Wall Street's "AI Stock God" Falls, Done in by Leverage
Original: Odaily Planet Daily (@OdailyChina)
Author: Qin Xiaofeng (@QinXiaofeng888)
In the final week of July, Wall Street witnessed the meteoric fall of a "legendary" hedge fund.
Situational Awareness, helmed by Leopold Aschenbrenner, a former OpenAI researcher and German-born investor in his mid-20s, suffered significant losses as a sharp pullback in AI-related stocks collided with high leverage. The fund was forced to unwind large portions of its public market positions, selling most of its equity portfolio to Citadel, the fund managed by Ken Griffin.
The event quickly became a focal point for global financial media, bringing the young manager—once dubbed the "AI Stock Picker" in Chinese crypto circles—crashing down from myth to reality.
The Rise of a Prodigy
Leopold Aschenbrenner's resume is itself the stuff of legend. Born in Germany in 2001, he entered Columbia University at age 15, double-majoring in economics and mathematics-statistics, and graduated at the top of his class.
After graduation, he briefly worked at FTX's Future Fund before joining OpenAI's Superalignment team in 2023, focusing on the safety and alignment of superintelligence. In April 2024, he was dismissed for alleged information leaks (which he denies), a move widely attributed to his repeated submission of safety memos to the board questioning the company's practices.
Just two months after his dismissal, Leopold self-published a ~165-page paper, "Situational Awareness: The Decade Ahead". In it, based on deep learning trendlines, he predicted AGI could arrive around 2027, emphasizing that the true bottleneck isn't the algorithms themselves, but the explosive demand for compute clusters, electricity, data center space, and energy infrastructure.
The essay spread rapidly through Silicon Valley and investment circles, earning Leopold the reputation of an "AI Prophet." In September 2024, he founded the hedge fund Situational Awareness LP (named directly after the paper), focusing on publicly traded companies tied to AI infrastructure—chips, storage, cloud computing, power equipment, and data center operators—while shorting traditional software or outsourcing businesses potentially disrupted by AI. The fund attracted marquee investors including Stripe founders Patrick and John Collison, former GitHub CEO Nat Friedman, Daniel Gross, and Jane Street.
Starting with relatively limited capital, the fund ballooned its AUM to roughly $20 billion in under two years; in the first half of 2026, returns soared to approximately 439%, with heavy positions in CoreWeave, Bloom Energy, and storage/compute-related companies. The market crowned him the new "AI Stock Picker."
The July Storm: Caught Between Leverage and Pullback
But the good times didn't last.
As the market reassessed AI capital expenditure payback periods, AI hardware and infrastructure stocks saw concentrated selling in July. The fund's top Q1 holdings—Nebius Group, Sandisk, Micron, CoreWeave—all fell over 35% that month; Asian names like SK Hynix also weakened sharply. Meanwhile, the software stocks the fund shorted (like Adobe) rebounded, creating a pincer move against both long and short positions.
More critical was the leverage. The Financial Times reported that Situational Awareness used borrowing to amplify returns (with reports suggesting nearly 4x leverage)—a common hedge fund strategy that generates spectacular gains in bull markets but dramatically magnifies losses on the way down.
On July 24, in his semiannual letter to investors, Leopold acknowledged the fund "did not escape unscathed" from recent market turmoil (especially in Asia), while still framing it as the most attractive investment opportunity since early 2025, offering investors a window to add capital on August 1.
But conditions worsened. Over the following days, positions continued to slide, some dropping 30% or more. Margin calls came in rapid succession, forcing the fund to either raise capital quickly or sell assets under duress.
Ultimately, Situational Awareness engaged with existing investors and lenders before swiftly pivoting to large buyers. Around July 30, the fund sold the bulk of its ~$16 billion public stock portfolio (both long and short) to Citadel, with Goldman Sachs, JPMorgan, and Bank of America acting as prime brokers facilitating the deal. The fund's assets shrank dramatically to roughly $10 billion, with July losses hitting around 67%. The fund retained some private investments (including approximately $5 billion in Anthropic equity) and plans to continue operating as a private investment firm.
After the deal closed, several related AI stocks rebounded sharply, widely interpreted as the release of forced selling pressure and a temporary easing of the crunch. Market sentiment quickly recovered.
Deep Dive: Why Can a Correct Narrative Still Fail?
The core lesson here isn't that Leopold Aschenbrenner's long-term thesis is necessarily wrong. His repeated emphasis on compute, power, and infrastructure bottlenecks aligns directionally with current fundamentals: Microsoft Azure's continued strong growth, persistent supply-demand tightness for Nvidia's high-end chips, and AMD's repeated upward revisions of AI server market forecasts. These signals all point to one fact: the macro demand curve for AI infrastructure is still sloping upward.
Many in-depth analyses note that the long-term trend in AI infrastructure demand hasn't been invalidated—only short-term valuation froth and capex pacing have seen significant corrections. The real risk was never about "being right on direction," but about the precise alignment of leverage, position concentration, and entry timing. If any one of these three falls out of balance, even the most accurate macro narrative can collapse rapidly under a liquidity shock.
History is full of cases where investors saw the big picture but died too early. Long-Term Capital Management (LTCM) had Nobel laureates and sophisticated mathematical models, yet imploded in 1998 during the Russian debt crisis due to extreme leverage. Peter Thiel's Clarium Capital similarly made a correct but premature bet on oil prices, suffering massive drawdowns before the market caught up, ultimately proving unsustainable. These precedents repeatedly remind investors: being right isn't a moat—survival is.
Business Insider attributed Situational Awareness's failure to a brutally precise summary: "Even if you can accurately identify the biggest fundamental shift of a generation, assemble the right portfolio, and achieve astonishing returns in the short term, once you add sufficient leverage and conviction, the market no longer judges whether your ideas are smart—it judges whether you can survive when liquidity dries up." That line could almost serve as the epitaph for high-leverage macro strategies.
Additionally, the fund's rapid expansion from a small operation to tens of billions of dollars in AUM massively amplified operational difficulty and systemic fragility. Highly concentrated public positions stacked on high leverage left almost no room to maneuver when faced with the dual shock of sudden Asian market volatility and a global selloff in AI stocks. Large-scale redemption pressure combined with instantaneous liquidity evaporation forced the fund to liquidate passively at the worst possible moment, fueling a loss spiral that turned what might have been a temporary correction into an irreversible exit.
Aftermath and Outlook
For now, Situational Awareness retains some private positions and remaining assets. Leopold Aschenbrenner's personal wealth remains substantial, and he is planning to marry this weekend (around August 1).
Whether this forced liquidation marks a definitive turning point in the fund's fate, or merely a pause and restart after deleveraging, remains an open question.
For the broader market, the AI narrative can be compelling—even exhilarating—but leverage is forever a double-edged sword. Between short-term violent volatility and long-term grand trends lies the cruelest of gaps: the margin call.
Leopold Aschenbrenner once rewrote his destiny with a single long essay, then used high leverage to amplify that essay's conviction to the extreme. July's liquidation may well become the most profound—and most expensive—lesson of his career. It is also a highly representative cautionary tale for the entire AI investment boom.
The market doesn't care how smart you are. It only cares about one thing: whether you can still stand firm while the storm rages.
