In Pershing Square's materials, the rhetoric is relentless: high quality, profit growth, free cash flow generation, operating leverage, capacity to return capital, and a preference for models where recurring usage and supply density enhance platform attractiveness; in short, exactly what powers a network effect. Even the Uber example - now Bill Ackman's top position - is presented as a marketplace where growth is driven by acquiring new customers and increasing frequency amongst existing ones, with robust free cash flow enabling massive returns to shareholders.
If we apply this Bill Ackman framework to Airbnb, the question is whether the company ticks the boxes of a high-quality compounder with a cumulative advantage and credible levers to further improve return on capital.
Airbnb today: a marketplace already highly profitable
The quarter is solid and, above all, very "Ackman-compatible" in the structure of the numbers. Airbnb reported Q3 2025 revenue of $4.1bn (+10% y-o-y), GBV of $22.9bn (+14%), and 134 million nights & seats (+9%). Profitability impresses: adjusted EBITDA of $2.1bn, a 50% margin, and net income of $1.4bn (EPS: $2.21), despite a one-off tax item of $213m. Over 12 months, free cash flow reached $4.5bn, a 38% margin, with $11.7bn in cash and investments. And the company is not just accumulating: it repurchased $857m of shares over the quarter, $3.5bn over 12 months, and says it has reduced diluted shares outstanding by 8% since 2022 ($6.6bn of remaining authorization).
For a long-only logic, this is key: profitability is not a distant promise; it already funds strategic optionality.
The engine's core: supply density, trust, conversion
Management at Airbnb says there is no silver bullet, only hundreds of improvements. Q3 counts 65, some directly tackling conversion and friction. The most emblematic is "Reserve Now, Pay Later" (RNPL) in the United States: among users offered the option, 70% adopt it, and management acknowledges an increase in cancellations... but says the net effect on bookings remains positive.
This is typical of a mature platform that knows where growth hides: not in a slogan, but in the microeconomics of the booking funnel. Note the re-acceleration of nights in North America and an ADR up ($171, +5% in Q3 2025), helped by mix and FX. They also note app bookings rising to 62% (a record), a sign that the direct relationship (and thus margin) is strengthening.
Airbnb reminds that 90% of its traffic is direct, a precious characteristic in a world where AI could shift the top of funnel and make acquisition more expensive. In other words, even if distribution is reshaped, Airbnb starts with a base of proprietary traffic that few travel players can claim.
| Fiscal Period: December | 2020 | 2021 | 2022 | 2023 | 2024 | 2025 | 2026 | 2027 |
|---|---|---|---|---|---|---|---|---|
| Net sales 1 | 3,378 | 5,992 | 8,399 | 9,917 | 11,102 | 12,174 | 13,424 | 14,767 |
| Change | - | 77.37% | 40.18% | 18.07% | 11.95% | 9.65% | 10.27% | 10.01% |
| EBITDA 1 | -250.7 | 1,593 | 2,903 | 3,653 | 4,041 | 4,277 | 4,757 | 5,346 |
| Change | - | 735.65% | 82.21% | 25.82% | 10.62% | 5.85% | 11.22% | 12.38% |
| EBIT 1 | -3,379 | 429.3 | 1,802 | 1,518 | 2,553 | 2,618 | 3,041 | 3,551 |
| Change | - | 112.71% | 319.74% | -15.76% | 68.18% | 2.56% | 16.14% | 16.79% |
| Interest Paid 1 | -171.7 | -437.6 | -24 | -83 | - | -69.2 | -23.33 | -20 |
| Earnings before Tax (EBT) 1 | -4,682 | -300.2 | 1,989 | 2,102 | 3,331 | 3,258 | 3,710 | 4,266 |
| Change | - | 93.59% | 762.54% | 5.68% | 58.47% | -2.19% | 13.87% | 15% |
| Net income 1 | -4,585 | -352 | 1,893 | 4,792 | 2,648 | 2,585 | 2,993 | 3,439 |
| Change | - | 92.32% | 637.73% | 153.14% | -44.74% | -2.4% | 15.79% | 14.93% |
| Announcement Date | 2/25/21 | 2/15/22 | 2/14/23 | 2/13/24 | 2/13/25 | - | - | - |
The three free options: international expansion, "beyond stays" and hotels
- First lever: international expansion, described as a multi-year strategy. Over 12 months, expansion markets are growing at twice the pace of core markets, with first-time bookers up +20% in Japan and nearly +50% in India. It looks like a country-by-country replay of the historical densification playbook.
- Second lever: Services and Experiences, launched in May. Ratings are very high (4.93/5), supply is flowing in (110,000 host applications), and usage extends beyond travel with lodging: nearly half of experience bookings are not paired with a stay, and 10% of services customers had never booked on Airbnb. This is exactly the kind of frequency extension that fuels a network effect: more reasons to come, more often, which attracts more supply, which boosts attractiveness (the classic virtuous cycle of marketplaces).
- Third lever: hotels, in pilot phases in Los Angeles, New York, and Madrid. The goal is not to become just another OTA, but rather to fill supply gaps, notably in urban markets constrained by regulation. Management stresses the incremental nature (low cannibalization), ease of onboarding, and the ace: demand is already there. The share of hotels remains small but is growing faster, which could matter in the growth algorithm in 2026-2027.
AI as a margin catalyst and a moat
Airbnb is integrating AI to modernize the user journey. The customer support AI assistant has already reduced the need to contact a human agent in the US by 15%, and the company is targeting over 50 languages next year. Then comes conversational search: first free text (what box), then a multi-turn experience, followed by an integrated concierge from top of funnel through post-stay.
On potential AI competition, management argues the advantage will not be having the best model (all are accessible via API), but building the interface, product data, messaging, verified identity, and payments. In short, the proprietary architecture that turns a generic AI into a specialized product. That is how a moat is modernized.
A compounder with visible levers
In this reading, Airbnb looks less like a travel stock than like a trust and distribution infrastructure. It is a capital-light marketplace already printing a 35% adjusted EBITDA margin annually (and 50% in Q3), converting massively into cash, buying back shares, and stacking growth options funded from internal resources. Debates exist on the pace of ramp-up (3 to 5 years for Services/Experiences to become material), and some fear a relative lag in the AI race versus competitors partnering with OpenAI. But Airbnb answers exactly where a long-only investor listens: execution discipline, international densification, incremental diversification (hotels) where supply is lacking, and AI as a lever on conversion and costs.
All this while the valuation is historically attractive: 2025e P/S of 5.7x versus a historical average of 12.2x, 2025e EV/EBITDA of 16.2x versus an historical average of 29.3x, and a current estimated FCF yield of 6.6%.
If Bill Ackman were looking for a new name, he might see the combination he favors: a global brand, a network effect fueled by usage frequency, a cash engine already mature, and potential to improve margins and return on capital as AI, monetization beyond stays, and geographic expansion increase value per user. The kind of company that, once installed in a concentrated portfolio, no longer needs to tell a story: it compresses it into numbers, quarter after quarter.
