Key Takeaways
- • Hedge funds rotated out of the largest technology stocks in October 2026, with the Nasdaq Composite down about 5% and the S&P 500 down about 1%, while the S&P 500 Equal Weight Index gained about 5%, according to Data Insights, a division of Hazeltree.
- • Short fund participation rose across Amazon, Apple, Nvidia and Meta, while Tesla was the only Magnificent Seven name where short holders declined and it saw a moderate increase in long holders.
- • Apple's long-to-short fund count ratio fell to 0.65x, its 2026 low and a third consecutive month below 1.0x, while Meta, Amazon and Microsoft each hit their own 2026 lows on the same measure.
- • The SOX semiconductor index fell roughly 16% in October, and AMD flipped from long-biased to short-biased, with its net long exposure dropping from 77.9% to 44.4%.
- • CoreWeave showed AI infrastructure credit stress, with its 2032 bonds yielding close to 15% and its stock down about 37% during October.
Summary
The Data Insights Crowding Report for October 2026, formerly the Hazeltree Crowding Report, shows a sharp rotation by hedge funds out of the largest U.S. technology stocks. According to Data Insights, the institutional hedge fund market intelligence platform and a division of Hazeltree, the Nasdaq Composite fell about 5% and the S&P 500 fell about 1% in October, while the S&P 500 Equal Weight Index gained about 5%. Non-technology sectors performed well during the month. Tim Smith, managing director of Data Insights, described the month as one that "ushered in a sharp rotation out of some of the largest technology stocks as some of the major indices fell, but in an interesting contrast non-technology sectors performed well."
Within the Magnificent Seven, short fund participation rose notably across Amazon, Apple, Nvidia and Meta, with Microsoft seeing a moderate increase and Alphabet a modest one. Tesla stood out as the only name where short holders declined, and it posted a moderate month-over-month increase in funds holding long positions. Nvidia and Alphabet posted modest gains in long holders, while Apple declined moderately. The long-to-short fund count ratio for Apple fell to 0.65x, its low for 2026 and a third consecutive month below 1.0x. Meta, Amazon and Microsoft each reached their own 2026 lows on the same measure, and Alphabet held just above parity at 1.02x.
Semiconductor sentiment turned less bullish. The share of PHLX Semiconductor Sector Index constituents showing net long positioning slipped to 70.0%, and the SOX index fell roughly 16% during the month. AMD flipped from long-biased to short-biased, with its long-to-short fund count ratio falling from 1.18x to 0.94x and net long exposure dropping from 77.9% to 44.4%. Crowding on the long side is now most pronounced in Nvidia, followed by Broadcom and Applied Materials, while on the short side Entegris is the most crowded name, followed by ON Semiconductor and AMD.
AI infrastructure credit also came under pressure. Smith said CoreWeave saw its default insurance costs widen and its 2032 bonds yield close to 15%, approaching levels typically associated with distressed debt. CoreWeave's long-to-short fund count ratio declined from 0.60:1 in September to 0.53:1 in October, and its stock fell about 37% during October, including a 15% drop on October 14. Net long exposure measured in quantity weakened from about 80% in the first four months of the year to 55.9% as of October.
The report covers hedge fund long and short crowding across the Americas, EMEA and APAC, based on anonymized data covering approximately 16,000 securities and more than 700 global funds on Hazeltree's securities-finance platform. Data Insights defines its crowding score as a relative metric normalizing the number of funds long or short a security within a region and market-cap group against its peers.
Within the Magnificent Seven, short fund participation rose notably across Amazon, Apple, Nvidia and Meta, with Microsoft seeing a moderate increase and Alphabet a modest one. Tesla stood out as the only name where short holders declined, and it posted a moderate month-over-month increase in funds holding long positions. Nvidia and Alphabet posted modest gains in long holders, while Apple declined moderately. The long-to-short fund count ratio for Apple fell to 0.65x, its low for 2026 and a third consecutive month below 1.0x. Meta, Amazon and Microsoft each reached their own 2026 lows on the same measure, and Alphabet held just above parity at 1.02x.
Semiconductor sentiment turned less bullish. The share of PHLX Semiconductor Sector Index constituents showing net long positioning slipped to 70.0%, and the SOX index fell roughly 16% during the month. AMD flipped from long-biased to short-biased, with its long-to-short fund count ratio falling from 1.18x to 0.94x and net long exposure dropping from 77.9% to 44.4%. Crowding on the long side is now most pronounced in Nvidia, followed by Broadcom and Applied Materials, while on the short side Entegris is the most crowded name, followed by ON Semiconductor and AMD.
AI infrastructure credit also came under pressure. Smith said CoreWeave saw its default insurance costs widen and its 2032 bonds yield close to 15%, approaching levels typically associated with distressed debt. CoreWeave's long-to-short fund count ratio declined from 0.60:1 in September to 0.53:1 in October, and its stock fell about 37% during October, including a 15% drop on October 14. Net long exposure measured in quantity weakened from about 80% in the first four months of the year to 55.9% as of October.
The report covers hedge fund long and short crowding across the Americas, EMEA and APAC, based on anonymized data covering approximately 16,000 securities and more than 700 global funds on Hazeltree's securities-finance platform. Data Insights defines its crowding score as a relative metric normalizing the number of funds long or short a security within a region and market-cap group against its peers.
Go deeper than the headline
Research every firm in this story, track the ones that matter and get the daily briefing — free for 7 days.