Key Takeaways
- • Situational Awareness, the AI-focused hedge fund founded by Leopold Aschenbrenner in July 2024, collapsed within four weeks after reaching a peak of $45 billion in assets and approximately 367% returns between August 2025 and June 2026.
- • NYU Stern valuation professor Aswath Damodaran attributes the fund's destruction not to a flawed AI thesis but to the mechanics of conviction-driven position sizing combined with leverage—a predictable and well-understood capital destruction mechanism.
- • The fund's dual-leg strategy—long AI infrastructure suppliers and short AI-disrupted software/services businesses—faced simultaneous adverse moves, triggering margin calls at approximately 400% leverage and forcing liquidations into falling markets.
- • Damodaran's analysis challenges the finance industry's treatment of "conviction" as an unambiguous virtue, arguing that high conviction paired with aggressive position sizing and leverage creates systematic risk regardless of thesis quality.
- • The four-week collapse illustrates how prime broker collateral demands can accelerate losses in levered portfolios, converting paper losses into forced selling that compounds drawdowns.
Summary
Aswath Damodaran, the NYU Stern valuation professor known for his authoritative annual data updates, published an analysis on August 10, 2026, examining the rapid collapse of Situational Awareness, the AI-focused hedge fund that imploded over a four-week period in early summer 2026. Rather than condemning the fund's artificial intelligence conviction, Damodaran uses the episode to interrogate how conviction itself—treated as an unambiguous virtue across asset management—becomes a mechanism for systematic capital destruction when combined with aggressive position sizing.
Situational Awareness was founded in July 2024 by Leopold Aschenbrenner around a 167-page thesis proposing a dual-leg strategy: long positions in companies supplying AI infrastructure and short positions in software and services businesses vulnerable to AI disruption. The fund attracted prominent technology backers, with Jane Street among notable investors, and held a private stake in Anthropic alongside its public equity book. Performance was extraordinary: according to Portfolios Lab data cited by Damodaran, the fund returned approximately 367% between August 7, 2025 and its high-water mark on June 19, 2026, with cumulative inception-to-date returns substantially higher.
The reversal compressed into four weeks. Semiconductor and AI infrastructure names reversed sharply in tandem with deterioration across the short book, creating simultaneous losses on both portfolio legs. With leverage reported at as much as 400%, prime brokers issued collateral calls that forced the fund to liquidate positions into a falling market. These forced sales generated additional losses, triggering further margin calls in a self-reinforcing cycle that obliterated the $45 billion in assets peaked in early July 2026.
Damodaran's framing shifts focus from thesis quality to position management mechanics. The Situational Awareness collapse demonstrates that conviction—however analytically sound—paired with aggressive leverage and concentrated position sizing creates predictable and devastating outcomes when correlated losses occur. Prime broker dynamics amplify this destruction: margin calls eliminate the fund's ability to weather temporary drawdowns, converting the initial reversal into forced liquidations that accelerate losses far beyond fundamental valuation changes. This is not a story about youth or hubris, nor is it fundamentally about the validity of the AI infrastructure trade. It is instead a story about how the mechanism linking conviction to position sizing to leverage to prime broker collateral management destroys capital with mathematical inevitability when market conditions move against concentrated, levered positions simultaneously.
Situational Awareness was founded in July 2024 by Leopold Aschenbrenner around a 167-page thesis proposing a dual-leg strategy: long positions in companies supplying AI infrastructure and short positions in software and services businesses vulnerable to AI disruption. The fund attracted prominent technology backers, with Jane Street among notable investors, and held a private stake in Anthropic alongside its public equity book. Performance was extraordinary: according to Portfolios Lab data cited by Damodaran, the fund returned approximately 367% between August 7, 2025 and its high-water mark on June 19, 2026, with cumulative inception-to-date returns substantially higher.
The reversal compressed into four weeks. Semiconductor and AI infrastructure names reversed sharply in tandem with deterioration across the short book, creating simultaneous losses on both portfolio legs. With leverage reported at as much as 400%, prime brokers issued collateral calls that forced the fund to liquidate positions into a falling market. These forced sales generated additional losses, triggering further margin calls in a self-reinforcing cycle that obliterated the $45 billion in assets peaked in early July 2026.
Damodaran's framing shifts focus from thesis quality to position management mechanics. The Situational Awareness collapse demonstrates that conviction—however analytically sound—paired with aggressive leverage and concentrated position sizing creates predictable and devastating outcomes when correlated losses occur. Prime broker dynamics amplify this destruction: margin calls eliminate the fund's ability to weather temporary drawdowns, converting the initial reversal into forced liquidations that accelerate losses far beyond fundamental valuation changes. This is not a story about youth or hubris, nor is it fundamentally about the validity of the AI infrastructure trade. It is instead a story about how the mechanism linking conviction to position sizing to leverage to prime broker collateral management destroys capital with mathematical inevitability when market conditions move against concentrated, levered positions simultaneously.
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.