The S&P 500 is trading less than 1% below its all-time closing high, yet fewer than 30% of its member stocks are above their own 50-day moving averages — a divergence that has only appeared a handful of times since 1990, according to index data tracked by S&P Dow Jones Indices.

The gap between the index and its average component is the widest since early 2000, when the dot-com bubble was at its peak. Back then, the S&P 500 fell 49% over the following 30 months.

Nvidia, Microsoft, Apple, Amazon, Meta and Alphabet now account for roughly 30% of the S&P 500's total market capitalization, up from about 18% five years ago. When those six stocks rise, the index rises — regardless of what the other 494 companies are doing.

That math explains how the headline index can approach record territory while the typical stock is underwater. Equal-weight versions of the S&P 500, which give every company the same influence, are trailing the cap-weighted index by more than 10 percentage points over the past 12 months, according to Invesco, which runs the largest equal-weight ETF.

The Russell 2000, which tracks 2,000 smaller U.S. companies, is down slightly year-to-date even as the S&P 500 has gained. Small-cap stocks tend to carry more floating-rate debt, and the Federal Reserve has held its benchmark rate in a range of 5.25% to 5.50% since July 2023 — the highest level in 23 years.

Market breadth — the share of stocks participating in a rally — matters because narrow advances are more fragile. If the handful of leaders stumble, there is no broad base of other stocks to cushion the fall. The S&P 500's 10 largest holdings now trade at an average forward price-to-earnings ratio above 30, compared with a long-term median near 18 for the index as a whole.

Bank of America's fund manager survey, released Tuesday, showed institutional investors' cash allocations at 4.2% of portfolios, below the 5% threshold that typically signals caution. The same survey found 62% of managers naming "long mega-cap tech" as the most crowded trade on the planet.

None of this guarantees a reversal. Concentration can persist for years — the Nifty Fifty era of the early 1970s ran for more than a decade before collapsing. But the current setup means the S&P 500's next 1% move will depend far more on six companies than on the 494 others combined.