The Nasdaq is at a record high, but there's talk that more individual stocks are hitting new lows than new highs. Isn't that exactly the same signal as right before the 2000 dot-com bubble burst?

Looking at the numbers, this really is a familiar pattern

As today's report's Reddit section noted, the S&P500 is within 0.8% of its record high, while the equal-weight S&P500 is down 5%, small caps down 6.4%, and mid caps down 7%. As of September 18, out of 4,742 total stocks, only 51 hit 52-week highs versus 267 that hit 52-week lows — new lows exceeded new highs by 216 names (compiled market data, 2026-09-18). The index is climbing while most individual stocks are quietly hitting new lows.

Concentration is also at a historic extreme. The Magnificent Seven's weight in the index has swelled from 12.5% in 2016 to 34.8% as of May 2026 (compiled market data, 2026). Multiple analyses agree this is the first time index gains have been this concentrated in a handful of names since late 1999-early 2000 (Bloomberg, Morningstar, 2026).

How did 2000 actually end

The setup in late 1999 was strikingly similar. The S&P500 kept setting record highs thanks to a small number of internet and telecom stocks, while the majority of listed stocks had already slipped into a quiet bear market. When that narrow leadership rolled over in early 2000, trillions of dollars in market value evaporated over the following two and a half years. The Nasdaq fell 77% from its peak, bottoming at 1,139.90 in October 2002, and the S&P500 dropped nearly 50% over 2000-2002. It took the Nasdaq 15 years to reclaim that peak (not until a new record in April 2015, per compiled historical data).

But there's a clear counterargument that this time is different

There's also a strong case that 2026 differs from 2000 in one decisive way. The 2000 bubble was built on speculation in a wave of internet companies with no viable path to future profit, whereas today's rally is led by the Magnificent Seven and large-cap semiconductor names that are generating substantial actual earnings. "The fact that most of today's high-flying stocks are AI-related draws comparisons to the 1999-2000 mania, but to those who lived through it, it doesn't feel the same" is a view that has been raised (TheStreet, 2026). In short, the concentration figure (34.8%) matches the historical extreme, but the quality of earnings behind that concentration is different — that's the core counterargument.

Scenario branches

This week's hawkish FOMC dot plot (16 of 18 members signaling further hikes this year) tilts the odds toward the first or third scenario rather than the second (rotation). The longer rates stay elevated, the harder it becomes for companies outside the small group of mega-caps to improve earnings given higher funding and borrowing costs.

So what should investors do

Don't take comfort just because the index headline reads "record high." A practical approach is to track weekly whether the gap between new-high and new-low counts, and between the equal-weight and cap-weighted indices, is widening or narrowing. If the gap keeps widening, the next earnings reports from a handful of leaders (9/24 Meta Connect, followed by semiconductor and Big Tech earnings) become a single point of risk for the entire index — so rather than holding only index-tracking products, it's worth considering blending in an equal-weight product (like RSP) or some exposure to beaten-down small/mid caps to diversify concentration risk. Conversely, if this gap starts narrowing alongside earnings improvement, that's a sign of a healthy rotation rather than a 2000-style collapse.