Hedge funds dumped US technology stocks at the fastest pace in more than a decade last week, a four-standard-deviation event that signals a structural unwinding of the AI trade.
Hedge funds dumped US technology stocks at the fastest pace in more than a decade last week, a four-standard-deviation event that signals a structural unwinding of the AI trade.

Hedge funds recorded the largest net selling of US information technology stocks in over a decade last week, with the selloff spanning semiconductors, mega-cap names and sector ETFs in a four-standard-deviation event.
The net selling of US IT stocks in the week through June 25 registered a z-score of minus 4.0, Goldman Sachs Prime Brokerage data show, driven by both long and short selling at a ratio of 1.3-to-1. Semiconductors and semiconductor equipment accounted for more than half of the dollar amount, marking eight consecutive sessions of net selling. The so-called Magnificent Seven — Apple, Microsoft, Amazon, Nvidia, Alphabet, Meta and Tesla — were net sold for a fifth straight week, with combined net exposure falling to the 6th percentile of the past three years.
"This is not the end of the AI investment cycle — we remain in the midst of a historic investment frenzy — but the market's judgment on net winners and losers is undergoing a profound reassessment," said Mark Wilson, a partner in Goldman Sachs' sales and trading division.
US equity funds posted $8.5 billion in net outflows for the week, the first weekly withdrawal since March 2026 and a sharp reversal from the prior week's record $119.2 billion inflow, EPFR data show. The S&P 500 fell about 2% on the week, while large-cap technology stocks dropped roughly 6%. The Russell index's annual rebalancing amplified trading volumes, though asset managers executed the selling in an "orderly" fashion without signs of panic, according to Goldman's trading desk.
Of the 11 S&P 500 sectors, eight recorded net selling. Information technology led the declines, followed by communication services, industrials and healthcare. Consumer staples, energy and real estate were net bought. The materials sector posted its largest weekly net sell in more than three months at minus 1.9 standard deviations, with metals and mining stocks seeing their long-short ratio fall to 1.21 from this year's high of 1.69 in early June.
The crowding problem behind the volatility
The selling coincided with extreme positioning signals across the hedge fund industry. Overall fund leverage reached all-time highs, while the correlation between US and Asian AI-related positions hit the 99th percentile on a historical basis, Goldman's prime brokerage strategy team found. Since the start of the year, hedge funds have been large net sellers of US assets while increasing exposure to Asia, primarily Japan and Korea.
That crowding became visible in Korea, where the Kospi index triggered circuit breakers twice in a single week — accounting for 20% of all such events this century. About 60% of the index's weight is concentrated in just two stocks, Samsung and SK Hynix, underpinned by tens of billions of dollars in leveraged ETF replication products. Recent capacity expansion plans from both companies have alerted the market that the storage industry's historically high profit margins will eventually face supply-driven pressure.
The 10-year US Treasury yield fell 8 basis points to 4.37% during the week, while crude oil dropped about 9%, providing some support to consumer-related sectors. The VIX, while not explicitly cited in the data, reflected elevated volatility as the rotation accelerated.
For investors, the key takeaway from the current turbulence is that the AI trade can no longer be treated as a monolith. The market is now drawing sharp distinctions between companies spending heavily on AI infrastructure with uncertain returns, those facing margin compression from rising costs, and those benefiting directly from that capital expenditure. That differentiation will be a defining variable for equity allocation in the second half of the year.
This article is for informational purposes only and does not constitute investment advice.