Citi: The Mag 7 Label Is Dead — AI Winners Need to Be Redefined
Claire Weston
Citi strategist Scott Chronert declares the Mag 7 framework "dead" and proposes a broader growth cluster covering over 50% of S&P 500 market cap — the lens for picking AI winners is shifting from seven star names to an entire value chain.
How did the Mag 7 "die"?
The core evidence: price correlations among the seven stocks have broken down — they no longer move in lockstep.
Microsoft and Meta fell as markets questioned the payoff on massive capex; Apple, by contrast, rallied 23% year-to-date precisely because it sat out the data-center arms race.
This means → bundling seven companies under one label no longer reveals who is winning and who is losing. The framework has lost its analytical value.
What is the "growth cluster" — and why does it replace Mag 7?
The growth cluster — a new basket Citi proposes — covers the largest-cap tech names plus most AI-infrastructure plays, accounting for over 50% of S&P 500 market cap and roughly 48% of index earnings.
In plain terms = instead of "seven giants vs. the other 493," the frame is now "the full AI supply chain vs. the rest" — finer-grained and wider.
Chronert argues this grouping more accurately reflects the S&P 500's fundamentals and price action than the old binary.
Is the valuation stretched? Can earnings hold up?
The growth cluster trades at a 12-month forward P/E in the 66th percentile of its 30-year range — not cheap, but far from historical bubble territory.
2027 earnings expectations remain firmly supported, the foundation that keeps the valuation standing.
This reflects a market still optimistic on AI commercialization — only the criteria for picking winners have changed.
Where is the AI trade rotating?
Chronert forecast as early as last December that the AI investment theme would shift from "enablers" (the pick-and-shovel sellers — chipmakers, cloud providers) to "adopters" (companies using AI to boost their own businesses).
That rotation has played out through 2026: semiconductor stocks led in the first half, then came under pressure on stretched valuations.
This means → the next leg of outperformance may belong not to those building AI, but to those making money with it.
What does the FAANG precedent tell us?
Chronert draws a direct parallel with FAANG — the old acronym for Facebook, Amazon, Apple, Netflix, and Google: "When was the last time you thought about FAANG?"
In plain terms = market labels have a natural shelf life. Once an industry evolves, the old tag can't keep up — FAANG faded, and Mag 7 is following the same path.
His conclusion: the Mag 7's exit is not an endpoint — it is the starting gun for a new map of AI winners.
Content is for reference only, not financial advice.