US hedge funds completely unwound their 2025 artificial intelligence stock purchases during an intense five-day selling streak in October, according to prime brokerage data from Goldman Sachs. The rapid liquidation contributed to a $2 trillion selloff across the Nasdaq-100 and major tech stocks, even as broad equity indexes like the S&P 500 experienced little overall movement through late October.
Scale of the Hedge Fund Deleveraging
According to Morgan Stanley, hedge funds have cut more than 8% of their total equity positions since June 2026, marking their largest retreat from stock bets since at least 2017. Data from Goldman Sachs reveals that during the five-day exit, funds sold existing long shares 3.5 times faster than they opened new short positions against stocks, netting approximately $20 billion in sales.
This aggressive unwinding heavily impacted semiconductor and hardware manufacturers. Memory chipmakers Samsung, SK Hynix, Micron, and SanDisk collectively lost over $700 billion in market value in less than two weeks. On October 22, Goldman's basket of hedge funds' favorite stocks fell 4.2%, recording its worst single-day performance since 2011.
Key Takeaways
- Complete AI Reversal: Hedge funds erased all net 2025 AI stock purchases over just five consecutive trading sessions in October.
- Record Deleveraging: Morgan Stanley reported an 8% reduction in overall fund equity exposure since June, the largest drawdown since 2017.
- Severe Sector Losses: Memory chipmakers lost more than $700 billion in under two weeks during the broader US tech stock selloff.
- Elevated Exposure Remains: Despite the massive liquidation, Goldman Sachs notes institutional AI holdings still rank in the top 17% of the past three years.
Leverage Unwind vs. Underlying Demand
Market observers emphasize that the drop was driven by leverage liquidation rather than declining end-user demand. Macro investor Jordi Visser noted that the accumulated leverage took 18 months to build up but only 75 days to unwind. Similarly, Fundstrat's Tom Lee attributed market pressure to forced institutional liquidation rather than macro headwinds.
Fundamental metrics remain sturdy across the sector. Broad corporate profits are up approximately 20%, and SK Hynix forecasts that high-bandwidth memory supply shortages could persist until 2030. However, Goldman Sachs warned that the hedge fund unwind may be closer to its beginning than its conclusion, as fund positions in AI equities remain high relative to historical ranges.
Why It Matters
The sudden liquidation of institutional AI exposure highlights how concentrated leverage can trigger severe equity drawdowns even while corporate earnings and underlying computing demand remain strong. As hedge funds rapidly de-risk across traditional markets, cross-asset liquidity conditions can tighten, indirectly affecting high-beta sectors including cryptocurrency. Investors should monitor whether hedge fund equity exposure stabilizes or if forced selling spreads into broader market benchmarks like the S&P 500.
