(Bloomberg) — The popular Magnificent Seven stocks moniker is “no longer relevant” in assessing how to play the US artificial intelligence trade, according to Citigroup Inc. strategists.
Most Read from Bloomberg
The team led by Scott Chronert said investors should instead focus on a broader swath of equities that have been the dominant driver of both earnings and share-price gains in the S&P 500 Index.
They recommend a so-called “growth cluster,” which comprises the biggest technology stocks as well as a majority of names tied to the AI infrastructure buildout. The cohort makes up more than half of the S&P’s market capitalization and contributes nearly 48% of its earnings, Chronert said.
“The Mag 7 is dead as a construct for assessing large-cap growth dynamics,” Chronert wrote in a note.
His views are underpinned by the rotation trade playing out in US stocks. After powering the S&P 500 to record highs in the past few years, the Bloomberg Magnificent Seven Index has underperformed in 2026 as investors favor sectors that stand to benefit from massive spending on AI.
The correlation in share prices of the seven stocks — Amazon.com Inc., Nvidia Corp., Meta Platforms Inc., Apple Inc., Microsoft Corp., Tesla Inc. and Alphabet Inc. — has also broken down. While Microsoft and Meta have declined on skepticism around the payoff from billions of dollars in capital expenditure, Apple has surged 23% on relief about its decision not to participate in the data center arms race.
Semiconductor-related stocks led the rally in the first half of 2026, but even they have begun to underperform in recent weeks amid concerns about lofty valutations. At the same time, some strategists including Societe Generale SA’s Manish Kabra have cautioned it’s too early to buy into the big AI spenders.
Chronert correctly predicted in December that the AI trade would shift away from the enablers to the technology’s adopters in 2026.
He said in his latest note that growth cluster valuations remain historically attractive. The group’s 12-month forward price-to-earnings is in the 66th percentile relative to the past 30 years and remains supported by strong earnings expectations through 2027, he said.
