Leopold Aschenbrenner’s Situational Consciousness AI hedge fund misplaced round 67% of its worth in July. This was the results of margin calls main the fund to promote its public portfolio. The collapse of such a fund poses a query for fairness buyers in regard to the second half of 2026. Regardless of precisely predicting the demand for AI, the fund virtually skilled a disaster, making this occasion vital for causes past a single hedge fund supervisor.
Whether or not or not the expertise sector will stay as unstable going ahead is among the central questions now dealing with markets heading into the tip of the yr. The identical names which have propelled inventory indexes upwards can fall rapidly as quickly as leverage and liquidity meet a downturn, as was demonstrated in July.
Why a 439% winner misplaced two-thirds in a month
Situational Consciousness was created by a former researcher from OpenAI and bagged returns of 439% in H1 2026 by way of about 4x leverage on concentrated bets on infrastructure and chips within the AI house, as per RCK Analytics’ paper on its collapse. Nevertheless, that leverage had its downsides. As its positions fell between 35% and 47% in July, three prime brokers, together with Goldman and JPMorgan, known as margin, which the fund was unable to honor, main it to promote its whole public equities e book to Citadel.
Experiences verified the 67% plunge in July and the sale of the fund’s majority public investments. Based on Aschenbrenner, the agency defined that the fund is positioned “to struggle one other day” and introduced the drop by way of threat as a substitute of thesis.
Breaking: Leopold’s full letter despatched to his LPs final evening
Leopold Aschenbrenner’s fund fell 67% in July however stays up 80% YTD and he introduced he’ll preserve investing in public equities https://t.co/n3Swjmp8Z6 pic.twitter.com/jswe9ROkTo
— Leopold Inventory Tracker (@LeopoldTracker_) July 31, 2026
“We embrace volatility. But it surely ought to by no means jeopardize the fund.” And the corporate didn’t exit the business, investing $400 million in Supply Foundry, a chip-making start-up with a valuation of just about $5 billion that makes a speciality of lithography, which is the bottleneck in chip manufacturing and the realm that’s beneath the monopoly of ASML.
The demand was actual, however the financing was not
It was not the guess that was inaccurate. Filings say that near half of the fund’s US fairness portfolio had truly come from investments in firms like SanDisk and Micron by the tip of June, simply when the demand for reminiscence chips had been trending up. A examine by Counterpoint Analysis indicated that enterprise solid-state disks accounted for 48 p.c of the world’s complete NAND gross sales in Q2 of 2026, throughout the shift of AI workloads from the coaching section to the inference section. Furthermore, TrendForce predicted that NAND costs would go up by 10 to fifteen p.c in comparison with the earlier quarter and that costs for DRAM chips would improve by 13 to 18 p.c quarterly.
Micron supported its place out there with statistics, reporting an unprecedented income of $41.46 billion for its fiscal third quarter, in comparison with $9.30 billion for the earlier yr. This, nevertheless, didn’t assist the hedge fund. The Philadelphia Semiconductor Index suffered an virtually 30% fall in comparison with the height recorded in June, and an aggressive place didn’t enable to attend lengthy for the restoration indicated within the demand figures.
Central banks flag the identical leverage build-up
The weak point in query had already been recognized by regulatory authorities earlier than the case of Situational Consciousness. The Monetary Stability Report of the Financial institution of England, printed in July 2026, recognized the danger of a “substantial improve in the usage of leverage in fairness markets,” with a outstanding emphasis on inflated estimations of a small variety of firms associated to AI. The Financial institution for Worldwide Settlements used the examples of the collapse of the Archegos funding fund in 2021 and the stress of liability-driven investments within the UK in 2022 to insist that non-bank funds ought to take measures to keep away from occasions such because the one within the case of Aschenbrenner. The best issues had been recognized by consultants as being leverage and focus dangers.
Primarily, the sample of conduct is unchanged: the correct concepts, however the improper type and method of realizing them with none liquidity buffer in place.
Whether or not the AI commerce stays institutional
The unanswered query that is still to be explored in 2026 is how the hole goes to be stuffed. JPMorgan means that the expertise commerce should be coping with the leverage that has contributed to its rise up to now, and that leveraged ETFs, choices and margin accounts may nonetheless be settling down. Such programming is predicted to additional modify how leveraged and unleveraged demand for expertise shares is balanced whereas the danger is being diminished by buyers on their half. Based on PitchBook, actual property leveraged fundraising associated to infrastructure, power and knowledge facilities reached the report degree of $206.6 billion throughout 2025, with allocators selecting the constructions with lengthy length and contracts over leveraged directional investments.
The funding funding in AI isn’t getting smaller. Goldman Sachs estimates spending of about $7.6 trillion on computing, knowledge facilities, and power from 2026 to 2031. What has modified in July is how keen the market will probably be to bear that threat and whether or not the subsequent downturn is greeted by affected person capital or extremely leveraged holders able to unload.
