No Hotels Owned in America, a 74 Percent Margin: Why Hilton Earns as Much From a Hotel as Its Owner — Without Putting Up a Dollar

Hilton does not own a single hotel in the United States. Of the 851,582 rooms flying a Hilton flag in America, the company owns none; 91 percent are franchised. Yet in 2025 Hilton earned a record $3.7 billion in adjusted EBITDA at a 74 percent margin — in a year when revenue per room rose just 0.4 percent. How that works, who actually carries the risk, and why a company with negative $6.3 billion of shareholders’ equity trades at twice the multiple of the people who own the buildings: an anatomy of arguably the most elegant business model in consumer stocks.

The short answer is that Hilton and Marriott are no longer hotel companies. They sell a sign over the door, a reservation system, and a loyalty program with a quarter of a billion members — and in return they collect a fixed slice of the revenue of a building that somebody else designed, financed, and insured. The long answer is more useful, because it shows where the returns in this model come from, where the risk was quietly relocated, and where, in 2026, the wall is showing its first cracks.

The Great Separation: How Hotel Companies Became Brand Companies

Until the 1990s, a hotel company was exactly what the name implies: a business that owned and operated hotels. The balance sheet was full of concrete, returns on capital were mediocre, and every recession hit with full force, because a fixed cost base suddenly met half-empty buildings. Marriott made the first cut in 1993, splitting into Host Marriott, which kept the real estate and today trades as the REIT Host Hotels & Resorts, and Marriott International, which kept only the brands and the management contracts.

Hilton came later and by a different route. Blackstone bought the company for roughly $26 billion in 2007, on the eve of the financial crisis, nursed it through the downturn, took it public again in 2013, and on January 3, 2017, broke it into three pieces: the owned hotels went into Park Hotels & Resorts, the timeshare business into Hilton Grand Vacations, and what remained is the Hilton Worldwide Holdings that trades today. In Europe, Accor made the same move in 2018 by selling 57.8 percent of its property company, AccorInvest, to sovereign wealth funds and institutional investors for €4.6 billion.

For investors, Hilton’s spin-off date is an unusually clean experiment. On January 4, 2017, the very same hotels began trading under two tickers: once as the brand, once as the buildings. Here is what has happened since.

Security	Role	Total return Jan 4, 2017–Sep 25, 2026	2020 trough vs. end-2019
Hilton (HLT)	Brand, franchise, management	5.63x (+463%)	−49%
Marriott (MAR)	Brand, franchise, management	4.65x (+365%)	−61%
S&P 500 (price index)	Benchmark	3.41x	—
Host Hotels (HST)	Owner (REIT)	1.72x (+72%)	−50%
Park Hotels (PK)	Owner, Hilton spin-off	1.13x (+13%)	−81%

The figures use dividend-adjusted closing prices and are our own calculation; the S&P 500 line excludes dividends and therefore understates the index slightly. None of that changes the picture. Buy the brand in 2017 and you have more than quintupled your money. Buy the buildings from which the brand collects its fees and, after almost a decade and every distribution included, you are up 13 percent. Same rooms, same guests, same nights — and a gap in returns of more than forty to one. The reason lies in the fee schedule.

The Real Product Is the Fee

An owner who wants to run a property under a major flag has two options. Under a franchise agreement, the owner operates the hotel directly or through a third-party management company and pays the brand ongoing fees. Under a management agreement, the brand also runs the hotel — it supplies the general manager and the leadership team — and the owner becomes a pure provider of capital. In both cases, the line staff are usually the owner’s cost, not Hilton’s.

Franchise fees are remarkably standardized. For Hampton by Hilton, the company’s single largest brand at roughly 360,000 rooms, the franchise disclosure document lists a 6 percent royalty on room revenue plus a 4 percent program fee for marketing, reservations, and technology — 10 percent in all. The consultancy HVS, surveying 93 U.S. brands, puts the average total franchise cost at 9.9 percent of room revenue over a ten-year term; the initial fee is just 2.3 percent of that total, while nearly 90 percent comes from ongoing revenue-based charges. Listed hotel owners such as Chatham Lodging disclose ranges of 2 to 6 percent in royalties plus 1 to 4.3 percent for marketing and reservations.

Management contracts add two layers: a base fee, typically a few percent of total hotel revenue, and an incentive fee tied to the hotel’s profit, usually paid only after the owner earns a minimum return. The incentive fee is the only part of the business that truly depends on the hotel making money — and, accordingly, the first thing to disappear in a downturn.

What matters most is the base on which fees are charged: revenue, not profit. The brand gets paid on the first room sold, whether or not the hotel ends the year in the black. Wages, utilities, property taxes, insurance, interest, and the mandated renovation every seven years are the owner’s problem. The brand carries the cost of its headquarters, its sales engine, and its technology — and even the program fees are explicitly structured as cost reimbursement, not as a profit center.

The Economics of One Hampton: Who Earns What From a Hotel?

To see how unevenly the risk is split, it helps to model a typical select-service hotel off a U.S. interstate exit. According to the franchise disclosure, a new Hampton costs between $15.2 million and $22.2 million to build. We assume 120 rooms and an $18 million investment, an average daily rate of $140 and 72 percent occupancy, or about $101 in revenue per available room. The owner finances 60 percent with a loan at 7 percent and puts up 40 percent in equity. We treat 60 percent of operating costs as fixed — core staffing, property tax, insurance, maintenance — with the rest flexing with occupancy.

Model hotel, 120 rooms ($M per year)	Normal year	RevPAR −20%	RevPAR −50%
Room revenue	4.42	3.53	2.21
Operating profit before brand fees	1.84	1.14	0.09
Fees to the brand (10%)	0.44	0.35	0.22
FF&E reserve (4%)	0.18	0.15	0.09
Owner net operating income	1.21	0.64	−0.22
Interest ($10.8M at 7%)	0.76	0.76	0.76
Cash flow on $7.2M equity	+0.46	−0.12	−0.98
Change: brand / owner	—	−20% / −126%	−50% / −313%

The model is rounded and deliberately simple, but its shape survives almost any set of assumptions. Three things stand out. First, in a normal year the brand makes $440,000 from this hotel — almost exactly what the owner makes, at $460,000. One party put up $7.2 million in equity and is on the hook for $10.8 million in debt; the other supplied a sign and a contract. Second, when revenue falls by a fifth, the brand’s fees fall by exactly a fifth. The owner, meanwhile, slips into the red, because the cost base stays put and the interest keeps accruing. Third, even in a disaster scenario with revenue cut in half, $220,000 still flows to the brand while the owner has to inject nearly a million dollars a year.

That is the heart of the model, and it should not be mistaken for a trick. The operating leverage that makes hotels so cyclical has not been abolished. It has been relocated — from the listed brand company to thousands of owners: family businesses, REITs, and private equity funds. In economic terms, the brand holds a share of the revenue without a share of the losses.

Hilton by the Numbers: 1.1 Percent of the Rooms, a 74 Percent Margin

How far Hilton has taken this model shows up in its year-end 2025 property table. Of 1.33 million hotel rooms in the system, the company owns or leases just 15,287, or 1.1 percent — almost all of them leased properties in Europe. It manages 265,000 rooms, or 20 percent, and franchises 1.05 million, or 79 percent. In the United States the picture is starker still: zero owned rooms, 9 percent managed, 91 percent franchised. Of 3,195 Hampton hotels, 3,143 are run by someone else.

Hilton, December 31, 2025	Rooms	Share	Of which U.S.
Owned or leased	15,287	1.1%	0
Managed	264,760	19.9%	79,351
Franchised and licensed	1,050,900	79.0%	772,231
Total hotels	1,330,947	100%	851,582

The income statement does not look like a 74 percent margin at first glance. Hilton reported 2025 revenue of $12.04 billion, but $7.09 billion of it — 59 percent — was cost reimbursement: money owners pay in for marketing, reservations, the loyalty program, and the hotel staff that are formally Hilton employees under management contracts, which Hilton says it passes through without a markup. Strip out those pass-throughs and $5.01 billion of real revenue remains, on which Hilton earned $3.73 billion of adjusted EBITDA. That is a 74.4 percent margin, up from 71.5 percent the year before.

The fees themselves broke down into $2.78 billion of franchise and licensing fees, $376 million of base management fees, and $313 million of incentive fees. Capital spending on property, equipment, and software came to $185 million, less than five percent of EBITDA. Outside of software and index providers, few businesses run on so little capital.

Marriott is larger and slightly less pure, but it follows the same logic. It counted about 1.81 million rooms across 10,082 properties at the end of June 2026, earned $5.4 billion in gross fees in 2025, and posted $5.38 billion of adjusted EBITDA. For 2026 it guides to $6.03 to $6.06 billion in fees and up to $6.03 billion in EBITDA.

Growth Without Pricing: 2025 as the Proof

The strongest case for the model came, fittingly, from a weak year. Hilton’s system-wide revenue per available room rose a meager 0.4 percent in 2025, and in the fourth quarter Marriott’s U.S. and Canada RevPAR actually slipped. A typical owner probably had a down year in real terms, because wages and insurance rose faster than room rates. Hilton’s adjusted EBITDA still grew 8.6 percent, from $3.43 billion to $3.73 billion.

The difference is net unit growth. Hilton expanded its system by 6.7 percent in 2025, and every new room starts paying fees on opening day without Hilton contributing a dollar to construction. The pipeline stood at 541,300 rooms at the end of June 2026 — equal to 40 percent of the existing system — and almost half of it is already under construction. Marriott reports 629,000 pipeline rooms, 44 percent of them under construction or pending conversion.

That produces a growth algorithm analysts can almost run on autopilot: six to seven percent more rooms, two to three percent more revenue per room, a bit of margin expansion because headquarters need not grow in step — and on top of that a four to five percent lower share count from buybacks. The result is earnings-per-share growth in the low-to-mid teens in an industry that itself is barely growing. Hilton guides 2026 adjusted EBITDA to $4.04 to $4.08 billion and adjusted diluted EPS to $8.89 to $9.01.

The second quarter of 2026 confirmed the pattern at a higher level: Hilton’s RevPAR rose 3.9 percent, and Marriott reported 3.4 percent worldwide and 5.0 percent in the U.S. and Canada — a quarter that also saw the World Cup kick off across the U.S., Canada, and Mexico. Internationally, though, Marriott’s RevPAR slipped 0.5 percent. The model is not immune to the economy. It simply transmits the economy into earnings through a much smaller lever.

The Second License: When the Credit Card Grows Faster Than the Hotels

One revenue stream deserves special attention because it has nothing to do with hotels at all. Hilton Honors counted about 243 million members at the end of 2025 and more than 260 million by spring 2026; Marriott Bonvoy is in the same range. The companies rent that membership base to banks: American Express issues the Hilton cards, JPMorgan Chase and American Express the Marriott cards. The bank buys the points its cardholders earn on spending and pays the hotel group a licensing fee.

Marriott discloses the numbers. Co-branded card fees rose just over 8 percent to $716 million in 2025. For 2026 the company expects a jump of roughly 35 percent to close to $1 billion, driven by higher card spending and a higher royalty rate, and the new U.S. agreements with Chase and American Express are expected to add another $100 to $125 million a year by 2028. Set that against the fee guidance and roughly $250 million — close to 40 percent — of the roughly $640 million in additional fees Marriott expects in 2026 comes from cards rather than from a single extra room night.

For investors that is both an opportunity and a warning. An opportunity, because the card business is even more capital-light than franchising and grows with consumer spending rather than travel volume. A warning, because a growing share of profit depends on a points currency the companies themselves control — and on owners who must house guests redeeming free nights from that currency while being reimbursed at only a fraction of the regular room rate. Owner associations have been complaining louder about this in recent years, and it previews the conflict described two sections below.

Negative $6.3 Billion of Equity: Why the Balance Sheet Runs Backward

Open Hilton’s balance sheet and you find an item that would set off alarms at most companies: shareholders’ equity stood at negative $6.3 billion at the end of June 2026, and Marriott’s at negative $4.5 billion. On paper, both companies owe more than they own. Both still refinance in the bond market without difficulty.

The paradox dissolves once you see where the negative equity comes from. It was not created by losses but by buybacks that, over many years, exceeded earnings. In 2025, Hilton repurchased 12.5 million shares at an average of $253.71 and, including dividends, returned $3.3 billion to shareholders. In the same year its net debt rose from $9.86 billion to $11.49 billion — an increase of $1.63 billion, or roughly half of what was paid out. For 2026, Hilton plans about $3.5 billion in capital returns and Marriott more than $4.5 billion.

The logic: a company that needs almost no capital can distribute all of its free cash flow, and as earnings grow, so does the debt it can carry. Hilton targets net debt of roughly three times EBITDA — 3.1 times at the end of 2025 — and each year borrows whatever is needed to hold that ratio as EBITDA rises. The additional debt goes entirely into buybacks. The result is a steadily shrinking share count: Hilton had about 297 million shares outstanding at the end of 2017 and 225 million in July 2026, a 24 percent reduction. Marriott’s count fell from 353 million to 261 million over the same span, down 26 percent.

Negative equity is therefore not a sign of weakness but a reflection of the fact that the most valuable assets these companies have — brands, contracts, loyalty programs — appear on no balance sheet. It is, however, leverage. A company carrying three times its operating profit in debt is less relaxed when that operating profit collapses. How relaxed, 2020 showed.

The 2020 Stress Test: What a Total Shutdown Does to Brand and Building

No stress test is harsher than a pandemic that grounds global travel. U.S. RevPAR fell 47.5 percent in 2020 according to CoStar — the worst year on record and nearly three times the 16.7 percent drop of 2009. Hilton’s system-wide figure fell 56.7 percent.

Hilton’s adjusted EBITDA dropped from $2.31 billion in 2019 to $842 million in 2020, a 64 percent decline — slightly worse than RevPAR, because incentive fees all but vanished, headquarters costs kept running, and the handful of leased hotels lost money. The bottom line showed a $720 million net loss. But the operating business stayed positive in every quarter, Hilton never had to issue equity, and the system actually kept growing — by 47,400 rooms, or 5.1 percent, because hotels financed before the pandemic simply got finished.

At Park Hotels, the spin-off that took the buildings, the same year looked very different. RevPAR fell 95.9 percent in the second quarter, and adjusted EBITDA was deeply negative at minus $122 million in the second quarter and minus $89 million in the third. The company suspended its dividend, renegotiated covenants, and issued expensive secured debt. Its stock fell 81 percent from the end of 2019 to the trough and has never recovered.

What the market did at the moment of panic is instructive, though. Marriott fell 61 percent to its low, more than the owner Host at 50 percent. In the fear, the market barely distinguished brand from building. The difference only emerged in the recovery: the brand companies were at new highs within about two years, while the owners took far longer or never came back. The lesson for investors: the asset-light model does not protect you from the crash. It protects you from permanent loss of capital.

Where the Wall Is Cracking: Key Money, Owner Margins, and New Distribution

As robust as the model is, it has a weak point built into the growth algorithm itself: the brands need tens of thousands of new rooms every year, and someone else builds them. As long as owners earn a decent return on a Hampton or a Courtyard, they sign up willingly. When they don’t, the brands have to sweeten the deal — and that is happening more and more.

The tool is called key money, or in corporate language contract acquisition costs: payments that persuade an owner to sign, often structured as interest-free loans that burn off over the contract term. Hilton spent a net $231 million on them in 2025, up from $105 million the year before — more than double, which the company attributes to the timing of certain strategic developments. Marriott plans $1.25 to $1.35 billion of investment spending in 2026, which includes such owner incentives alongside its own projects. Hyatt reported $49 million of key money payments in the first quarter of 2026, up 172 percent year over year. Marriott stresses that key money per deal remains below 2019 levels. But the direction is clear: brands increasingly have to bring money to the table in order to grow. That does not make asset-light asset-heavy — but it makes it less light than the balance sheet suggests.

The second pressure point is owner margins. CoStar and Tourism Economics now forecast U.S. RevPAR growth of 4.4 percent for 2026, but only 2.1 percent for 2027 — and gross operating profit per room growth of just 1 percent in 2027, because expenses keep rising faster than inflation. Since fees are levied on revenue, they grow faster than owner profit in that kind of environment. The brand’s slice of the pie grows precisely when the owner’s pie is shrinking. That is sustainable only as long as the brand delivers more in rate and occupancy than it costs.

The third pressure point is distribution. Much of a big brand’s value to an owner lies in guests booking direct with Hilton or Marriott instead of through online travel agencies that charge commissions of 15 to 20 percent. If AI assistants start booking trips and optimize for price and reviews rather than loyalty points, that edge could narrow. Then again, the largest loyalty programs, with hundreds of millions of stored profiles, may be exactly the data layer such assistants plug into. The outcome is open — and it is the most honest question mark in the whole model.

The pure franchisors in the economy segment show that asset-light alone is not a moat. Wyndham and Choice Hotels own even less real estate than Hilton, yet trade at only about 14 times EBITDA and sit 23 and 18 percent below their 52-week highs. Their guests are price-sensitive, their loyalty programs weaker, and their owners are the most exposed to rising costs. The moat is not the absence of real estate; it is the combination of network, loyalty program, and pricing power — and that is far stronger at the upper end of the market.

Valuation: What Is a Revenue Share Without a Loss Share Worth?

The stock market prices the gap between brand and building quite precisely. The table below shows valuations at the September 25, 2026 close. For Hilton and Marriott, EV/EBITDA is our own calculation based on company guidance for 2026; for the others it is trailing twelve months.

Company	Price	Market cap	EV/EBITDA	P/E	Off 52-wk high
Hilton	$313.96	$70.7B	20.4 (2026E)	35 (2026E)	−12%
Marriott	$352.03	$91.8B	18.0 (2026E)	30 (2026E)	−14%
Accor	€45.89	€10.5B	14.2	17 (fwd)	−12%
Wyndham	$69.99	$5.2B	13.9	13 (fwd)	−23%
Choice Hotels	$101.81	$4.6B	13.9	14 (fwd)	−18%
Host Hotels (owner)	$22.42	$15.6B	11.4	—	−13%
Park Hotels (owner)	$15.43	$3.1B	10.8	—	−5%

The brand costs nearly twice the multiple of the building. Is that too much? It depends on how you value the earnings stream. A hotel owner must plow a meaningful share of profit back into brand-mandated renovations every seven to ten years, so an owner’s EBITDA is worth considerably more than its free cash flow. At Hilton, almost all EBITDA converts to free cash flow, apart from interest, taxes, and the rising key money payments. Add structural room growth of six to seven percent a year — something an owner can only achieve with fresh capital — and an 80 to 90 percent premium on the multiple is not absurd.

At the same time, 35 times 2026 earnings is no bargain, and the stock has already fallen 12 percent from its $358 high without any deterioration in the numbers. Returns over the next few years depend less on the business than on the multiple the market is willing to pay. A scenario analysis for Hilton based on expected 2028 EBITDA, valued at the end of 2027, makes the point.

Hilton scenario	Assumptions	2028 EBITDA	Multiple	Implied price	vs. $313.96
Bear	U.S. recession in 2027, RevPAR −10%, 5% unit growth, slower buybacks	$4.0B	15x	~$221	−30%
Base	Rooms +6%, RevPAR +2.5%, $3.5B buybacks a year	$4.8B	18x	~$348	+11%
Bull	Rooms +7%, card fees accelerate, margin keeps expanding	$5.1B	21x	~$451	+44%

The math assumes net debt of $13.5 to $15.5 billion depending on the scenario and 203 to 210 million shares. The result is not spectacular, but it is revealing: in the base case, earnings per share keep growing at a double-digit pace, yet the stock gains only about 11 percent because the multiple drifts lower. The shares are not a bet on the business but on the market continuing to prize the quality of the fee stream this highly. In the bear case the price falls by nearly a third even though Hilton faces no operational crisis, just an ordinary recession. At the current price, the risk-reward is roughly balanced rather than obviously attractive. It becomes compelling when a growth scare pushes the stock well below 17 times EBITDA while the pipeline keeps filling.

What It Means for U.S. and International Investors

For U.S. taxable investors, Hilton and Marriott are unusually tax-efficient holdings, precisely because they pay so little in dividends. Hilton pays 60 cents a year, a yield of 0.2 percent; Marriott yields a bit over 0.8 percent. Nearly all of the return arrives through buybacks and therefore through the share price, which is taxed only when you sell — and then at long-term capital gains rates if you have held for more than a year. The 1 percent federal excise tax on buybacks is paid by the company, not by you. That makes these names natural candidates for a taxable brokerage account, where deferral compounds, rather than for an IRA, where the advantage is wasted.

The owners are the mirror image. Host Hotels and Park Hotels are REITs, and their dividends — 6.5 percent at Park — are mostly taxed as ordinary income, softened by the 20 percent qualified business income deduction under Section 199A. Those payouts are better sheltered in a tax-advantaged account. For investors outside the U.S., the usual 15 percent treaty withholding on dividends is almost irrelevant for Hilton and Marriott given their tiny yields, but it bites harder on the high-yielding REITs.

For a portfolio meant to be held through a full cycle, the record since 2017 points clearly to the brand. Investors who want the model with less U.S. concentration can look at Accor in Paris, which has run largely asset-light since selling AccorInvest and trades at about 14 times EBITDA with a 2.9 percent dividend yield, albeit with a weaker loyalty program and heavier exposure to Europe and the Middle East. Britain’s IHG, parent of Holiday Inn and InterContinental, follows the same playbook and also carries negative equity. For a tactical bet on a hotel upcycle after a sell-off, the buildings offer more leverage — but only for investors prepared to sit through another 2020.

Bottom Line: The Best Business in Hotels Is Not Owning One

Hilton and Marriott split the hotel business in two and kept the better half. Owners carry the capital, the debt, the staff, and the cycle; the brand collects about ten percent of room revenue, rents its loyalty program to banks on the side, and grows six to seven percent a year without laying a single brick. The result is a 74 percent margin, a shrinking share count, and returns that have outrun those of the buildings many times over since 2017.

The weak spots are not on the balance sheet, negative equity notwithstanding, but in the relationship with owners. Rising key money, squeezed owner margins, and the question of who owns the guest in a world of AI booking agents will decide whether the pipeline is as full in five years as it is today. As long as it is, this remains one of the most dependable growth businesses in the stock market. But at 20 times EBITDA, the price of that dependability is already high. The better entry point will not come in a good year — it will come in the next recession scare, when the market once again throws brand and building into the same basket.