Meta's Loudest AI Bull Just Cut His Own Stock Exposure and It Is Not Because of AI

Altimeter Capital's Brad Gerstner maintained a high-conviction bullish view on Meta, calling its Muse AI agent a potential "next 10x moment," while reducing the firm's overall market exposure from large to medium. The move was driven by the 10-year Treasury yield reaching 5.1% and mortgage rates exceeding 7%, which increase discount-rate pressure on long-duration growth equities and could weigh on advertising demand. Altimeter's latest June 30 filing showed a $763.3 million Meta position, representing 7.8% of its disclosed portfolio, with no quarterly share change; Meta was trading at $748.53, up 31% over the prior month but still 1% below its year-ago level.

Analysis

The relevant signal is not a single manager’s disclosed position but the emerging conflict between AI-led earnings expectations and long-duration equity valuation. META can sustain premium earnings revisions if AI improves ad targeting, conversion and engagement faster than incremental inference/depreciation costs; however, a higher real-rate regime reduces the multiple investors will pay for those out-year gains. The more important near-term variable is whether consensus raises 2026-27 EBIT faster than the Treasury curve reprices—without that, AI optimism becomes multiple-dependent rather than earnings-backed.

Advertising creates a second channel beyond discount rates. Persistent mortgage and refinancing pressure is most damaging to lower-funnel discretionary categories—retail, autos, travel and housing-related advertisers—which may cut budgets before broad consumer weakness is visible in headline ad spending. This would favor Alphabet (GOOGL), whose Search mix is typically closer to high-intent demand, over META’s more discretionary social-ad exposure; conversely, falling long yields could produce a sharp META catch-up because its monetization upside remains more operationally leveraged.

AI capex is also a margin-duration issue. Market enthusiasm should not be extrapolated from product demonstrations: the investable proof point is whether AI-generated ad tools drive enough incremental revenue per impression to offset rising depreciation and infrastructure costs. Over the next 6-18 months, META’s relative return will hinge on revenue growth net of capex intensity, while suppliers such as NVDA, AVGO and VRT face the opposite risk: any hyperscaler capex discipline would hit their order-duration assumptions before it materially damages META’s ad business.

Key Decisions for Investors

• Maintain a neutral-to-underweight META beta for the next 1-3 months unless the 10-year Treasury yield decisively retreats and META reiterates revenue growth with stable or improving operating-margin guidance. A yield-driven multiple contraction can overwhelm a favorable product narrative; reassess on the next earnings call using capex growth versus revenue growth as the key test.
• Express relative ad resilience via long GOOGL / short META in equal dollar size over a 1-3 month horizon if long yields remain elevated. The thesis is that intent-driven Search should hold up better than discretionary social advertising; exit if META shows accelerating ad-price growth and AI tools lift revenue growth materially above GOOGL.
• For existing META longs, use a 3-6 month collar rather than adding outright exposure: retain upside to AI monetization while protecting against a rates-led de-rating. The hedge should be revisited if management raises capital-spending guidance without a commensurate acceleration in revenue or margin expectations.
• Monitor NVDA, AVGO and VRT as a second-order confirmation basket rather than chase them on AI rhetoric. Any evidence of hyperscaler capex normalization, longer deployment schedules, or weaker power/cooling backlog conversion is a warning that AI infrastructure expectations have outrun end-market monetization.