Shiller CAPE ratio (S&P 500)
Cyclically adjusted price-to-earnings ratio using 10-year inflation-adjusted earnings.
Reading
Second-highest reading in 144 years of data. December 1999 was the only higher observation. Current level is being held at 4–5% on the long bond, not zero.
Thresholds
- watch38 — approaches the November 2021 peak
- alarm44 — matches or exceeds the December 1999 peak
Context
Why this matters
The Shiller CAPE smooths the earnings denominator across a full cycle, which removes the noise that ordinary trailing P/E carries. On 144 years of data it has the strongest correlation with subsequent ten-year real returns of any single price-based measure with that length of track record. The current reading sits at the second-highest level in the series — only December 1999 has exceeded it — and is being held there at long bond yields of 4–5%, not zero.
Who watches this
- Robert Shiller — the originator; still publishes the underlying series monthly
- John Hussman (Hussman Funds) — built MarketCap/GVA as a more rigorous successor; integrates CAPE in his monthly market comment
- Jeremy Grantham (GMO) — uses CAPE as one input to GMO's published 7-year asset-class forecasts
- Cliff Asness (AQR) — papered the empirical case that CAPE works on a multi-decade horizon
- David Einhorn (Greenlight) — references CAPE in his quarterly letters, particularly when paired with concentration metrics
Recent history
Crossed 40 in late 2024 and has held there. The November 2021 peak of ~38.6 was lower and was held with zero rates; the current 42 is held with positive real yields. The structural defence ("this time is different — AI") echoes 1999 closely.
What would change my read
A meaningful earnings catch-up across the broad index (not just the top-7) on the order of 30%+ real over 3–4 years would bring CAPE back inside its long-run band without a price drawdown. The historical base rate for that is very low; it has happened twice in 144 years.