Trading Giant Jane Street Reports $15 Billion Loss in July, First Monthly Loss in Nearly a Decade
Trading giant Jane Street recorded a loss of approximately $15 billion in July, marking its first monthly loss in nearly a decade, primarily due to exposure to the AI hedge fund Situational Awareness and sell-offs in AI-related stocks.
The fund, operated by former OpenAI researcher Leopold Aschenbrenner, had an initial investment of about $2.5 billion, which grew to nearly $10 billion in the first half of the year due to strong performance, but after a significant pullback in July, it dropped to $3 billion to $3.5 billion, directly causing a loss of nearly $7 billion.
Internal memos described July as a "bad month," with stocks exposed to AI experiencing widespread declines, including drops of about 50% in some storage and semiconductor stocks. Non-AI long positions in Asia also faced pressure, as the same investment portfolio that had strong excess returns in the second quarter became a major source of losses.
Despite the monthly setback, Jane Street's trading revenue year-to-date still exceeds $40 billion, surpassing the full-year record for 2025, continuing to lead Wall Street's market makers and large banks.
Situational Awareness triggered margin calls during the AI sell-off in July, forcing it to sell off most of its stock portfolio to institutions like Citadel, which caused simultaneous losses for Jane Street as a major investor.
Market mechanisms saw funds rapidly flowing out of high-leverage AI-themed positions, driven by a style switch in crowded trades; beneficiaries included buyers like Citadel that absorbed the sold-off assets, while those under pressure were quantitative market makers and hedge funds heavily betting on AI and semiconductors.
Source: Public Information
ABAB AI Insight
Jane Street rarely allocates to external managers but invested heavily in Situational Awareness and allowed its scale to expand, continuing its concentrated risk preference in AI infrastructure and tech stocks, previously making multi-billion dollar trades with AI cloud service providers like CoreWeave.
In terms of capital flow, the company directly funneled internal trading profits into high-leverage AI-themed funds, motivated by capturing the benefits of computing power and model competition, but this amplified losses during market style reversals; concurrently, it pushed for about $14.6 billion in debt refinancing, attempting to shift public debt to private investors to reduce disclosure pressure.
Similar cases can be seen in the 2021 Archegos collapse that triggered a chain of bank losses, and LTCM's massive pullback in 1998 due to failed concentrated bets; the current quantitative and market-making industry is in an expansion phase from low-volatility arbitrage to high concentration thematic bets.
The structural judgment pertains to the technical substitution risk following capital concentration: as AI trading crowding increases, single thematic positions become systemic vulnerabilities, with the mechanism being that mismatches in leverage and liquidity rapidly amplify losses during style switches, forcing market makers to reduce risk exposure.
ABAB News · Law of Cognition
The same portfolio of excess returns is the source of pullbacks
External fund leverage amplifies internal risks
Monthly losses can end a decade-long myth of no losses