5.21% Against 5.26% (29 September 2026): The S&P 500's Earnings Yield Has Caught the 10-Year Treasury — and Earnings, Not the Multiple, Are Carrying the Index

At 19.2 times forward earnings, the S&P 500 now earns about 5.21% a year on its price — and the 10-year Treasury closed at 5.26% on Tuesday 29 September. On 30 June the stock market's forward earnings yield sat 46 basis points above the bond; three months of rising yields have closed that gap to zero. The index is still up on the quarter only because analysts raised earnings estimates by 8.9% while the multiple they are willing to pay fell by about 6%.

What actually happened: the numbers side by side

The equity figures come from FactSet's Earnings Insight of 25 September 2026, which prices the index as of the 24 September close of 7,704.13. The bond figures are Treasury's own daily par yield curve and real yield curve. The earnings yield is simply the P/E turned upside down: 1 ÷ 19.2 = 5.21%.

The 29 September equity column is an estimate, and it is worth being explicit about how it was built: FactSet's P/E and price imply a forward EPS estimate of about 401 index points. Holding that fixed and applying the 29 September close of 7,670.84 — reported by AP after a 0.2% decline — gives a multiple near 19.1. Estimates move week to week, so the sign of a three-basis-point gap is not the point. The point is that a 46-point cushion has gone.

Two further records sit behind the table. The 10-year's 5.26% matches its close of 12 June 2007 on Treasury's series. The 30-year's 5.59% is the highest close on the 30-year constant-maturity series since 5.61% on 17 December 2001.

Price equals earnings times multiple: how the index rose while the P/E fell

The cleanest way to read the quarter is as an identity. An index level is its expected earnings multiplied by the price the market pays for each unit of them. FactSet reports that between 30 June and 24 September the forward 12-month EPS estimate rose 8.9% and the index rose 2.7%. The multiple therefore had to fall — from 20.4 to 19.2, about 5.9%.

That split matters because the two halves are driven by different things. Earnings revisions are driven by company results and guidance — FactSet has CY 2026 earnings growth at 32.0% and Q3 at 29.1%, the third straight quarter above 25% if it holds. The multiple is driven largely by what investors can earn elsewhere, and the thing they can earn elsewhere just got a great deal more attractive. Through the third quarter the earnings line won that tug of war comfortably. What changes when the cushion is gone is not that it must now lose, but that it has to keep winning by more to hold the index where it is.

The bond side: a real-yield move, not an inflation scare

The obvious reading of a 10-year at 5.26% during a war that has lifted oil is that markets fear inflation. Treasury's own curves say otherwise. From 30 June to 29 September the 10-year nominal yield rose 82 basis points; the 10-year TIPS yield — the return after inflation — rose 71. The breakeven inflation rate, the difference between them, went from 2.24% to 2.35%: about 11 basis points.

That distinction runs straight into equities. If yields were rising because investors expected more inflation, some of that inflation would eventually show up as higher nominal earnings, cushioning the hit. When the real yield does most of the rising, there is no such offset — the market is demanding a higher real return to lend for ten or thirty years, and that is a higher hurdle for every other long-duration asset, stocks included. The storyline behind the long end — Treasury's buybacks, Japanese yields, and the term premium — is traced in detail in the site's long-end selloff piece.

Two versions of the same comparison — and why they disagree

Setting the earnings yield against the nominal 10-year is sometimes called the "Fed model", and on that version the equity premium has vanished. But the comparison has a well-known flaw: a bond pays a fixed nominal coupon, while earnings are a claim on businesses whose revenues and profits tend to rise with the price level over time. An earnings yield therefore behaves more like a real yield than a nominal one.

Compare it with the TIPS yield instead and the picture changes: the gap was 2.70 percentage points on 30 June and about 2.32 on 29 September. Narrower, by roughly 38 basis points, but still positive and still wide by the standard of a decade in which the 10-year real yield spent most of its time below 1%.

Neither version is a forecast, and neither says stocks are cheap or expensive. What both say is the same thing from different angles: the compensation for holding equities rather than government bonds has shrunk this quarter, and the only reason the index has not fallen with it is that expected earnings kept rising. The trailing 12-month P/E of 25.8 — above its 5- and 10-year averages of 24.4 and 23.6, per FactSet — is a reminder that the forward multiple looks moderate only because the forward earnings line is so steep.

Where the pressure lands: duration inside the index

Rising discount rates do not hit every stock equally. FactSet's sector table shows Consumer Discretionary (22.9x) and Industrials (22.8x) on the highest forward multiples and Energy (13.3x) and Financials (14.5x) on the lowest. The higher the multiple, the more of the price rests on profits far in the future, and the more a given rise in long yields removes.

Energy is the mirror image this quarter. FactSet reports its Q3 dollar earnings estimate up 18.0% since 30 June, with oil 36% higher over the window ($69.50 to $94.61), and the sector's expected Q3 growth rate at 111.4%. A low multiple on a rising earnings line is the least rate-sensitive combination in the index; a high multiple on a flat one is the most.

For index traders the practical channel runs through concentration. NAS100 and the largest US500 weights are dominated by technology names whose valuations rest heavily on long-dated growth — the Information Technology sector's Q3 earnings estimate rose 4.1% since 30 June, led by Nvidia, according to FactSet. That is the earnings line doing its job. The question for the fourth quarter is whether it keeps outrunning a discount rate that has already moved 82 basis points.

What would change the picture

Three things, each checkable against data rather than opinion:

• Real yields reverse. A fall in the 10-year TIPS yield would widen both versions of the gap at once. The September jobs report on 2 October is the nearest catalyst for the front end, and the site's jobs report preview sets out the scenarios for yields and the dollar.
• Earnings revisions stall. The quarter's gain rests on an 8.9% rise in the forward EPS estimate. Q3 reporting season tests it directly; Micron, which reports on 30 September, is covered in the site's earnings preview.
• The demand side cracks. The Conference Board's consumer confidence index fell 6.7 points to 81.9 in September, and AP reported it as the lowest level in 12 years. High real yields plus a weakening consumer is the combination that would pressure the earnings line and the multiple at the same time — the one scenario in which the two halves of the identity stop offsetting each other.

The currency angle

For FX traders the same split explains why the dollar's response to the long end has been inconsistent. A rise in real yields at the front of the curve tends to support the dollar through the interest-rate factor. A rise concentrated in the long end, driven by term premium and accompanied by softer equities, can instead weigh on risk sentiment — and the site's coverage of real yields and currencies explains why those two channels can pull in opposite directions. The US dollar page shows how both are currently scoring.

The takeaway

The equity market has not been cheapened by falling earnings; it has been re-priced by a rising discount rate, and it has held up only because earnings expectations kept climbing faster. On 30 June a buyer of the S&P 500 earned 46 basis points more in forward earnings yield than a buyer of the 10-year note. On 29 September that premium is roughly zero against the nominal bond and about 2.3 points against the real one. How that gap is read — and whether earnings can keep carrying the index — is what Q3 season is about to test.

To learn how Pip Theory builds its fundamental currency-strength scores, see the methodology overview.

Educational macro context only — not investment advice.