Three Hotel Tapes, One Trade: MAR Up, HLT Down, H Sideways

I spent the first week of November with three earnings transcripts on my desk and a single ratio sheet between them. By Friday close on November 7th I had circled the same three numbers in red three different times — Marriott’s RevPAR (+0.5%), Hilton’s RevPAR (-1.1%), Hyatt’s RevPAR (+0.3%) — and convinced myself the consensus framing of “U.S. business travel is soft, everybody’s the same trade” was wrong by at least one full multiple of EV/EBITDA. These three tapes are not one story. They are three.

The contrarian thesis, stated plainly so you can disagree with it from the second paragraph: the asset-light economics favor Marriott into 2026 — better fee mix, better capital return, the only one of the three to print positive U.S. RevPAR. But the pipeline data, on a forward-look basis that the sell-side keeps under-weighting, slightly favors Hilton. That tension — asset-light says MAR, pipeline says HLT — is the trade. Hyatt is a separate problem with a Caribbean accent.

The Three Tapes, Side by Side

Strip out the management narrative and Q3 leaves you with a ratio sheet that fits on a napkin.

Marriott printed system-wide RevPAR up 0.5%, Q3 adjusted EBITDA of $1.349B (+10% YoY), and ended the quarter with a development pipeline of approximately 3,923 hotels and 596,000 rooms — pipeline up 5% YoY. Net rooms growth came in around 5%. The U.S. & Canada segment did the heaviest dragging, with RevPAR roughly flat-to-negative on a comp basis, but International more than offset it. Management’s framing — covered in our Nov 4 piece on the Marriott print — was unusually confident on unit growth visibility into 2026. The buyback program ran hard.

Hilton printed system-wide RevPAR down 1.1%, Q3 adjusted EBITDA of $976M (+8% YoY), and a pipeline of 515,400 rooms — up roughly 5% YoY. Net rooms growth was approximately 7.5%. They also used the quarter to announce their 25th brand, Outset Collection, an upper-midscale conversion play we covered the day it dropped. Read carefully: Hilton grew EBITDA 8% while RevPAR contracted. That’s the fee model working at scale, and it is the single most important sentence in this column.

Hyatt is its own animal in Q3 2025 because of the Playa integration. System-wide RevPAR was up 0.3%. Reported net rooms growth was 12.1% YoY — but that headline number is doing a lot of work, because it includes the inorganic step-up from the Playa all-inclusive portfolio acquisition. The Q3 release showed a Q3 GAAP net loss of approximately $49M, with management pointing to integration costs and asset-disposition timing. Pipeline came in around 141,000 rooms, up 4.4%. Our Nov 6 read on the print noted that “noise” was the operative word.

So: one company grew RevPAR, fees, and EBITDA. One company shrank RevPAR but grew fees and EBITDA. One company barely moved RevPAR, took a GAAP loss, and bolted on a different business model. Three tapes.

Why Asset-Light Math Still Favors MAR

The argument for Marriott as the cleaner 2026 owns-the-fee-stream story comes down to four numbers and one structural observation.

First, RevPAR direction. A 0.5% beat is not heroic, but it is the only positive sign in the U.S.-exposed comp set, and it matters disproportionately because incentive management fees (IMFs) are non-linear on RevPAR. IMFs kick in once owner returns clear hurdle rates. Even modestly positive RevPAR in the U.S. resort and luxury segments — where IMFs are denser — pulls the fee line up faster than a sell-side model with linear assumptions will pick up. Hotel Dive’s recap of the Marriott print quoted CFO Leeny Oberg explicitly on incentive-fee leverage; it is not a small piece of the 10% EBITDA beat.

Second, capital return. Marriott has the cleanest balance sheet of the three and used Q3 to return roughly $1.0B to shareholders between buybacks and dividends — and management reaffirmed full-year guidance of $4.0B+ in return. On a market cap of roughly $70B that is a ~5.7% capital-return yield, of which the buyback portion is the dominant share. Hilton’s program is comparable in absolute dollars but on a larger denominator; Hyatt’s is constrained by the Playa absorption and the disposition pipeline.

Third, mix. Marriott’s fee mix sits at approximately 60% base, 20% incentive, 20% franchise/other across the segments — and the incentive bucket is where you get convexity. In a “flat-ish” RevPAR world, the company with the highest IMF gearing wins.

Fourth, unit growth. 596,000 rooms in the pipeline is the largest absolute pipeline in the public hotel C-corp universe, and 5% YoY pipeline growth is matched only by Hilton. Conversions — a particular Marriott strength in 2025 — are running at a pace that pulls forward fee revenue without taking the construction-financing risk owners are increasingly unwilling to absorb.

The structural observation: when you’re long the fee model in a soft-RevPAR environment, you want the company whose contractual fee floors hold up best and whose incentive fees can re-rate fastest on the way back. Marriott’s portfolio mix — heavier luxury and full-service share — gives it both. That is why the asset-light play favors MAR for 2026.

Why the Pipeline Slightly Favors HLT

Now I’m going to argue against my own conclusion, because the pipeline data tells a different story.

Hilton’s pipeline grew approximately 5% YoY to 515,400 rooms, but the more important number is the ratio: pipeline as a percentage of existing system. On Hilton’s roughly 1.27M-room base, that is a 40-41% pipeline-to-system ratio. Marriott’s pipeline-to-system ratio is closer to 32-33% on a roughly 1.74M-room installed base. Hyatt’s is approximately 22% on its post-Playa base.

What does that ratio mean operationally? It means Hilton has more committed forward fee revenue per existing key than Marriott does. In a fee-stream DCF, that translates into a higher terminal-value contribution from new units, holding RevPAR assumptions constant. The market is currently paying roughly 17-18x forward EBITDA for HLT and roughly 19-20x for MAR — that is, the market is paying a premium for MAR’s near-term IMF gearing and discounting HLT’s longer-duration unit-growth annuity.

Add the Outset Collection launch. A 25th brand is not a vanity exercise; it is a deliberate move into upper-midscale conversion supply, which is the fastest-growing pipeline segment in the U.S. right now because owners cannot finance new-builds at current cap rates. Hilton signaled in Q3 that conversions are running at a record share of openings — call it 35-40% of new keys. Conversion economics for the franchisor are excellent: no construction risk, faster fee onset, and a much lower hurdle for the owner to clear before the IMF kicks in.

Now the inconvenient bit. Hilton printed RevPAR -1.1% while still delivering 8% EBITDA growth. That is the model working. The sell-side will spend the next two quarters arguing about whether the RevPAR weakness is U.S.-specific (it is, mostly) and whether it persists (probably one more quarter, then improves on easy comps). But the pipeline says: it doesn’t matter. The fee annuity is being built regardless.

So the pipeline math whispers HLT and the in-quarter operating math shouts MAR. Reasonable people can hold both views and pair-trade them — long MAR / long HLT against a basket short — but the cleaner directional read is: own MAR for 2026 IMF leverage, own HLT for 2027-2028 unit-growth optionality.

The Hyatt Problem: Read the Footnotes Twice

Hyatt is the position I have not put on yet, and I want to explain why even though the surface-level metrics — net rooms +12.1%, pipeline +4.4% — look attractive.

The 12.1% net rooms number is mostly the Playa absorption. Strip it out and organic net rooms growth is somewhere in the 6-7% range, which is still good — better than MAR’s 5% — but it is not what the headline implies. The $49M Q3 GAAP net loss reflects integration costs, asset-disposition timing, and purchase-accounting noise that should normalize over the next two-to-three quarters. None of that is fatal.

What concerns me is the asset-disposition execution. Hyatt’s strategy is to recycle Playa’s owned all-inclusive real estate into a managed/franchised fee stream — to do, in effect, what Marriott did with Starwood’s owned hotels a decade ago. That worked for Marriott because the buyers were there and the cap-rate environment was friendly. The cap-rate environment in 2025-2026 is not friendly. If Hyatt cannot dispose of the Playa real estate at the prices they underwrote, the math on the deal gets harder fast.

The optionality is real. If they execute, Hyatt re-rates from a hybrid owner/franchisor multiple toward a pure-franchisor multiple, and that’s worth several turns of EBITDA. If they don’t, you’ve paid for a Caribbean real-estate book at the wrong point in the cycle. I am not yet convinced enough either way to take the position.

What I’m Watching Into Q4 and 2026

Four things on the watch list, in order of importance.

One: the U.S. RevPAR comp in Q4. If Hilton’s -1.1% widens, the pipeline thesis gets harder because owners stop signing new deals when current-system RevPAR is contracting. If it narrows, the HLT setup into 2026 gets meaningfully better. The Smith Travel weekly data through mid-November is showing stabilization, not further weakness, which is mildly positive for HLT.

Two: Marriott’s incentive-fee disclosure in the Q4 print. If IMFs grew double-digits in Q3 on flat U.S. RevPAR, the operating leverage into any U.S. RevPAR recovery is dramatic. The technology and loyalty investments — which we’ll return to in a forthcoming May piece on Marriott’s AI deployment — are a separate flywheel that should compound the fee-mix story over a multi-year horizon.

Three: Hyatt’s asset-disposition cadence. Watch the 8-K filings between now and February. Each tranche sold tells you what the next tranche is worth.

Four: brand launches and conversions. Hilton’s Outset Collection is the model. Expect Marriott to respond with something analogous in the next two quarters — they have white space in upper-midscale conversion that they have publicly acknowledged. A Marriott counter-launch would compress the HLT pipeline-ratio advantage I described above.

Bottom line: three earnings tapes, three different stories, one trade. Own MAR for the IMF convexity and the cleanest 2026 fee math. Layer HLT for the pipeline annuity and the conversion-supply structural tailwind. Leave Hyatt on the watch list until the Playa disposition prints. The market is currently pricing all three closer together than the fundamentals justify, and that is where the alpha is.

— Oliver writes The Bottom Line for TableTransfers. Tips: [email protected].