In Q1 2026, top-tier institutions have not abandoned AI—but they are no longer buying “the same AI.” The portfolio adjustments made by six major hedge fund managers have been revealed. This is more than just a list of holdings; it’s effectively a map of opposing positions among Wall Street’s elite capital. Although the 13F filings are inherently lagging, they clearly illustrate how institutions re-evaluated the market’s primary narrative in the previous quarter.
For several years, the U.S. equity market operated under a clear, shared narrative: buy the “Magnificent Seven,” buy AI, buy platform leaders, and buy high-quality tech. This time, however, that consensus among Wall Street’s top funds has begun to fracture. Take Google, for example: some funds aggressively added exposure, while others nearly liquidated their entire positions. Similarly, Amazon saw one fund fully divest while another maintained a heavy position. Traditional SaaS stocks were massively unwound by Bridgewater, yet AI hardware and compute infrastructure were concurrently concentrated buys by another cohort of investors. This divergence signals growing disagreement among institutions on AI capital flows, reassessments of competitive moats, and whether valuations have become overstretched.
Berkshire Hathaway—entering its post-Buffett era—has significantly streamlined its portfolio while adding to Google, targeting the cash-flow giant currently doubted by the market. Pershing Square, by contrast, positioned itself almost diametrically opposite Buffett: it nearly exited Alphabet entirely and established a new position in Microsoft, arguing Microsoft holds higher odds of winning at the enterprise AI entry point. Bridgewater demonstrated a clear supply-chain pivot—selling traditional software names while buying NVIDIA and TSMC, reasoning that regardless of which application-layer players win, underlying capex will flow overwhelmingly to hardware.
Apalutza’s David Tepper doubled down on “hardware no one can bypass,” concentrating positions in Amazon, Micron, and TSMC—core infrastructure plays. Duquesne Family Office embodied the anti-crowded-trade philosophy, rapidly adjusting exposures and sidestepping the hottest AI-themed names. Egerton Capital pursued a more complex strategy: exiting Microsoft while adding Google, NVIDIA, and industrial hard assets—constructing a diversified portfolio spanning financial infrastructure, AI platforms, industrial hard assets, and energy.
Horizontally, Google has emerged as the single largest source of divergence this cycle, with institutions holding starkly opposing views on its moat. Microsoft is no longer an uncontested consensus asset—the question of whether its elevated valuation has already priced in too much future growth has moved front-and-center. Traditional SaaS, once viewed as a safe-haven asset, has now become an “audit asset,” prompting the market to re-examine the mid-layer value proposition of its business models. Even financial infrastructure names like Visa and Mastercard are now showing divergence—no longer automatic, no-brainer consensus picks for institutions.
The real signal from this round of 13F filings is that AI investing has transitioned from a broad, monolithic theme into a layered, stratified phase. Institutions are no longer betting on a single AI narrative; instead, they’re deconstructing the stack—platform layer, application layer, hardware layer, and financial tollbooth layer—and re-pricing each accordingly. For investors, the key insight is this: divergent assets merit deeper scrutiny than consensus assets. When elite institutions begin acting as counterparty to one another, investors must understand the underlying assumptions behind every trade—reassessing moats, cash-flow expectations, and risk-adjusted valuation metrics—not simply copy-paste positions.
[MacTong MSX]
The Great Divergence: How Institutional Fragmentation in AI Equities Reshapes Crypto’s Investment Landscape
The recent 13F filings revealing starkly opposing positions among Wall Street’s elite investors signal a fundamental shift in market dynamics. When Berkshire Hathaway doubles down on Google while Pershing Square exits nearly all Alphabet positions, and Bridgewater simultaneously sells traditional SaaS while loading up on NVIDIA and TSMC, we’re witnessing more than portfolio rebalancing—we’re observing the fracturing of a consensus narrative that has dominated markets for years.
This institutional divergence carries profound implications for crypto markets, particularly as traditional finance increasingly intersects with digital assets. The era of monolithic investment theses is giving way to a more sophisticated, stratified approach where understanding counterparty assumptions becomes as crucial as the positions themselves.
The AI Stack Decoupling and Crypto’s Parallel Evolution
The most significant revelation is how institutions are deconstructing the AI ecosystem into distinct layers—platform, application, hardware, and financial tollbooth—and developing independent theses for each. This mirrors the evolution we’re witnessing in crypto, where projects are increasingly evaluated based on their specific layer within the broader digital infrastructure stack.
For crypto investors, this presents both opportunities and challenges. On one hand, it validates the layered approach that many sophisticated crypto investors have already adopted. On the other hand, it demands even more rigorous research as the simple “blockchain will revolutionize X” narratives give way to more nuanced investment theses.
Key Crypto Opportunities Emerging from This Shift
-
AI-Infrastructure Convergence: As institutions bet heavily on “hardware no one can bypass” like NVIDIA and TSMC, crypto projects providing decentralized alternatives to centralized AI infrastructure stand to benefit. Decentralized compute networks (Bittensor, Fetch.ai) and GPU-sharing protocols are no longer experimental concepts but potential infrastructure plays in an increasingly AI-dependent world.
-
Tokenized Real-World Assets: Berkshire’s strategic shift toward cash-flow generative assets like Google suggests a growing appetite for real-world value capture. In crypto, this translates to projects successfully tokenizing tangible assets or generating sustainable cash flows—particularly those in the cross-section of AI and physical infrastructure.
-
DePIN 2.0: The divergent approaches to hardware infrastructure (from Tepper’s concentration in Amazon/Micron/TSMC to Bridgewater’s NVIDIA/TSMC play) validate the decentralized physical infrastructure network thesis. Crypto projects offering ownership or access to physical infrastructure through token mechanisms may attract institutional capital seeking exposure to both digital assets and real-world utility.
-
Sophisticated DeFi+AI Hybrids: Just as institutions are distinguishing between application-layer and infrastructure-layer AI plays, crypto investors are differentiating between basic DeFi and AI-enhanced financial protocols. Projects incorporating machine learning for risk assessment, automated portfolio management, or sophisticated market analysis represent the next evolution of financial primitives.
Risks and Market Implications
The institutional divergence carries significant risks for crypto markets. First, the increasing fragmentation of investment approaches could lead to greater volatility as different theses battle for dominance. Second, the sophisticated research frameworks being developed for traditional assets will increasingly be applied to crypto, potentially exposing weaker projects with unsustainable narratives.
Perhaps most concerning is the potential for valuation decoupling. As institutions adopt increasingly divergent views on traditional tech assets, we may see similar divergence in crypto valuations—where projects with strong fundamentals and clear utility command significant premiums over hype-driven alternatives. This could accelerate the market’s maturation process but also create periods of extreme volatility as different investor groups reconcile their conflicting views.
Strategic Implications for Crypto Investors
For experienced crypto investors, the key insight from this institutional fragmentation is clear: consensus positions are becoming less valuable, while divergent assets with strong underlying merit deserve deeper scrutiny. When elite institutions begin trading against each other, it signals that markets are entering a more sophisticated phase where understanding the “why” behind positions becomes as important as the positions themselves.
This suggests a shift toward more active research-driven investing in crypto, where investors must develop and defend their own theses about the future of technology, adoption, and value capture. The era of following “smart money” is giving way to the era of understanding the smart money’s reasoning.
The Post-Consensus Crypto Investment Environment
We’re entering a post-consensus phase in both traditional and crypto markets. In this environment, success will come not from identifying the next narrative but from understanding which narratives have real substance. For crypto investors, this means:
- Deep-dive research: Understanding tokenomics, competitive advantages, and real-world utility beyond the marketing hype
- Layer-specific analysis: Evaluating projects based on their specific function within the broader crypto ecosystem rather than generic “blockchain” potential
- Counterparty awareness: Recognizing when different investor groups are taking opposing positions and understanding the assumptions driving each side
- Valuation discipline: Applying more sophisticated valuation frameworks that account for token utility, network effects, and real-world cash flows
The institutional fragmentation in AI equacies isn’t just a Wall Street story—it’s a harbinger of the more sophisticated, research-driven investment environment that crypto markets are beginning to embrace. For those willing to put in the work, this evolution presents significant opportunities. For those still chasing simple narratives, it signals a potentially challenging period ahead.
As the lines between traditional and crypto investments continue to blur, the ability to think independently and develop nuanced investment theses will separate successful investors from those who follow the crowd into oblivion.