Booking Holdings (BKNG) · Deep analysis

Report previewBKNG.USReport date Last updated
← Back to reports
Report contents

At a glance

  • Business model: Booking Holdings earns accommodation commissions from hotels and alternative-accommodation hosts, strongest at the fragmented end of supply where independents cannot generate their own demand. Major chains steering guests to their own sites cap the blended take rate, and because Booking owns no rooms it carries no unsold-inventory risk.
  • Financial quality: Operating cash flow has covered net income 1.4x to 2.4x every year since FY2021, and free cash flow reached $9.54 billion in the twelve months to June 2026 on capex of 1.1% of revenue. The constraint is the balance sheet: equity was negative $10.78 billion and $10.5 billion of repurchases exceeded that free cash flow.
01

Business model

Business model: commissions still come from the search auction

Booking Holdings sells rooms other people own. Booking.com, its largest brand, aggregates hotel and alternative-accommodation inventory and sells it to leisure travelers, with the biggest demand pool in Europe; Priceline serves U.S. demand, Agoda covers Asia-Pacific, KAYAK runs travel metasearch, and OpenTable sells restaurant reservations and software. The property usually pays: Booking keeps a commission on the room rate where it acts as agent and takes the spread where it is merchant of record. Accommodation commissions generate most revenue and nearly all the profit, so three variables decide the economics: room nights, the commission rate (the take rate, meaning revenue per dollar of gross bookings), and the cost of getting a traveler to the site. That makes this a two-sided platform, mature in hotel distribution and still expanding in alternative accommodations, flights and payments.

MetricLatestComparison
Revenue (USD bn)26.92 FY202510.96 FY2021
Gross bookings, Q2 2026~51+9% year over year
Room nights, Q2 2026+5% y/y; Q3 2026 guided +3–5%
Adjusted EBITDA margin, Q2 202636.0%
Marketing / gross bookings, Q2 20264.7%4.6% in Q2 2025
Direct room nightsmid-50s %trailing four quarters

Revenue went from $10.96 billion in FY2021, as reported in that year's filing, to $26.92 billion in FY2025, per the 10-K filed in February 2026. The starting year was pandemic-depressed, so the four-year rate says little about underlying demand. The current split is more informative: room nights grew 5% in Q2 2026 while gross bookings grew 9% to about $51 billion, which implies higher average booking values from room rates, length of stay or destination mix. Booking does not decompose that gap, and with a single reportable segment it does not report revenue or profit by brand, so mix has to be inferred from commentary.

Pricing power holds where supply is fragmented

Booking's leverage comes from the shape of its supply base. Independent hotels and small regional chains cannot generate enough direct demand to fill rooms, and a listing on Booking.com reaches buyers in countries they cannot afford to advertise in. That is why a commission on the room rate survives even as properties complain about it. The major chains are the exception: they have loyalty programs and the balance sheet to buy their own search traffic, and they have spent a decade steering guests to their own sites. Where a chain can fill a room itself, the commission becomes a marketing expense to avoid. Pricing power is strongest at the fragmented end of the market and weakest at the top, and the mix between the two moves the blended take rate as much as any negotiation.

The merchant model is the other moving part. When Booking takes the payment it holds cash until the stay and absorbs chargebacks, fraud and cancellation risk; the agency model leaves that risk with the property but gives Booking less control of the customer relationship and no payment fee. Which way that mix shifts affects revenue per booking, the size of the customer-prepayment float, and how much of the transaction Booking actually owns.

Acquisition cost is set in Google's auction, not in Amsterdam

Travelers shop by searching, and the paid-search auction prices a click before anyone knows whether it converts. Marketing was 4.7% of gross bookings in Q2 2026 against 4.6% a year earlier; on roughly $51 billion of quarterly bookings, that near-flat ratio is still an enormous check, most of it written to Google. The defense is habit. Direct and app bookings have held in the mid-fifties percent of room nights over the trailing four quarters, and repeat guests arrive without an auction. That conversion is the asset behind the 36.0% adjusted EBITDA margin Booking reported for Q2 2026. The $650 million of annual run-rate savings management targets from its Transformation Program by the end of 2027 buys margin but does not make traffic cheaper.

Flights and payments are the expansion bets. They add booking events that can feed accommodation demand, but neither is broken out enough in the filings to test whether the cross-sell actually happens.

Alternative accommodations are the contested flank

Airbnb built a supply base Booking cannot easily copy: individual hosts with one or two properties, often in leisure destinations where hotels are thin. Booking's answer is to list those hosts beside hotels so a traveler comparing a city hotel with a coastal villa does not leave. Management discusses alternative-accommodation nights on earnings calls but does not publish a standard series, so the contest shows up mainly in Booking's total room nights against Airbnb's disclosed nights. Room nights growing 3–5%, as guided for Q3 2026, keep Booking in the game; whether it is gaining share in rentals is not something the filings settle.

Booking owns no rooms, which cuts both ways. It carries no unsold-inventory risk and needs little capital, so a demand shock hits revenue almost immediately without leaving unused capacity behind. It also cannot set the room rate. When demand softens, the levers Booking controls are its own commission and its marketing spend.

What would overturn this

The judgment is that Booking is a mature commission business with real pricing power over fragmented hotel supply and a structural cost of demand set by search advertising. Two observations would break it. If growth of 3–5% in room nights starts requiring marketing well above 5% of gross bookings, the commission is being taxed at the point of acquisition and the take rate overstates what the platform keeps. If direct and app mix keeps climbing while marketing intensity stays near 4.7%, Google is less of a toll booth than the paid-search exposure implies, and habit is carrying the margin. The disclosure that would settle either point, such as take rate by brand or region, repeat-guest share, or rental economics, is not published, and that gap is itself part of the risk.

Questions worth following

  • Will Booking's mix shift toward Agoda and alternative accommodations hold the blended take rate below its historical level?

    Agoda and alternative accommodations are the fastest-growing parts of the supply base but typically earn less per room night than the European hotel core. If that mix shifts faster than core economics improve, revenue grows more slowly than room nights and margin compresses.

  • Can Booking hold its adjusted EBITDA margin near 36% as marketing and cloud costs rise?

    Adjusted EBITDA margin was 36.0% in Q2 2026. Holding that depends on marketing intensity staying near 4.7% of gross bookings and on about $650M of Transformation Program savings offsetting higher cloud computing, software, and foreign-exchange costs.

  • Can Booking's alternative-accommodation supply keep pace with Airbnb?

    Alternative accommodations are the one segment where Booking's hotel-led funnel meets its strongest rival. If host supply and rental nights do not keep pace, travelers who want a villa book elsewhere and Booking's share of leisure demand stays capped at the hotel end of the market.

02

Financial quality

Financial quality: strong cash conversion, stretched payout

Booking Holdings turns profit into cash well above one-for-one and earns high returns on a small asset base. The strain sits in the capital structure: repurchases now edge past free cash flow, shareholders' equity is negative $10.8 billion, and much of the $17.2 billion cash balance is money held for hotel partners. The earnings look high quality. The buyback is the part that depends on everything continuing to go right.

Five years of cash-backed returns
Metric (USD bn unless noted)FY2021FY2022FY2023FY2024FY2025LTM Jun 2026Assessment
Revenue10.9617.0921.3723.7426.9228.24Strong
Revenue growth (%)—55.925.011.113.412.9Strong
Operating income2.505.105.837.559.289.74Strong
Operating margin (%)22.829.827.331.834.534.5Strong
Net income1.173.064.295.885.407.21Strong
Net margin (%)10.617.920.124.820.125.5Strong
ROA (%)5.112.517.322.618.524.3Strong
ROE (%)n/mn/mn/mn/mn/mn/mNot meaningful
Total liabilities17.4622.5827.0931.7334.8440.46Weak
Total equity6.182.78-2.74-4.02-5.58-10.78Weak
Liabilities/assets (%)73.989.0111.3114.5119.1136.3Weak
Cash11.1312.2212.1116.1617.2017.21Strong
Operating cash flow2.826.557.348.329.419.86Strong
Free cash flow2.526.197.007.899.099.54Exceptional
OCF/net income (x)2.412.141.711.411.741.37Exceptional
Capex/revenue (%)2.72.11.61.81.21.1Strong
Interest expense0.330.390.901.291.621.10Adequate

n/m = not meaningful. The last column covers the twelve months to June 2026, built from four reported quarters so the flows compare with full years; balance-sheet ratios use period-end assets.

ROE carries no information here. Buybacks pushed equity below zero, which turns the ratio into noise (22,574% for FY2023, negative for FY2024). ROA is the usable measure and has held between 17% and 24% since FY2023.

One figure needs explanation. In FY2025 operating income rose 23% to $9.28 billion while net income fell 8% to $5.40 billion. Interest expense, up $0.33 billion, and a flat tax charge account for only a small part of that $2.2 billion swing, so roughly $1.5 billion of it must come from an item outside those lines. Until it is identified, FY2025's 20.1% net margin reads as depressed, not as a new base.

Quarterly cash follows the booking calendar
USD bnQ1 2025Q2 2025Q3 2025Q4 2025Q1 2026Q2 2026AssessmentTrend
Revenue4.766.809.016.355.537.35StrongDecelerating
Operating income1.062.253.942.031.272.50StrongMargin steady near 34%
Net income0.330.892.751.431.081.95StrongImproving
Net margin (%)6.913.130.522.519.526.5StrongImproving
Total assets27.1930.6828.7529.2627.7229.68AdequateFlat
Total liabilities33.3037.3433.4934.8436.4440.46WeakRising
Total equity-6.11-6.66-4.74-5.58-8.72-10.78WeakEroding
Cash15.5817.5916.5117.2016.0217.21StrongFlat
Goodwill2.822.852.672.672.662.67AdequateStable
Operating cash flow3.283.201.441.493.213.72StrongSeasonal
Free cash flow3.163.141.371.423.113.64StrongSeasonal
OCF/net income (x)9.93.60.521.043.01.9AdequateWorking-capital driven

Net PP&E and a current-asset split are not broken out for these quarters; at 1.1% of revenue, capex makes them immaterial.

Six quarters are enough to show the pattern. Operating cash flow ran well ahead of net income in the first half of each year, when customers prepay stays, then dropped in Q3 as prepayments unwound: $3.20 billion of operating cash against $0.89 billion of net income in Q2 2025, then $1.44 billion against $2.75 billion in Q3 2025. Over four quarters the ratio is 1.37x and it has sat between 1.4x and 2.4x every year since FY2021. Annual conversion is genuine; the quarterly swings are timing.

Returns, float and the cost of the buyback

Capex at 1.1% of revenue means nearly all operating cash becomes free cash flow, $9.54 billion or 34% of revenue in the twelve months to June 2026. That is why ROA looks high: the asset base is receivables and customer balances, not plant, and goodwill of $2.67 billion has barely moved since FY2024.

Negative equity is a buyback artifact more than a solvency signal. Liabilities of $40.46 billion against $29.68 billion of assets sounds extreme until the composition is clear. Current liabilities of $21.07 billion are largely supplier and customer balances, with $17.21 billion of cash sitting against them. But that cash is float: it moves with bookings, and coverage of current liabilities has slipped from 0.94x in June 2025 to 0.82x in June 2026, so a demand shock would tighten reported liquidity before it reached earnings.

The funding mix is what changed. Booking reported $10.5 billion of repurchases in the twelve months to June 2026 against $9.54 billion of free cash flow, and $7.4 billion in the first half of 2026 against $6.75 billion. Equity fell $5.2 billion in six months and total debt rose to $20.7 billion from $19.0 billion. The added debt stays cheap: interest expense of $1.10 billion over the last twelve months is covered 8.9 times by operating income, better than FY2025's 5.7 times. A 15% fall in free cash flow would push the current pace of repurchases onto the cash balance, and a second year of that without slower buying would weaken the assessment.

What needs checking

The FY2025 profit gap. Roughly $1.5 billion of the year-over-year net income decline is unexplained by operating income, interest, and tax. The FY2025 10-K income statement and the tax footnote should show whether a legal accrual, an investment loss, or a prior-year gain produced it.

The debt footnote. Long-term debt stops appearing in the summary figures after September 2025 even as the company reported $20.7 billion of total debt and quarterly interest expense declined. The maturity schedule, the mix of notes and any commercial paper, and the average rate paid are the items to read.

The float. Whether the $17.21 billion of cash is unrestricted, and whether partner payment terms have shortened, decides how much of it can fund buybacks. Falling cash-to-current-liability coverage alongside continued repurchases would be the first sign the payout is being funded from the balance sheet.

Questions worth following

  • Can Booking hold a 20%-plus return on assets as growth slows?

    Return on assets is the only usable return measure while buybacks hold equity below zero. Holding it above 20% would show the margin gains are durable and that a flat asset base still supports growth; a slide toward the mid-teens would mean earnings lean on a margin peak.

Generated by DeepVal AI. Data and models may contain errors or be outdated. For research only; not personalized investment advice.

Go beyond the preview

Continue reading the full report

  1. Key risksWhat could break the investment case
  2. Valuation and margin of safetyWhat the current price assumes, and how much room for error it leaves
  3. SummaryOur overall judgement and what to watch next

Free to sign up · 20 research points · unlocking this report uses 10

20 points cover 2 full reports. No card needed.

Research reflects its generation date. Preview the first two modules.

More on Booking Holdings Inc