Las Vegas Sands

IThe business

Las Vegas Sands builds and runs integrated resorts: a hotel, a casino, a shopping mall, a convention centre and a theatre or arena in one complex. It has two clusters. Marina Bay Sands in Singapore it owns outright. In Macau it has five properties through Sands China, a Hong Kong-listed company it owns 75.56% of.⁵⁴ It owns nothing in Las Vegas. It sold its Las Vegas properties and operations in 2022, and in May 2026 the buyer repaid the last $1.26B of seller financing early.⁵ The head office is still at 5420 South Durango Drive in Las Vegas, and the shares trade in New York.⁴

I picked it because it is listed as the largest public company in Nevada by market value, and it did not come from any screen.⁶ That site shows $30.5B; I count 626,990,133 shares on 21 October times $40.74, which is $25.5B C (calculated).⁵¹ All figures are in US dollars. The year is the calendar year. The third-quarter report for 2026 is the latest, and the price is the close on 4 November, $40.74.³¹²

In 2025 revenue was $13.0B, net income attributable to the company was $1.63B and diluted earnings were $2.35 a share R (reported).⁴ The measure management steers by is adjusted property EBITDA, profit at each resort before corporate costs, development spending, stock pay and depreciation. It was $5.23B in 2025, 44% from Macau and 56% from Singapore R (reported).⁴ Over the last twelve months it is $5.07B C (calculated).³⁴

How it earns money

Gambling is three-quarters of revenue. In the first nine months of 2026 casino revenue was $7.45B of $9.96B, or 75%; rooms $1.11B; food and beverage $0.52B; the malls $0.60B; conventions, retail and other $0.27B C (calculated).³ The malls matter more than their size. They are rented to luxury brands, and the four Cotai malls (Venetian, Four Seasons, Londoner and Parisian) earned $114M of operating profit in the third quarter on $131M of revenue, a margin of 87% R (reported).³

Governments take a large cut of the gambling. In 2025 gaming taxes were $3.92B, 40% of casino revenue C (calculated).⁴ In Macau it is a 35% special gaming tax plus a 5% contribution on gross gaming revenue.⁴ In Singapore the rate is tiered, and premium play moved from 8% to 12% once the threshold was reached in June 2026, a month earlier than in 2025.⁵ Payroll was another $2.05B in 2025 R (reported).⁴ What is left is about 37 cents of every revenue dollar before head-office costs in 2026, down from 41 cents in the first nine months of 2025 R (reported).³

Two businesses with different economics

Macau is the bigger business and the thinner one. In 2025 its five properties had revenue of $7.43B and EBITDA of $2.31B, a margin of 31% C (calculated).⁴ Singapore had revenue of $5.58B and EBITDA of $2.92B, a margin of 52% C (calculated).⁴ One property in Singapore earned more than all of Macau, with 25% fewer revenue dollars. That is the reason this company’s value rests more on Singapore than the headline count of resorts suggests.

What changed in the last two years

Four things. First, Singapore finished a multi-year renovation of all three hotel towers into 1,844 rooms and suites, and EBITDA there rose 42% in 2025, to a record $2.92B R (reported).⁴ Second, in Macau it converted the Sheraton towers into the Londoner Grand, with 2,405 rooms, in the second quarter of 2025; Macau EBITDA stayed flat, at $2.31B, because of higher marketing spend and payroll R (reported).⁴ Third, it stopped chasing new games: in April 2025 it dropped its bid for a New York casino licence and part of its online gaming effort, and booked $191M of impairments in 2025. Development spending was $269M that year, $193M of it on digital gaming R (reported).⁴ Fourth, it began the $8.0B expansion of Marina Bay Sands in May 2025 and started buying back shares in the fourth quarter of 2023.⁵³

The state of play now

The third quarter was weak. Revenue fell 3.5% to $3.22B, adjusted property EBITDA fell 17% to $1.12B, and earnings were $0.48 a share against $0.61 R (reported).³ Singapore EBITDA fell 21% to $590M and its margin dropped 6.1 points to 45.6%; Macau EBITDA fell 12% to $529M on 1.8% more revenue C (calculated).³ Management called the Singapore quarter aberrational. Patrick Dumont said fewer high-value customers visited, mostly in July and September, and that the win rate on mass-market tables was unfavourable; he said he could not point to one cause.⁷ The shares fell 2.6% the next day, to $36.53, and two days later reached a closing low of $35.68.¹ They rose 6.5% on 4 November, from $38.26 to $40.74. I found no company news for that day and do not know why.¹²

The tape outside the company is soft too. Macau’s gross gaming revenue was MOP21.68B in October, down 10% on a year earlier, the fifth monthly fall in a row, and the first ten months are up 1.6%.¹⁰ In the third quarter, mass baccarat fell 7.9% across the market while VIP baccarat was flat.¹¹ On 3 November, Sands China sold $1.0B of new bonds at 6.400% and 6.700% to repay a 2.300% note due in 2027.¹²

IIThe moat

The edge here is a permit. Macau has six gaming concessions, and Sands has one, running to 31 December 2032.⁴ Singapore has two casino licences, and the regulator must not grant a third before 1 January 2031.⁴ A rival cannot open a casino down the road in either place. That is worth a lot, and it is why the margins are high. It is also an edge a government gives and can take back, so I treat it as the thing to worry about first.

What protects it

The number of rivals is fixed. There are six operators in Macau: Sands, SJM, Wynn, Galaxy, MGM and Melco. There are two in Singapore: Marina Bay Sands and Resorts World Sentosa.⁴ Nobody new can enter before the rules change.

Scale on one strip of land. Four of the five Macau properties are next to each other on Cotai: the Venetian, Londoner, Parisian and the Plaza with Four Seasons. The Parisian is connected to the Venetian and the Plaza, and the Londoner is across the street.⁴ In the third quarter, management says, Sands China took 26.1% of Macau’s mass-market gaming revenue, the highest since the fourth quarter of 2023, and its slot revenue grew about double the market’s rate.⁷ I have only management’s word for the share, as the government does not publish it by operator.

Non-gaming income that rivals find hard to copy. About 2.8 million square feet of meeting and exhibition space, the Londoner Arena with 6,000 seats, and from 2031 a second arena of about 15,000 seats at Marina Bay Sands.⁴⁵ Mall rent and room income were $2.2B in 2025 C (calculated).⁴ They do not depend on luck at the tables.

A balance sheet that can carry an $8B project. Net debt is $12.7B, 2.5 times EBITDA; the Singapore delayed-draw term loan has SGD 5.85B, about $4.57B, still undrawn and the interest rate on the Singapore loans is 2.4% R (reported).⁵

Free cash flow, ten years

The record is honest about the cycle. Free cash flow was $3.8B in 2018, fell to $1.8B in 2019, and was negative for three years while Macau was closed to most visitors. Since 2023 it has run between $1.6B and $2.4B a year.¹³ The business can lose cash for a long time when a government closes the border; it did.

On book capital the return is good. Operating income over the last twelve months was $2.85B. Against $15.0B of buildings, land rights and goodwill on the balance sheet, that is 19% before tax and about 16% after, using the 17.9% tax rate of the first nine months C (calculated).⁵³ That is about six points above the 9.5% I use to discount the business. The book value of buildings put up between 2004 and 2012 is less than what they would cost to build now, so the real return on replacement cost is lower.

Where it is thin

The permits end. If the Macau concession is not extended, the gaming areas go back to the government without payment, and Sands cannot run casinos in Macau.⁴ Singapore’s casino licence has a three-year term and is next up for renewal in April 2028; the company paid SGD 101M, about $75M, for the current one.⁴ The development agreement runs to August 2036, and the company must give notice to renew by August 2031.⁴ Singapore’s exclusive period ends on 31 December 2030.⁴ The bond market has noticed some of this: Sands China just sold notes that mature on 8 November 2033, ten months after the concession ends, at 6.700%, against 2.300% on the note being repaid.¹²

The government also sets the price. To win the Macau concession Sands promised to invest at least MOP35.84B, about $4.43B, by 2032, of which MOP33.39B must go to non-gaming projects. About $1.03B counts so far: $717M confirmed for 2023 and 2024, and $312M spent in 2025 that is still being audited C (calculated).⁵ In late October it committed MOP2.30B, about $285M, to a government-guided economic diversification fund, which counts toward that promise.⁵ Its shareholders owe a 12% tax on dividends from the Macau business unless a substitute agreement is in place. The last one covered 2023 to 2025, the company has asked to extend it to the end of 2027, and its 2026 tax rate assumes the extension.⁵⁴

The rivals match what Sands does. Same game, same tax, same customers. In the first nine months of 2026 Sands’ Macau revenue rose 7.7% and its EBITDA fell 6.5%. The 10-Q gives the reason in plain words: more spending on patron reinvestment and more pay for extra table hours and service levels.⁵ That is how a fixed number of rivals still competes on price. The permit keeps the number of players steady; it does not protect the margin.

IIIThe competitive advantage

I can measure the advantage in three places: share of the market, margin against the market, and the cost of staying ahead. The filings give me Sands’ own numbers. They do not give me rivals’ numbers, and I do not guess at them.

Share, and what it costs

Management says Sands China’s share of mass-market revenue was 26.1% in the third quarter, the highest since the end of 2023, and that it came with the Londoner Grand and more table hours. Growth in the third quarter against the year before: rolling-chip volume up 31%, non-rolling drop up 11% and slot and electronic-table handle up 26% R (reported).⁷ What did that buy? More volume and no more profit. In the quarter, Macau casino costs rose $60M, of which $23M was higher gaming tax, which comes with higher revenue, and $20M was pay for more table hours and staff.⁵ The company says it will not cut those hires, that fixed costs were about flat from the second quarter, and that it expects no big headcount growth next year; it aims for Macau EBITDA of more than $700M a quarter. In the third quarter it was $529M.⁷³

Win rates add noise. In the third quarter, rolling-chip play (big bettors) won a little above the expected rate in Macau, and management says EBITDA would have been $518M at the expected rate, $11M less than reported.⁷ In Singapore it won a little below, and EBITDA would have been $594M, $4M more.⁷ So luck on rolling play explains almost none of the fall. Mass-market tables at Marina Bay Sands won 18.7% of drop against 24.5% a year before, which does explain a lot, and is the item I cannot judge.³

Where the rival is ahead

In Macau, the evidence is that a rival can match Sands at a cost: the market grows but Sands’ profit does not. MGM China reported a record share in the third quarter, which I covered on my MGM Resorts page.¹⁴ In Singapore there is one rival and no published numbers. What I can say is that Resorts World Sentosa is owned by Genting Singapore, that it holds the other of the two licences, and that Marina Bay Sands earns margins of about 50%.⁴³ The edge Sands has is one location, the Marina Bay skyline, which no rival can build.

What would make the advantage wider. The Singapore expansion adds a hotel tower, premium gaming space and a 15,000-seat arena; the company says it will meaningfully increase suite capacity and make the property the most important entertainment venue in the region.³ If Marina Bay Sands alone has that scale in Singapore, the other licence holder will find it hard to match. I assume that adds about $1.1B of EBITDA by 2036 in my base case, and nothing in a filing supports that figure; it is my estimate E (estimate).

IVManagement

Who controls it

The Adelson family does. Dr. Miriam Adelson, trusts for her family and the entities around them beneficially owned 386,657,840 shares on 16 March 2026, 58.2% of the stock.⁹ Buybacks since then have shrunk the share count, so if the family sold nothing, its stake is now about 62% of the 627.0M shares C (calculated).⁹⁵ That is about $15.8B at today’s price. The company says openly that it is a controlled company under New York Stock Exchange rules.⁹ The next largest holder in the proxy is Vanguard, with 6.4%. No other holder is named at more than 5%.⁹

Patrick Dumont became chairman and chief executive on 1 March 2026. He joined the company in 2010, was chief financial officer from 2016 to 2021 and was president and chief operating officer until this year. He is Dr. Adelson’s son-in-law, and since December 2023 he has been governor of the Dallas Mavericks, which the family owns a majority of.⁹ Robert Goldstein, who held the top job from 2021, became a senior adviser on the same day. Randy Hyzak is chief financial officer and Zachary Hudson is general counsel.⁹ A family-controlled company with a family chief executive can be run for the long term without quarterly pressure. It also has no outside check if the family wants something the other holders do not, and a minority holder cannot vote it out. I have no evidence that has happened. I note that the structure allows it.

How they are paid, and what that pays them to do

Now follow the incentive. Both performance measures go up if the company spends borrowed money on a big project that adds EBITDA, because neither charges for the capital spent. Operating cash flow per share also goes up when the company buys back stock, because the share count falls. The company is doing both: an $8.0B Singapore project and $2.5B of buybacks in nine months.⁹⁵ I do not say that is wrong. I say the pay plan pays him for growth and for a smaller share count, and does not pay him for return on the capital he puts in. It is also true that the plan is new, the stockholder feedback behind it was real, and the board cut the chief executive’s salary and raised the share of pay at risk.⁹

His pay at target rises from the $18.3M he was paid for 2025 to about $27M, and the equity part is four times larger. His own money is modest next to the family’s. He holds 510,367 shares outright, about $21M at today’s price, plus options on 1,561,740 more.⁹ The company’s rule asks the chief executive to own eight times salary, $20M, so he is at the line and not far past it. The real alignment is the family’s $15.8B, which sits behind the same share price.⁹

What they did with the money

Over a full cycle I read this as a team that pays out nearly all it earns, sometimes more, and has kept net debt near 2.5 times EBITDA. It cut the dividend when it had to, in 2020, which was the right call, and it restarted payouts once the border opened. The buybacks look worse than the dividends. It bought 145M shares at an average of $48.03, and the price is now 15% under that. The count of 19% fewer shares is real, and it raises earnings per share by about a fifth if profit is flat. It is a good use of cash only if the shares were bought under what they are worth. My base value is $46, under the $48.03 average, so on my numbers the average purchase was made above value, and the third-quarter purchases at $45.31 a share were about at it E (estimate).³

Asked on the third-quarter call whether the company might go private, management said it had no interest, and said buying back stock was its focus over acquisitions.⁷ The unproven part is the Singapore project. It will cost $8.0B, 3.3 times the last twelve months of free cash flow, and it was committed in a market that has just had a soft quarter. Marina Bay Sands and the Cotai Strip are the track record, and both now earn a great deal. I have no filing that compares their first budgets with what they cost, so I cannot say they came in on budget. Nor does any filing give the return management expects on the new project, so I cannot test it. I use my own estimate.

VFinancials

Ten years of numbers, with the changes of structure marked. Years 2016 to 2018 include the Las Vegas properties as first reported. From 2019 the figures are restated without them, because they were sold in February 2022. Net income for 2019 to 2022 includes results from the discontinued Las Vegas operations, and 2022 includes a gain on the sale of $2.9B R (reported).¹³

Revenue and operating income, ten years

The pattern is plain. Revenue recovered to its pre-closure level in 2025, with a smaller company. Operating income of $2.8B in 2025 is below the $3.4B of 2019, because costs moved up as fast as revenue: payroll, gaming taxes and what the company spends to win customers. The margin was 21.6% in 2025, 27.7% in 2019 and 22.3% in 2023 C (calculated).¹³

What an owner can really take out

Free cash flow overstates what an owner can take. It is before stock pay, which the filings add back as a non-cash cost, before the share of cash that goes to the minority holders of Sands China, and before the capital needed just to keep the buildings as they are. I take those off in the table.

So an owner could take out about $3.94 a share before paying for growth, 62% more than reported earnings of $2.44. The gap is depreciation, $1.41B a year, against the $0.70B I think is needed to stand still. If I am wrong and the true figure is nearer depreciation, owner cash falls to $2.81. I do not know which is right, because the company gives the Macau figure and gives none for Singapore. I carry both.

Then comes growth that is already promised. Of the $8.0B Singapore project, $3.1B is spent and $4.9B remains. The Macau investment plan has about $3.4B to go by 2032 in the filing, though the company says some of it may be moved to other projects C (calculated).⁵ Together that is about $8B, or $13 a share, over six years. In my model, owner cash after interest and minority dividends and after this spending is only $0.8B in 2027 and 2028 E (estimate). The 2027 dividend of $1.44 a share costs about $0.9B. This is the tight spot in the story: in my base the dividend is slightly more than owner cash in 2027 and 2028, so it is partly paid from debt, and there is no room for $3B-a-year buybacks.

In nine months of 2026 the company paid out $3.2B, against $1.5B of free cash flow. The $1.26B repayment of the Las Vegas sale loan in May covered most of the gap, and cash fell by the rest. That repayment will not come again. At the same pace, next year’s payout has to be funded from cash flow and debt.

The weakest line is the Parisian, with a 19.9% margin and 69.5% of its mall let. The strongest recent move is the Plaza and Four Seasons, up 15% in EBITDA on 16% more revenue. Management put the Londoner’s fall down to a bad run of luck in the very top of premium mass play.⁷ The Venetian is mid-renovation: about 400 rooms are out of service, rising to about 500 next quarter and staying out through the fourth quarter of 2027, with the refreshed hotel and new premium gaming rooms due by Chinese New Year 2028.⁷³

The balance sheet

Total debt is $15.58B with finance leases, against cash of $2.86B, so net debt is $12.72B, or 2.5 times last-twelve-month EBITDA C (calculated).⁵ The weighted cost of borrowing is 4.7%, and interest cost was $186M in the quarter, covered 6.0 times by EBITDA.³ The covenants have room: leverage was 1.74 times at the US parent, 3.29 at Sands China and 1.54 at Singapore, against limits of 4.00, 4.00 and 4.50 R (reported).⁵ The Sands China figure is the one to watch, because a fall in Macau EBITDA moves it fastest. The undrawn bank lines are $4.27B plus $4.57B for the Singapore project.³

The refinancing bill is rising. The 2027 Sands China note carried 2.300%; the replacement carries 6.400% to 2032. That is 4.1 points more on $700M, about $29M a year. About $7.5B of notes and bonds mature from 2027 to 2030 at an average coupon of 4.8%, and the market has just asked 5.3% to 6.7% for new ones C (calculated).⁵¹²

Dilution and stock pay

Stock pay is small: $66M over the last twelve months, 0.3% of the market value, and 1M shares of dilution C (calculated).⁵³ What matters is the count of shares. It fell from 765M in 2023 to 640M in the third quarter of 2026. The company holds 215M shares in treasury at a cost of $11.5B, and total stockholders’ equity is a deficit of $344M, because buybacks were charged against it R (reported).⁵ A negative book value is a result of paying out more than book profit. It means the balance sheet says nothing about value, and the cash flows have to.

The share price over ten years

Ten years ago the price was $56.74. It is $40.74 now, 28% lower. Add the $12.89 a share of dividends declared from 2017 and the holder is about 5% behind, before inflation C (calculated).¹¹³⁵ The shares ended 2025 at $65.09, a closing high for 2026 of $65.21 on 2 January, and have fallen 37% since R (reported).¹ In the third quarter the company bought 21M shares at an average of $45.31, 11% above today’s price.³

At 7.5 times EBITDA before the minority and 8.5 times with it, the market is not paying a growth price. At 16.7 times earnings and a 9.4% free cash flow yield on market value it is paying for a business whose earnings carry two drags: a project that costs $8.0B before it earns, and the cost of staying level in Macau C (calculated).

VIValuation

First, how this goes badly wrong

Before the numbers, I ask what would most likely hurt an owner, and work back from there. There are three ways, and one of them is the whole story.

The Macau concession ends and is not renewed, or is renewed on worse terms. It expires on 31 December 2032. If it is not extended, the gaming areas go back to the government without payment.⁴ Macau is 52% of the enterprise value in my base case E (estimate). If I set the chance of renewal to zero, my base value falls from $46 to $27, a loss of 41%.

The high end stays weak in both markets while costs stay high. The third quarter showed fewer high-value visits and a heavy reinvestment bill. If Macau earns 20% under my base every year and Singapore 15% under, the value is $30. That is the case in which this quarter is the new level.

The balance sheet meets the build. Net debt is 2.5 times EBITDA now, and it could rise as the company spends $4.9B on Singapore while paying a dividend that grows 20%. At Sands China, covenant leverage is 3.29 times against a limit of 4.00 R (reported).⁵ If its EBITDA fell 20% with its debt unchanged, simple arithmetic puts the ratio near 4.1 times E (estimate). I do not know the exact covenant definition, so I treat that as a warning only.

The first of these is the one I cannot analyse with data. Whether the Macau government renews is a decision, and I use a probability and show the range.

What an owner can really take out

Last twelve months, before growth spending, an owner could take about $3.94 a share, or $2.81 if the buildings need depreciation-sized spending to stay as they are, against reported earnings of $2.44 (Section V). That is the gap I value. In my base case the cash left to owners after interest, minority dividends and the promised growth spending is $0.8B in 2027, $1.9B in 2030 and $4.0B in 2036, which is $6.38 a share in the last year E (estimate).

What I cannot judge

Why Singapore was soft in the third quarter. It is either bad luck, or a change in how rich Chinese customers behave. Management could not say. Analysts at J.P. Morgan, Bank of America and Citigroup all asked on the call whether China’s tax rules on offshore trusts were behind the high-end softness.⁷ The rules were overhauled in late July, with a 20% tax on gains in assets moved into offshore trusts and a 20% annual tax on their income.¹⁵ I could not verify a source that ties it to casino play, so I do not build it into the base. The bear case does.

What the Singapore expansion earns. The company gives no return. I assume it adds $1.1B of EBITDA by 2036 in the base, $0.5B in the bear and $1.7B in the bull E (estimate). If it earns nothing and still costs what it costs, the base value is $35.

The chance of renewal in Macau. I use 85% in the base, 60% in the bear and 95% in the bull. Those are my guesses. Between 60% and 100% my base value runs from $41 to $50.

The cash Sands China holds. It is not disclosed apart from the group’s. I assume $1.0B. Each $0.5B of difference moves the value of the minority holders’ share by about $0.12B, or $0.20 a share. It is small, and I say so.

How the cases are built

I value the business in three streams: Macau, Singapore and head office. For each, cash is adjusted property EBITDA less cash tax, capital to stand still and growth spending; head office costs, development spending and stock pay come off as a separate stream. I discount at 9.5%, the 10-year Treasury yield of 4.90% on 4 November 2026 plus 4.6 points for two regulated licences, one of which ends in 2032, and for cash that depends on Chinese customers E (estimate).⁸ Everything Macau earns after 2032 is multiplied by the chance of renewal. I take off the debt, $15.6B less $2.9B of cash, and the minority holders’ 24.4% of Sands China. The shares are 628M, the 627.0M outstanding plus 1M of dilution.⁵

Look at what the base asks. Macau EBITDA goes from $2.20B over the last twelve months to $2.65B in 2030, a rise of about 4.8% a year over four years, which is below management’s own aim of more than $700M a quarter, $2.8B a year.⁷ Singapore goes from $2.87B to $3.03B in 2030, 1.3% a year, a little over 2025’s record. Then the expansion opens in January 2031 and, in my estimate, adds $1.1B by 2036. Total EBITDA is $5.68B in 2030 and $7.64B in 2036, against $5.23B in 2025.

What the price needs

The market is asking for less than my base on every line. The price needs property EBITDA to reach $5.4B in 2030, 6% above the last twelve months and 2% above 2025’s record C (calculated). Or it needs the chance of renewal to be 60% and not 85%. I prefer to say it this way: the price is what my base case is worth if the market thinks Macau is a six-in-ten chance after 2032 instead of an eighty-five in a hundred. I cannot tell which view is better. That is the sort of bet where the honest answer is a range.

The margin of safety, and the price against the range

My margin-of-safety price is $32, 70% of the $46 base. I take off 30% because the largest input is a government decision I cannot forecast. At $40.74 the price is 11% below the base and 27% above $32. In plain words: the price is under my base value and above the point where I would call it a margin of safety. It is 39% of the way from my bear to my bull. The probability-weighted value, with 25%, 50% and 25% on bear, base and bull, is $47 E (estimate).

Quality and price are separate questions. As a business, this is two high-margin resorts protected by permits, with strong cash generation and a Macau permit that ends in six years. As a purchase at $40.74, it is priced for a middling result with little protection if Macau goes wrong. A good business can be a poor purchase at the wrong price. Here the price is neither very high nor very low against my range, and the reason is the permit.

The market may be right. It has seen the third quarter, a fifth monthly fall in Macau gaming revenue and an $8B project that will not earn until 2031. My base needs the high end to recover and Macau to grow 4.8% a year. If those do not happen, the price is not too low.

The next best use of the same dollar

The comparison is not only Treasury bills. A 10-year Treasury pays 4.90% with no permit risk.⁸ At the price, my base case gives about a 12% yearly return over ten years, my bear 3% and my bull 19% E (estimate). That is a return on the share price after interest and minority dividends, unlike the 10.1% in the answer, which is a rate on cash before debt. It is a wide spread over the Treasury in the base and a thin one in the bear. Among my own pages, the closest comparison is MGM Resorts, which has the same Macau cycle and a Las Vegas business that pays most of its profit to landlords: its price of $29.96 is 21% below its base of $38 and above its margin-of-safety price of $25.¹⁴ The widest gap between price and base I have on this site as I write is Booking Holdings, whose $148.13 is 34% under its $223 base, a different risk altogether.¹⁶ LVS has the smallest gap of the three, and its main risk, a permit that ends, is not shared by Booking.

Where this could be in 5 and 10 years

In the base, the yearly return is 12% over ten years because the price is below the value and because cash flow grows after the Singapore build. In the bear it is 2.6%. The return from today’s price is higher than the 9.5% I require only if the base is right.

VIIThe case for the price

What most people miss

Three facts cut in the company’s favour. First, the lowest-cost money in the group is where the growth is. The Singapore delayed-draw loan has SGD 5.85B, about $4.57B, still to draw at 2.423%, while Sands China just paid 6.4% to 6.7% for new bonds.⁵¹² Second, about $2B of the $8.0B Singapore bill is land premium, which management says is amortized over time, and the 10-Q says the $8.0B also includes financing fees and interest. Management says a good part of the $4.9B still to spend is locked in.⁷⁵ Third, customers are still coming. In the third quarter Macau rolling-chip volume was up 31%, non-rolling drop up 11% and slot handle up 26%, and retail tenant sales rose 10% R (reported).⁷ The problem in Macau is margin.

The bull case

In plain words: the third quarter was a bad run. Patrick Dumont called the Singapore quarter aberrational and said the company is confident in its long-term strategy in both markets.⁷³ Grant Chum, who runs Sands China, said the long-term goal for Macau is more than $700M of EBITDA a quarter, against $529M now, and that renovated Venetian rooms are doing better than expected.⁷ The dividend rises 20% for 2027 to $1.44, and $5.06B is left on the buyback plan.³ CreditSights, as reported in early November, thinks the weak Macau data may be cyclical and not structural, which is about the market, not about this company.¹⁰ I found no other named view on Las Vegas Sands that I could verify, and checked no 13F filing. The bull case here is management’s.

Against reality

If Macau gets back to about $2.65B of EBITDA in 2030 and Singapore stays near $3.0B before the expansion, the model gives my base, $46. The price of $40.74 is 11% below that, so it is possible and asks for less. To be worth $40.74, property EBITDA has to reach about $5.4B in 2030, against $5.07B over the last twelve months and a best-ever $5.23B in 2025 C (calculated). That is a rise of 6% in four years, in a market that fell 10% in October. My bull case needs $6.5B, 29% above today, and four things at once: renewal on good terms (about 90%), Macau near $3.15B (about 30%), Singapore near $3.4B before the expansion (about 35%) and the expansion adding $1.7B (about 25%). Independent, that is about 2%; with overlap, perhaps 8% E (estimate).

What a 10x would take. Ten times $40.74 is $407.40 a share, a market value of about $255B on 627M shares. With net debt added, and the Sands China minority taken as about a tenth of the total as in my model, the business would be worth about $300B. At 12 times EBITDA that needs about $25B a year, five times the $5.07B of today C (calculated). That is four-fifths of the $30.9B all Macau casinos took in as gaming revenue in 2025.⁴ The highest close in the ten years of data, $81.27 in June 2018, valued the company at about $64B on 786M diluted shares, so a tenfold move from here is four times that C (calculated).¹¹³ In ten years a tenfold move is 26% a year. Over the last ten years the shares fell 3.3% a year, from $56.74 to $40.74. A tenfold gain for a company this size, from this market value, is not plausible, and I do not model it.

Where I come down

What moved me is the third quarter: it puts my starting points for Macau and Singapore below 2025’s, and I put Macau at $2.65B in 2030, under management’s own aim. What did not move me is the permit risk, which was always the largest input. My values are $18, $46 and $76. What would raise them: two quarters in a row of Macau EBITDA above $600M and Singapore above $700M, a clear route to the renewal of the Macau concession, and a firm cost for the Singapore build.

VIIIRisks, and what would change my view

What a bad year looks like in the model

In my bear, total property EBITDA is $4.65B in 2028, 8% under the last twelve months, and the Singapore build is at its heaviest. Owner cash after interest and minority dividends is $0.03B that year, so the dividend of about $0.9B is paid out of borrowing E (estimate). That is how a bad year looks for this company: profit slips a little, spending does not, and the gap comes from debt. Net debt would rise by about $1.4B over 2027 and 2028, to roughly 3.0 times EBITDA.

Six things that could go wrong, rerun through the model

Only a failure to renew takes the value below $35 on its own, to $27. Two together do much more: Macau and Singapore both under base, as in the high-end case, gives $30. The risks overlap. They share customers, regulators and a calendar.

Other risks I weigh

China. Beijing changes the rules for its citizens with little notice, on tax, travel and spending. The offshore-trust tax rules were overhauled in July.¹⁵ Singapore customers are not governed by Macau’s regulator, but many are from the same pool of wealthy Chinese visitors.

Singapore’s calendar. The casino licence is up for renewal in April 2028, exclusivity ends on 31 December 2030, and the company’s own estimate for completing the expansion, June 2030, is 11 months past the date it agreed with the government, 8 July 2029. Any extension needs the government’s approval.⁵⁴ I do not know how that will be treated.

Litigation. A Macau case, brought in 2012 and now claiming MOP96.45B, about $11.9B, has been dismissed twice. The plaintiff was dissolved in 2025 and its shareholders took its place. I count nothing for it in my values. The company’s note says that, in management’s opinion, its litigation will not have a material effect on its finances.⁵

Control. The family controls the vote. A minority holder cannot change the board, the pay plan or the capital plan without the family’s consent.⁹

Currency. Macau revenue is in patacas and Hong Kong dollars, which move closely with the US dollar. Singapore revenue is in Singapore dollars and its debt is too. The company hedges part of it, with $1.80B of net-investment hedges on Marina Bay Sands.⁵ A fall in the Singapore dollar lowers reported profit.

What would change my view