Archive · 15 Aug 2026
Institutional investors cautious on tech favorites in quarterly 13F filings
Big investment companies, which manage money like retirement savings, must tell the government every three months which stocks they own.
These reports are called 13F filings.
The newest reports showed that many big investors did not buy more shares of the biggest technology companies.
Almost as many investors sold these stocks as bought them.
This means investors cannot agree on what will happen next with these companies.
Some investors already own as much stock as their rules allow them to own.
Computer chip companies were still popular — more investors bought them than sold them.
Many investors also bought shares of companies that work on artificial intelligence.
Oil and energy companies were not very popular this time.
These reports help us understand what the most powerful investors in the world are thinking.
Nearly 44% of institutional investors trimmed holdings of the Magnificent Seven megacap tech firms, while 42% initiated or expanded positions.
Semiconductor stocks kept a bullish tilt, with 48% of filers net buyers versus 34.5% net sellers.
Tiger Global Management cut stakes in Microsoft, Nvidia and Meta and reduced its Alphabet exposure by 45.4% to 5.8 million shares.
AI-themed stocks drew net buying from 36% of institutions, including CoreWeave, Arista Networks and Broadcom.
Energy drew limited interest, with 40.3% of filers net sellers of major oil firms against 28% net buyers.
- Who
- Institutional investors, including pension funds, hedge funds and wealth managers such as Tiger Global Management and OnyxPoint Global Management.
- What
- Institutional investors slightly pulled back from megacap technology, semiconductor and AI infrastructure stocks, with buys and sells closely matched in quarterly 13F filings.
- When
- For the quarter ended June 30, based on filings reviewed as of early Friday afternoon (August 14).
- Where
- United States, based on Securities and Exchange Commission filings analyzed by Reuters and reported from Providence, Rhode Island.
- Why
- Risk parameters limited how much more large holders could buy, and investors disagreed about which AI-related companies would ultimately profit.
Signals of Uncertainty
Risk Limits and Crowded Trades
What the balanced buy-sell gap means for big tech
Signals of Uncertainty
The closely matched numbers of buyers and sellers signal the absence of consensus, with disagreement about which companies will ultimately profit from heavy AI spending.
Risk Limits and Crowded Trades
Large firms may simply be long as much as they want to be or should be under their risk parameters and investment policies, so the data does not reflect a changed fundamental outlook.
Cause of the July tech selloff
Signals of Uncertainty
Hot AI names had moved from a fundamental growth story into a highly leveraged momentum trade, making technology stocks vulnerable to sharp reversals.
Risk Limits and Crowded Trades
The selloff was less a rejection of the long-term AI thesis and more a classic crowded-trade unwind amplified by leverage and inadequate risk controls.
- Filings analyzed
- 6,371 pension funds, hedge funds, wealth managers and other institutions
- Magnificent Seven trims
- Nearly 44% of filers
- Magnificent Seven adds
- 42% of filers
- Semiconductors
- 48% net buyers vs 34.5% net sellers
- AI-themed stocks
- 36% net buyers
- Tiger Global Alphabet cut
- 45.4% reduction, to 5.8 million shares
- Oil sector
- 40.3% net sellers vs 28% net buyers
- Data centers
- 24.3% net buyers and 24.3% net sellers
