Key Takeaways
- • Gavin Baker, CIO of Atreides Management and early SpaceX investor, declares the AI "bottleneck trade" has ended, signaling a potential inflection point in the market's AI-driven rally that has dominated investor positioning since 2024.
- • Unity Software reported Q4 2026 revenue of $609 million, up 35% year-over-year and beating consensus estimates of $562.71 million, though the company posted a quarterly loss of $0.66 per share, reflecting ongoing profitability challenges in the AI software space.
- • The broader AI-fueled market rally is losing momentum, with U.S. stocks heading for their first weekly decline since March as investor enthusiasm around AI infrastructure plays cools following months of sustained outperformance.
- • SpaceX IPO trading has reached "bonkers" levels with new ETF inflows, while semiconductor and AI-adjacent equities (NVIDIA, AMD, Intel) faced pressure amid shifting allocation strategies away from bottleneck-dependent trades.
- • Baker's thesis directly challenges the premise that AI chip scarcity and power constraints remain the primary driver of equity valuations, suggesting investors must reassess portfolio positioning as the market transitions away from this crowded trade.
Summary
The timing of Baker's thesis aligns with measurable weakness in the AI-focused equity rally. U.S. stocks are positioned for their first weekly decline since March 2026, marking a notable shift after months of sustained performance driven primarily by AI infrastructure plays. Semiconductor stocks that benefited most from bottleneck narratives—including NVIDIA, AMD, and Intel—experienced meaningful pressure on the session, with Intel declining over 8% and AMD falling more than 7%.
Against this backdrop, Unity Software reported fourth-quarter 2026 results that underscore the uneven health of the AI software ecosystem. The company posted revenue of $609 million, representing 35% year-over-year growth and exceeding consensus estimates of $562.71 million. However, Unity reported a quarterly loss of $0.66 per share, illustrating that revenue growth alone has not translated to profitability for companies dependent on AI monetization. This disconnect between top-line performance and bottom-line results mirrors broader challenges facing software vendors reliant on AI infrastructure assumptions.
Baker's call suggests that investor positioning has become overly concentrated in plays predicated on persistent scarcity—whether in chips, power capacity, or computational access. As these bottlenecks ease through increased supply, competitive dynamics shift, and alternative AI business models emerge, the fundamental rationale for current valuations in hardware and infrastructure stocks may require substantial recalibration. The SpaceX IPO trading surge and emergence of new AI-focused ETFs indicate that capital is actively reallocating, though the direction and conviction behind these flows remain uncertain as the market enters a transition phase.
Go deeper than the headline
Research every firm in this story, track the ones that matter and get the daily briefing — free for 7 days.