Booking Holdings (BKNG)

Booking Holdings is an extremely high-quality, asset-light compounder: the world's largest OTA, with a model funded by its customers' float (negative working capital → negative net invested capital and effectively infinite ROIC), capex well below depreciation, and an operating margin of ~33% and expanding. The moat — two-sided network effect, marketing and data scale, and brand — is wide, and the company returns nearly all its free cash flow to shareholders.

It is valued on EV/EBIT, the clean measure because EBIT already expenses stock-based compensation (which at Booking is barely ~2.2% of revenue → the SBC trap does not apply). On year-5 EBIT (~US$13.7bn) at an exit multiple of 17× — realistic for an asset-light compounder — and adding the accretion from buybacks, the five-year base value is ~US$366 per share.

The buyback is the central engine of value, not an accessory: Booking returns ~100% of free cash flow to shareholders, mostly by repurchasing its own shares (~US$8bn a year), which reduces the share count by ~−4% a year. That reduction accounts for ~+22% of the value per share; without modelling it, the business is undervalued. At ~US$214 (post 25-for-1 split), the base total return is ~+12% a year.

At ~US$214 per share, the base total return of ~+12% annually (appreciation plus a young, fast-growing dividend) exceeds the method's 15% hurdle. Measured against intrinsic value brought to today at the risk-free rate, the discount reaches the required margin of safety. The verdict is Undervalued: Booking combines an exceptionally high-quality business with a price that still leaves room to run, something uncommon in a compounder of this caliber. The source of the discount is market perception of AI disintermediation risk and the temporary drag on room nights from the Middle East conflict, both discounted harshly.

The main disconfirmer is capital-allocation discipline: the thesis rests on Booking continuing to return ~100% of free cash flow via buybacks at reasonable prices — buying back at expensive prices would destroy the per-share value engine. It's also worth watching interest coverage (~7.8×): the company borrows to fund buybacks and, although net debt is ~0, that leverage makes it sensitive to a rate increase or a drop in cash flow.

In the business, the focus is twofold: that the Middle East conflict does not extend the drag on room nights beyond the second half of the year (guidance assumes recovery), and — the longer-term risk — the evolution of AI disintermediation on the take rate and the direct channel. The quality of the business is not in question; the technological disruption risk and capital discipline are.