Public Pension Exposure to AI-Related Companies Climbs as Firms Grow

As major tech companies continue to pour money into artificial intelligence resources, public pension funds invested in these firms are increasingly exposed to the sector.

This is highlighted by a recent report from Equable Institute that estimates that the 25 largest pension systems hold between $578 billion and $707 billion worth of assets in firms it deems AI-related companies.

According to the report, approximately 85% worth of those assets are concentrated in 15 companies: Apple, NVIDIA, Microsoft, Alphabet, Amazon, Meta, Tesla, Broadcom, Micron, Advanced Micro Devices, Intel, Lam Research, Applied Materials, KLA Corp., and SpaceX. The first seven, also known as the “Magnificent Seven,” accounted for 63.3% of the exposure, with the other eight accounting for approximately 21.8%.

Much of the exposure, however, is likely due to the fact that state and local pension funds often use market-cap-weighted passive index strategies and allocate a significant portion of their portfolios to public equities, particularly those that are included in the S&P 500 Index. All of the companies on the report’s top-15 list are included in the index, except for SpaceX, which is not yet eligible to be on the index because its stock has not yet traded for 12 consecutive months.

In conducting the report, Equable said it reviewed Form 13F reports, quarterly disclosures filed by investors with the Securities and Exchange Commission, for the 25 largest public pension funds. For funds that do not file the form, AI exposure was projected from annual listings of public equity holdings.

Equable also said the report does not fully account for private equity investments in AI-related companies, corporate debt for AI-related companies, or other private debt investments in the companies.

While Equable did not define what it considers to be an AI-related company, it’s fair to assume that the more than 50 companies listed in the report fall into this category by its standards.