Mag 7 faces $1 billion ETF exodus as chip stocks surge past software

The Magnificent Seven trade is fracturing as investors pull $1 billion from the group's ETF and rotate into surging chip stocks.

The S&P 500's Magnificent Seven trade is fracturing as investors pulled $1 billion from the group's ETF in June, rotating into surging chip stocks.

"The market has largely assumed that all seven companies benefit equally from artificial intelligence. That theory will prove to be one of the biggest investment misconceptions of this decade," said Nigel Green, CEO and founder of deVere Group.

The Roundhill Magnificent Seven ETF saw $1 billion in outflows during what traders called a black June for the group. Apple shares fell more than 6% after the company acknowledged it could no longer shield consumers from rising memory and storage costs tied to AI data center demand. Memory and storage costs have quadrupled in the past three quarters as suppliers prioritize production for high-bandwidth memory used in AI servers.

Green predicts that within five years, markets will conclude that only three of today's mega-cap tech giants will command premium valuations as AI infrastructure owners, with the others viewed as consumers of AI computing power. The seven stocks — Apple, Microsoft, Nvidia, Alphabet, Amazon, Meta and Tesla — have accounted for a substantial share of S&P 500 gains in recent years, making any rotation a potential headwind for the broader index.

Chip Stocks Surge as Software Lags

The divergence within technology is stark. Chip stocks have experienced a historic surge while software shares continue to underperform, according to a Barron's report. Nvidia, the dominant AI chipmaker, has benefited directly from the infrastructure buildout, while companies spending heavily on AI capabilities face growing scrutiny over returns. Amazon, the worst-performing Mag 7 member with a 33% gain over five years, plans $200 billion in capital expenditures this year — a figure that has drawn skepticism from Wall Street.

The Infrastructure vs. Consumer Divide

Green's thesis draws a sharp line between companies that extract economic rents from the AI ecosystem and those that consume them. "The companies controlling chips, memory, computing power and AI infrastructure will ultimately possess the greatest pricing power," he said. "The companies buying those inputs will continue to thrive, but they may increasingly face margin pressure and lower valuation multiples."

Amazon Web Services, the most profitable part of the e-commerce giant, posted $37.58 billion in first-quarter revenue, up 28.4% from a year earlier, with a 37.6% profit margin. CEO Andy Jassy projected AWS will reach $600 billion in annual sales within a decade, doubling his previous estimate. But the $200 billion capex commitment has left some investors questioning whether the returns will materialize.

For portfolio managers, the fragmentation of the Mag 7 trade carries significant implications for concentration risk in the S&P 500. If the rotation broadens, capital could shift from mega-cap software and consumer tech into semiconductor and infrastructure names. The next test comes in July, when the group begins reporting second-quarter earnings, offering the first hard data on how the AI spending divide is affecting margins.

This article is for informational purposes only and does not constitute investment advice.