The note that circulated on trading desks this week carried an uncomfortable title: Last Man Standing.
Its author, Jonathan Krinsky, the chief market technician at BTIG, was describing a market in which the Magnificent Seven technology giants are the last large group of stocks still making fresh highs while most of the index they dominate has already turned down. The index itself sits within a few percentage points of a record. Beneath it, roughly 60% of S&P 500 constituents have fallen more than 20% from their own all-time highs, according to data compiled by Michael Cembalest of JPMorgan Asset Management.
A decline of that size, sustained, is the common definition of a bear market. On that measure the average American large-cap stock has been in one for months, even as the headline index has not.
Breadth this poor is rare. Strategists who have run the numbers say the participation rate was last this weak in two periods: the bear market of 1973 and 1974, and the peak of the technology bubble in 1999. Both were followed by losses that the index-level numbers did not anticipate.
The mechanics are easy to state and hard to trade. Apple, Microsoft and Nvidia together account for about 22% of the index’s market capitalisation, up from roughly 14% in late 2022, according to market data. Passive funds that track the S&P 500 therefore buy those three names in proportion to their weight, whatever the investor’s view. The same crowding appears in active portfolios, where the same seven stocks are held for growth, for quality, for AI exposure and increasingly for the perception of safety, as Mr. Krinsky wrote in his note to clients.
When a stock is owned for four reasons at once, weakness in any one of them has fewer willing buyers to absorb it. That is the risk the note describes, and it is why the phrase “Something’s gotta give” recurred in the research that followed it.
The gap between the index and its components shows up in the products built to track it. The equal-weighted version of the S&P 500, which holds every constituent at the same size, has trailed the capitalisation-weighted index by a wide margin this year, a reversal of the pattern that prevailed when money was cheap and a small number of balance sheets could carry the market. Investors who bought breadth as a hedge against concentration have instead paid for it.
Any rotation would start from a low base. With three-fifths of the index already more than 20% below its own high, the pool of stocks that could lead a broader advance is the same pool that has spent the year falling. Strategists describe that as the uncomfortable part of the setup: a broadening that requires beaten-down shares to recover is a different event from a broadening that arrives with the leaders still rising, and the two produce very different index outcomes.
The earnings calendar offers the near-term test. The Magnificent Seven are expected to have grown third-quarter earnings by 20.3% from a year earlier, according to estimates compiled by Russell Investments, while the other 493 companies in the index are expected to grow by 27.7%. That would invert the pattern of the past two years, when the largest technology companies doubled profits while the rest of the market stalled. If the forecast is right, the case for owning only the giants weakens without anyone having to sell them.
Two other threads run through the same argument.
The first is credit. The capital spending that has driven the leaders’ earnings is increasingly financed with borrowed money. The International Monetary Fund said this week that hyperscale issuers now take about 10% of global investment-grade bond supply, up from less than 1% in 2024, and that hedge funds’ share of the Treasury market has more than doubled since 2022. Concentration in equity ownership has a cousin in concentration among lenders, and the two would tighten together in a disorderly move.
The second is the semiconductor complex, which has spent the past two sessions falling while the index held. Memory names led the decline after a Chinese chipmaker disclosed progress on a denser DRAM design and a Nvidia-backed data-centre operator withdrew a planned multibillion-dollar listing in Australia. The leaders’ earnings depend on that chain continuing to spend; a wobble there reaches the Magnificent Seven’s income statement long before it reaches their share price.
None of this is a forecast of a decline. Bad breadth has persisted for long stretches before, and narrow markets can grind higher for quarters while the average stock goes nowhere. Analysts who track positioning note that the largest companies’ balance sheets are stronger than the index’s average constituent, which is part of why capital gathered in them in the first place.
What the numbers describe is an index that has stopped summarising its own market. A portfolio manager who owns the S&P 500 today is running a concentrated bet on seven companies and calling it diversification, and the hedge against being wrong is smaller than it looks. The strategists who flagged the pattern are not predicting what triggers a reversal. They are noting that a market carried by one cohort has less room to absorb a surprise, and that a small push is enough to produce a large move when nothing else is holding the index up.
The back half of October will be read for one thing: whether earnings from the other 493 companies broaden the rally or confirm that the index’s strength and the market’s health have stopped being the same fact.
