Diversification: your 'diversified' index fund is a 40% bet on ten stocks

Ten companies hold about 40% of the S&P 500 — Nvidia alone is about 8%. The index still owns 500 names, but its fortunes now ride on a handful of AI-linked giants.

Five hundred names, ten drivers

The S&P 500 is the world's most popular definition of diversification: 500 leading American companies, covering about 80% of the available United States market capitalization. As of May 29, 2026, Investopedia counts 503 constituents worth an estimated $67 trillion in total. One fund, hundreds of companies, no stock picking required — that is the pitch, and for decades it was roughly true.

But the index does not give each company one vote. It weights every member by market value, so the biggest companies get the biggest slices. Morningstar's Brett Arends puts the result plainly: the ten largest companies in the S&P 500 now account for 40% of the entire index's value. Put $100 into a fund tracking the index and $40 of it goes to ten companies, with the other $60 spread across the remaining 490. Morningstar's Wei Dai and Mamdouh Medhat give the longer view: at the end of 2025 the top 10 held 41% of the index, up from 19% a decade earlier — and as of June 23, all ten, about 37% of the index, are technology and AI-focused. Information technology alone is 38.6% of the index.

One name can move the whole index

The concentration is not abstract. Nvidia alone is about 8% of an investor's money in a standard S&P 500 fund, according to Arends. This week, Nvidia shed $573 billion in market value — the largest weekly loss for any company on record, according to The Wall Street Journal and MarketWatch. When one name is 8% of a 500-stock index, a bad week for that name is a bad week for the whole index, and the other 499 names cannot vote it down.

Concentration cuts both ways, and 2024 showed the upside of the same arithmetic. Dai and Medhat note that the S&P 500 returned 25% that year, driven heavily by the Magnificent Seven, which returned 48%, while the other stocks in the index collectively returned 16%. The giants carried the index up — the same weight that now pulls it down when they fall. That is the point of this lesson: a fund can hold 500 names and still behave like a bet on a handful of them.

Priced ahead of the rest

The top ten are not only heavy; they are expensive relative to the rest. Dai and Medhat report that the ten largest companies command a weighted average price-to-earnings ratio of 37x, or 40% higher than the remaining index constituents. The premium is the market's judgment that these companies will keep growing faster — a judgment that has mostly been right, and that leaves less room for error if it stops being right.

History offers a useful comparison. Research by Hendrik Bessembinder of Arizona State University, cited by Arends, found that over the 55 years from 1971 through 2025, an equal-weight strategy — every stock given the same weight — returned 12.6% a year, against 11.3% for the traditional value-weighted approach, a gap of 1.3 percentage points a year. Bessembinder's research also shows that the highest-value stock in the market has usually ended up underperforming the rest of the index over time. None of that makes the index fund a bad product; it makes the label on the tin worth reading.

What 'diversified' still means

None of this changes the basic fact that the index still owns 500 companies. What has changed is what drives them. The earlier lesson in this series was that diversification works on behaviour, not on headcount — and the index is now the largest-scale example of that lesson. When the ten largest holdings all answer to the same artificial-intelligence story, the offsetting math of diversification weakens: a broad-market fall and a mega-cap fall become nearly the same event.

Dai and Medhat argue that investors can lower that concentration by looking beyond the United States: the ten largest stocks in the MSCI All Country World ex USA index made up roughly 9% of its total value in 2015 and 13% in 2025 — a smaller share, rising more slowly. Their broader point is the one to take away: diversification is a property of what a portfolio holds and what moves it, not of how many names appear on the statement. The number worth watching is the top ten's share of the weight — not the count of companies in the fund. Educational content, not investment advice.

Sources

• Morningstar, 2026-11-14
• Morningstar, 2026-11-14
• Investopedia, 2026-11-14
• The Wall Street Journal, 2026-11-14
• MarketWatch, 2026-11-14