Macro ReviewHow to read
Cross Asset · Fair value · Valuation history

Unproven, not disproven: four decades cannot test the 42% gap against simpler measures

Has this valuation gap predicted ten-year returns?

Shiller's data to Nov 2026Real-time release Nov 2026 · Valuation inputs 16 Nov 2026Updated 17 Nov 2026
()+3.50ppon this history's rules 2.41pp, median 5.55pp · Nov 2026
Gap to the −41.50%on this history's own inputs −52.09% · Nov 2026
Months in today's decile4733 followed by a below-average decade · since 1975
RecordUnproven4 independent decades · Nov 2026
Earnings yield less the real yield+1.65ppthe benchmark, record since 1926 · Nov 2026

In brief

What the record says, in a few lines; what would change the read beneath
  • After the 47 months in today's decile the next ten years' real return was below average in 33 of them.→ Exhibit 3
  • Since 1995 a wider gap preceded lower ten-year returns, but 4 independent decades cannot separate it from the earnings yield less the real yield.→ Exhibit 10
  • from 1995, the gap's ten-year forecasts did worse than the historical mean on 2 non-overlapping windows.→ Exhibit 8
  • Since 1926, the earnings yield less the real yield cut the historical mean's ten-year forecast error by 26%, on 9 non-overlapping windows.→ Exhibit 5
  • Recomputed from what was known each month, the two-stage gap reads −52.1% in Nov 2026, against the Valuation page's −41.5%.→ Exhibit 1
What would change this view
What would change this viewTriggerTodayDecided by
The gap starts adding to the earnings yield less the real yieldten-year t above 1.96 and encompassing p below 0.05t 0.36, p 0.44 (Nov 2026)not crossedShiller's monthly update: each adds a realised month of returns
The gap's forecasts start beating the historical meanten-year out-of-sample R² above 0−0.38 (origins to Nov 2016)not crossedShiller's monthly update: each adds a realised ten-year return
The benchmark's long record failsthe earnings yield less the real yield, ten-year out-of-sample R² since 1926 at or below 0+0.26not crossedShiller's monthly update
Real-time data change the verdictthe revised-data gap passes the encompassing rule at a horizon where the real-time gap fails, or the reversethe same verdict at every horizonnot crossedthe Philadelphia Fed's next real-time release, mid-November 2026
The reasoning behind each point
  • What followed
    After the 47 months in today's decile the next ten years' real return was below average in 33 of them.

    Today's two-stage gap, −52.1% (Nov 2026), sits in decile 1 of the 607 months since 1975 (−80.2% to −47.2%). Of its 61 months, 47 have a known ten-year real return: median +0.5% a year against 6.65% for every month since 1871. A handful of overlapping decades, not a forecast: the record is unproven, not disproven.

  • Encompassing
    Since 1995 a wider gap preceded lower ten-year returns, but 4 independent decades cannot separate it from the earnings yield less the real yield.

    With both in one regression of forward real returns, the gap's Hodrick t-statistics at 1, 3, 5, 10 years are 0.56, −0.06, 0.00, 0.36; the encompassing test of forecasts (the earnings yield less the real yield against the gap) gives p-values of 0.59, 0.70, 0.73, 0.44. The rule set before the estimates were made asks for t above 1.96 and p below 0.05 together. The one-stage gap, at ten years, has t 0.87 and p 0.001: it passes only the out-of-sample half. No horizon passes, but the test has little power: with 4 non-overlapping ten-year windows a gap as strong as its own estimate would often fail the rule, and at this horizon the out-of-sample encompassing test's small-sample correction breaks down, so its p-value is a rough guide only. The two measures move together (both are earnings over price less a rate). The revised-data history gives the same verdict (ten-year t 0.39).

  • Out of sample
    Out of sample from 1995, the gap's ten-year forecasts did worse than the historical mean on 2 non-overlapping windows.

    Each month's forecast uses only returns already known then. On the same months the earnings yield less the real yield scores −0.18, the cyclically adjusted earnings yield −0.15, the one-stage gap −0.30. 248 monthly forecasts overlap heavily: behind them are 2 independent ten-year windows.

  • Long record
    Since 1926, the earnings yield less the real yield cut the historical mean's ten-year forecast error by 26%, on 9 non-overlapping windows.

    Out-of-sample R² at one, three, five and ten years: +0.04, +0.06, +0.13, +0.26; the cyclically adjusted earnings yield +0.03, +0.05, +0.08, +0.15. Trained from 1881, each forecast uses only returns known when it was made. The predictor is nearly a random walk (monthly autocorrelation 0.990) and its shocks move against returns (correlation −0.50), so the in-sample slopes are biased up; the bootstrap removes it.

  • Today
    Recomputed from what was known each month, the two-stage gap reads −52.1% in Nov 2026, against the Valuation page's −41.5%.

    The one-stage gap reads −56.5% against −27.6%. The history's rules differ from that page's on purpose, because they must work in 1975 as well as today: long-run growth is the surveys' 4.4% rather than the Committee's 4.0%; near-term growth is the earnings model's 5.5% rather than 5.8%; and the premium compared with is the median of every month's implied premium since Jun 1970, 5.55pp.

  • Expected return
    Model-conditional: the gap's regression implies +5.8% a year real over the next ten years.

    The same arithmetic on the earnings yield less the real yield (1.64pp today), fitted since 1881, gives +4.5% with an error range of −1.2 to 10.7; its record is longer, so it is the better guide of the two, though on the gap's own months it too lost to the historical mean (out-of-sample R² −0.18), and before 1997 its real yield is the 10-year less trailing inflation, not a TIPS yield. Across the payout range the gap's figure runs from 5.6 to 6.0. Neither is a forecast: each is a fitted line read at today's value. The slope is shallow: at +0.038pp a year per point of gap, today's −52.1% takes 2.0pp off the line's zero-gap 7.7%, which leaves +5.8% only 0.9pp below the 6.65% average of every month since 1871.

Next: Range and scenarios — the range around today's prices

Computed here from Shiller's data, the Philadelphia Fed's real-time data and surveys, and the Valuation page's identity.cross.valuation_history-read

What premium has the market implied, month by month?

The recomputed from what was known each month since 1970, today and the full-history median marked; the gap to fair value as the other view

Exhibit 1 · Implied equity risk premium since 1970, as known each month

Premium over the long rate that returns each month's price through the two-stage model, pp, monthly since 1970; its median so far; dashed, Damodaran's.
About this exhibit
Monthly line since 1970: the implied equity risk premium recomputed from what was known each month, its median so far, and Damodaran's (dashed); a toggle shows the gap to the model's value.
The market's implied premium, 2.41pp, against a 5.55pp median
The premium over the long rate that returns each month's price through the two-stage model, from what was known then; percentage points, monthly since Jan 1970. Shaded: the episodes, and the 2008–09 write-downs through the months with no reading. Dashed: Damodaran's monthly premium.
Shaded "write-downs": Jun 2007 – Dec 2010, from the last high of trailing reported earnings to the end of the months the model cannot price; trailing earnings fell 92% to their trough. The fall in 2009 and the gap in Jan 2010 – Dec 2010 are the trailing reported earnings of 2008–09: the write-downs cut the earnings this history starts from, and the near-term growth its earnings model then expects fell so far below zero that the two-stage model cannot price the month (its bracket, (1 + g_L) + H(g_S − g_L), is not positive). Damodaran's series starts from dividends plus buybacks, which the write-downs did not cut, so it runs through; the levels differ by construction, the movements are the cross-check. His ERP (T12m) was 3.70% on 1 Oct 2026.
Source: Robert J. Shiller data; Federal Reserve Bank of Philadelphia; Aswath Damodaran, NYU Stern; Macro Review calculations · data to Nov 2026cross.valuation_history-1

Exhibit 2 · Inputs and gaps as known each month against the Valuation page's

Each input and gap, %, at Nov 2026 on what was known then, beside the Valuation page's, with why they differ.
InputThis historyValuation page, 16 Nov 2026Why they differ
Two-stage gap to the model's value−52.1%−41.5%revised data: −52.4%
One-stage gap to the model's value−56.5%−27.6%the Valuation page's one-stage gap prices forward earnings and consensus growth
Long-run growth g_L4.40%4.00%the surveys' ten-year real GDP 2.1 + CPI 2.3, against the Committee's longer-run medians
Near-term growth g_S, a year+5.48%+5.77%the earnings model on index EPS, against the profits fan bridged to EPS
Long rate4.88%5.12%Shiller's long rate (the month's average), against the Treasury 10-year par yield
Trailing earnings yield E0 ÷ P3.18%3.99%earnings to Mar 2026 (six months back), against the latest trailing earnings
Premium the price implies (two-stage)2.41pp–the premium that returns today's price through the identity
Premium the gap is measured at5.55pp5.17ppthe median of every month's implied premium so far, against the yield-gap premium (today's implied premium plus the yield gap's distance to its median in months like these)

At the Valuation page's own inputs the identity returns its published −41.504% to within rounding.

About this exhibit
Table: each input and gap as the monthly history computes it against the Valuation page's, with why they differ.
Today, Nov 2026: the history's rules against the Valuation page's
The history must work in 1975 as well as today, so its rules differ from the Valuation page's where that page uses something with no past. Both gaps come from the same identity; the rows say where the inputs part.
Source: Computed here; the Valuation page's published inputs (copied and checked).cross.valuation_history-2

What followed readings like today's?

Every month's gap against the real return of the next ten years, today's reading marked; the months in today's decile and how often the decade after fell short

Exhibit 3 · Each month's valuation gap against the next ten years' real return

Two-stage gap, %, against annualised real return over the next ten years, monthly 1975–2016; vertical, the latest gap (−52.1%); horizontal, the average 6.7%.
About this exhibit
Scatter: each month's two-stage gap against the real return of the next ten years, the latest gap and the average return as lines.
After readings like today's, 33 of 47 decades returned below average
Each month's gap against the real return of the ten years that followed, 1975–2016; today's reading (decile 1, -80.2% to -47.2%) is the line. A description of the record, not a forecast.
Source: Robert J. Shiller data; Federal Reserve Bank of Philadelphia; Macro Review calculations · data to Nov 2026cross.valuation_history-17

What do the plain benchmarks say over the long record?

The earnings yield, the earnings yield less the real yield and the earnings yield less the bond yield since 1881, and their forecast record since 1926

Exhibit 4 · Simple valuation measures over the long record

Percentage points, monthly since 1881, as known each month: earnings yield less real yield, cyclically adjusted earnings yield, and earnings yield less bond yield.
About this exhibit
Three lines, monthly since 1881: earnings yield less real yield, cyclically adjusted earnings yield and earnings yield less bond yield.
The plain benchmarks since 1881, each read with what was known at the month
Earnings six months old over the month's price; the cyclically adjusted yield averages ten years of earnings in the prices of the last known CPI; the real yield is the long rate less trailing ten-year CPI inflation, so it needs no index-linked bonds.
Source: Robert J. Shiller, Irrational Exuberance data (shillerdata.com); Computed here.cross.valuation_history-10

Exhibit 5 · Forecast record of simple valuation measures since 1926

Out-of-sample R² (a ratio) against the historical mean at each horizon, and the ten-year regression on the full record since 1881, bias removed.
MeasureR², 1yR², 3yR², 5yR², 10yClark–West p, 10y10y slope less biast, HodrickWindows fitted · forecast
Earnings yield less real yield+0.04+0.06+0.13+0.260.010+0.5193.0313 · 9
Cyclically adjusted earnings yield+0.03+0.05+0.09+0.150.016+0.5252.3813 · 9
Earnings yield less bond yield+0.01−0.01−0.03−0.100.220+0.4222.0714 · 9
About this exhibit
Table: out-of-sample R² at four horizons, the Clark–West p, the slope less its bias and the Hodrick t for three benchmark measures since 1926.
The benchmarks' long record: forecasts from 1926, trained from 1881
Out-of-sample R² against the historical mean at each horizon, and the ten-year regression on the whole record with its bias removed. Windows: non-overlapping ten-year spans.
Source: Robert J. Shiller, Irrational Exuberance data (shillerdata.com); Computed here.cross.valuation_history-11

What does today's reading imply for the next ten years? ()

The ten-year real return the history's regressions imply at today's readings, with the error of each rule as its range

Exhibit 6 · Ten-year real return each fitted line implies at the latest reading

Model-conditional: each measure's fitted line read at Nov 2026, implied ten-year real return, % a year; ranges are each rule's out-of-sample misses.
RuleReading todayFitted lineImplied, % a year realRangeKindIts record
Two-stage gap, real-timefitted May 1975 – Nov 2016, 4 windows−52.097.72 + 0.038 × reading+5.8-7.7 to 6.3out-of-sample recordR² −0.382 windows forecast
Earnings yield less real yieldfitted Mar 1881 – Nov 2016, 13 windows+1.653.57 + 0.589 × reading+4.5-1.2 to 10.7out-of-sample recordR² +0.269 windows forecast
Cyclically adjusted earnings yieldfitted Jun 1881 – Nov 2016, 13 windows+2.421.07 + 0.769 × reading+2.9-1.8 to 10.6out-of-sample recordR² +0.159 windows forecast
Two-stage gap, payout k across its declared range5.6 to 6.0–assumed valuespayout k from 0.75 to 0.95, each end refitted

the 10th and 90th percentiles of the rule's out-of-sample errors, 1995–2016 (2 non-overlapping windows) for the gap; the 10th and 90th percentiles of the rule's out-of-sample errors, 1926–2016 (9 non-overlapping windows) for the benchmarks. On its few windows the record cannot tell the gap from the earnings yield less the real yield, whose record is longer, so that row is the better guide of the two.

About this exhibit
Table: for three measures, the latest reading, the fitted line, the implied ten-year real return, its error range and record; plus the gap at the ends of the payout range.
Model-conditional: the ten-year real return each fitted line implies at today's reading
Model-conditional, not a forecast: the last open section reads each regression line at today's value. The line is fitted on a record of 4 non-overlapping ten-year windows for the gap and 13 for the earnings yield less the real yield; its range is the rule's own out-of-sample error, not a probability. Each row: the regression of ten-year returns on the measure, read at Nov 2026; the range is the rule's own out-of-sample misses.
Source: Computed here on Robert J. Shiller, Irrational Exuberance data (shillerdata.com) and Federal Reserve Bank of Philadelphia, Real-Time Data Set for Macroeconomists and Survey of Professional Forecasters.cross.valuation_history-12

The tests: slopes, forecasts , , the inputs and the payout assumption

Overlap-robust slopes with the small-sample bias removed, forecasts scored out of sample, whether the gap adds to the earnings yield less the real yield, what moved the gap, and each January's estimate

Exhibit 7 · How strongly each measure predicted ten-year real returns

Slope of annualised real return over the next ten years on each measure, % a year, with t-statistics, bias and R²; records to Nov 2016.
About
MeasureSlopet, Newey–Westt, HodrickBias under the nullSlope less biasBootstrap pR²Windows
Two-stage gap, real-timeMay 1975 – Nov 2016, per point of gap+0.0381.851.61+0.004+0.0330.0510.184
Two-stage gap, revised dataMay 1975 – Nov 2016, per point of gap+0.0391.881.62+0.004+0.0350.0470.194
One-stage gap, real-timeMay 1962 – Nov 2016, per point of gap+0.0633.722.13+0.006+0.0570.0200.185
One-stage gap, revised dataMay 1962 – Nov 2016, per point of gap+0.0633.992.06+0.007+0.0560.0230.185
Cyclically adjusted earnings yieldJun 1881 – Nov 2016, per point of gap+0.7695.112.38+0.244+0.5250.0710.2413
Earnings yield less real yieldMar 1881 – Nov 2016, per percentage point+0.5895.083.03+0.070+0.5190.0210.2813
Earnings yield less bond yieldJun 1872 – Nov 2016, per percentage point+0.4921.892.07+0.070+0.4220.0640.1214

Windows: non-overlapping 10-year spans in the sample — the honest count behind each number.

About this exhibit
Table: for seven measures, the slope of forward real returns on the measure with its t-statistics, small-sample bias, a significance level and R²; a selector chooses the horizon.
Forward real returns on each measure: the slope, its errors and its bias
Annualised log real total return over the next 10 years, regressed on the measure at the month. Newey–West errors at 120 monthly lags and Hodrick's; the bias is the mean slope a measure with no power produces in 1,000 bootstrap samples like this one.
Source: Computed here on Robert J. Shiller, Irrational Exuberance data (shillerdata.com) and Federal Reserve Bank of Philadelphia, Real-Time Data Set for Macroeconomists and Survey of Professional Forecasters.cross.valuation_history-6

Exhibit 8 · Out-of-sample forecast record of each measure against the historical mean

Out-of-sample R² (a ratio) by horizon: one less the measure's squared forecast errors over the historical mean's; forecasts from Apr 1986.
About
Measure1 year3 years5 years10 yearsClark–West p, 10y
Two-stage gap, real-time−0.05−0.30−0.40−0.380.539
Two-stage gap, revised data−0.05−0.30−0.41−0.390.569
One-stage gap, real-time−0.05−0.28−0.37−0.300.416
One-stage gap, revised data−0.05−0.29−0.39−0.280.507
Cyclically adjusted earnings yield−0.14−0.39−0.74−0.150.467
Earnings yield less real yield−0.04−0.19−0.23−0.180.847
Earnings yield less bond yield−0.04−0.15−0.17−0.030.605
Non-overlapping windows forecast391162
About this exhibit
Table: out-of-sample R² at four horizons for seven measures against the historical mean, on the gap's months.
Forecasts made with what was known: each measure against the historical mean, on the gap's months
Campbell–Thompson R²: one less the measure's squared forecast errors over the historical mean's. Every measure is scored on the same months, forecasts from Apr 1986 at one year to Apr 1995 at ten. Negative: worse than the mean.
Source: Computed here on Robert J. Shiller, Irrational Exuberance data (shillerdata.com) and Federal Reserve Bank of Philadelphia, Real-Time Data Set for Macroeconomists and Survey of Professional Forecasters.cross.valuation_history-7

Exhibit 9 · Where each measure won and lost against the historical mean

Summed squared forecast error of the historical mean less each measure's, squared pp, ten-year returns, monthly since Apr 1995; rising: the measure beat the mean.
About this exhibit
Lines by forecast month: cumulative squared-error advantage over the historical mean for four measures; a selector chooses the horizon.
Where each measure won and lost against the historical mean
Cumulative sum of the mean's squared forecast error less the measure's, 10-year returns (squared points a year), by forecast month. Rising: the measure beat the mean over that stretch; below zero at the end: worse overall.
Source: Computed here on Robert J. Shiller, Irrational Exuberance data (shillerdata.com) and Federal Reserve Bank of Philadelphia, Real-Time Data Set for Macroeconomists and Survey of Professional Forecasters.cross.valuation_history-8

Exhibit 10 · Gap and earnings yield less real yield tested together, by horizon

Gap and benchmark in one regression by horizon: the gap's slope and t, its weight (a ratio) and encompassing p-values; forecasts from Apr 1986.
HorizonGap slopet, Hodrickt, Newey–WestBenchmark t, HodrickWeight on the gap (λ)Encompassing pClark–West pAdds?
1 year50 windows in sample+0.0230.560.540.16-0.220.5900.829no
3 years16 windows in sample−0.002-0.07-0.070.34-0.420.7030.704no
5 years9 windows in sample±0.0000.00-0.010.50-0.660.7290.858no
10 years4 windows in sample+0.0100.361.350.940.030.4360.569no
About this exhibit
Table by horizon: the gap's slope and t-statistics alongside the benchmark in one regression, the weight on the gap and the encompassing p-values; a selector chooses the gap.
Does the gap add anything to earnings yield less real yield?
In one regression with both, and out of sample by the Harvey–Leybourne–Newbold test. The rule, set before the estimates: the gap adds something at a horizon only if its Hodrick t exceeds 1.96 and the test's p is below 0.05.
Source: Computed here on Robert J. Shiller, Irrational Exuberance data (shillerdata.com) and Federal Reserve Bank of Philadelphia, Real-Time Data Set for Macroeconomists and Survey of Professional Forecasters.cross.valuation_history-9

Exhibit 11 · Long-run and near-term growth inputs, as known each month

% a year, monthly to Nov 2026: long-run growth as published at the time and revised, and the model's near-term earnings growth.
About
About this exhibit
Lines: long-run growth (real-time and revised) and the model's near-term earnings growth, as known each month.
The growth rates the identity used, as known each month
g_L: twenty years of nominal output growth as published at the time (the first real-time release Dec 1965; before it, today's figures), and from Mar 1992 the Survey of Professional Forecasters' ten-year medians. g_S: the earnings model's expected growth over two years from the last known quarter.
Source: Federal Reserve Bank of Philadelphia, Real-Time Data Set for Macroeconomists and Survey of Professional Forecasters; Robert J. Shiller, Irrational Exuberance data (shillerdata.com); the model computed here.cross.valuation_history-3

Exhibit 12 · Implied premium against its median so far, and the long rate

Percentage points, monthly since 1970.
About this exhibit
Lines since 1970: the implied premium, its median so far, and the long rate.
What moves the gap: the premium the price implied against its median so far
Percentage points. When the implied premium sits above its median the model price at the median exceeds the price and the gap is positive; the gap is close to a transform of this distance, which is why the record must beat the plain earnings yield less the real yield.
Source: Computed here on Robert J. Shiller, Irrational Exuberance data (shillerdata.com) and Federal Reserve Bank of Philadelphia, Real-Time Data Set for Macroeconomists and Survey of Professional Forecasters.cross.valuation_history-4

Exhibit 13 · Gap on real-time data less gap on revised data, since 1965

Points of gap (pp), monthly since 1965; two-stage and one-stage.
About this exhibit
Two lines: the two-stage and one-stage gap on real-time data less on revised data, points of gap.
What the look-ahead was worth: the gap on real-time data less the gap on today's data
Points of gap. Before Mar 1992 the anchor differs by the revisions to twenty years of output growth (−0.13pp on average); after, the surveys are the same in both runs and only the premium's median, which reaches back before Mar 1992, differs. Average size before: two-stage 1.0, one-stage 1.8; after: 2.3 and 2.0.
Source: Computed here on Federal Reserve Bank of Philadelphia, Real-Time Data Set for Macroeconomists and Survey of Professional Forecasters and Robert J. Shiller, Irrational Exuberance data (shillerdata.com).cross.valuation_history-5

Exhibit 14 · The record at the ends of the payout range

The payout ratio k is an assumption (declared range 0.75–0.95); the whole history is recomputed at each end and scored again at ten years.
IdentitykGap today10y slopet, HodrickR², out of sampleImplied 10y return
two-stage0.75−53.5%+0.0291.57−0.50+6.0%
one-stage0.75−57.6%+0.0562.13−0.45+3.1%
two-stage0.95−51.1%+0.0451.63−0.33+5.6%
one-stage0.95−55.1%+0.0662.05−0.23+2.8%
two-stage0.85−52.1%+0.0381.61−0.38+5.8%
one-stage0.85−56.5%+0.0632.13−0.30+2.9%
Source: Computed here.cross.valuation_history-13

Exhibit 15 · The earnings model's parameters estimated each January

Real-time run, Jun 1970 to Jan 2026, newest first: long-run growth, % a year; persistence φ; shock sizes; jump quarters; log-likelihood.
EstimatedQuartersg_Lφσ_ηρ1ρ2σ_νσ_εJumpsLog-likelihood
Jan 20262624.480.95 fixed0.0000.86-0.386.950.00Oct 2008, Apr 2009, Oct 2009, Jan 2010-880.8
Jan 20252584.230.95 fixed0.0000.86-0.397.000.00Oct 2008, Apr 2009, Oct 2009, Jan 2010-869.2
Jan 20242544.400.95 fixed0.0000.86-0.397.040.00Oct 2008, Apr 2009, Oct 2009, Jan 2010-857.2
Jan 20232505.230.95 fixed0.0000.86-0.397.070.00Oct 2008, Apr 2009, Oct 2009, Jan 2010-844.6
Jan 20222464.800.95 fixed0.0000.87-0.417.070.00Oct 2008, Apr 2009, Oct 2009, Jan 2010-831.2
Jan 20212424.120.95 fixed0.0000.93-0.456.730.00Oct 2008, Apr 2009, Oct 2009, Jan 2010-805.6
Jan 20202384.190.95 fixed0.0000.95-0.476.580.00Oct 2008, Apr 2009, Oct 2009, Jan 2010-787.3
Jan 20192344.360.95 fixed0.0000.95-0.476.630.00Oct 2008, Apr 2009, Oct 2009, Jan 2010-775.8
Jan 20182304.650.95 fixed0.0000.95-0.476.680.04Oct 2008, Apr 2009, Oct 2009, Jan 2010-764.4
Jan 20172264.490.95 fixed0.0000.95-0.476.740.00Oct 2008, Apr 2009, Oct 2009, Jan 2010-752.9
Jan 20162224.650.95 fixed0.0000.81-0.377.170.00Oct 2008, Oct 2009, Jan 2010-753.3
Jan 20152184.800.95 fixed0.0000.81-0.377.220.00Oct 2008, Oct 2009, Jan 2010-741.3

Near-term growth g_S. The Earnings page's unobserved-components model on index EPS — its state space, bounds and procedure (estimate freely; if φ reaches a bound or σ_η collapses, fix φ at 0.95; quarters beyond |z| > 3 become jump dummies; re-estimate), steps one to three from the default start as on that page's trailing-EPS fit, on quarterly growth from 1960, with μ̄ = g_L ÷ 4. Re-estimated each January (and at the first month with 40 quarters, 1970) on the quarters known then — 57 estimates in the real-time run — and filtered every month with the last estimate's parameters; a new quarter whose one-step error exceeds the jump cut is treated as a jump rather than moving the state, as re-estimating would treat it (23 such quarters in the real-time run). g_S is the model's expected growth over the next 2 years from the last known quarter — the long-run state reverting to μ̄, the cycle following its AR(2) — as a compound annual rate. Months the identity cannot price. The two-stage gap needs (1 + g_L) + H(g_S − g_L) above zero and r − g_L above 0.5pp. In the real-time run 12 months fail the first (2010: the January estimate after the 2008–09 write-downs put near-term growth so far below zero that the bracket turned negative), and 0 the second; they are left out of every statistic, never filled.

About this exhibit
Table: the earnings model's long-run growth, persistence, shock sizes, jump quarters and log-likelihood, estimated each January on what was known then.
The earnings model, estimated each January on what was known then
The Earnings page's model and procedure on index EPS growth, real-time run, newest first. φ is fixed at 0.95 when the long-run state is not identified; jumps are the quarters given their own dummy.
Source: Computed here on Robert J. Shiller, Irrational Exuberance data (shillerdata.com).cross.valuation_history-14

Exhibit 16 · The history every January

The gaps and their inputs at each January, newest first, with the ten-year real return that followed where it is known (log, % a year).
JanuaryTwo-stage gaprevisedOne-stage gaprevisedg_Lrevisedg_SReference premiumLong rateEarnings yieldNext ten years
2026surveys−46.4−46.8−50.3−50.84.484.48+4.65.584.213.21–
2025surveys−49.9−50.2−54.7−55.24.234.23+5.75.624.633.29–
2024surveys−36.0−36.2−41.6−42.64.404.40+5.45.664.063.76–
2023surveys+4.1+2.3−0.8−2.95.235.23+3.95.693.534.85–
2022surveys+8.2+5.5+2.0−1.24.804.80+2.35.681.763.47–
2021surveys−12.2−14.3−24.3−26.84.124.12+4.65.701.082.62–
2020surveys+10.8+8.6±0.0−2.74.194.19+3.95.671.764.13–
2019surveys+5.1+4.6−5.7−8.44.364.36+5.05.662.714.70–
2018surveys−10.8−12.4−18.7−20.34.654.65+3.85.672.583.73–
2017surveys−5.2−6.1−17.7−19.74.494.49+4.95.662.433.82–
2016surveys+39.8+37.3+21.8+18.44.654.65+3.95.622.094.95+11.4
2015surveys+74.0+71.1+42.5+37.84.804.80+5.55.561.885.08+9.5
Source: Computed here on Robert J. Shiller, Irrational Exuberance data (shillerdata.com) and Federal Reserve Bank of Philadelphia, Real-Time Data Set for Macroeconomists and Survey of Professional Forecasters.cross.valuation_history-15

Appendix

The rules for what was known when, the anchor, the model, the identity, the statistics, sources and checks36 checks, all pass
About this page

The Valuation page prints today's gap to the model's value. This page recomputes that gap at every month since the data allow, from what was known then — long-run growth from the Philadelphia Fed's real-time data and surveys, near-term growth from the earnings model re-estimated each January — and asks whether it ever told a reader more about the returns that followed than the plain earnings yield less the real yield.

The predictive regression and its tests

12h∑j=1hrt+j=ah+bh xt+εt+h,h=12,36,60,120 months\frac{12}{h}\sum_{j=1}^{h} r_{t+j} = a_h + b_h\,x_t + \varepsilon_{t+h}, \qquad h = 12, 36, 60, 120 \text{ months}
ROS2=1−∑t(rt+h−r^t+h∣t)2∑t(rt+h−rˉt+h∣t)2,Clark–West: ft=e0,t2−e1,t2+(r^0−r^1)2R^2_{OS} = 1 - \frac{\sum_t (r_{t+h} - \hat r_{t+h|t})^2}{\sum_t (r_{t+h} - \bar r_{t+h|t})^2}, \qquad \text{Clark–West: } f_t = e_{0,t}^2 - e_{1,t}^2 + (\hat r_{0} - \hat r_{1})^2
SymbolMeaningFromAssumed or estimated
rtr_tthe month's log real total return; the left side is its mean over the next h months, a year (× 12)Shiller's real total-return index–
xtx_tthe measure at month t: the two-stage gap, or a benchmarkthis page–
bhb_hthe slope; its errors Newey–West (h lags) and Hodrick (1992)estimatedestimated
ROS2R^2_{OS}Campbell–Thompson out-of-sample R² against the historical meanestimatedestimated

Overlapping monthly origins: the count that matters is the non-overlapping ten-year windows. The Stambaugh bias is removed by a bootstrap under the null.

Exhibit 17 · Definitions

  • Model-conditional
    What the last open section is, and is not.

    Model-conditional, not a forecast: the last open section reads each regression line at today's value. The line is fitted on a record of 4 non-overlapping ten-year windows for the gap and 13 for the earnings yield less the real yield; its range is the rule's own out-of-sample error, not a probability.

  • When
    What was known at each month.

    What was known when. Each month's price is Shiller's average of daily closes. Earnings are the four-quarter total at the last quarter-end at least 6 months back — from Mar 1926; before that the figures are annual and the last December at least 6 months back is used, since the months between are interpolated towards the next December. CPI is read 2 months back. The long rate is the month's own from Apr 1953; before that Shiller's annual series is interpolated between Januaries, so the January at least 12 months back is used. The cyclically adjusted earnings yield is the average of ten years of earnings to the last known month, each in the prices of the last known CPI month, over the price. A check asserts every rule at every month.

  • Anchor
    Long-run growth known at the time.

    Long-run growth g_L known at each month. From Mar 1992: the Survey of Professional Forecasters' ten-year medians, real GDP (asked in the first-quarter survey) plus CPI inflation (every survey) — 2.1 + 2.3 = 4.4% in Nov 2026. Before that: nominal output's compound growth over the last 20 years (at least 10) in the Philadelphia Fed's real-time data as published at the time, the first from Dec 1965; before that, today's figures, and the page says so. A survey or real-time release of a quarter is used from the month after the quarter's middle month. Nominal output is GNP in releases before 1992 and GDP after, as published. CPI inflation runs a few tenths above the GDP deflator, so the survey anchor sits a little above nominal GDP growth.

  • Real time
    Two runs, and what separates them.

    Two runs. Real-time: every output figure as published at the time. Revised: the same quarters as published today. The surveys are never revised, so from Mar 1992 the anchor is the same in both, and from Jan 1993 so is every estimate of the earnings model; before, they differ by the revisions to twenty years of output growth (−0.13pp on average in g_L, real-time less revised). The gaps still differ a little after Mar 1992 (2.26 points on average, +0.26 in Nov 2026) because the reference premium is a median over every earlier month, including those before it.

  • The model
    Near-term earnings growth, re-estimated each January.

    Near-term growth g_S. The Earnings page's unobserved-components model on index EPS — its state space, bounds and procedure (estimate freely; if φ reaches a bound or σ_η collapses, fix φ at 0.95; quarters beyond |z| > 3 become jump dummies; re-estimate), steps one to three from the default start as on that page's trailing-EPS fit, on quarterly growth from 1960, with μ̄ = g_L ÷ 4. Re-estimated each January (and at the first month with 40 quarters, 1970) on the quarters known then — 57 estimates in the real-time run — and filtered every month with the last estimate's parameters; a new quarter whose one-step error exceeds the jump cut is treated as a jump rather than moving the state, as re-estimating would treat it (23 such quarters in the real-time run). g_S is the model's expected growth over the next 2 years from the last known quarter — the long-run state reverting to μ̄, the cycle following its AR(2) — as a compound annual rate.

  • Identity
    The Valuation page's identity, month by month.

    The identity is the Valuation page's, the site's one implementation: P = k·E0·[(1 + g_L) + H(g_S − g_L)] ÷ (r − g_L), with k 0.84 and H 3 (6-year fade) as that page sets them, E0 ÷ P the trailing earnings yield known at the month, and r the long rate plus a reference premium. The one-stage gap sets g_S = g_L. At the Valuation page's own inputs the same function returns its published −41.5% — a check.

  • Premium
    The premium the gap is measured at.

    The reference premium. At each month the identity implies the premium that returns the price exactly; the reference is the median of that implied premium over every earlier month and this one (at least 5 years of them), and the gap is the model price at the reference premium over the price, less one. Positive: the index sits below fair value. The Valuation page uses a median over months with the same sign of the stock–bond correlation since 2005, which has no earlier history. A gap made this way is close to a transform of the earnings yield against the bond yield — which is why the encompassing test is compulsory.

  • Record
    Slopes, errors and the bias.

    The record. For each measure and horizon h (one, three, five and ten years), forward log real total returns (Shiller's real total-return index, a year) regressed on the measure at the month. Monthly origins overlap, so the errors are Newey–West with h monthly lags and Hodrick's (1992) — one-month returns under the null times the sum of the last h readings of the predictor. The slope is biased in small samples when a persistent predictor's shocks move against returns (Stambaugh, 1999): a bootstrap under the null (1,000 samples, the predictor's own AR(1) with its shocks paired to the returns', a fixed seed) gives the bias at each horizon and a one-sided p-value. Non-overlapping windows are counted for every number.

  • Out of sample
    Forecasts made with what was known.

    Out of sample. At each month a regression on the pairs whose returns were already known then, against the historical mean of the same returns, once 120 such pairs exist; Campbell and Thompson's R² = 1 − the model's squared errors ÷ the mean's; Clark and West's test for a nested model; Goyal and Welch's cumulative difference of squared errors, which rises when the measure beats the mean. The gap and every benchmark are scored on the same months (ten-year forecasts from Apr 1995); the benchmarks also on their long record, forecasts from 1926, trained from 1881.

  • Encompassing
    The test the gap must pass.

    The encompassing test. The gap adds something beyond the earnings yield less the real yield at a horizon only if both hold: in one regression with both, the gap's Hodrick t-statistic exceeds 1.96; and out of sample the Harvey–Leybourne–Newbold test rejects that the benchmark's forecasts encompass the gap's (p below 0.05). The rule was set before the estimates. The earnings yield less the real yield is the earnings yield less the long rate plus trailing ten-year CPI inflation — a real yield that needs no index-linked bonds and so runs from 1881.

  • Gaps
    Months the identity cannot price.

    Months the identity cannot price. The two-stage gap needs (1 + g_L) + H(g_S − g_L) above zero and r − g_L above 0.5pp. In the real-time run 12 months fail the first (2010: the January estimate after the 2008–09 write-downs put near-term growth so far below zero that the bracket turned negative), and 0 the second; they are left out of every statistic, never filled.

  • Limits
    What the record can bear.

    What the record can bear. The gap exists from 1975 (two-stage) and 1962 (one-stage): 4 and 5 non-overlapping ten-year windows. The benchmarks run from 1881: 13. Payout policy changed over the century (lower dividend payout after 1950, buybacks after 1990), which moves the right k; the detail fold shows the record at k = 0.75 and 0.95.

  • Dates
    Data dates.

    Data dates: Shiller's workbook read on 17 Nov 2026 (prices to Nov 2026, earnings to Sep 2026); the Philadelphia Fed's files read on 17 Nov 2026 (the last real-time release Nov 2026, surveys to Nov 2026); BEA's national accounts read on 16 Nov 2026; the Valuation page's inputs 16 Nov 2026; made from data read on 17 Nov 2026.

  • Licence
    Official data, Shiller's index series, the rest calculated on this site.

    The Philadelphia Fed's data and BEA's accounts are free to use. Shiller's series is shown with its citation. The histories, the model and the statistics are computed here.

Source: Computed here; Robert J. Shiller, Irrational Exuberance data (shillerdata.com); Federal Reserve Bank of Philadelphia, Real-Time Data Set for Macroeconomists and Survey of Professional Forecasters.cross.valuation_history-16

Before you read

Macro Review is a personal research site, still being built.

Nothing on it is investment advice, a recommendation or a solicitation.

Check every figure for yourself. Each names its publisher — confirm it there before you rely on it.

Figures can be revised, delayed or wrong.

Cones, fans, scenarios, fair values and “what followed” tables are conditional on the model each page names: they describe, they do not predict.

No warranty is given; use is at your own risk.