The market looks strong at the index level, but fewer stocks are participating. Here’s why concentration beneath the S&P 500 matters.
KEY TAKEAWAYS
The S&P 500 reached a record above 7,800, but participation underneath the index has recently weakened. The headline index is increasingly telling a different story from many of the stocks beneath it.
Breadth data is flashing an unusual divergence. On Monday's Nasdaq record close, 57 stocks made new 52-week highs while 243 made new lows.
Mega-cap concentration allows a very small number of companies to dominate index performance. The ten largest S&P 500 companies now represent roughly 40% of the index, making today's “500-stock” portfolio considerably less diversified than the name suggests.
Even the AI trade is becoming more selective. Winners such as Ciena and Nebius surged while Seagate and Western Digital fell sharply, showing investors are beginning to discriminate among companies exposed to the same theme.
High Treasury yields make concentration more important. With the 10-year Treasury above 5%, investors have a substantial risk-free alternative competing against richly valued equities.
MY HOT TAKES
A record high by itself tells increasingly little about the health of this market. In a capitalization-weighted index this concentrated, a small group of enormous companies can disguise considerable weakness underneath.
The most interesting signal isn't that mega-cap technology is strong—it is that the rest of the market is beginning to behave differently. Divergence becomes much more important when it appears while indexes are setting records.
This is not yet a “sell everything” signal. It is a portfolio-awareness signal telling investors to calculate how much exposure they actually have to the same handful of companies across multiple funds and accounts.
The fragmentation inside AI may ultimately matter more than the broad AI narrative. Markets appear to be moving from “buy anything connected to AI” toward demanding proof that massive spending will actually produce profits.
A 5%+ Treasury yield changes the math. Mega-cap equities no longer compete against nearly free money, which means earnings, valuations, financing costs, and execution matter more than they did during the first phase of the AI boom.
You can quote me: “When the headline number and the colors underneath it stop agreeing with each other, believe the colors.”
Colorful. You know me by now. I am up before the sun–well, before OUR sun, anyway. It's already up ON OTHER CONTINENTS, which is precisely why I'm awake. By the time the first espresso hits the cup, the screens are on, and in my office there are a LOT of them–enough that a visitor once asked whether I was monitoring the markets or preparing to launch something. I have spent the better part of four decades bathed in that glow, and I'll let you in on a little secret, I don't actually read the screens first. I read the COLORS. Before my brain processes a single number, my eyes (green ones, which my family insists are my signature, and which I insist are simply well-suited to the job) have already taken the temperature of the room. A wall of green tells you one thing. A wall of red tells you another. But a patchwork–a patchwork tells you to put the espresso down and look closer.
Yesterday, the big number at the top of the screen said exactly what this morning's headlines are shouting–the S&P 500 closed above 7,800 for the first time EVER. A record. Champagne. Green, green, green. And yet, as my eyes drifted down from that headline number, the colors didn't line up the way a record day is supposed to look. Seagate was down more than 9%. Western Digital was down nearly 7%. The small caps–trailing the big indexes on a day when everything was supposed to be celebrating. Up at the top, a handful of giants doing all the heavy lifting. Down below…a patchwork. If an almost-lifetime of staring at screens has taught me anything, it's this 👉when the headline number and the colors underneath it stop agreeing with each other, believe the colors.
So let me do exactly that–put the espresso down and look closer. First, a quick word about what a "record" actually measures. The S&P 500 is a capitalization-weighted index (meaning each company counts in proportion to its total stock-market value, so a $5 trillion company moves the needle roughly a thousand times more than a $5 billion one). That's not a flaw–it's the design, and it has served investors well for decades. But it means a record close tells you that the WEIGHTED AVERAGE went up. It does not tell you that most stocks went up. I know you probably know this, but I think that it’s important to underscore it. On most record days those two things travel together. Lately they've been traveling separately–and that's the patchwork my eyes caught yesterday. Wall Streeters call it breadth (simply, how many stocks are participating in a move), and right now breadth is telling a very different story from the headline.
Here's an interesting point that didn't make many headlines. On Monday–the day the Nasdaq Composite set its own record close at 27,477–the Nasdaq logged 57 new 52-week highs. That’s right, fifty-seven. Against 243 new 52-week LOWS. Read that again. 👀 On a record day for the index, more than four times as many Nasdaq stocks hit a one-year low as hit a one-year high. The S&P 500 told a quieter version of the same story–13 new highs against 20 new lows. Now, one day of breadth data is a snapshot, not a trend, and I won't pretend otherwise. But when the champagne is flowing at the top of the screen and four times as many names are visiting the basement as the penthouse, you don't need a Doctorate in finance 😉 to see that not everyone on the guest list showed up to the party.
Then came Tuesday. By late morning, it became pretty clear that the market was essentially being held up by three stocks–NVIDIA, Apple and Microsoft. Three. Out of…um, five hundred. And look at what was happening INSIDE the AI trade itself, which is supposed to be the one thing everybody agrees on. Ciena jumped more than 11% and Nebius more than 9%, while Seagate and Western Digital–two names that had been among the hottest on the board as investors chased anything connected to data-center memory and storage–got taken to the woodshed as traders reassessed how much demand and profit that boom would actually deliver. Same theme. Same day. Wildly different outcomes. Even the leadership isn't marching in lockstep–it's a patchwork within the patchwork.
Why can a handful of names carry so much weight? Because over the past several years they have grown into giants. By most measures, the ten largest companies now make up close to 40% of the S&P 500's total value. Think about what that means. For every dollar you put into an S&P 500 index fund, roughly forty cents goes into just ten companies, and the other sixty cents gets spread across the remaining 490. Back in 1985, the top ten accounted for about 21% of the index. In 1995, about 18%. Not gonna lie, I got those numbers from Claude–but I knew it was less at the time, BECAUSE I WAS THERE–at least in ‘95. 🤣 Today's share is roughly double that–one of the highest levels of concentration in modern market history. Plenty of investors believe they own "the market" because they own an index fund. They do. They just own a far more concentrated version of it than the words "five hundred" suggest.
Now, in fairness–and I always want to give you the whole picture, not just the scary half–this has NOT been a year in which only the giants won. As of the close on October 1, the small-cap Russell 2000 was up 11.90% for the year, essentially neck and neck with the S&P 500's 11.78%. The rest of the market has been participating. What I'm describing is a narrowing that has shown up recently, and most conspicuously on the very days the headlines are celebrating. That distinction matters. A market that is narrow all year is one thing. A market that narrows as it makes new highs is something else entirely–and it's worth watching.
So why does this matter now, and not just as a bit of trivia? Because the backdrop has changed. Last Thursday, the yield on the 10-year Treasury Note touched 5.31%, its highest level since May 2002–this morning, it’s kissing 5.34%. Stocks are setting records while a risk-free government bond pays north of 5%–that is a much higher hurdle than investors faced when the giants began their run. And the AI buildout that fuels so much of that leadership isn't being paid for entirely with cash. AI-related companies have issued more than $1.5 TRILLION in corporate debt this year–trillion, with a T. I'm not predicting anything. But I spent the late 1990s doing tech M&A during the CLEC telecom boom, and I remember what it felt like when a handful of shooting stars carried the averages while more and more of the market quietly stopped coming along for the ride. The champagne tasted great. The hangover……was memorable. Nobody rang a bell at the top, and almost nobody looked at the breadth until it was far too late to matter.
None of this is a reason to panic, and it's certainly not a call to sell anything. It is a reason to ask yourself a few honest questions. How much of my portfolio actually rides on a handful of companies–once I add up my index fund, my tech fund, my 401(k) target-date fund AND the individual shares I bought on my own? Would I buy those few names individually, at today's prices, in the amounts I effectively already own them? And if the giants had a rough quarter while everything else was merely flat, how would that feel–not in theory, but on my actual account statement? Keep an eye on the clock today, too. The Fed releases the minutes from its September meeting at 2 PM Wall Street Time, and with the 10-year sitting near 24-year highs, anything that moves rates moves the giants–and therefore moves "the market."
Tomorrow I'll be up before the sun again–well, before OUR sun–with the espresso in one hand and the screens already glowing. I'll glance at the big number at the top, like everyone does. And then–like always–I'll let my eyes drift down. The headline number will tell you where the market closed. The colors will tell you who actually showed up. And when those two stop agreeing with each other–I will–once again, believe the colors. 😉
YESTERDAY’S MARKETS
The S&P 500 rose 44 points, or 0.58%, to 7,818 yesterday, its first close above 7,800, while the Nasdaq Composite added 0.45%, to a record 27,599. The Dow Jones Industrial Average gained 253 points, or 0.49%, to 51,521, while the small-cap Russell 2000 was the only major index to decline, slipping 0.57%. Treasury yields eased across the curve, with the 10-year yield falling about 3.6 basis points to roughly 5.28% heading into the close. Crude eased yesterday.
NEXT UP
This afternoon at 2:00 PM Wall Street Time, the Fed will release minutes from its September 16th meeting. YES–the hiking meeting. Wanna get a clue about what the committee was thinking and whether or not we might expect more hikes? Read the minutes!
Important earnings today: Levi Strauss and Applied Digital.