#AI fund leverage risk
AI Fund's 439% Gain Wiped Out by 67% Leverage Collapse
WooFun2026-08-17 16:22
Key Takeaways
Situational Awareness AI fund lost 67% in July due to margin calls despite accurate AI bets. The collapse highlights systemic leverage risks, prompting regulatory warnings and shifting investor strategies toward patient capital over aggressive leverage.
Woofun AI reports that the Situational Awareness AI hedge fund, owned by Citadel and founded by former OpenAI researchers Leopold Aschenbrenner, experienced a catastrophic 67% value erosion in July, driven by forced liquidations rather than strategic failure.
The structural fragility of the fund's position was exposed when equity positions declined between 35% and 47% during the month, triggering margin calls from major brokers including Goldman Sachs and JPMorgan. Because the fund had utilized approximately 4 times leverage to amplify its returns, it lacked the liquidity buffer necessary to withstand the drawdown, forcing the immediate sale of its entire public stock portfolio to meet obligations.
Despite the public market collapse, the fund's underlying thesis remained intact, leading to a strategic pivot toward private infrastructure. Aschenbrenner announced a $400 million investment in Source Foundry, a chip startup valued at nearly $5 billion that specializes in lithography technology. This move targets a critical manufacturing bottleneck currently dominated by ASML, signaling a shift from volatile public equities to foundational hardware capabilities.
The accuracy of the fund's initial macroeconomic bets was validated by industry data from June, which showed that nearly half of the U.S. stock portfolio was allocated to memory chip leaders like SanDisk and Micron. Counterpoint Research confirmed that in the second quarter of 2026, enterprise solid-state drives accounted for 48% of global NAND sales, reflecting the industry's transition from AI training to inference workloads.
Woofun AI data shows that market fundamentals further supported these positions, with TrendForce forecasting month-on-month price increases of 10% to 15% for NAND and 13% to 18% for DRAM. Micron reported record third fiscal quarter revenue of $41.46 billion, a substantial rise from $9.3 billion in the same period the previous year.
However, the Philadelphia Semiconductor Index fell nearly 30% from its June peak, preventing the leveraged fund from holding through the temporary dislocation.
Regulatory bodies had previously flagged these systemic vulnerabilities. The BOE's financial stability report released in July 2026 warned of excessive leverage in the stock market and inflated AI valuations. Citing the 2021 Archegos fund collapse and the 2022 UK Liability-Driven Investment (LDI) crisis, the BIS urged non-bank funds to implement stricter risk controls to prevent similar liquidity shocks.
JPMorgan analysts note that the core issue remains the mismatch between correct directional views and flawed execution via leveraged ETFs, options, and margin accounts. The firm argues that tech trading is still adjusting to the deleveraging cycle, with investors actively reducing exposure to mitigate the risks associated with high-leverage positions in 2026.
Capital allocation trends are shifting away from short-term leverage toward long-term physical assets. Leverage financing for infrastructure, energy, and data centers reached a record $206.6 billion in 2025, as allocators prioritized long-term contracts over speculative directional bets. This structural change reflects a broader institutional preference for stability and predictable cash flows in the AI supply chain.
Goldman Sachs estimates that spending on computing, data centers, and energy will total around $7.6 trillion from 2026 to 2031, indicating sustained demand. The July collapse serves as a stark warning that while the AI investment thesis remains robust, the market's tolerance for leverage has diminished, favoring patient capital over aggressive financial engineering.
Comments
No comments yet.