The Magnificent Seven are Apple, Microsoft, Amazon, Alphabet, Meta, Nvidia and Tesla. The label was popularized in 2023, succeeding FAANG, and it caught on because it described something true at the time: a handful of very large companies were rising together and dragging the index up behind them.

The case for retiring it is that the first half of that description has stopped being true.

What a grouping is for

A label like this is a claim that the members share a fate. It is useful precisely when knowing one company's direction tells you something about the others.

That is what has broken down. Bloomberg reported in July that the correlation within the group had fallen to its lowest since 2017, well down from the spring. AllianceBernstein put the same point plainly in a research note titled "no longer a monolith," arguing that these are seven companies with distinct economic exposures rather than one trade.

The dispersion is visible in returns. CNBC reported that the group as a whole had fallen into negative territory for the year by the start of July even as the wider market rose, with individual members far apart.

The reason they diverged

The split is not random. It follows a line through the middle of the group, and the line is capital spending on artificial intelligence.

Four of the seven are hyperscalers building enormous data center capacity. That spending runs through the accounts in a specific and painful way: it converts cash into fixed assets now against revenue expected later. It depresses free cash flow, and free cash flow is what funds share buybacks.

The others are differently placed. Tesla's capital intensity is about vehicles and autonomy rather than data centers. Nvidia is on the receiving end of everyone else's spending, which is a fundamentally different position: it books as revenue what the hyperscalers book as capital expenditure. Apple has taken a lighter approach to building its own AI infrastructure.

So one company sells the shovels, several are digging, and the market has begun pricing them accordingly. Bloomberg noted that a semiconductor index has substantially outperformed the seven this year, which is the same observation from another angle: the AI trade has not weakened, it has moved to a different part of the supply chain.

Analysts at Goldman Sachs have argued the group will underperform an equal-weighted version of the S&P 500 this year, according to Bloomberg's reporting.

Why it matters to people who never picked a stock

The reason this is not merely a naming argument is concentration.

By Forbes' account, these seven companies account for around a third of the S&P 500 by weight, a degree of concentration in a single cohort without modern precedent in the index. Anyone holding a standard index fund holds that concentration whether or not they intended to.

While the seven moved together, that was a straightforward bet on large-cap technology. Now that they have separated, the index outcome depends on which of them does well, which is a subtler exposure than most passive investors realize they have taken on. Some have shifted toward equal-weighted funds, which reduce the effect.

The case against retiring it

The counterargument is reasonable and should be stated.

These remain the largest companies in the index, and all seven are exposed to the same broad technological shift, even if at different points in it. A shorthand does not need to imply identical performance to be useful; it can simply denote the set of companies whose scale makes them systemically important to the index. On that reading, the label still picks out a real thing, and the dispersion is just what happens inside any group over time.

The more practical objection is that shorthand of this kind tends to outlive its accuracy. FAANG survived for years after Netflix had stopped belonging in the same category as the others. The risk is not that the phrase is wrong, but that it keeps being used as though the seven still move as one, when the market has already stopped treating them that way.

Nothing here is investment advice. Figures are as reported by the cited sources on the dates given.