LVS — Q3 2026 Financial Report Analysis

Q3 2026: Luck is no longer the excuse. property EBITDA fell 17% even though VIP luck was close to normal

Las Vegas Sands' net revenues fell 3.5% to $3.22 billion in the third quarter of 2026, from $3.33 billion a year earlier. Profit fell further. Operating income was down 14.0% to $618 million, net income attributable to the company fell 27.4% to $304 million, and diluted EPS fell 21.3% to $0.48 from $0.61. Consolidated adjusted property EBITDA, the company's main measure of resort profitability, fell 16.7% to $1.12 billion.

Last quarter, a large part of the decline came from bad luck at Macao's VIP tables. That explanation does not work this quarter. Management's own estimate says luck at the VIP tables cost almost nothing ($7 million of EBITDA). Even after that adjustment, EBITDA fell about 14%, worse than the roughly 6% hold-adjusted decline in Q2. The two weak spots were Marina Bay Sands in Singapore, where casino revenue fell 15%, and Macao, where gaming volumes grew but margins kept shrinking.

Where this comes from

This analysis uses the Q3 2026 earnings release (Form 8-K, Exhibit 99.1, filed October 21, 2026). The full 10-Q, with management's line-by-line explanation of costs, had not been filed when this was written. So where the release gives figures but no reasons (for example, why mass-market win rates fell), we say so rather than guess.

Key figures

A few terms used throughout. Adjusted property EBITDA is each resort's earnings before interest, tax, depreciation, and corporate, development and pre-opening costs. Operating margin is the share of revenue left after running the business, before interest and tax. Drop is the cash and chips customers bring to the tables. It measures how much mass-market (non-VIP) customers are playing, and win % (or "hold") is the share of it the casino keeps. Rolling Chip is the company's term for VIP play. RevPAR is revenue per available hotel room. Operating margin, the Macao drop total and the hold-adjusted line are our calculations from the company's reported figures.

Takeaway: In Q2, about two-thirds of the EBITDA decline was VIP luck. In Q3, almost none of it was. VIP luck cost a net $7M, and hold-adjusted EBITDA still fell about 14%. Volumes rose in both markets (Macao mass drop +10.6%, Singapore +12.6%), but LVS kept a smaller share of each dollar played and spent more to earn it. Buybacks, not operations, are holding up the per-share numbers.

Separating luck from the underlying trend

High-stakes VIP play can swing a casino's quarterly results by tens of millions of dollars on luck alone. So the release estimates what EBITDA would have been if VIP win had matched the expected rate. That rate is 3.3% in Macao. At Marina Bay Sands it is 4.6% this year, raised from 4.2% a year ago.

Applying those adjustments (our calculation):

• Macao: about $518M vs. $599M, -13.5%
• Marina Bay Sands: about $594M vs. $700M, -15.1%
• Consolidated: about $1,112M vs. $1,299M, -14.4%

This quarter, VIP luck was roughly normal. Every Macao property with VIP play won more than the 3.3% expected rate: The Venetian Macao 3.97%, The Londoner 3.43%, The Plaza/Four Seasons 3.76%. Marina Bay Sands won 4.53%, just under its new 4.6% expected rate. Last year's Singapore result had about $43M of extra VIP luck in it. That explains only about a quarter of the $153M Singapore EBITDA decline.

One caveat matters a lot here. Management's hold adjustment covers VIP play only. Mass-market win rates also fell sharply this quarter, and the release does not adjust for that. Mass-market hold usually moves much less than VIP hold. The 10-Q should say whether the drop reflects luck, game mix, or more promotional play (for example, free-play and rebates that come out of reported win).

Marina Bay Sands: more players, much less kept

Singapore is still the most profitable property by far: a 45.6% EBITDA margin vs. 27.3% for Macao. But its casino side weakened sharply this quarter:

• Casino revenue fell $161M (-14.9%) to $916M. Mass-market drop rose 12.6% to $2,874M, but mass win fell to 18.7% from 24.5%. As a rough check, drop × win % gives about $537M of mass win vs. about $625M a year ago, roughly $90M less before rebates and taxes. VIP volume fell 7.1% to $8.43B, and VIP win slipped to 4.53% from 4.84%.
• Non-gaming kept growing. Room revenue rose 5.2% to $162M, ADR (average daily room rate) rose 5.7% to $1,038 at 95.8% occupancy, and food and beverage rose 9.5% to $104M. Rooms, dining, mall and other combined grew 5.3% to $378M. That's steady growth, but small next to the swing in casino revenue.
• EBITDA margin fell 6.1 points to 45.6%. Even on the hold-adjusted figures, the margin is about 45.7% vs. about 50.8% (our calculation). So costs grew while casino win shrank. The release does not break out those costs; the 10-Q will.

Macao: volume up everywhere, margin stuck lower

Macao net revenues rose 1.8% to $1,940M, but adjusted property EBITDA fell 12.0% to $529M. Margin was 27.3%, down from 31.5%. Management said investments in "enhanced service and hospitality offerings" contributed to "growth in volumes across all gaming segments." The volume figures support that:

• VIP volume across the five properties rose about 31% to roughly $5.74B (our sum). The Londoner led, up 39% to $3.22B.
• Mass-market drop rose 10.6% to about $7.18B (our sum), led by The Londoner (+23.6% to $2,803M) and Sands Macao (+29.4% to $480M).
• Slot handle (total amount bet on slot machines) rose about 26% to roughly $6.6B (our sum).

But the casino kept less of it. Weighted across the five properties, mass-market win fell to about 20.5% from about 23.1% (our calculation from each property's drop and win %). The biggest drop was at The Londoner Macao, 18.6% vs. 23.4%. As a result, The Londoner's casino revenue fell 2.1% despite 24% more mass drop and 39% more VIP volume, and its EBITDA fell 18.7% to $178M (margin 26.0% vs. 31.9%).

Property by property:

Margins in brackets are Q3 2025.

• The Venetian Macao's renovation is still weighing on it. Room revenue fell 15.4% to $44M. The release says about 500 rooms on average will be out of service through the fourth quarter of 2027, so this drag continues for another year.
• The Plaza/Four Seasons recovered from Q2's negative VIP win. Its VIP win was 3.76% (vs. 1.84% a year ago), and RevPAR rose 6.4% to $485.
• The Parisian grew revenue 6.0%, but its EBITDA fell 13.2%. That's the same pattern as the Macao segment overall: more business, lower profit on it.

Sands China, the Macao subsidiary, reported net revenues up 1.4% to $1.93B and net income down 29.8% to $191M. LVS bought 62 million more Sands China shares during the quarter for about $101M, taking its stake to 75.56%.

What the headline numbers hide

• The GAAP vs. adjusted gap points the other way from usual. Adjusted EPS fell 30.8% ($0.54 vs. $0.78), more than GAAP EPS fell (-21.3%). That's because last year's GAAP figure was held down by one-off items that adjusted EPS excludes: a $68M loss on disposal or impairment of assets and $72M of development spending. This year those items were $4M and $37M. This quarter's adjusted figure adds back $37M of development expense, $5M of pre-opening, $4M of impairment and $1M of other expense, net of $7M of tax.
• Buybacks cushioned EPS by about 6 points. Diluted shares fell 6.6% to 640 million from 685 million. That is why net income attributable to LVS fell 27.4% but EPS fell only 21.3%.
• A higher tax rate cost roughly $31M. The effective tax rate rose to 22.5% from 15.6%. The release attributes it mainly to the 17% statutory rate in Singapore. Macao gaming profits are exempt from corporate income tax, so the blended rate rises whenever Singapore makes up a larger share of profit. Pre-tax income fell 22.0% to $454M, and net income fell 28.3% to $352M. At last year's rate, tax would have been about $31M lower (our estimate).
• Less interest income. Interest income fell to $23M from $39M, with less cash on hand after buybacks and the early repayment of the Las Vegas sale loan in Q2. Interest expense was flat at $186M.
• VIP luck was neutral this time, but mass-market hold was not. As noted above, the company's luck adjustment covers only VIP play. Mass-market win fell 5.8 points at Marina Bay Sands and about 2.6 points across Macao. That is a large unexplained swing, and the hold-adjusted EBITDA figure does not capture it.
• Cash conversion can't be checked yet. The release has no cash flow statement. The balance sheet points in one direction, though. Unrestricted cash fell to $2.86B from $3.38B at June 30, and debt (excluding finance leases) rose to $15.30B from $15.11B. So net debt rose about $0.7B in the quarter, while LVS spent $950M on buybacks, about $101M on Sands China shares, its regular $0.30 dividend and $281M on capital spending. Shareholder payouts are running ahead of what the business generated this quarter.
• No numerical guidance, now or in Q2, so there's no guidance change to compare.

Did last time's read hold up?

Our Q2 report listed four things to watch. Macao hold-adjusted EBITDA margin: it was about 27% in Q2 and about 27.0% in Q3 (hold-adjusted EBITDA $518M on hold-adjusted revenue of about $1,921M). So it stabilized rather than falling further, but did not recover toward the 31% of a year earlier. Venetian renovation: still a drag, with room revenue down 15% and a stated end date of Q4 2027. MBS VIP hold vs. the 4.2% expected rate: Q3 VIP win was 4.53%. Management has now raised its expected rate to 4.6%, so Singapore no longer shows the large "above-normal luck" we flagged. Its weakness came from mass-market win instead. Dividend-tax agreement in Macao: the release does not mention it; the 10-Q should. Our broader Q2 call, that Macao's problem was cost and reinvestment rather than luck, held up. We expected Singapore to slow on tougher comparisons, but we underestimated how much.

Capital returns and balance sheet

• Buybacks: $950M in Q3 (about 21 million shares at a weighted average $45.31). Since buybacks resumed in Q4 2023, LVS has retired about 145 million shares, or 19.0% of the count, for $6.98B. $5.06B of authorization remains.
• Dividend: $0.30 per share paid in the quarter, and the next $0.30 is payable November 10, 2026. The board raised the recurring dividend 20% for calendar 2027, to $0.36 per quarter ($1.44 a year).
• Liquidity: $2.86B of unrestricted cash and $4.27B of undrawn revolving credit. A further $4.57B is available under a delayed-draw term loan for the Marina Bay Sands expansion.
• Capital spending: $281M in the quarter, including $154M in Macao and $121M at Marina Bay Sands. Management calls the MBS expansion one that will "meaningfully increase our suite capacity and entertainment offerings," including a new arena.

Outlook and our read

Management gave no numerical guidance. Its forward-looking comments were the 2027 dividend increase and confidence in "growth prospects in Singapore both today and upon the completion of our ongoing expansion project." It also pointed to "growth in volumes across all gaming segments" in Macao.

Our read:

• The volume story is real, the profit story isn't yet. In both Macao and Singapore, customers are playing more than a year ago. But LVS is keeping less of each dollar played and spending more to attract it. After Q1, Q2 and Q3 2026, the pattern in Macao looks structural: a margin in the high 20s rather than the low 30s.
• Singapore's mass-market hold is the swing factor. A 5.8-point fall in mass-market win in one quarter is unusual. If it reverses, Singapore EBITDA should recover a good part of the $153M decline. If the 10-Q shows the cause was more promotional play (rebates and free-play that come out of win), it may not.
• The dividend raise and buyback pace rely on the balance sheet. Net debt rose about $0.7B in one quarter, while large expansion spending in Singapore is still ahead. At this pace of payouts, per-share results will keep looking better than the business. As in Q2, we'd judge the company on property EBITDA, not EPS.

What to watch in Q4: whether Marina Bay Sands' mass-market win returns toward the low-to-mid 20s; the 10-Q's explanation of Macao and Singapore cost growth; the Venetian Macao renovation drag; the Macao dividend-tax agreement; and whether net debt keeps rising alongside the buybacks.

Source: Las Vegas Sands Q3 2026 earnings release (Form 8-K Exhibit 99.1, filed October 21, 2026). The full Form 10-Q for the quarter had not been filed at the time of writing. Operating margin, hold-adjusted EBITDA, property-level YoY changes and margins, Macao volume totals, weighted win rates and the tax-rate effect are our calculations from the company's reported figures.