MGM has aggressively shrunk the share count, but the underlying business is a cyclical, capital-intensive casino operator now buried under triple-net rent obligations and a 50% stake in a still-unprofitable digital venture. At $38.50 versus a base IV of $19.51, the market is paying nearly 2x intrinsic value for hope.

Plain English

MGM owns famous Las Vegas hotels like Bellagio and MGM Grand, two casinos in Macau, regional casinos, and a slice of an online sports-betting business called BetMGM. The catch: a few years ago they sold the buildings under their Vegas casinos and now rent them back forever. That makes their balance sheet look better than it really is, because they still owe billions in rent. The stock is $38.50 but a careful estimate of what it is worth is about $19.50. So you would be paying nearly twice fair value to own a cyclical business. Better to wait for a much lower price.

Thesis

MGM Resorts is a U.S./Macau casino operator that has executed an 'asset-light' transformation: it sold the underlying real estate of Bellagio, MGM Grand/Mandalay Bay, Aria, and The Cosmopolitan to REITs (VICI, Blackstone) and now leases those buildings back under long-dated triple-net leases. The proceeds funded an enormous buyback program — share count is down 6.6% over the last decade per the scorecard, with the cadence accelerating ($494M in Q1 2025 alone, plus $1.5B remaining authorization at March 2026). Owner-earnings TTM stand at roughly $564M.

The pitch from bulls: the company controls premier irreplaceable real estate on the Las Vegas Strip, owns 56% of MGM China, holds 50% of BetMGM (the U.S. #2/3 in online sports betting and iGaming), is building a generational integrated resort in Osaka, and is shrinking the float at a low double-digit FCF yield. Composite score is 62, capital allocation is 20, and the business throws off cash (5-yr FCF conversion of 1.32).

The problem is price. The deterministic scorer pegs base IV at $19.51 and high IV at $28.13. At $38.50 the stock trades at a P/IV ratio of 1.97 — nearly double base intrinsic value, and above even the optimistic case. The reverse-DCF implies only 3.9% growth is needed to justify today's price, but the scorer was forced to clamp the trailing CAGR from -15.3% to -5.0%, which means MGM is currently shrinking owner-earnings, not growing them. ROIC averages just 8.96% over a decade — below most reasonable hurdle rates and well below the 15%+ Buffett wants from a compounder. With base IV roughly half the market price, no margin of safety exists and even bull-case IV ($28.13) is 27% below the current quote. Owning MGM at this price is a bet on multiple expansion in a cyclical industry, not on business value compounding.

Moat

Casinos look moaty from a distance — premier Strip real estate, regulated licenses, brand recognition, loyalty databases — and crumble when you press on each pillar. I will work through the five canonical moat sources.

1. Brand / intangibles. MGM owns Bellagio, Aria, MGM Grand, Mandalay Bay, Luxor, Park MGM, New York-New York, Excalibur and The Cosmopolitan; the MGM Rewards loyalty program crosses gaming and non-gaming spend; LeoVegas adds a European digital brand; BetMGM is a top-three U.S. sportsbook brand. These are real assets, but they are not Coca-Cola-grade. Damodaran's framing is useful: a brand creates value only when it commands premium pricing or share that competitors cannot replicate [1]. In Las Vegas, customers comparison-shop hotel rates across the Strip in a single Expedia query; Caesars, Wynn, Venetian/Sands successor (Apollo), Resorts World, Fontainebleau, and a future Hard Rock all compete for the same convention nights and weekend leisure dollars. Buffett's See's Candy test asks whether you would walk an extra block, or pay more, for the brand [3]; for a typical Vegas guest, the answer at MGM Grand vs. Caesars Palace is no.

2. Switching costs. Loyalty programs create modest stickiness — accumulated tier status, comped rooms, MGM Rewards status matches — but a determined gambler simply opens a Caesars Diamond account or signs up for Resorts World's program. In digital, BetMGM faces near-zero switching costs: bettors routinely arbitrage promos across DraftKings, FanDuel, BetMGM and Fanatics. Switching cost moat: weak.

3. Network effects. None in the gaming sense. Online sports betting has a mild liquidity-pool benefit at scale (better lines, deeper markets), but that favors the #1 (FanDuel) and #2 (DraftKings), not BetMGM, which has slipped to #3 in U.S. iGaming and lower in OSB.

4. Cost advantages. This is where the asset-light transformation actively destroyed the moat. By selling the real estate of Bellagio, MGM Grand, Mandalay Bay, Aria, and The Cosmopolitan to VICI and Blackstone REIT and leasing them back under triple-net leases, MGM converted owned hard assets into very long-dated rent expense — escalators, CPI ratchets, and decades of inflexibility. In a downturn, owners of real estate can defer maintenance and ride it out; tenants must pay rent regardless. Buffett's framing of BNSF and MidAmerican is the opposite of MGM: he prizes 'huge investment in very long-lived, regulated assets' that the owner holds, with 'earning power that even under terrible business conditions amply covers their interest requirements' [4]. MGM now has the leverage profile of a regulated utility but without the regulated return — the worst of both. Net debt/EBITDA of 1.80x looks fine, but that figure excludes ~$25B+ of operating-lease liabilities (Bellagio lease and the VICI/BREIT master leases). Adjusted for capitalized rent, true financial leverage is materially higher.

5. Efficient scale / regulation. Macau is genuinely scarce: only six concessions, MGM China holds one of them through 2032. Las Vegas Strip licenses are also de facto scarce. This is the strongest moat element — but it is shared with five Macau peers and a half-dozen Strip operators, so it bounds the industry's profit pool rather than awarding MGM a unique slice. Buffett would note that 'a moat that must be continuously rebuilt will eventually be no moat at all' [6], and Macau concessions are renegotiated periodically with concession fees, market-share allocations, and gaming-tax adjustments at the regulator's discretion.

The ROIC numbers tell the story: 10-year average ROIC of 8.96% is below most credible cost-of-capital estimates for a leveraged, cyclical, lease-heavy business. A true wide-moat business should compound capital at 15%+ for decades; MGM does not.

Moat verdict: NARROW

Management

CEO Bill Hornbuckle, CFO Jonathan Halkyard, and Chairman Paul Salem have pursued a coherent — if highly debatable — capital allocation playbook. I will grade each of the five capital-allocation choices.

1. Reinvestment in the existing business. MGM has reinvested heavily in property refreshes (MGM Grand, Mandalay Bay rooms, Aria spa tower, The Cosmopolitan integration after the 2022 acquisition) and in BetMGM (cumulative funding contributions in the billions across the JV). The Cosmopolitan deal closed at roughly $5.65B (real estate + opco), and the operating company piece has been productive. Reinvestment quality is mixed: domestic resort capex is largely defensive maintenance in an industry where the consumer demands constant freshness, and BetMGM has yet to post sustained equity-method profits despite years of investment.

2. Acquisitions. Cosmopolitan (2022) is largely a win — the property fits the portfolio and management has driven margins higher. LeoVegas (2022, ~$607M) gave MGM a global online platform and was modest in size relative to the balance sheet. Push.

3. Debt management. Net debt/EBITDA of 1.80x looks conservative on the surface and is far better than the pre-asset-light leverage profile. Credit-rating trajectory has improved. The catch — and it is enormous — is the operating-lease liabilities created by the sale-leasebacks. Including the VICI master lease, the Bellagio BREIT lease, the Mandalay Bay/MGM Grand lease, and the Cosmopolitan lease, MGM owes tens of billions of dollars of contractually fixed rent. Counted as debt-equivalent (which it economically is), leverage is materially higher than the headline ratio.

4. Buybacks. This is where management has been most aggressive — and most expensive in retrospect. Two consecutive $2B authorizations (Nov 2023 and April 2025), with $1.5B remaining at March 2026. Q1 2025 alone: 15M shares for $494M (roughly $33/share average). Q1 2026: 2M shares for $90M (roughly $45/share average). Share count is down 6.6% over a decade and the recent pace has been faster. The Buffett-Munger test is whether buybacks are accretive at the price paid: MGM has been repurchasing at $33-$45 against a deterministic IV of $19.51 (high IV $28.13). Even granting that the scorer's clamp on shrinking owner-earnings is conservative, a buyback program executed at roughly 1.5-2.0x intrinsic value destroys value rather than creating it. Capital that could have de-leased the balance sheet, or accumulated for the cyclical bottom, has instead been spent above fair value because the alternative — building cash for the next downturn — is unfashionable.

5. Dividends. A token quarterly dividend of $0.0025 per quarter — essentially symbolic. Management has chosen buybacks as the return-of-capital vehicle, which is at least tax-efficient if executed well; they have not.

Communication. Filings are clear and segment disclosure is reasonable. Forward guidance on Macau and BetMGM has been measured. The Osaka project is honestly characterized as a long-dated capital commitment with peer JV participation (April 2026 ownership decreased from prior level) — sensible risk-sharing on a multibillion-dollar build.

Incentives. Compensation is heavy on equity, but a meaningful slug is RSU-based with TSR overlays — better than pure options, but does not punish overpaying for one's own stock. Insider ownership is modest in absolute terms; this is a hired-hand management team, not a founder operator.

The summary verdict: management is competent and strategic, but is buying back stock at roughly twice intrinsic value while the company carries hidden lease leverage and a stake in a still-unprofitable digital venture. That is not the discipline of a great capital allocator.

Capital allocator: C

Industry

The U.S. casino-resort and global gaming industry is structurally average at best. I'll work through Porter's Five Forces.

1. Rivalry — HIGH. On the Las Vegas Strip, MGM competes head-to-head with Caesars Entertainment, Wynn Resorts, Sands successor (now Apollo), Resorts World (Genting), Fontainebleau, Treasure Island, and the Venetian/Palazzo, plus the looming Hard Rock conversion of the old Mirage site. New entrants are constant, room counts grow, and convention space is being added. In Macau, six concessionaires share a market that is recovering but well below 2019 GGR. In U.S. online sports betting and iGaming, BetMGM faces FanDuel (Flutter) and DraftKings as entrenched leaders, plus aggressive promotional spending from Fanatics, ESPN BET (Penn), and Caesars. Promotional intensity in OSB has compressed margins industry-wide.

2. Threat of new entrants — MODERATE. Brick-and-mortar casinos require gaming licenses, which are scarce — Macau concessions are capped at six; Strip resort licenses require Nevada Gaming Control Board approval; new states open one or two licenses at a time. So entry into existing markets is hard, but every new state legalization (NY, MA, OH, PA, IL) creates new entrants by definition. Online sports betting has lower entry barriers — software-and-license — and the market has seen Fanatics enter in the last two years.

3. Bargaining power of customers — HIGH and rising. Las Vegas guests use OTAs to compare hundreds of room offers in one query; promotional sportsbook bonuses train bettors to switch among BetMGM, FanDuel, DraftKings each season; convention buyers run RFPs across Caesars and MGM. The $50 resort-fee era has made customers more price-sensitive, not less. Macau VIP business has been hollowed out by the junket crackdown, leaving mass-market players who are also price-comparison shoppers.

4. Bargaining power of suppliers — MODERATE. Labor is a major and rising input cost; the Culinary Workers Union won material wage increases in the 2024 Strip contract. Slot-machine and table-game suppliers (IGT, Aristocrat, Light & Wonder) have pricing power. Most importantly, MGM's REIT landlords (VICI and BREIT) are now massive 'suppliers' of real estate with contractually escalating rents; MGM has very little power in those negotiations because the leases are ~30 years with extension options and triple-net economics.

5. Threat of substitutes — HIGH. Streaming entertainment, sports streaming, online gaming, esports, and experience travel all compete for the consumer's discretionary dollar. iGaming partly cannibalizes brick-and-mortar gaming. Cruise lines and Cancun-style resorts compete with Las Vegas for leisure trips. International tourism shifts (Asia traveler patterns, U.S. dollar strength) reshape demand quarterly.

Layered on top: the industry is highly cyclical (gaming revenue fell 35-50% in 2008-09 and 70%+ during COVID), highly regulated (gaming taxes, problem-gambling rules, AML compliance), and increasingly subject to political risk in Macau. Capital intensity is enormous — even after asset-light transformation, MGM spends hundreds of millions a year on property maintenance and refresh capex.

The 10-year-average ROIC of 8.96% reported by the scorer is the industry's verdict: even a top-three operator earns barely above the cost of capital across a full cycle. This is what Buffett means when he says some industries are simply harder than others — being the best operator in a structurally average industry is not a path to compounding wealth.

Industry Verdict: Average

Inversion

I have to make the strongest credible case that MGM is a value trap and the next decade will be worse than the consensus. Here are the five mechanisms by which I lose money owning this stock at $38.50.

1. The thesis is wrong because the consumer cycle turns and the lease structure transforms a fixable problem into a permanent impairment. Las Vegas RevPAR and Strip gaming win are statistically correlated with U.S. consumer discretionary spending, which is in turn correlated with employment and asset prices. From 2008-2010, Strip gaming revenue fell roughly 20% and RevPAR fell 35%; high-end resort EBITDA fell 50%+. In 2020 the drop was even more violent. In both downturns, MGM owned its real estate and could ride it out by deferring maintenance, drawing on real-estate-secured credit, and waiting. After the asset-light transformation, the picture is different: VICI master-lease rent is contractually fixed and escalates with CPI; the Bellagio BREIT lease is similar; the Mandalay Bay/MGM Grand lease is similar; the Cosmopolitan lease is similar. In a 30% revenue drawdown scenario with rent and interest fixed, EBITDA-after-rent could go negative at the property level. The asset-light story sold to Wall Street as 'unlocking capital' is actually a sale of optionality. Equity holders bear the timing risk; the REITs collected the option premium up front.

2. The capital allocation is wrong because management is buying back stock at roughly twice intrinsic value. Per the deterministic scorer, base IV is $19.51; the company has been repurchasing at $33-45. Q1 2025 saw 15M shares retired for $494M ($33 average) and Q1 2026 saw another 2M for $90M ($45 average), with $1.5B authorization remaining. Buying back stock above intrinsic value is the textbook way to destroy shareholder value while making EPS look better. The right move at $40+ would be to accumulate cash for opportunistic deployment in the next downturn (when buybacks at $15-20 would be massively accretive), or to pay down lease equivalents. Management's choice to keep buying signals either (a) they believe IV is much higher than the screener says — a possible but unsupported claim — or (b) they are responding to incentive structures that reward EPS growth rather than per-share intrinsic value compounding. Munger would call this 'incentive caused bias' in capital allocation.

3. The moat is wrong because BetMGM's competitive position has deteriorated. When BetMGM was conceived as the JV with Entain in 2018, the bull case was a top-two finish in U.S. OSB and iGaming behind FanDuel. Reality after seven years: FanDuel and DraftKings command roughly 70% combined U.S. OSB share; BetMGM is solidly third and faces fresh competition from Fanatics, ESPN BET, and Caesars. The promised cross-sell synergy — convert a Bellagio guest into a digital bettor — has been real but smaller than anticipated. BetMGM remains in or near losses on an equity-method basis after years of capital injections. Bulls extrapolating BetMGM as a $5-8B equity stake are paying for a future that may not arrive; the market for U.S. OSB is increasingly looking like a duopoly with fringe players, and MGM owns 50% of a fringe player.

4. The bull narrative is extrapolating Macau and Osaka through rose-colored glasses. Macau has recovered from COVID lows but remains well below 2019 peak GGR, and the political reality of a Chinese government that views gaming as a tolerated industry rather than a favored one creates persistent overhang. The 2022 concession renewals included a steeper non-gaming investment commitment and shorter tenure. Osaka is a 10-year, multibillion-dollar bet on a culture that has historically resisted casinos; the gross IRR even in a base case has been characterized as low double-digit, which after currency risk and partner dilution is closer to a coin flip. MGM China is 56% owned, so a meaningful slice of the 'good case' is owed to minority holders; Osaka was 50% owned and decreased further in April 2026 per the 10-Q.

5. Valuation trap and multiple compression. P/E TTM of 16.4x is above the 10-year average of 14.5x. EV/FCF of 13.2x looks reasonable until you remember the EV calculation excludes lease liabilities; on a true-EV basis (debt + lease PV) the multiple is materially higher. The reverse-DCF implies 3.9% perpetual growth, which sounds modest until you see the trailing CAGR was clamped from -15.3% to -5.0% because the actual recent direction is down, not up. P/IV ratio of 1.97 means the market price is essentially double base IV. Even on the high IV ($28.13), the stock is 27% over-valued. There is no margin of safety. In a market drawdown that compresses cyclical multiples, MGM could trade to $20-22 (10-12x trough EBITDA after rent) — roughly a 45-50% drawdown from current levels.

If I am right, the stock could be worth $20 within 3 years.

Lollapalooza Bias Check

I should examine which biases are operating on me as I evaluate MGM, because Buffett-Munger investing is as much about bias-management as it is about business analysis.

Anchoring is active. The deterministic scorer hands me a base IV of $19.51, and once that number is in my head every other piece of analysis defers to it. That number was generated by a clamp ('CAGR clamped from -15.3% to -5.0%') and a reverse-DCF that may not reflect the underlying economics of post-asset-light MGM. I should be epistemically humble that the IV could be materially higher — say $30-35 in a normalized environment — even though I don't have the work to prove it. My recommendation should reflect this uncertainty by being conviction-medium, not conviction-high.

Confirmation bias is active. I came into this analysis with a structural skepticism of casino operators (cyclical, capital-intensive, lease-heavy) and I have arranged the evidence to support that conclusion. The bull case — premier irreplaceable real estate, Macau recovery, BetMGM optionality, Osaka long-tail value, aggressive buybacks of an undervalued (in their view) stock by competent management — is real and I have given it short shrift in the inversion section. A Wynn or Sands shareholder would tell me my framing is uncharitable.

Recency bias is active. The post-COVID Strip recovery has been spectacular, and I am partly extrapolating it as the new normal. If I roll the clock back to March 2020 or December 2008, the same MGM business produced very different cash flows; I should weight those scenarios more heavily than the last 36 months.

Authority bias is active. The scorecard composite is 62 (mediocre), and I am letting that anchor my recommendation. A high-quality analyst with a different model could legitimately produce a composite in the 70s. The scorer is deterministic and consistent — useful properties — but it is not omniscient.

Deprival super-reaction is the bull's bias. People who own MGM at higher prices, or who 'missed' the post-2022 rally, will resist selling because they are anchored to a higher price than the current quote. Symmetric: I should not avoid the stock just because I would feel foolish if it ran to $50.

Incentive-caused bias is active in the management I'm evaluating: equity comp tied to TSR rewards buybacks regardless of price paid; this explains why MGM keeps buying at 1.5-2x IV. I should not let my distaste for that incentive structure spill over into a more negative read of the underlying business than the facts support.

The net effect: my conviction should probably be downgraded to medium, and I should frame the call as 'Avoid at this price; reconsider at $20-22.' That is more honest than a strong-sell at $38.50.

10-Year Outlook

Same fundamental business model in 10 years? Mostly yes. MGM will still operate Las Vegas Strip resorts, Macau properties, regional U.S. casinos, and digital gaming through BetMGM and LeoVegas. The asset-light/leased real estate structure will persist (the master leases run to ~2050+). Osaka will be open by 2030-31. The mix will tilt more digital and more international, but it will still be a casino-and-hospitality operator, not a transformed business.

Customer base larger? Probably yes, modestly. Online sports betting and iGaming will continue legalizing across U.S. states (perhaps 3-5 more by 2035), expanding the addressable market for BetMGM. Macau visitation should normalize and grow with Chinese GDP. Las Vegas will continue benefiting from the structural shift of leisure spending toward experiences. Osaka opens a new market entirely. Customer-base growth: low-to-mid single digits annually.

Profit per customer higher? Uncertain. On the brick-and-mortar side, rising labor costs, REIT rent escalators, and competitive intensity argue against margin expansion. On the digital side, BetMGM's path to sustained profitability remains unproven and the duopoly dynamic favors FanDuel/DraftKings. Profit per customer could easily be flat to down.

Moat wider? No, narrower. The asset-light transformation removed the real-estate ownership moat. New Strip supply (Fontainebleau, Hard Rock conversion, future projects) erodes scarcity. BetMGM's competitive position has weakened, not strengthened. Macau concession terms favor the regulator, not the licensee. The MGM Rewards/loyalty program is competitive but not unique. The only widening element is brand reach via LeoVegas internationally and the Osaka resort.

Single biggest threat? Two-way tie. (a) Consumer recession in the U.S. that exposes the now-fixed cost structure (rent + debt) to operating-leverage pain; a 25%+ EBITDA drawdown is plausible in a real recession. (b) Long-term Chinese policy shift that further constrains Macau gaming — gaming-tax hikes, mass-market caps, or geopolitical decoupling.

Confidence in 10-year prediction: I can describe the business with reasonable confidence, but I cannot put a tight error band on the cash flows because three of four segments depend on macro variables (U.S. consumer cycle, Chinese policy, Japanese launch success) that are individually low-confidence and combine multiplicatively. This is exactly the type of business where Buffett's 'too hard' bin is the honest answer for a concentrated investor — but as an analytical exercise the numbers say overvalued.

CONFIDENCE: medium

Position Guidance

• Recommendation: Avoid
• Conviction: Medium
• Target buy price: $15.00 (roughly 75% of base IV of $19.51, providing a 25% margin of safety)
• Target trim price: $28.00 (just below high-IV of $28.13, where even the optimistic case is fully reflected)
• Position sizing: Zero today. If price reverts to the $15-18 range during a cyclical drawdown and the lease-adjusted balance sheet still passes a stress test, a 1-3% portfolio position would be reasonable. Above $25, no new capital. Existing holders should consider trimming above the $28 trim price.