These three names are compared constantly and almost always on feeling: Four Seasons is reliable, Aman is rarefied, Rosewood is characterful. That is brand positioning repeated back. Three things about them can actually be checked, and all three are more interesting than the positioning: what the independent scoreboards say, what the companies are now built to earn from, and what happens when something goes wrong.
What follows is drawn from company filings, the current editions of the three ratings systems that matter, and published financial reporting. Where a number does not exist, we say so rather than borrow one.
The Short Answer
Four Seasons has by far the broadest reach and a lead that is measurably narrowing. Aman has the highest strike rate on Michelin Keys and not one confirmed Forbes Five-Star hotel. Rosewood has the best hotel in the world two years running and an owner currently selling assets. Rank them by Forbes and you get one answer, by The World's 50 Best a different one, and neither system is wrong.
First, what none of them publishes
Four Seasons, Aman and Rosewood are all privately held. None discloses average daily rate, revenue per available room, occupancy, average length of stay or repeat-guest rate. Not one of them. Any article that gives you a per-brand average nightly rate has invented it or inherited it from somewhere that did.
The last audited operating data Four Seasons ever published came from its life as a listed company. In the first half of 2006, across a core managed portfolio of 56 hotels, it reported an average daily rate of US$369.76, RevPAR of US$262.88 and occupancy of 70.2 per cent. That is the most recent verified figure in the public record, and it is twenty years old. The company went private in 2007 in a buyout worth US$3.8 billion including debt.
For context, the luxury tier of the US hotel market ran an average daily rate of US$419.39 and RevPAR of US$290.42 in the first half of 2026, against US$173.76 and US$120.97 across all US hotels. These three sit above that segment, not within it, but by how much is not a matter of public record.
The scoreboards disagree, and that is the finding
The luxury hotel world runs on three independent assessments: Forbes Travel Guide, which sends anonymous inspectors against a published standard; The World's 50 Best Hotels, which polls an academy of industry voters; and the Michelin Keys, introduced in 2024 and now covering 2,832 hotels worldwide, of which 155 hold the top Three Keys.
They do not agree, and the disagreements are not marginal.
Rosewood Hong Kong has been the number one hotel in the world in both 2025 and 2026, the first property ever to take the top place in consecutive years. Forbes Travel Guide lists it as Recommended. Not Five-Star, not Four-Star. Unrated.
Amangiri and Aman New York each hold Three Michelin Keys, a distinction given to 155 hotels on earth. Forbes rates both Four-Star. Across the Aman properties we checked against Forbes directly, including Aman Tokyo and Aman Venice, every one is Four-Star and none is Five-Star.
Four Seasons holds 42 Forbes Five-Star hotels, more than any other company for the eleventh consecutive year, and takes only nine Three Keys from 91 hotels with any Key at all.
| Measure | Four Seasons | Aman | Rosewood |
|---|---|---|---|
| Properties open, own figure | 135 | 36 | 43 |
| Countries | 47 | 21 | 21 |
| Hotels with a Michelin Key | 91 | 27 | 39 |
| Three Keys | 9 | 7 | 6 |
| Three-Key strike rate | 10% | 26% | 15% |
| Forbes Five-Star hotels | 42 | none confirmed | 11 |
| In the World's 50 Best | 4 | 3 | 4, including No. 1 |
| In the extended 51 to 100 | 2 | 2 | 3 |
| Points programme | none | none | none |
Read that table twice, because the ranking flips depending on which column you weight. On breadth of competence, Four Seasons wins and it is not close. On the probability that any given hotel in the portfolio is exceptional, Aman leads at roughly one property in four holding Three Keys, against one in ten at Four Seasons. On peak achievement, Rosewood holds the single highest-ranked hotel in the world.
Four Seasons is still the broadest, and its lead is narrowing
135 hotels and resorts in 47 countries, 61 residential properties, around 60 projects in development. Nothing else in the category operates at that scale while holding that standard, and the Forbes count says so.
A word on that first figure, because it is the kind of thing that gets repeated without checking. 135 hotels in 47 countries is Four Seasons' corporate boilerplate. Its own destinations directory lists more: we counted 143 properties one by one when we built the full Four Seasons directory. The table above uses the boilerplate figure so that all three columns sit on the same footing, but the property-by-property number is 143.
The trend inside that count says something else. Four Seasons took 73 Forbes Five-Star awards across 51 hotels in 2023. In 2024 it was 66 awards and 45 hotels. In 2025, 65 and 44. In 2026, 62 and 42. It has held the number one company position throughout and lost nine Five-Star hotels while doing it.
The World's 50 Best tells a similar story faster. Four Seasons had seven properties in the top 100 in 2025 and five in 2026, and every returning property fell: Bangkok from second to third, Firenze from ninth to twelfth, Astir Palace Athens from seventeenth to twenty-seventh. The Surf Club and Madrid dropped off the list entirely.
One more number, easy to miss. Four Seasons has been in exactly 47 countries since September 2021, when it operated 121 hotels. It now operates 135. Every net addition over five years has been infill in a market it was already in. The company has said it wants 180 properties by 2033, which from 135 requires roughly six and a half net openings a year. Recent pace has been one to five.
Aman's scarcity is real, and partly accidental
Aman's positioning rests on staying small, and on the numbers it has. Doronin puts the resort average at 35 keys and urban properties under 90. The brand went from 26 hotels in 2014 to 36 in 2026, compound growth of under three per cent a year, at a time when the luxury segment has been the fastest-growing part of the hotel market.
What the positioning does not say is how much of that restraint was chosen. Aman Miami Beach has moved from 2024 to 2026 to 2027 and now sits at 2028 on Aman's own site. Aman Beverly Hills has moved from 2026 to 2028. Janu, the sister brand announced in March 2020 with three hotels promised for a 2022 debut, has one hotel open in late 2026. Aman Sveti Stefan was closed for five years over a public beach access dispute and reopened in July 2026 on terms that give the Montenegrin state ten per cent of resort profits. Amangani has been closed since April 2025. With 34 hotels actually trading, two closures take about six per cent of the network out of service.
The pipeline is where the scarcity argument gets difficult. Aman's own website describes ten announced projects. Bloomberg reported 23 hotel projects underway and a further 39 in pre-development, against 34 open hotels. Both figures trace back to the company. A third brand has been in development since at least 2022.
There is also a structural point a candid advisor should raise, and it is not hidden, merely unusual. Vladislav Doronin is majority owner and chief executive of Aman Group, and separately owns OKO Group, the property developer behind Aman New York and Aman Miami Beach. Aman earns management and licensing fees on hotels his other company builds. Saudi Arabia's Public Investment Fund, Cain International, Mubadala Capital and Alpha Wave have all invested alongside, at a reported US$3 billion valuation. Four Seasons, by contrast, is a pure fee business that has never franchised and holds no material real estate.
Rosewood is growing fastest, and its owner is selling
Rosewood is growing faster than either of the others, and its own published counts cannot quite keep up with it. Its January 2025 fact sheet describes 34 hotels, resorts and residences in 21 countries. Its 2026 fact sheet says 43 in one sentence and 44 in another, and that figure folds in standalone residences. Meanwhile 39 Rosewood properties were recognised in the 2026 Michelin Key selection, six of them at Three Keys. The counting is loose because the portfolio is moving, and the movement is the point: from 26 properties in 2021, roughly twelve to thirteen per cent a year, comfortably the fastest of the three.
The awards have kept pace with the openings rather than lagging behind them. Rosewood Hong Kong became the first hotel to hold the world number one position twice running, and Rosewood São Paulo climbed from twenty-fourth to sixth.
Then there is the owner question, which almost every account of this gets wrong, so it is worth stating precisely.
New World Development, the listed Hong Kong property company, is in genuine difficulty. For the year to 30 June 2026 it reported a loss of HK$26.8 billion, net gearing of 68.3 per cent, HK$23.2 billion of debt due within twelve months, and dividends and perpetual coupons still suspended. In June 2025 it closed an HK$88.2 billion refinancing, reported as a Hong Kong record.
Rosewood Hotel Group is not inside that company. New World Development sold it to Chow Tai Fook Enterprises, the Cheng family's private holding vehicle, in July 2015. The operator is one corporate step removed from the distressed listed entity, and Rosewood has said on the record that its brands are not for sale and its properties remain fully operational.
The real exposure sits at the level of individual buildings, and it is not nothing. Rosewood London was put up for sale by Chow Tai Fook Enterprises in December 2025. Rosewood Hong Kong's building carried an HK$15.9 billion valuation inside a proposed debt swap. The Cheng family has been reported exploring a partial sale of Rosewood hotel assets. A guest booking a specific Rosewood three years out has a real chance the building has a different owner by arrival. Whether a new owner keeps the Rosewood management contract is not disclosed for any of these properties.
All three are now, substantially, property businesses
This is the least discussed and most consequential change in the category.
Aman has sold approximately US$9 billion of branded residences in four years, against a last-reported group valuation of US$3 billion. Doronin himself bought the US$135 million penthouse at Aman New York, the most expensive unit in a building his own development company built. Four Seasons recorded US$2.1 billion of gross residential sales in 2024, up from more than US$1.5 billion in 2023, and 65 per cent of its development pipeline now carries a residential component. Rosewood's residential sales rose 135 per cent in 2023 and it has begun building standalone residences with no hotel attached.
The market explains why. Savills counted 910 branded residence schemes globally at the end of 2025, up 19 per cent in a year and nearly triple the number in 2015, carrying an average price premium of 33 per cent over comparable unbranded stock, rising to 39 per cent in resort markets. Hotel brands account for around four fifths of completed stock.
For a guest this is not abstract. On a mixed-use property the pool, the spa, the gym and the restaurants are shared with residents who live there permanently and who paid a premium for access. The industry's own analysis puts it plainly: a project optimised primarily for residential sales does not necessarily produce the strongest hotel.
The failure mode nobody mentions
All three are management companies. They operate buildings other people own, under contracts that in Four Seasons' case have historically run around fifty years, against a European market norm of ten to twenty. The owner funds everything: operating costs, working capital, and a capital reserve typically set at three to five per cent of gross revenues.
That arrangement works until it does not, and when it fails the guest is the one who finds out.
Four Seasons Hotel New York closed in March 2020 and did not reopen until November 2024. The extended closure was attributed in multiple reports not to renovation but to a dispute between the owner, Ty Warner, and Four Seasons over management fees. A flagship in the brand's most important city was dark for roughly four and a half years. In 2009 the owner of Four Seasons Aviara in California attempted to terminate the management agreement; Four Seasons refused to vacate, put the matter into arbitration, and contemporaneous reporting describes barricades and checkpoints at the resort. It is a Park Hyatt today.
Aman lost Sveti Stefan for five years and Amangani is closed now. Rosewood London is on the market.
None of this makes any of the three a bad booking. It makes the specific property, its owner and its capital position a more useful question than the badge above the door.
None of them has points, which changes how you book
Four Seasons, Aman and Rosewood operate no points programme, no elite tiers and no free-night redemption between them. This is deliberate. Four Seasons frames it as the difference between recognition and rewards, and in practice it means there is no status to accumulate and nothing to spend.
The consequence is that the booking channel, rather than your history with the brand, determines what you receive on arrival. Each of the three runs an advisor programme for exactly this purpose: Four Seasons Preferred Partner, Rosewood Elite, and at Aman the equivalent recognition runs through Virtuoso. A first-time guest booking through one of these arrives with more standing than a tenth-time guest who booked direct, which is an odd fact about the top of the market but a consistent one.
Aman also sells access outright. Aman Club membership at Aman New York is reported at US$200,000 to join plus US$15,000 a year.
What actually decides it
Book Four Seasons when the trip has several stops and you need each one to be predictable, when you are travelling with people whose tolerance for surprise is low, or when you want a city hotel that will simply work. Check the individual property rather than trusting the badge: the range between the best and the weakest Four Seasons is wider than at either of the others, and the brand's own Forbes count has been falling for three years.
Book Aman when the hotel is the destination and you are staying put. The strike rate argument is real: one Aman property in four holds Three Michelin Keys, and the seven that do are Aman Venice, Amangiri, Amanemu, Aman New York, Amanpuri, Amanoi and Amanjiwo. Accept that you are paying a premium the ratings do not fully corroborate, that the resorts are genuinely small and the city hotels are not, and that the brand's next decade will look less like its first three.
Book Rosewood when you want a building with a history and a hotel that could not be lifted and set down somewhere else. It has the best hotel in the world, the fastest growth of the three, and the consistency risk that goes with that growth. Ask about the specific property's ownership if you are booking far ahead.
And in every case, book it through a programme. None of the three will reward you for loyalty, so the only lever left is the one that costs nothing to pull.