Hedge funds and mutual funds are taking increasingly different approaches to the artificial intelligence trade, with hedge funds retaining greater overall exposure to AI-related stocks while the two investor groups have diverged sharply on some of the sector’s biggest names, according to a report by Seeking Alpha.
Goldman’s analysis of institutional portfolios found that hedge funds continued to have a strong link to the broader AI trade, while mutual funds remained underweight the theme. Mutual funds have increased their exposure to companies benefiting from AI infrastructure spending this year, but their allocations remain below benchmark levels.
The biggest divergence emerged among major technology companies. Hedge funds increased their holdings of Microsoft and Amazon during the quarter, while mutual funds reduced exposure to both stocks.
At the same time, hedge funds reduced positions in a range of other prominent AI names, including Alphabet, Meta Platforms, Nvidia, Broadcom, Lam Research, Marvell Technology, Cisco Systems, Hewlett Packard Enterprise and Applied Materials.
Mutual funds took a more constructive view of several semiconductor and memory companies. A majority of managers increased holdings of Advanced Micro Devices, Micron Technology and Sandisk, even as hedge funds reduced their exposure to those stocks.
There was greater agreement around the infrastructure required to support the expansion of AI. Goldman identified a group of 12 companies that were bought by both hedge funds and mutual funds during the second quarter.
The list included power companies American Electric Power, NiSource and Xcel Energy, as well as data-centre and technology infrastructure businesses including CoreWeave, Flex, Sanmina, SiTime, Seagate Technology and Talen Energy. AXT, Bloom Energy and Legence were also among the companies attracting buying from both investor groups.
However, the two groups also cut exposure to several AI-related stocks during the quarter. Those included Viavi Solutions, Digital Realty Trust, Argan, MasTec, Corning and EQT, highlighting the increasingly selective nature of institutional positioning across the AI ecosystem.
Goldman said the contrasting trades came despite hedge funds maintaining a much stronger overall relationship with the AI theme. Returns generated by hedge funds and their most widely held stocks have remained closely tied to movements in AI-related equities, while a basket of stocks favoured by mutual funds has shown considerably less exposure to the theme.
The analysis covered 991 hedge funds with about $5.4tn in gross equity positions and 504 large-cap active mutual funds with roughly $4.6tn in equity assets. Together, the portfolios represented around $10 trillion of institutional equity positioning at the beginning of the third quarter.
The findings suggest that the AI trade is becoming increasingly fragmented. Rather than simply adding exposure to the largest technology companies, investors are increasingly distinguishing between the chipmakers, software groups, data-centre operators and power providers expected to benefit from the next phase of AI investment.
