Hedge funds pulled back from U.S. technology stocks at a record pace over the past two months, according to Goldman Sachs Group’s Prime Services desk. FinancialMediaGuide views the retreat as one of the clearest signs yet that sophisticated investors, not just retail traders, are growing genuinely uneasy about how far this year’s AI-driven rally has run.
The cohort was a net seller of the technology sector in six of the past eight weeks, according to the desk led by Vincent Lin. The cumulative reduction in market value totaled about 10%, marking the largest such retreat from the sector since Goldman’s data series began more than a decade ago.
“Amid continued volatility and sharp selloff across the semis/memory/AI infrastructure complex, the persistence and magnitude of selling since early June point to significant length reduction by tech investors, and some signs of capitulation are starting to emerge,” the desk wrote. FinancialMediaGuide notes that Goldman’s use of the word “capitulation,” typically reserved for describing panic selling near a market bottom, suggests the bank sees this unwind as more than routine profit-taking.
The S&P 500 Information Technology Index has dropped some 10% since early June as investors take profits and grow more skeptical that AI-fueled valuations are sustainable. Investors have been rotating into other sectors, including consumer stocks, amid mounting concern that the biggest technology companies could eventually pull back on the flood of spending underpinning the AI infrastructure buildout.
Goldman noted that technology was the worst-performing and by far the most net-sold U.S. sector last week, as investors cut long positions and increased short bets. Most subsectors were net sold, led by technology hardware, storage and peripherals and IT services, with semiconductors and software seeing comparatively less selling pressure. FinancialMediaGuide points out that the unevenness of the selling, concentrated in hardware and services rather than spread evenly across the sector, suggests investors are making increasingly selective bets on which parts of the AI trade are overextended rather than abandoning the theme wholesale.
Goldman strategists led by Ben Snider said the “painful volatility in popular AI infrastructure stocks” has renewed investor interest in other sectors. “History, positioning, and lack of a favorable catalyst point to continued near-term challenges for the AI infrastructure Momentum trade despite solid fundamentals,” they wrote, distinguishing between weak near-term positioning and the sector’s still-intact long-term earnings power.
The gap between Goldman’s acknowledgment of “solid fundamentals” and its warning of “continued near-term challenges” captures the core tension now facing tech investors: the earnings case for AI infrastructure spending remains largely intact, even as positioning and sentiment have turned sharply negative. Financial Media Guide concludes that until a “favorable catalyst” emerges, in Goldman’s own words, hedge funds appear more inclined to keep de-risking their AI exposure than to bet on an imminent rebound, regardless of how strong the sector’s underlying fundamentals remain.
