Stats › Concentration and breadth › Index concentration
S&P 500 concentration and the Magnificent 7
The S&P 500 holds five hundred companies, but it doesn't hold them equally. Each one counts in proportion to its market value, so the biggest names carry the index and the smallest barely register. Right now the ten largest hold 38.5% of it. If you own an S&P 500 fund, roughly two fifths of your money is in ten companies, and that's true whether or not anyone ever told you.
This page tracks that weight, how it compares to earlier periods, which companies make up the top of the index, and how much of the index's return those few names actually produced. For three years the answer was most of it. The Magnificent 7 drove 62% of the return in 2023 and 53.7% in 2024. Then the first half of 2026 flipped: the same seven stocks subtracted 1.96 percentage points from the index, and the share of companies beating the index jumped from 30.5% to 46.3%.
I built the page partly because that reversal is the most interesting thing in the dataset and it got much less attention than the concentration itself did, and partly because there are two widely circulated numbers here that are simply wrong. I've dealt with both in the note at the bottom rather than repeating them. Six sources feed this, each with its own date, all listed underneath. Where two of them count differently, I've explained why instead of picking the tidier number. Check them if something looks off.
The ten largest companies made up 38.5% of the S&P 500 as of 30 April 2026, against roughly 19% in 1990 and 23% at the end of 2000. The Magnificent 7 alone were 33.4% at the end of June 2026. But the concentration story reversed this year: after driving 62% of the index return in 2023 and 53.7% in 2024, the Magnificent 7 *subtracted* 1.96 percentage points from the index return in the first half of 2026, and the share of constituents beating the index jumped from 30.5% to 46.3%.
Members only
The rest of this page is for members
Below this point there are 10 sections, 3 charts, 1 table and 6 named sources, roughly 1400 words of it. Every figure carries the source it came from and the date the data is from.
You can keep browsing every statistic in the library for free. The intro and the summary are always open.