Leopold Aschenbrenner AI fund suffers fresh losses amid AI stock rally

Leopold Aschenbrenner AI fund re-entered the AI market last week, only to watch six of its newly-acquired positions tumble between 5% and 8% by Monday’s open. The rapid decline follows a viral essay by Anthropic CEO Dario Amodei predicting an AI takeover of the internet within a year, a narrative amplified by Elon Musk and Sam Altman. While the commentary sparked a broad AI rotation, the fund’s own risk-management choices—leveraged exposure and a shift to “flex options”—raise deeper concerns for institutional investors about market structure, capital flows, and operational resilience.

Leopold Aschenbrenner AI fund re-enters AI market

• September 2-10: Situational Awareness purchased options on AMD, Bloom Energy, CoreWeave, SanDisk, SK Hynix, and the Roundhill Memory ETF.
• Friday, Sep 11: The fund’s activity was reported, prompting heightened scrutiny.
• Monday open: All six tickers opened lower, with declines ranging from 5% (AMD) to 8% (SK Hynix, Roundhill Memory ETF).
• Key data point: AMD closed at $516.13 on Friday; it opened around $486 on Monday, a 6% slide.

The speed of the price drop underscores how quickly sentiment can translate into price action when a high-profile fund is perceived to be “back in the game.” Institutional traders monitor such moves for clues about liquidity and potential cascade effects.

Shift from legacy leverage to flex-option model

• Historical leverage: Aschenbrenner previously employed up to 400% leverage, inflating a $45 billion AUM to a $225 billion notional exposure before the July crash.
• Current approach: The fund now uses fully paid “flex options,” which cap losses at the premium paid rather than exposing the fund to margin calls.
• Operational implication: Flex options eliminate forced liquidation risk, but they require upfront capital outlay and still expose the fund to premium erosion if the underlying assets decline.
• Risk-model critique: The shift appears conservative on paper but may be ill-timed; the premium cost can be substantial in a volatile AI market, eroding returns even if positions do not breach loss caps.

Capital flow dynamics after the July implosion

• July fallout: A sharp AI correction erased billions from the fund, prompting Citadel to acquire the bulk of the portfolio at a deep discount and JPMorgan to cut off leverage financing.
• Prime brokerage switch: Aschenbrenner moved from Citadel to boutique Clear Street, signaling a search for more flexible terms but also raising questions about counterparty risk.
• Follow-on financing: Despite the loss, the fund secured fresh capital from backers such as Stripe co-founders Patrick and John Collison, Nat Friedman, and Daniel Gross, indicating continued belief in Aschenbrenner’s AI thesis.
• Market-structure note: The rapid re-allocation of capital from large prime brokers to smaller boutiques can fragment liquidity, potentially amplifying price swings in thinly traded AI equities.

Viral AI-doom narrative as a market catalyst

• Anthropic essay: Dario Amodei’s claim that AI will dominate the internet within 6-12 months sparked a wave of sell-offs across AI-heavy stocks.
• High-profile endorsements: Elon Musk and Sam Altman echoed the sentiment, lending credibility and widening the sell-off beyond niche traders.
• Regulatory angle: While no regulator has acted on the essay, the episode illustrates how non-financial commentary can trigger market-wide risk events, prompting compliance teams to monitor media sentiment as part of systemic risk assessments.
• Operational consequence: Hedge funds may need to integrate real-time media analytics into trade-execution pipelines to anticipate sentiment-driven volatility.

Institutional exposure to AI-centric ETFs and memory-chip plays

• Roundhill Memory ETF (RMEM): Down ~8% on Monday, reflecting broader pressure on memory-chip makers.
• SK Hynix and SanDisk: Both suffered 7-8% declines, highlighting the sector’s sensitivity to macro-AI sentiment.
• Strategic lesson: Institutional investors with concentrated exposure to niche AI supply-chain components should diversify across broader technology baskets or employ dynamic hedging to mitigate sentiment spikes.
• Reference: For a broader view of market movements, see the latest data on CoinDesk Markets.

Implications for other market participants

• Prime brokers: The shift from Citadel to Clear Street may encourage other funds to renegotiate terms, potentially reshaping the prime-brokerage landscape.
• Retail investors: The high-visibility nature of the fund’s trades can create “copy-cat” behavior, driving retail inflows into the same AI names and magnifying volatility.
• Liquidity providers: Market makers must price wider spreads on AI stocks during sentiment-driven sell-offs, affecting execution quality for both institutional and retail orders.
• Technology stack: The episode underscores the need for robust risk-engine APIs that can handle rapid changes in exposure, especially when using exotic instruments like flex options.

What to watch next: risk signals and market cues

• Option expiry calendar: Upcoming AI-related options expiries could trigger further price pressure if large positions need to be unwound.
• Sentiment metrics: Track mentions of AI-doom narratives across social platforms; spikes often precede short-term sell-offs.
• Regulatory monitoring: Watch for any SEC commentary on the use of high-leverage and exotic option structures in hedge fund portfolios.
• Infrastructure readiness: Institutions should ensure their execution venues can handle sudden order-book imbalances without excessive slippage.
• Layer-2 activity: A healthy L2 ecosystem can provide alternative liquidity channels; see the latest activity on a layer-2 activity board.

For additional context, the original Protos report provides a detailed timeline of the fund’s re-entry strategy (original Protos report).

The re-emergence of Leopold Aschenbrenner AI fund illustrates that even a shift to ostensibly safer instruments does not immunize a portfolio from market-wide sentiment shocks. Institutions that adapt their market-structure monitoring and execution frameworks accordingly will be better positioned to navigate the next wave of AI-driven volatility.

Related coverage

• Pump.fun tokenized stocks: Creators Launch Coins Priced in Tokenized Stocks
• MetaMask wallet protections: AI-Enhanced Safeguards Against Crypto Scams
• Fly Brain Crypto Trading Experiment Shows Only 1% Loss

Explore more on this topic

• More hedge funds coverage