Third Point's Aggressive Pivot: Decrypting Loeb's Multi-Billion Dollar Vote of Confidence in AI

Em um sinal definitivo ao mercado, o investidor ativista Dan Loeb e seu hedge fund, Third Point LLC, ampliaram agressivamente sua exposição à camada de infraestrutura de inteligência artificial (artificial intelligence), aumentando significativamente posições na Microsoft Corporation (NASDAQ: MSFT) e na Meta Platforms (NASDAQ: META). Essa realocação estratégica, revelada nos últimos documentos regulatórios e comunicações aos investidores divulgadas esta semana, ocorre em meio a um período de alta volatilidade no setor de tecnologia, sugerindo que uma das vozes mais proeminentes de Wall Street acredita que o "Superciclo da IA (AI Supercycle)" está longe de terminar.

Loeb’s move stands in stark contrast to the growing murmurs of an "AI bubble" that characterized market sentiment in late 2025. By nearly tripling his position in Microsoft and increasing his Meta holdings by almost half, Loeb is effectively doubling down on the thesis that hyperscalers—the massive tech giants building the physical and digital backbone of AI—remain the most attractive risk-adjusted vehicles for capitalizing on the next phase of generative AI: the transition from experimentation to agentic deployment.

Breaking Down the "Smart Money" Rotation

The scale of Third Point’s portfolio adjustment is substantial, reflecting a high-conviction bet rather than a mere hedging strategy. According to the latest 13F filings covering the period ending September 30, 2025, and subsequent investor updates, the fund has executed a decisive rotation of capital into what Loeb terms "digitally savvy, capital-efficient winners."

The most headline-grabbing figure is the massive increase in the Microsoft position. Third Point raised its stake in the Redmond-based giant by approximately 175%, bringing its total holding to roughly 1.1 million shares. This move positions Microsoft as a cornerstone of the fund’s equity portfolio, underscoring confidence in the company's ability to monetize its Azure OpenAI Service and the expanding Copilot ecosystem.

Simultaneously, Loeb increased the fund's exposure to Meta Platforms by 47%. This decision appears to validate Mark Zuckerberg’s heavy capital expenditure strategy, which had previously drawn ire from some corners of Wall Street. Third Point’s analysis seemingly aligns with the view that Meta’s open-source Llama models and massive data center build-out will yield long-term dominance in social AI integration and advertising efficiency.

Strategic Rationale: Infrastructure over Speculation

To understand the weight of this decision, one must look at the comparative data of Third Point’s recent maneuvers. The following table outlines the fund's major adjustments in the AI sector, highlighting the specific investment theses driving these allocations.

Third Point LLC: Key AI Sector Adjustments (Q3/Q4 2025)

The table illustrates a clear preference for incumbents with fortress balance sheets. Unlike venture capital flows that chased speculative application-layer startups in 2024, Loeb’s strategy in 2026 focuses on the "pick-and-shovel" providers—companies that generate revenue regardless of which specific AI applications ultimately succeed.

Navigating the "Capex Wall" Fears

One of the prevailing narratives in late 2025 was the fear of a "Capex Wall"—the idea that Big Tech’s spending on NVIDIA GPUs and custom silicon would outpace revenue growth, leading to margin compression. Loeb’s letter to investors addresses this head-on, dismissing the skepticism as short-sighted.

Third Point argues that the "relentless commitment to AI capex" is not a drag on earnings but a necessary moat-building exercise. For Microsoft, the accelerated earnings seen in its Intelligent Cloud segment suggest that demand for AI compute is outstripping supply, justifying the heavy investment. Similarly, Meta’s ability to integrate AI into its core advertising products has resulted in higher engagement and conversion rates, effectively paying for the infrastructure upgrades in real-time.

This perspective is bolstered by recent industry reports from Hazeltree and other market analysts, which note that while hedge fund crowding in tech is high, the distinction between "high-quality" AI plays (MSFT, META) and "speculative" ones is widening. Loeb is positioning Third Point on the side of quality, betting that the companies with the deepest pockets will be the ones to solve the critical bottlenecks of power consumption and data center capacity.

The Broader Market Context: 2026 as the Year of "Agentic AI"

The timing of these investments coincides with a technological shift often referred to as "Agentic AI"—systems that can reason, plan, and execute tasks autonomously rather than just generating text or images. Microsoft Research has flagged 2026 as the year AI moves "beyond conversation to collaboration," a sentiment that likely underpins Loeb’s bullishness.

If 2024 was about chatbots and 2025 was about integration, 2026 is poised to be the year of autonomy. Microsoft’s roadmap, heavily reliant on its partnership with OpenAI, envisions AI agents that can navigate enterprise software to perform complex workflows. By increasing his stake now, Loeb is betting that Microsoft will capture the lion’s share of the value created by this productivity revolution.

Furthermore, the volatility mentioned in market reports—stemming from geopolitical tensions and regulatory scrutiny—has created entry points that value-oriented activists like Loeb find irresistible. While some investors fled the sector during the "Liberation Day" sell-offs and recession scares of mid-2025, Third Point utilized these dips to accumulate shares at what they perceive to be a discount relative to future earnings power.

Risks and Counter-Narratives

Despite the optimistic positioning, the strategy is not without risks. The concentration of capital in the "Magnificent Seven" (or the surviving members thereof) leaves the fund exposed to regulatory headwinds. Both Microsoft and Meta face ongoing antitrust investigations in the EU and the US, which could theoretically force structural changes or hefty fines.

Moreover, the "AI Bubble" theorists argue that the Return on Invested Capital (ROIC) for Generative AI has yet to materialize at a scale justifying trillion-dollar valuations. If enterprise adoption stalls, or if "Agentic AI" proves more difficult to engineer than anticipated, the multiples on MSFT and META could contract significantly.

However, Loeb’s track record suggests he views these as manageable risks. His re-entry into Nvidia, after previously trimming the position, indicates a belief that the hardware cycle has legs. The "training compute" demand is now being supplemented by "inference compute"—the processing power needed to run the models once they are deployed—which provides a secondary tailwind for the entire semiconductor and cloud ecosystem.

Conclusion: A Vote for the incumbents

Dan Loeb’s latest moves serve as a bellwether for institutional sentiment in early 2026. The message is clear: the AI trade is evolving, not ending. By boosting stakes in Microsoft and Meta, Third Point is signaling that the winners of the next decade will likely be the same as the winners of the last—provided they continue to execute on their AI roadmaps.

For investors, this reinforces the importance of looking past short-term volatility and focusing on the structural advantages of data, distribution, and capital. As the AI landscape shifts from hype to hard engineering, the "smart money" is betting on the builders.