Index of One

There is a comforting story that tens of millions of people tell themselves about their money, and it goes like this. I am not a stock-picker. I do not gamble. I own a low-cost index fund that tracks the S&P 500 — five hundred of the largest companies in America — so my savings are spread across the whole economy. Banks and railroads and drugmakers and retailers and software firms. If one stumbles, the others carry me. I am diversified.

It is a good story. It is also, as of 2026, largely untrue.

The index those tens of millions believe spreads their money across five hundred businesses now concentrates a third of it in seven. The "Magnificent Seven" — Nvidia, Apple, Microsoft, Alphabet, Amazon, Meta and Tesla — make up roughly 34% of the entire S&P 500. Stretch the lens to the ten largest companies and you capture nearly 40% of the index in ten names, leaving the other 490 to divide the remaining 60% among themselves. The market-capitalization-weighted index fund, the single most popular investment vehicle ever created, sold to the public for half a century as the very definition of prudent diversification, has quietly transformed into something close to its opposite: a concentrated, leveraged wager on a handful of technology companies riding one story.

The story is artificial intelligence. And the thing about owning the S&P 500 today is that you are far more exposed to it than you think.

How the spread disappeared

Concentration of this kind is not unprecedented, but its current degree is. The ten largest companies accounted for about 22% of the S&P 500 as recently as 2020. They are pushing 40% now. The Magnificent Seven's share has climbed from around 12% in 2015 to roughly 34% today — nearly tripling in a decade. By some measures, the five largest companies now constitute the greatest share of the US market in half a century, and the phenomenon is not confined to America: the same handful of names has swelled to around a fifth of the entire MSCI World index, so that an investor in a "global" fund in Tokyo or Frankfurt is, increasingly, buying the same seven California-adjacent stocks as everyone else.

At the center of the center sits Nvidia, which became the first company in the history of capitalism to be worth $5 trillion, and which alone accounts for more than a fifth of the Magnificent Seven's combined value. A single chipmaker is now a large enough share of the global equity market that its fortunes move the retirement accounts of people who have never heard its name and could not say what it makes.

This did not happen because a cabal rigged the index. It happened because the index is doing exactly what it was designed to do — weight companies by size — at a moment when a few companies have grown monstrously large. That is the trap hidden in the machinery of passive investing: the more a stock rises, the bigger its weight becomes, so the index fund is compelled to hold more of it precisely as it gets more expensive. Indexing does not buy low and sell high. It buys whatever is winning, in ever-greater proportion, automatically, with your money, until the winners are nearly the whole thing.

The word that stopped meaning anything

"Diversified" is doing an enormous amount of unearned work in the sales pitch for index funds, and it is worth being precise about why it no longer means what savers think.

Diversification protects you when the things you own are not all exposed to the same risk — when a bad quarter for banks can be offset by a good one for healthcare. But the Magnificent Seven are not seven independent bets. They are, to a striking degree, one bet wearing seven jerseys. Nvidia sells the chips; Microsoft, Amazon, Alphabet and Meta are its largest customers, spending hundreds of billions of dollars buying those chips to build AI data centers; their stock prices rise and fall together on the same news about the same technology. When the AI trade is in favor, all seven levitate as one. When it falls out of favor — as a single disappointing earnings guidance from one chip company demonstrated when it cracked the entire Korean market in a day — they will fall as one, and they will take the "diversified" index, and your "prudent" retirement fund, down with them.

This is not a hypothetical. The last time this group of leaders broke, in 2022, the lesson was delivered in full. As the S&P 500 fell 20% that year, the Magnificent Seven fell about 41% — roughly twice as hard. The concentration that supercharges the index on the way up does precisely the same thing in reverse. The passive investor who enjoyed the seven stocks dragging the whole index higher has unknowingly signed up to have the same seven stocks drag it lower, with leverage, exactly when it hurts most.

The quiet exodus to "equal weight"

The professionals have noticed, and their response is the tell.

Across 2025 and into 2026, one of the most consistent recommendations from strategists has been a migration from the standard, capitalization-weighted S&P 500 to its equal-weighted sibling — a version of the same five hundred companies in which Nvidia counts no more than the five-hundredth-largest name. Goldman Sachs has forecast that the Magnificent Seven will underperform the rest of the market in 2026, and Goldman's asset-management clients have been raising the same structural worry in meeting after meeting: that as a handful of mega-caps swell, the index's volatility becomes hostage to a few stocks, and the diversification that justified owning it in the first place quietly erodes to nothing.

When the smart money starts paying extra to strip the biggest winners out of the index — to own the boring 493 instead of the glamorous 7 — it is making a statement about where it thinks the risk now lives. It lives in the part of your portfolio you were told was the safe part.

The mirage, and the morning after

None of this means the Magnificent Seven are bad companies, or that the index will crash tomorrow. These are, by almost any measure, the most profitable enterprises in human history, and their dominance of the index partly reflects genuine, earned, extraordinary earnings power. An equal-weighted index has its own cost: for years it lagged badly precisely because it underweighted the winners. The bull case — that these companies are so good they deserve their weight — is not foolish, and concentration alone has never been a timing signal. Markets can stay concentrated, like they can stay irrational, far longer than the cautious can stay patient.

But the savers being sold the S&P 500 as a diversified holding deserve to understand what they actually own, because almost none of them do. They own a portfolio in which a single trade — long artificial intelligence, financed increasingly with debt, priced for a future that must arrive on schedule — has grown to a third of the whole. They own an index that has stopped being a cross-section of the American economy and become a concentrated position in seven correlated stocks, dressed in the language of prudence and sold with the promise of safety.

The label still says "diversified." The contents say otherwise. And the gap between the two is not a problem on the way up, when the seven stocks carry everything before them. It is a problem on exactly one morning — the one nobody schedules — when the trade that became the index turns out to have been a single trade all along.

You think you own five hundred companies. It would be worth finding out, before that morning, how few you really own.