalty on the Strip; premium-mass execution in Macau; marketing spend and product in online.

Customers’ switching costs? (I) Meaningful for convention/group (multi-year contracts, loyalty) and premium play; negligible in online (multi-homing).

Financial Condition & Balance Sheet

Assets not fully recognized on the balance sheet? (I) The BetMGM 50% JV (carried below zero), the Osaka JV, the residual value of remaining owned real estate, and brand value — all under-recognized. Offsetting: the ~$25.3B lease ROU asset/liability is fully on-balance-sheet.

Off-balance-sheet liabilities? (F) ~$6.0B of shortfall guarantees on Bellagio and Mandalay Bay/MGM Grand landlord debt; ~$2.8B remaining Osaka equity commitment.

How conservative is the accounting? (I) Mixed — SBC is genuinely low ($90.5M, ~0.5% of revenue, a positive), but GAAP net income is heavily distorted by impairments, property-transaction gains/losses, and minority consolidation. Cash flow is clean; the P&L is not.

How CapEx-hungry is the business? (F/I) Moderate: ~$1.07B/yr capex (~6% of revenue) for the resorts, but plus ~$2.8B of Osaka equity through 2030 — the growth-capital call is heavier than the maintenance run-rate.

Capital Allocation & Management

How much FCF does the business generate, how does management use it, what is the philosophy? (F) ~$1.3–1.6B real owner FCF, plus large asset-sale proceeds. Philosophy: monetize assets at high multiples and buy back stock at a low multiple. ~$9.4B repurchased 2021–25.

Significant acquisitions recently? (F) The Cosmopolitan operations (2022), LeoVegas (2022), Tipico US tech. Net a seller of casinos, not a buyer.

Buying back shares? (F) Aggressively — ~47% of the count in five years; $2.0B authorization (April 2025); reaccelerating post-Northfield.

Issuing large amounts of new shares to insiders? (F) No — SBC is only $90.5M/yr; dilution is negligible.

Compensation policy of directors/management? (F/I) CEO FY2025 comp $25.3M; bonus 75% Adjusted EBITDAR; LTI 50% relative-TSR PSU / 50% RSU; say-on-pay 89.3%. Reasonable but lacks a per-share/FCF/ROIC/leverage gate — a gap for a per-share-compounding thesis.

Motivations of management? (I) Share-price-aligned via TSR/RSUs and consistent with the buyback agenda, but no personal open-market buying to corroborate the “cheap stock” argument; activist directors (Corvex, IAC) reinforce the capital-return bent.

Valuation & Market Data

Is the stock an ADR, MLP, or K-1 issuer? (F) No — a US C-corp common stock on the NYSE; standard 1099 treatment.

Dividend policy? (F) Suspended entirely since February 2023 (was a token $0.01). All capital return is via buyback.

How profitable is the business? (I) Cash-generative and stable; GAAP-unprofitable-looking due to the distortions above — see Section 6.

Is net income diverging from cash from operations? (F) Sharply — $206M GAAP NI vs. ~$2.5B CFO in 2025. The divergence is structural (minority, impairment, non-cash items), and cash is the reliable figure.

Risks & Downside

What factors would cause the stock to decline? (I) A Las Vegas consumer/high-end downturn against fixed rent; a Macau/China shock; a BetMGM stall; a buyback pause from Osaka/UAE capital calls; multiple compression on a 1.28-beta name in a market drawdown.

Risk of a catastrophic loss? (I) Low-to-moderate. Real cash generation, modest corporate net debt (~$4.2B ex-lease), staggered maturities, and monetizable assets protect solvency; the tail risks are the ~$6.0B landlord-debt guarantees and the fixed-rent covenant structure in a severe, prolonged downturn.

Chance of a total loss? (I) Remote. The COVID stress (negative EBITDA) was survived without equity wipeout. But the equity is a levered residual — the realistic downside is a large drawdown, not zero.

Recent News & Events

Has the business environment changed recently? (F) Yes at the margin: Vegas top-line turned positive in Q1’26 for the first time in six quarters; Macau printed its first YoY GGR decline of the year in June 2026; BetMGM cut FY26 guidance; self-insurance/litigation costs stepped up.

Significant acquisitions? (F) None recently on the buy side; the notable transaction is the sale of Northfield Park ($546M, ~6.6x, closed April 2026).

Change in accounting policies? (F) None material; the 2025 optics reflect the Empire City impairment, not a policy change.

Recent changes — new markets, facilities, management? (F) Osaka construction underway (2030 opening); UAE Dubai project underway; MGM China brand fee raised to 3.5%; MGM Grand room remodel completed; management team stable (CISO search underway).

APPENDIX B — Source Appendix

MGM Resorts International (NYSE: MGM) · Report date: 2026-07-03

All non-obvious facts in this memo trace to the sources below. Primary (filings) prioritized over secondary; third-party aggregated/statistical data (ROIC.ai, FactorsToday, AZI) treated as a starting point and reconciled to filings.

Primary — SEC filings (EDGAR, CIK 0000789570)

• MGM Resorts International Form 10-K, FY2025 (filed 2026-02-11; mgm-20251231). Segment revenue & Adjusted EBITDAR (Note 17); triple-net lease terms, rent, and capitalized lease liabilities (Note 11); debt & maturities (Note 9); BetMGM/unconsolidated affiliates (Note 9); Osaka commitment (commitments footnote); Empire City impairment (MD&A); cybersecurity; risk factors; share repurchases (Item 5).
• MGM Resorts Forms 10-K, FY2021–FY2024 (mgm-20211231 … mgm-20241231) — multi-year trend for revenue, EBITDAR, buybacks, rent, disposals (Mirage 2022, Gold Strike Tunica 2023).
• MGM Resorts Forms 10-Q (2021–2026) — quarterly segment and cash-flow detail.
• MGM Resorts Form 8-K filings 2021–2026 — Northfield Park sale, financings/authorizations, quarterly earnings releases, guidance.
• MGM Resorts DEF 14A (proxy), 2026 (filed 2026-03-27) — executive compensation metrics, CEO/CFO pay, say-on-pay, incentive design.
• MGM Resorts Form 4 corpus, 2025–2026 (EDGAR) — insider transaction read: no code-P open-market purchases; director sales; Corvex/Meister and IAC holdings; CEO ownership.

Primary — company disclosures

• MGM Resorts Q1 2026 earnings call transcript (2026-04-29; via ROIC.ai) — Las Vegas top-line inflection, self-insurance/BI items, MGM China brand fee, BetMGM/MGM Digital, Osaka funding, Northfield multiple, buyback commentary. Treated as management hypothesis, cross-checked against filings.
• MGM Resorts investor relations (investors.mgmresorts.com) — investor presentations, press releases.

Quantitative aggregators (reconciled to filings)

• ROIC.ai MCP — income statement, balance sheet, cash flow, enterprise value, valuation multiples, profitability ratios, company profile, earnings-call transcripts (FY2020–FY2025).
• Public market-data feeds — 5-year adjusted-close price history (price arc, 52-week range, beta) and own-history valuation percentile ranks (P/E 89.6th, P/B 99.3rd, P/S 18.8th, composite 69th).
• FactorsToday factor model — stock loadings (market beta ~1.28), risk-adjusted leaderboard (m3/m6/y1/y5/lifetime returns, drawdowns, Sharpe), related-stocks (CZR 0.90, RRR 0.88).

Industry & market data (public, cited inline with access date 2026-07-03)

• Las Vegas Convention and Visitors Authority (LVCVA) — 2025 visitation (~38.5M, −8%), convention attendance (~6M), ADR/RevPAR; via reviewjournal.com and casino.org.
• American Gaming Association, State of the States 2026 — US commercial gaming GGR ($78.7B, +9.2%), land-based vs. iGaming split.
• GGRAsia / GGB Magazine — Macau 2025 GGR (~$30.8B, +9.1%), June 2026 YoY decline, FY26 forecast, six-concession structure; MGM Osaka cost (~¥1.51T / ~$10.2B) and stake reporting.
• asgam / Tribuna — MGM China market-share history (9.5% 2019 → 15.8% 2024 → ~16–17% 2026).
• casino.org / bonus.com / sportbotai / covers.com / Gaming America — US OSB and iGaming market share (FanDuel/DraftKings duopoly, BetMGM #3); BetMGM 2025 financials and FY26 guidance cut.
• CDC Gaming; stockanalysis.com — Caesars take-out multiple (~6.6–7.5x forward EV/EBITDAR); Wynn (~11x EV/EBITDA).
• Yahoo Finance (2282.HK) — MGM China market capitalization for the sum-of-the-parts.
• InsuranceBusiness / Institute for Legal Reform — federal tort-filing trends (litigation cost context).

Peer & landlord context

• DraftKings (DKNG) public filings — US online sports betting / iGaming competitive structure and BetMGM positioning.
• VICI Properties (VICI) public filings — MGM’s landlord; triple-net lease economics and gaming REIT context.

Frameworks

• Competition Demystified (Greenwald & Kahn) — moat taxonomy, share-stability and ROIC tests.
• Capital Returns (Marathon Asset Management) — supply-side capital-cycle and asset-monetization lens.