Billionaire Bill Ackman Dumps Alphabet and Loads Up On a Mega-Cap Stock Down 15% From Its Peak

Pershing Square exited its Alphabet position in Q2 after initiating it in Q1 and increased its Meta Platforms stake by about 20%. Since the end of Q2, Meta shares have returned more than 19%, while Alphabet shares have underperformed, supporting Ackman's relative-value shift. The article argues Meta remains attractively valued at roughly 19.5x next-year earnings, while Alphabet could offer upside from future AI product momentum and a potential Anthropic-related windfall.

Analysis

The relevant signal is not Pershing’s disclosed rotation itself—13F data are stale and cannot establish current conviction—but the relative setup: META’s AI narrative is being rewarded before a direct revenue model is observable, while GOOG’s valuation is more exposed to proving that AI preserves, rather than dilutes, Search economics. META can sustain a premium only if engagement gains translate into higher ad load, conversion, or pricing; otherwise incremental inference and infrastructure costs could compress the advertising EBITDA margin before monetization arrives.

Over the next 1-3 months, the key relative catalyst is quarterly evidence of AI-driven ad-product ROI at META versus Search query/share and Gemini monetization disclosures at GOOG. A positive META ad-pricing print should reinforce momentum, but expectations after the recent rerating leave little tolerance for a capex step-up without commensurate revenue acceleration. Conversely, GOOG is a cleaner catch-up candidate if management demonstrates stable Search margins alongside Cloud backlog growth, because the market is likely assigning an excessive probability to AI cannibalization rather than monetizable query expansion.

The contrarian view is that an agent product without disclosed active users, retention, or unit economics should not drive a durable multiple expansion. META’s distribution advantage is real, but agent utility may also raise privacy, liability, and trust friction that slows commercial deployment; GOOG retains superior intent data and default-distribution economics. This is better expressed as a relative valuation/earnings trade than a directional bet on consumer AI adoption.

Falsify the relative-long GOOG thesis if Search revenue growth decelerates materially while traffic-acquisition costs rise, or if META shows sustained ad-revenue acceleration with stable operating margins despite higher AI capex. For the next 6-18 months, monitor whether AI capex converts into durable free-cash-flow growth; both stocks risk multiple compression if inference costs rise faster than advertising yield.

Key Decisions for Investors

• Initiate a 3-month market-neutral pair: long GOOG/GOOGL, short META in equal beta-adjusted dollars. Target 10-15% relative outperformance as GOOG’s earnings-risk discount narrows; stop if META outperforms by 12% from entry or GOOG reports a clear Search-margin/traffic deterioration.
• Do not chase META momentum ahead of the next earnings print without verified disclosures on AI-product engagement and advertising lift. For existing META longs, reduce tactical exposure into results unless ad-revenue consensus revisions are rising; the downside scenario is capex-led margin compression and a 10-15% multiple reset.
• Use GOOG call spreads dated 3-6 months out only if implied volatility remains below META’s and Cloud/Search estimate revisions stabilize. A defined-risk upside structure is preferable to outright stock for an event-driven rerating, with the thesis dependent on preserved Search monetization.
• Set an earnings watchlist: META ad-price growth, impressions, operating-margin guidance, and 2027 capex; GOOG Search growth, TAC as a percent of ad revenue, Cloud operating income, and AI-related query commentary. Treat any claimed consumer-agent success without MAU, retention, and monetization data as non-investable narrative.