Half of S&P 500 stocks sit at least 16% below their highs as the 10-year yield reaches 5.27%

Goldman says a 16% median drawdown from 52-week highs is similar to the bubble era of the early 2000s ^[10]. For the median stock to be that far down while the index is within a point of its record ^[10], the largest weights have to be close to their own highs. The source attributes this to large AI-related stocks with heavy weightings ^[11]. It does not name them or give their combined share of the index.

The multiple has already fallen. The 12-month forward P/E is 19, down from 23 a year earlier ^[12], so a buyer pays about 17% less for each dollar of expected earnings ^[2]. The index set its record in mid-August and sits about 1.5% under it by the source's own figure (Goldman's is 1%) ^[3][10], so it is at worst 1.5% below where it stood a year ago. Put that beside the lower multiple and forward earnings must be at least about 19% higher than a year earlier ^[3], consistent with State Street's 30%-plus growth forecast for this year ^[7]. Investors have not sold ^[3]. The forward earnings yield is about 5.26%, one over 19 ^[4], against a 10-year that touched 5.27% ^[1].

Axios's decomposition puts about 92% of the past month's five-year move in real yields ^[1]. The source notes that this sits somewhat at odds with the account of oil-driven inflation worries pushing Treasury yields up ^[5]. Axios links the real-rate rise to jobless claims at a 57-year low and to large-scale corporate investment centered on AI ^[6]. The prevailing view reported in the source is that the same investment is the main driver of corporate results ^[8].

The outcomes split three ways. Should earnings at the lagging half of the index start rising toward that forecast, breadth repairs itself with yields still above 5%. On the rates path, Bank of America's 7% threshold ^[9] sits 1.73 points above the intraday high ^[5]. A slowdown in AI investment would probably ease real yields, and it would also remove the earnings growth behind the stocks holding up the index ^[6][11].

I think the third outcome is the one the index is least prepared for. The median stock has already absorbed its 16% ^[10]. The heaviest weights carry an earnings outlook tied to the AI spending that Axios says is pushing real rates up ^[6][11]. A change in that spending would move the discount rate and the earnings being discounted at the same time. Bank of America's 7% figure is the counter-thesis: on its reading, a 10-year in the 5% range is well short of the level at which equity investors turn ^[9].