Stock Markets May 15, 2026 01:18 PM

Institutions Pile Into AI Infrastructure Stocks in Q1 2026

13-F filings show widespread buying of data-center and AI infrastructure names while some tech giants see mixed institutional flows

By Sofia Navarro MSFT ORCL DLR META

Institutional investors increased exposure to companies tied to artificial intelligence infrastructure during the first quarter of 2026, according to an analysis of 13-F filings with the U.S. Securities and Exchange Commission. Thousands of funds added or started positions in a set of nine AI infrastructure-related companies, while data-center equities and utilities also attracted net buying. At the same time, some large-cap AI leaders saw more measured demand, and software-as-a-service names suffered net liquidation in aggregate.

Institutions Pile Into AI Infrastructure Stocks in Q1 2026
MSFT ORCL DLR META

Key Points

  • Over 4,000 institutional filers increased or initiated positions in nine companies tied to AI infrastructure, including Oracle, Arista Networks and Vertiv.
  • Data-center stocks such as Digital Realty and utilities attracted net buying, with nearly 3,500 filers reporting purchases and no recorded sellers of utilities by the reporting cutoff.
  • Investors were selective toward large AI-focused technology giants like Meta and Microsoft, where sellers slightly outnumbered buyers; SaaS stocks overall experienced more liquidation than accumulation.

An analysis of quarterly 13-F filings submitted to the U.S. Securities and Exchange Commission indicates that institutional investors were broadly buyers of firms connected to artificial intelligence infrastructure in the first quarter of 2026. The filings, which capture holdings reported by nearly 6,000 hedge funds, pension funds, college endowments and other asset managers, show extensive accumulation of stock in a cohort of companies central to AI rollouts and deployments.

More than 4,000 institutional filers either increased existing stakes or established new positions in a group of nine companies identified as prominent players in AI infrastructure. That group includes Oracle, Arista Networks and Vertiv. By contrast, only 146 of the reporting institutions reduced or sold holdings in that specific cluster - roughly 2.5% of the filers that had reported positions by the reporting cutoff.

These quarterly 13-F reports provide a window into how major investors adjusted portfolios over the three-month period, revealing where buying interest concentrated and where selling prevailed. Institutional investors that meet the reporting threshold are required to disclose any portfolio changes within 45 days after the end of each calendar quarter. The dataset used for this analysis reflects filings that had been submitted to the SEC’s database as of mid-morning on Friday and therefore does not capture moves made after April 1.

In addition to AI infrastructure names, the filings also show substantial buying of data-center companies and utilities. Data-center firms, a category that includes Digital Realty, attracted notable inflows, and the filings show utilities experienced consistent demand: there were no reported sellers of utilities stocks during the quarter among those filers, while nearly 3,500 reported net buying.

Market-level snapshots included a mix of performances and flows. Tickers printed alongside the filings in trading summaries showed Microsoft up 4.28%, Oracle down 0.16%, Digital Realty down 2.09%, Meta unchanged, Arista Networks down 2.04%, Shopify up 4.12%, Vertiv down 1.25% and Palantir up 0.98% in intraday quotes embedded in reporting screens.

While the AI infrastructure group drew broad institutional participation, the filings suggest that investors exercised greater selectivity toward some of the largest AI-focused technology companies often grouped among the so-called Magnificent Seven. That subset, which includes Meta and Microsoft, experienced more balanced flows; according to the 13-F data, sellers narrowly outnumbered buyers for those large-cap AI leaders during the quarter as institutions assessed whether those companies could maintain heavy AI-related spending and growth trajectories.

Palantir drew interest from a subset of institutions. Among the 143 investors that initiated positions in Palantir during the quarter was Mubadala Capital, the UAE sovereign wealth fund, which recorded a small holding valued at $9.9 million in its filing. Mubadala also initiated a position in Shopify, a software-as-a-service company that faced heavy selling pressure during the quarter amid concerns that AI could disrupt SaaS business models and impair profitability.

More broadly across the U.S.-listed SaaS cohort analyzed, institutional behavior skewed toward selling: for a group of 20 such stocks, 397 filers liquidated one or more positions over the quarter. By contrast, in the semiconductor sector institutional demand was robust, with more than 4,100 investors either adding to or initiating positions as chip stocks began to rally sharply.


What the filings show

  • Widespread institutional accumulation of AI infrastructure names: over 4,000 filers increased or began positions in nine key companies.
  • Data-center and utilities stocks saw net buying, with utilities showing no sellers among the filers recorded by the cutoff.
  • Differentiated flows across technology: some mega-cap AI firms saw more measured buying, while SaaS names experienced net liquidation.

Risks

  • Uncertainty about whether large AI-focused technology companies can sustain the pace of AI-related spending and growth, which influenced more cautious institutional demand in that cohort - affecting large-cap tech and AI-related sectors.
  • Potential disruption from AI to software-as-a-service business models contributed to heavy selling in the SaaS group, creating execution and profitability risk for companies in the software sector.
  • File submission timing limits: the 13-F dataset only includes filings submitted by the mid-morning on the reporting cutoff and does not capture portfolio moves made after April 1, introducing a reporting lag that may obscure subsequent reallocations across sectors such as semiconductors and data centers.

More from Stock Markets

Coinbase Retreats After Regulatory Win Fails to Translate Into Certainty May 15, 2026 Lincoln International IPO Sees Strong Demand as Bookbuild Nears Close May 15, 2026 DexCom Shares Leap After Elliott Stake, $1 Billion Buyback and Ambitious 2030 Targets May 15, 2026 West Pharmaceutical Services reports broad operational recovery after cyberattack May 15, 2026 Moody's Assigns Aa2 to Alphabet's New Yen Notes as Ratings Firms Affirm Debt Quality May 15, 2026