Key points
- The Nasdaq closed at a record. The S&P finished 0.4 percent below its own.
- 30 S&P 500 stocks hit new 52-week lows. Only seven hit new highs.
- Up 1 percent near a record with more new lows than highs: only 1929 and 1999 fit.
- My read: respect it, don't sell on a sample of two.
On the surface, Monday was a great day for stocks. The Nasdaq climbed 2.3% to a record 27,122.09. The S&P 500 rose 1.49% to 7,764.70, finishing just 0.4% below its own high.
Then you looked underneath.
Thirty S&P 500 stocks made new 52-week lows Monday. Seven made new highs. On a day the index nearly touched a record, more of its companies were breaking down than breaking out. CNBC has the alarming version. The numbers earn it.
Jason Goepfert at SentimenTrader ran the history. The S&P has only twice risen at least 1% to within 1% of a new high while more stocks made new lows than new highs: July 23, 1929, and December 21, 1999. "We've never in almost 100 years seen breadth this bad," he posted. His firm's price projection just went negative for the first time since they started publishing it about six months ago.
Now, two dates are not a model. They are a coincidence with a great publicist. And neither date was a sell-by-Tuesday signal. The 1929 market did not peak until early September, weeks later, before the October crash. The S&P did not top in 1999 until late March 2000, months later.
Both dates came before later market peaks. That deserves respect. It does not tell you to panic.
So I'm not running for the exits on a two-day pattern. What I am respecting is Goepfert's other number, the one that got less attention. When he loosened the screen to pull in more history, he found 83 days with participation this weak or worse, and by his count the S&P rallied over the next year only 16 percent of the time. That's not a fluke of two dates. That's a base rate, and it's an ugly one.
You don't need the history books to feel it. Monday the Nasdaq closed at a record, and the Dow barely got off the mat, up less than 1 percent. The index's gains were concentrated in a small group of leaders. On the surface it was a great day for the AI trade: AMD tagged $1 trillion for the first time and Intel jumped double digits. Breadth is the part of that day nobody put on the highlight reel. Art Hogan, chief market strategist at B. Riley Wealth, put it plainly to CNBC: "The leadership's battling against weaker performance in the near term, and what's selling off has been selling off, so the creation of new lows has an easier glide path than the creation of new highs with today's leadership." That's the whole story in one sentence. The winners keep winning, and everything else has an easy road lower.
Here's where I land. Narrow markets can stay narrow longer than the people shorting them can stay solvent. I have watched that movie enough times not to short a record high just because the internals look thin.
But thin internals do change what I want to own. On a day when 30 S&P stocks make 52-week lows, and seven make new highs, I want to know whether my names are acting like leaders or passengers. If I cannot answer that quickly, that is my warning. I do not need a chart from 1999 to tell me.
Hogan's list was conditional: a war that persists, energy prices that stay high, and a Fed that may have to keep hiking. None of that helps the hundreds of stocks outside megacap tech.
I am not selling Tuesday morning. I am also not adding to a tape this narrow and calling it brave. Sometimes the honest move is to hold what is working, know why you own it, and let the leaders carry the market until they cannot.
Just do not confuse a record on the screen with a healthy market. Monday, those were two different things.
Index levels are as of the market close on September 21, 2026. Percentage and breadth figures are as reported by CNBC and SentimenTrader, or by our arithmetic from closing quotes where noted.
