BKNG — Booking Holdings Inc.

Near-term market sentiment is clearly bearish: as of 2026-09-13, BKNG was down 10.0% in a week and 18.06% in a month, retracing from roughly $212 to ~$174 and giving back most of the summer rally, with RSI at 30.04. The triggers were mostly exogenous - an early-September 2026 spike in oil and jet fuel plus renewed Middle East tensions that spooked long-haul travel demand, a European court ruling upholding the block on the eTraveli acquisition (capping one inorganic expansion path), and continuing anxiety that agentic AI search tools disintermediate OTAs. That fear is running directly against the fundamentals and against the sell-side: the Q2 2026 print beat, the annual structural cost-savings target was raised to $650M from $550M, $4.1B was returned to shareholders in the quarter (of which $3.7B of buybacks), and analyst consensus remains Buy (45 buy / 25 hold / 0 sell in the FMP breakdown; the brief cites ~31 buy / 8 hold / 0 sell) with targets at $233.47 average and $231.05 recent. Price action and narrative are bearish; estimates and ratings are not.

Growth is real but visibly decelerating, and that is a trend, not a blip. Revenue rose 12.7% YoY in Q3 2025 ($9.01B), 16.0% in Q4 2025 ($6.35B), 16.2% in Q1 2026 ($5.53B) - then only 8.1% in Q2 2026 ($7.35B vs. $6.80B), with management guiding Q3 2026 to just 4-6% growth. The sequential build confirms the slowdown: the Q2-over-Q1 revenue delta was +$1.82B in 2026 versus +$2.04B in 2025, which is exactly what the deterministic score's 'seasonality-adjusted acceleration' leg is picking up. Margins, by contrast, keep grinding higher: operating margin was 34.0% in Q2 2026 vs. 33.1% a year earlier, 43.7% in Q3 2025 vs. 39.8%, 23.0% in Q1 2026 vs. 22.3%, and TTM operating margin is 34.5% versus 32.8% a year ago (+170bp). TTM revenue of $28.24B is +12.9% and TTM operating income of $9.74B is +18.5%. One caution: reported net income looks spectacular (+118% YoY in Q2 2026 to $1.95B) but that is heavily flattered by a depressed 2025 base - Q2 2025 net income was only $895M on $2.25B of operating income, and Q1 2025 was $333M on $1.06B, implying large below-the-operating-line charges in 2025 that are not repeating. Underlying quality is confirmed by cash: TTM free cash flow of $9.54B (Q2 2026 alone was $3.64B) and an accruals ratio of -5.96%, meaning cash earnings are running ahead of reported earnings - the healthy direction. Net: mid-teens top-line is gone, mid-single-digit is the new guide, but margins are still expanding modestly and cash conversion is excellent.

Cheap on every line in the dataset, as of 2026-09-13. The stock traded at 19.19x TTM EPS of $9.387 against a 5-year average P/E of 40.25; 16.6x the next-FY consensus EPS of $10.453; 12.74x EV/EBITDA; 4.77x sales versus a 5-year average of 6.64x; and roughly 14.2x TTM free cash flow of $9.54B, i.e. a 7.08% TTM FCF yield. PEG is 0.343. For a business running a 34.5% TTM operating margin, a 50.8% ROIC and ~$9.5B of annual FCF on a $134.8B market cap, that is a real discount both to its own history and to quality-compounder peers. The one honest caveat: don't touch book-value-based screens here - P/B of -12.39x, D/E of -1.92x and ROE of -96.9% are all mathematical artifacts of stockholders' equity sitting at -$10.78B because of buybacks, not evidence of distress (cash of $17.21B against $18.18B of long-term debt).

This is multiple compression, not estimate cuts - and that distinction matters. Estimates went up (Q2 2026 beat by ~$0.10 on EPS and 4.1% on revenue, room nights +5%, gross bookings +9%, cost-savings target raised to $650M) while the stock fell 18% in a month. The market is not repricing near-term earnings; it is repricing the terminal value of an OTA in an agentic-AI world, plus adding a geopolitical travel discount and a regulatory discount (the upheld eTraveli block). Practically: at 16.6x forward EPS the market is already paying for a mid-single-digit grower. If room nights hold at the guided 3-5% and the $650M cost program lands, low-double-digit EPS growth stacked on top of ~2-3% per quarter of buyback-driven share shrinkage makes the current multiple look too low; if AI agents start eating into the roughly two-thirds direct-traffic mix, 16x is not a floor. The gap between $174.33 previous close and a $233.47 average target is the market's way of saying it doesn't yet know which of those two worlds it is in.

Oversold and technically broken, in that order, as of 2026-09-13. RSI(14) was 30.04 - right at the oversold threshold - with the price 12.75% below its 20-day SMA, 10.26% below its 50-day and 6.28% below its 200-day, after a 10.02% one-week and 18.06% one-month decline; yet the stock is still +5.44% over three months and +2.52% over six, and sits in the lower third of its 52-week $150.14-$224.996 range. That is the signature of a fast, violent pullback within a still-positive intermediate trend: stretched far below every moving average with momentum washed out. Implied levels from the gaps put the 200-day near ~$186 and the 50-day near ~$194, so a reclaim of ~$186 is the first evidence selling has exhausted, and ~$194 would confirm. Entry implication: scale in on weakness rather than chasing; the risk is that this becomes a value trap that keeps bleeding if AI-disruption headlines persist into the Q3 print.

• Oil and jet-fuel price spike plus renewed Middle East tensions (early September 2026) hitting long-haul travel appetite; management has already noted a mix shift toward intra-regional and domestic routes rather than intercontinental.
• Agentic AI search and native LLM trip planners as a long-term disintermediation threat to OTA take rates and marketing efficiency; Booking is countering with Priceline's 'Penny' assistant and agentic experiences while still owning roughly two-thirds direct traffic.
• European regulation: the court upholding the block on the eTraveli acquisition caps inorganic expansion in flights/Connected Trip, and hotel rate-parity/DMA-DSA scrutiny remains a structural overhang on cross-border growth.
• Consumer discretionary travel normalization after the 2023-24 post-pandemic peak, with airline capacity growth and post-summer booking velocity as the main swing factors into the Q3 2026 guide of 4-6% revenue growth.
• FX translation: the overwhelming majority of gross bookings are generated outside the US, so a stronger dollar mechanically compresses reported revenue and EPS even when underlying room nights grow.

• Q3 2026 earnings, expected late October / early November 2026: room-night growth, late-summer and holiday booking velocity, and evidence of progress against the raised $650M structural cost-savings target.
• Buyback cadence - $3.7B of repurchases in Q2 2026 alone out of $4.1B total returned; sustained repurchases at these prices shrink the share count roughly 2-3% per quarter and mechanically support EPS.
• Any tangible evidence that Booking's own AI products (Priceline 'Penny', Connected Trip cross-sell) are defending direct traffic - the single biggest swing factor in whether the multiple re-rates or de-rates further.
• Oil price relief or Middle East de-escalation, which feeds directly into the long-haul travel demand that underpins the 4-6% Q3 revenue guide and the FY guidance frame.

• AI disintermediation: if LLM-based agents capture search and booking intent, Booking's two-thirds direct-traffic advantage erodes and the marketing-efficiency model that produces a 34.5% operating margin compresses.
• Growth deceleration is already visible - Q2 2026 revenue +8.1% versus +16.2% in Q1 2026, with Q3 guided to 4-6% - and could prove structural rather than a temporary Middle East air-pocket.
• Balance-sheet optics: stockholders' equity of -$10.78B at 2026-06-30 (versus -$5.58B at 2025-12-31), long-term debt up to $18.18B from $12.22B at 2023-12-31, and buybacks that would exceed annual FCF if Q2's $3.7B pace persisted; no near-term solvency issue with $17.21B of cash and $9.5B of TTM FCF, but the cushion is thinner than it looks.
• Macro and geopolitical shock: an oil spike or widening conflict hits discretionary long-haul travel first, and Booking's international revenue base makes it more exposed than a domestic US consumer name.
• Regulatory and M&A constraint: the upheld eTraveli block removes an inorganic path to deepening flights and the Connected Trip, leaving execution risk squarely on organic build-out.

Booking is a cash machine currently being priced like a melting ice cube, and I think that is wrong - but not wrong enough to be an unqualified buy. The business earned a 34.5% TTM operating margin and 50.8% ROIC, converted $9.54B into TTM free cash flow (a ~7.1% FCF yield, or ~14x FCF), and its cash earnings exceed reported earnings (accruals ratio -5.96%); yet as of 2026-09-13 it trades at 19.2x trailing and 16.6x forward EPS against a 5-year average trailing P/E of 40.25, after an 18% one-month slide that estimates did not cause. My own work puts fair value around $228, implying meaningful upside. The honest counterweights are that growth has decelerated to a 4-6% Q3 guide, the agentic-AI threat to OTA economics is genuinely unresolved, and the market may be structurally de-rating the group rather than mispricing it. Read this as a scale-in-on-weakness situation with a real 30%+ upside case and a real value-trap tail, not a screaming bargain to back up the truck on.

Blended two independent methods as of 2026-09-13: (1) 22x next-FY consensus EPS of $10.453 = $230, a multiple I justify as below the 5-year average trailing P/E of 40.25 and below where BKNG has historically traded for ~11% EPS growth, but above the current 16.6x because that multiple already embeds excessive AI-disruption pessimism given the 34.5% operating margin and two-thirds direct-traffic mix; and (2) 18.5x TTM free cash flow per share of $12.31 (= $228), versus the ~14.2x the stock trades at today, implying a 5.4% normalized FCF yield. A 9.5%-WACC DCF (9% FCF growth for five years, 3% terminal) yields roughly $250, so I haircut the DCF for terminal-value risk from agentic AI and regulation and land at $228 - about 31% above the $173.92 close, independently derived but converging with the $233 sell-side average.