Excess CAPE Yield

The CAPE earnings yield minus the ten-year real interest rate — what stocks yield over bonds after inflation, on Shiller’s own convention. Low is expensive.

+0.87pts of yield over bonds
p9.1 · deep red — thin vs bonds zero +0.87 pts away
In its deep red — thin vs bonds level since Sep 2026 (1 month).
Low ECY = expensive — a thin yield edge over bonds is the stretched state; this lens’s scale inverts the other three.
Tier boundaries today: p10 at +0.95 · p25 at +1.73 · p75 at +5.29 · p90 at +7.62.

Excess CAPE Yield since 1963

1987-111991-031994-081998-012001-062004-122008-072012-022015-082019-022022-082026-09−30369
  • Excess CAPE Yield
  • thin vs bonds (the expensive extreme)
  • wide vs bonds
  • zero

Low ECY = expensive — the inversion is the point of this lens.

The zero line is absolute: below it, bonds out-yield stocks. One period in the computed record — May 1999 → November 2000.

The zero-crossing register

the periodS&P +1y from the crossingS&P +3y
May 1999 → Nov 2000 · 19 months below zero+11.9%−20.3%

One period in the computed record. A named state, not a timing signal. The windows open the month the crossing month's reading was published, with its CPI.

The tier ladder

tierperiodsS&P med +1ymed +3yworst +3y
deep red — thin vs bonds TODAY9+12.8%+26.7%−8.9%
pale red25+12.2%+30.6%−30.5%
neutral14+15.6%+52.0%−19.6%
pale green8+14.6%+54.8%+17.0%
deep green — wide vs bonds7+18.8%+50.0%+21.0%

Counts and medians of the collapsed register — the same periods listed below; medians are records, not rankings (the one-year and three-year columns do not order the same way). Low ECY = expensive: the red rungs are the thin-margin states.

Tier occupancy (post-stabilization, Mar 1963 → Sep 2026, n=762)

tiermonthssharelongest period
deep red — thin vs bonds15019.7%30m (1996-03..1998-08)
pale red11515.1%9m (2025-05..2026-02)
neutral39251.4%108m (2009-05..2018-04)
pale green618%23m (1975-03..1977-01)
deep green — wide vs bonds445.8%15m (1977-02..1978-04)

Shares of the 762 post-stabilization months (Mar 1963 → today), each month ranked only against the months before it — the expanding-window rule; the shares move as history accrues.

The tier-entry register — 63 periods since 1963

entrytierECYpctmonthsS&P +1yS&P +3y
Apr 2026pale-red+1.66p235openopen
May 2025pale-red+1.49p1710+25.0%open
Oct 2024pale-red+1.52p176+15.1%open
Aug 2024neutral+1.82p262+18.5%open
Feb 2024pale-red+1.65p206+11.4%open
Dec 2023neutral+1.95p282+26.1%open
Sep 2023pale-red+1.75p243+37.9%+89.2%
Nov 2018neutral+1.92p2958+26.2%+91.7%
May 2018pale-red+1.73p246+7.0%+62.5%
May 2009neutral+5.51p71108+19.5%+51.8%

Periods of at least two consecutive months in a tier (the VE convention); the raw monthly census — 86 boundary events, 23 lasting a single month — is disclosed on the methodology page. Each period's S&P windows open the month its first reading was published — a month's ECY is known when that month's CPI is, about two weeks after the month.

The two real-rate renderings

Shiller’s convention reads the ten-year real rate at 1.60% this month, and the market’s TIPS read averages 2.64%. The convention sits −1.04 points below the market, because the 2021–22 inflation burst still rides inside its trailing ten-year window. On the market’s real rate, today’s ECY would be thinner still.

Latest-basis, labeled: the CAPE leg revises, and readings shown are today’s revised history. October 2025 has no reading — the BLS published no CPI for that month; this site publishes only observations that exist. Shiller’s convention real rate (GS10 minus trailing ten-year CPI inflation, his file’s own arithmetic) is shown beside the market’s TIPS read; the two are different instruments and their gap is stated, not reconciled. Counts, never probabilities; descriptive record, never a forecast. As of: CAPE 2026-10-31 · GS10 2026-10-01 · displayed month 2026-09. How it's measured →