MGM Resorts Q3 2026: Strip room rates fell 8% and adjusted EBITDA dropped 20% despite lucky table hold

MGM Resorts' third quarter of 2026 (three months ended September 30, 2026) looks like a recovery on the surface: the net loss attributable to MGM shrank to $17 million from $285 million, and diluted loss per share narrowed to $0.07 from $1.05. That improvement comes almost entirely from charges booked a year ago that did not recur. The casinos themselves earned less: Consolidated Adjusted EBITDA (earnings before interest, taxes, depreciation, amortization and one-off items, a rough measure of the cash profit from operations) fell 20% to $403 million, and MGM's own adjusted EPS swung to a $0.11 loss from a $0.24 profit. Revenue slipped 0.9% to $4.21 billion.

Takeaway: The smaller loss is an accounting comparison, not a turnaround. Every operating segment except digital earned less than a year ago, and the two biggest (the Strip and MGM China) only held up as well as they did because the casinos won an unusually high share of table bets.

Source: MGM's Form 10-Q for the quarter, filed October 28, 2026, the same day as the earnings release (Form 8-K, Exhibit 99.1), which supplies adjusted EPS and the BetMGM distribution figure.

Key metrics

"n/m" means not meaningful: a percentage change between two losses, or from a loss to a profit, doesn't describe anything useful. RevPAR is revenue per available room, which combines how full the hotels are with what each room sells for.

Segment by segment

Las Vegas Strip: cheaper rooms, higher costs, lucky tables

Strip revenue fell 2% to $1.95 billion. The 10-Q puts this down to "a decrease in room revenue due to lower ADR as well as a decrease in entertainment revenue from our venues, partially offset by an increase in casino revenue, which benefited from a higher table games win percentage." ADR is the average daily rate a room sells for.

Segment Adjusted EBITDAR is MGM's main profit measure for each segment. It is like EBITDA but also leaves out rent paid to the landlords who own most of MGM's buildings. The 10-Q lists three reasons it fell 12%: higher payroll costs, the lower revenue, and $23 million of business interruption insurance received in the prior-year quarter.

Casino revenue rising 8% looks like strength, but most of it is luck. "Hold" is the share of money bet at tables that the casino keeps, and it swings from quarter to quarter with the cards and dice. Our own estimate: at last year's 22.6% hold, this quarter's $1,435 million of table drop would have produced about $324 million of table win, not $363 million. That is roughly $39 million of revenue that a normal quarter would not repeat. Slot play, which behaves more steadily, rose only 1%.

Regional Operations: down 8% reported, up 1% on a same-store basis

Regional revenue fell 8% to $883 million because MGM sold the operations of MGM Northfield Park (Ohio) in April 2026. On a same-store basis, which removes Northfield from last year's figures, revenue rose 1% to $883 million from $873 million. Profit still fell: same-store Segment Adjusted EBITDAR dropped 7% to $243 million from $260 million. The prior-year quarter also had $9 million of business interruption insurance. Casino volumes at the remaining properties were soft, with table drop down 1% and slot handle down 9% reported (the reported slot figure still includes the sale of Northfield).

MGM China: record revenue from hold, not from more players

MGM China posted record revenue of $1.14 billion (+5%) and what the release calls a record 17.1% share of the Macau market. The 10-Q attributes the gain to "an increase in main floor table games win percentage, partially offset by a decline in table games volume." Main-floor table drop fell 10% to $3.68 billion, while hold rose to 27.7% from 24.7%. On our estimate, last year's hold applied to this year's drop would have produced about $910 million of main-floor win, not $1,021 million. That gap is roughly $110 million of revenue, before Macau's gaming tax.

Segment Adjusted EBITDAR fell 7% to $264 million. The largest stated reason is a $22 million increase in the branding fee MGM China pays its parent under a new long-term agreement ($41 million versus $19 million). That fee cancels out in MGM's consolidated accounts. Add it back and MGM China's segment profit was about flat ($305 million versus $303 million). Payroll and gaming taxes also rose. Because MGM owns about 56% of MGM China (it received $68 million of the $121 million September dividend), a larger fee moves profit from MGM China's minority shareholders to MGM's own.

Digital and BetMGM: the one area that improved

MGM Digital (LeoVegas and other online gaming outside North America) grew revenue 6% to $184 million and narrowed its Segment Adjusted EBITDAR loss to $4 million from $23 million. BetMGM, the North American online sports betting and casino venture MGM owns 50/50, is not consolidated. MGM's share of its operating income rose to $27 million from $24 million, and the release says BetMGM distributions and branding fees paid to MGM totaled $34 million in the quarter.

What the headline numbers hide

• The smaller loss is mostly last year's charges dropping out. Q3 2025 carried a $256 million goodwill impairment (a write-down of what MGM paid for acquired businesses) plus $93 million of write-downs tied to withdrawing Empire City's New York commercial casino license application. Neither recurred. Operating income swung from -$113 million to +$119 million, while Adjusted EBITDA, which excludes those items in both years, fell $103 million.
• About half of the EBITDA decline is one-offs and the asset sale, not trading. The prior-year quarter had $32 million of business interruption insurance ($23 million Strip, $9 million Regional), and Northfield contributed $36 million of segment EBITDAR: $68 million of the $103 million drop. The sale also cut annual cash rent to VICI by $53 million, and triple-net rent expense was $16 million lower than a year ago, which gives back part of that. The rest of the decline is lower Strip room rates and higher payroll.
• Adjusted EPS is worse than GAAP EPS this time. MGM's adjusted loss of $0.11 is larger than the GAAP loss of $0.07. A $44 million loss on debt and equity investments is added back, while a $24 million derivative gain and a $12 million currency gain are taken out. The GAAP figure is also helped by a $17 million tax benefit on $22 million of pre-tax income (a 79.1% benefit rate), which the 10-Q attributes to the mix of U.S. and Macau income. Macau gaming profit is exempt from complementary tax.
• Two segments rode on table luck. Our estimates put the higher hold at about $39 million of Strip revenue and $110 million of MGM China revenue. Underlying volume was weaker: Macau main-floor drop fell 10% and Strip slot handle rose only 1%.
• Cash conversion is softer, and credit losses rose. Year-to-date operating cash flow was $1.66 billion, down 11% from $1.87 billion. Year-to-date net income of $537 million is inflated by the $253 million Northfield sale gain, which is not operating cash. The provision for credit losses (mainly unpaid casino credit) nearly doubled to $74 million from $39 million year to date. Receivables were $1.07 billion against $1.12 billion at year-end, so they are not running ahead of revenue.
• A 7.5% lower share count, but no buybacks this quarter. Diluted weighted-average shares fell to 252 million from 273 million, the result of buybacks over the past year ($253 million in the first half of 2026 and $711 million in the first nine months of 2025). MGM bought back no shares in Q3. With a loss in the quarter, the lower share count adds little to EPS. MGM pays no regular common dividend; the 10-Q notes it suspended them.

On balance the quarter is flattered: the loss per share narrowed from $1.05 to $0.07 because last year's $256 million impairment and $93 million of Empire City write-downs did not repeat, while adjusted EPS fell from $0.24 to a $0.11 loss and Adjusted EBITDA fell 20%.

Did last time's read hold up?

Our Q2 2026 report named three things to watch in Q3, and all three played out:

• Strip room rates: we warned that if ADR kept falling without the high table hold, Strip EBITDAR would fall. ADR fell faster (-8% vs -4% in Q2), and Strip EBITDAR fell 12% even with higher hold.
• MGM China volume: main-floor drop fell 7% in Q2 and fell further, by 10%, in Q3. Hold again did the work.
• Digital losses: we said the losses needed to narrow. They did, to $4 million from $23 million.

Outlook

MGM gave no numerical guidance in the release or the 10-Q. The CEO said that "lower room rates during the third quarter impacted our results," but that "October has rebounded favorably." No October figures were given, so that claim can't yet be checked. The longer-term story is the same: MGM Osaka, where MGM has about $2.1 billion (¥333 billion) still to fund, and the digital businesses.

The new item is capital return. People Inc. withdrew its proposal to acquire MGM. Under the amended voting agreement, People and Barry Diller won't buy above 29.9% of the shares, but a stake that rises past that level only because MGM buys back stock is not a breach. They will also vote with the board on any change-of-control deal, subject to conditions. The CFO said this lets MGM "confidently pursue value-creating share repurchases," and $1.4 billion remains on the April 2025 authorization. Long-term debt fell to $5.67 billion from $6.23 billion at year-end, and cash was $2.40 billion.

Our read: three things to watch in Q4.

• Whether October's rebound shows up in Strip ADR. Q4 will also be compared against a year without unusual insurance receipts, so it is a cleaner test. Look for ADR at or above last year's level.
• Hold normalizing. If Strip table hold and Macau main-floor hold drift back toward 22–25%, revenue will fall unless volumes recover. Macau drop is the figure to check.
• How fast buybacks restart. With $1.4 billion authorized and the People constraint eased, a large Q4 repurchase would cut the share count further. It would not change the underlying trend in operating profit, which is down.