Booking, Expedia face margin pressure from AI agents, Bernstein says By Investing.com

Bernstein warns AI platforms could materially disrupt OTAs by compressing take-rates and marketing efficiency, eroding the informational 'supply moat' that favored Booking and Expedia. Booking's merchant pivot has left EBITDA margins more than 3% below 2015 levels, while Expedia's B2B now accounts for over one-third of revenue, reflecting strategic shifts. The firm maintains Market-Perform ratings on both but flags terminal value risk as the primary downside driver.

Analysis

AI agents that return a narrow, high-context set of travel options reprice the value of top-of-funnel marketing and will compress the effective take rate for intermediary platforms over a multi-year horizon. Expect a 150–400bp EBITDA-margin impact for pure-play OTAs if conversion shifts 10–25% of high-intent traffic back to direct or specialist channels within 12–36 months, because customer acquisition costs (CAC) rise while yield capture on each booking shrinks.

Second-order winners will be firms that own canonical supply or booking flows (large hotel chains with strong loyalty ecosystems, PMS/channel managers, and B2B consolidators) — these players can sell structured inventory to AI agents at lower distribution cost. Conversely, metasearch ad stacks and auction-driven CPC models are vulnerable; platforms that monetize via merchant spreads will see margin volatility as AI forces price transparency and smaller suppliers compete on parity.

Key catalysts and timing: 1) product integrations from major AI platform owners in the next 6–18 months (rapid adoption window), 2) OTA margin prints and guidance over the next 2–4 earnings cycles (realized effect), and 3) hotel chain responses — loyalty promotions and direct-booking incentives — which can re-steer supply economics inside 3–12 months. Upside reversal risks include slow consumer adoption of AI agents, poor conversion economics for agents (keeping a paid CPC/search model intact), or OTAs monetizing new AI distribution layers successfully within 12 months.

Key Decisions for Investors

• Initiate a directional short on Booking Holdings (BKNG) via a 12–18 month put-spread sized for 1–2% portfolio exposure: buy 1x 12-month ~25% OTM puts and fund by selling 1x ~45% OTM puts. Rationale: target a 15–25% downside re-rate if margins compress 150–300bps; payoff profile ~2–4x if catalysts materialize. Hedge with a 20–30% notional long in hotel franchisors to limit macro downside.
• Construct a pair trade: short BKNG (notional) / long Marriott (MAR) or Hilton (HLT) equal notional for 12–24 months. Thesis: capture reallocation of take-rate from OTAs to branded direct channels; expected asymmetric payoff if OTAs lose pricing power while branded chains maintain RevPAR. Risk: demand shock that hits hotels and OTAs together — size accordingly.
• Buy call spreads on major AI platform owners (GOOGL or MSFT) with 6–18 month expiries to express capture of search/agent monetization. Modest premium for upside if AI agents become primary discovery layer; expected payoff 1.5–3x if ad pricing or new marketplace fees scale within 12 months. Monitor regulatory headlines and integration demos as short-term gating items.
• Avoid unilateral long on Expedia (EXPE) absent evidence B2B margin mix accelerates; instead, consider a defensive long-expe idea by buying EXPE equity only after a confirmed 2-quarter improvement in B2B margin contribution or initiating a small long-dated call/short-dated put spread to benefit from a successful pivot while limiting downside exposure.