Expedia: the network beyond the booking

A travel network with two ways to reach the customer

Expedia distributes the same broad supply capability through its consumer brands and through partners that already have travelers. The direct route pays to obtain attention and maintain a relationship; the partner route pays the interface owner through its commercial terms. Reusing inventory connections, payments, and service can make both routes valuable. Their contribution has to be measured after their different demand-acquisition bills.

The trailing year through June 2026 generated $15.700 billion of revenue. B2C contributed $9.834 billion and B2B $5.362 billion, with the remaining $504 million in other revenue. Those figures use the full 2025 year plus current first-half results less the comparable earlier first half. They preserve seasonality instead of doubling a strong interim result.

The Rapid shopping documentation gives the infrastructure thesis a concrete job. A selected rate is checked again before booking; a changed price or unavailable rate changes the next step. Discovery is therefore not the completed transaction. The value lies partly in maintaining an executable connection to live supply and its conditions. The documentation describes that booking capability. Its competitive advantage depends on what customers choose and pay for.

The booking-management API then supports retrieval, changes, and cancellation of the itinerary. That continuing work is part of what a partner buys. An AI interface can make planning easier without taking over every supplier connection and service obligation. The contrary case remains serious: if partners can obtain comparable execution and service elsewhere, they can bargain away much of the benefit of Expedia’s scale.

The fastest-growing business is not producing the fastest profit growth

The latest quarter contains the central tension in this pitch. B2B revenue grew about 23.5%, but adjusted segment EBITDA grew only 11.5%. Its margin fell from 27.4% to 24.7%. Partner commissions and growth investment absorb part of the expansion. Applying the consumer business’s margin to those new partner dollars would overstate the long. E3

B2C shows the stronger current profit conversion. Revenue grew about 8.0%, while adjusted segment EBITDA grew 22.1%, lifting margin from 29.4% to 33.2%. In the release’s operating tables, B2C direct marketing rose only about 1% while its bookings grew about 8%. The consumer business kept more of the incremental revenue. E3

The current operating split is therefore economically coherent: consumer acquisition efficiency can improve while a wider partner network absorbs commissions and investment. The long needs those investments to mature into worthwhile contribution after partner payments. Shared infrastructure creates an opportunity to lower cost; it does not determine which party keeps the saving.

For B2B I would join growth, margin, and the cost of servicing completed trips. A lower percentage margin can be attractive if genuinely incremental business uses shared capacity and adds enough dollars. Persistent deterioration with a growing service burden would mean partners capture too much of the economics. The quarter cannot distinguish a productive investment period from that weaker steady state.

Booking growth has to become stayed revenue and usable cash

In the second quarter, booked room nights rose 6%, gross bookings rose 12%, and revenue rose 14%. Those measures have different timing and content. Bookings reflect transactions when booked and are adjusted for cancellations; much lodging revenue reflects completed stays. Pricing, mix, and the timing of travel affect the relationship. E3

The trip must still be served, changed when necessary and settled with the supplier. The filing describes both Expedia Collect and Hotel Collect, including a customer choice under Expedia Traveler Preference. More bookings need not bring the same advance funding if that mix changes. I would follow the payment path and remaining obligations before treating early collection or gross booking growth as distributable cash.

The latest full-year guide raised revenue to $16.05–16.22 billion and expected 150–175 basis points of adjusted EBITDA-margin expansion. Management issued that August outlook before the full-year result was reported. It supports an improving operating direction. My two-year case assumes growth will slow from the strongest quarter. E3

The GAAP EPS headline also needs care. Second-quarter EPS of $7.16 includes investment and tax effects, while the first half benefited from a $71 million reversal of earlier Canadian digital-services-tax liabilities. I do not annualize the quarter or treat that reversal as recurring operating leverage. Nor do I charge the same potentially noncash reversal against CFO again without a cash reconciliation. E2

Travel working capital lowers the cash I would underwrite

The trailing cash statement, built from the annual and interim filings, shows $5.216 billion of operating cash and $757 million of cash capex, leaving $4.459 billion of reported free cash flow. A large part of the difference between that number and recurring cash earning power comes from travel working capital.

My sensitivity removes the net of three cash movements: $1.958 billion added by deferred merchant bookings, $319 million added by merchant payables, and $878 million used by receivables. Together they supplied $1.399 billion. Subtracting only the deferred-booking increase while ignoring receivables would penalize one side of the travel cycle. Keeping all of it as permanent shareholder cash would make the opposite mistake.

Removing the $1.399 billion net source and $412 million compensation add-back leaves $2.648 billion. This conservative sensitivity holds growth flat and tests how much cash remains after working-capital needs. Receivables mix merchant and agency flows that the selected statements do not fully separate. That imperfection is a reason to keep the sensitivity explicit, rather than call the entire reported cash surplus permanent or remove one unfavorable movement in isolation.

A second compensation sensitivity uses $502 million credited to equity instead of the CFO add-back, reducing the anchor to $2.558 billion. The filing’s capitalized-compensation and parent-equity bases do not reconcile completely across the selected interim periods. I show that alternative because the proxies leave future dilution costs uncertain.

The $7.127 billion of June cash and short-term investments is not an independent valuation windfall against $15.426 billion of deferred merchant bookings and supplier obligations. I add no excess-cash bonus. Nor do I treat the comparison as proof of a shortfall payable tomorrow. Trips, receivables, supplier payments, and refunds have different dates; the appropriate object is a funded cash schedule. Interest already passes through the equity-cash anchor. E2

Payment-model mix matters to that schedule. A traveler paying the hotel at the stay does not provide the same booking-date financing as one paying Expedia in advance. The filing also describes the seasonal reversal as stays catch up with bookings. A larger network can consequently generate more revenue without the same working-capital source. I would protect enough liquidity for the payment mix before carrying the share-repurchase plan forward. E2

The stress I would run is a period in which new advance collections weaken while earlier trips, supplier bills, and refunds still settle. That can consume cash even while the stayed business remains profitable. My $1.399 billion working-capital sensitivity strips out the trailing financing benefit; it does not quantify that adverse timing path. I would reduce repurchases before treating the spring balance as surplus available through a travel downturn.

Airline loyalty financing provides a second cash-timing example: selling miles brings in cash before award travel and creates a collection stream that can secure debt.

The base forecast lets each channel grow at its own rate

I grow B2C revenue 5% annually, B2B 15%, and other revenue at zero. That produces approximately $18.437 billion in the second model year. The partner assumption is slower than the latest quarter’s revenue growth, and the consumer assumption does not require an exceptional travel boom.

The modeled 17% recurring cash conversion yields $3.134 billion, above the $2.648 billion starting sensitivity. This charges compensation and gives no extra financing benefit from net travel working capital. To earn it, consumer efficiency must persist and partner growth must leave sufficient contribution after commissions and service. The forecast is not obtained by lending the stronger B2C EBITDA margin to every B2B dollar.

The share count falls 3% annually from the latest disclosed July 24 count of 120.019 million, reaching 112.926 million. Before those purchases, the budget covers dividends and protects another $250 million annually for liquidity or debt needs. Assumed average purchase prices are $290 and $325. The model can fund the desired reduction from the remaining cash; gross compensation-neutralizing purchases do not become an extra return on top of the lower share count.

The cash budget prevents the same repurchase from becoming both a smaller denominator and an extra yield. More importantly, it subordinates the purchase to the travel obligation. I would prefer a slower share reduction to forcing buybacks through a period when bookings weaken or refunds and supplier settlements consume cash. Capital allocation can strengthen the operating opportunity only when the cash belongs to shareholders at the time it is used.

Getting to $360 requires some revaluation

Thirteen times second-year cash per share gives $360.82; I round the prospective price forecast to $360. The exact model implies about 34.4% price appreciation from the entry. A constant $0.48 quarterly dividend would add $3.84 over eight quarters, but future declarations are assumptions and the primary benchmark compares price returns.

At the entry’s multiple of the lower cash anchor, the same operating and share-count path gives about $338. The business case therefore supplies most of the modeled appreciation; some revaluation is still needed for the central target. An 11.5% net-margin assumption gives about $18.78 of second-year earnings per share, placing the target near 19.2 times that check. I use it to challenge the cash valuation; adding the two would count the same business twice.

The reverse test asks what cash would support a 10% annual price return under the same channel growth, purchase-price and multiple assumptions. It recalculates the share count as cash changes. Roughly $2.822 billion in second-year cash is required, 6.6% above the current lower anchor. The hurdle is achievable, but it still asks the business to improve. Expedia is less forgiving than a case that works through stagnation.

The competing cases are about demand and who keeps the margin

The bear case combines consumer revenue declining 3% annually, partner revenue growing only 8%, a 13% cash margin and no net share reduction. At ten times cash, the value is around $173. A travel downturn can turn operating leverage against the shareholder quickly, even when the partner channel continues growing.

The bull case has 8% consumer growth, 20% partner growth, an 18.5% cash margin and an affordable 4% annual share reduction. Fifteen times cash gives about $495. That result needs better economics as well as volume.

The strongest bear is a larger fulfillment network with weaker net economics: partners demand more commission, servicing becomes more expensive and payment choice reduces advance funding. Bookings and cash collected can each look healthy at a selected date while recurring shareholder cash disappoints. The bull uses the same connections positively—more useful trips through shared infrastructure, with lower incremental cost and margins that stabilize after investment.

The observations that would earn the target

Before the October 2028 price evaluation, I expect the latest available trailing B2B revenue to exceed $7.0 billion, B2B adjusted EBITDA margin to be at least 25%, and B2C margin to remain at least 31%. I also expect the conservative cash sensitivity to exceed $3.0 billion and the share count to be below 114 million. Those operating predictions are explicit author tests, independent of management guidance.

The target will be earned by B2B contribution stabilizing, consumer acquisition efficiency persisting and a full travel cycle confirming cash after obligations. I would also examine whether servicing cost grows faster than stayed revenue as distribution expands. Repurchases create value through the shares retired within that funded cash budget.

The original disconfirmation tests remain: B2B margin below 23% for two comparable quarters, B2C below 28% without a convincing temporary explanation, or the cash sensitivity below $2.2 billion over a trailing year. A major payment, acquisition or new capital need changes the share budget as well. I would abandon the dated long if distribution keeps expanding while the cash Expedia keeps fails these tests. The network is valuable when it can deliver the trip and still leave an attractive claim for its owners.

The network has to earn cash for its owners

I believe Expedia can build a more valuable business by combining a better consumer operation with broader partner distribution. The current evidence supports that direction, and the entry price leaves room for the resulting cash to compound per share. My $360 forecast depends on the cash Expedia keeps after serving those bookings. The dated forecast remains available for comparison with the original entry and SPY. If the network expands but shareholders receive less cash than I expect, the prediction has failed on its own terms.

Assumptions

• USD millions and millions of shares unless stated; prices are USD per original share.
• TTM = annual result + current comparable interim − prior interim. Forward cash margins include economic compensation and company-specific recurring reinvestment.
• Net share reduction is cash-budget constrained. Retained/protected cash has no terminal bonus; gross buybacks are not added to price returns. Cases are not probabilities.