CAPE 1881-2026: the major peaks and troughs of the U.S. equity market across 145 years of Shiller data

Seven inflection points structure the Shiller series since 1881: 1920 trough at 4.78, 1929 peak at 32.56, 1932 Depression trough at 5.57, 1966 peak at 24.1, 1982 Volcker trough at 6.64, 1999 dot-com peak at 44.20 (absolute record), 2009 GFC trough at 13.32, joined by the cycle that opened in November 2021 (38.58), which took the CAPE to 41.12 in August 2026.

Reconstructing each episode, identifying the macro conditions that produced them, and examining the forward 10-year returns that followed is the empirical reference exercise of the CAPE cluster. The counterpart of this page lives in CAPE explained: methodology and 10-year real earnings average.

Historical Troughs: Four Passages in the Deep Value Zone

The Shiller series since 1881 has documented four durable passages through the deep value zone (CAPE below 10), each produced by a major macroeconomic shock. The first is the post-WWI trough of December 1920 at 4.78 per Robert Shiller’s data. This zone of extremely low valuation is the combined legacy of the 1920-1921 depression (which was not recognized at the time as a major recession and lasted less than two years) and post-war monetary uncertainty. The real S&P 500 return over the following ten years was remarkable: 16.6% a year, dividends reinvested, per Eco3min calculations on Shiller’s data, the strongest decade to follow any of the seven inflection points. A related answer: reading CAPE alongside the forward multiple.

The second trough is that of the Great Depression, in June 1932 at 5.57, the lowest reading outside 1920-1921 (4.78 in December 1920). This value was reached at the trough of the most severe economic contraction of the 20th century: U.S. GDP down cumulatively about 25% between 1929 and 1933 per back-calculated BEA estimates, unemployment at 25% in 1933. The real S&P 500 return over 1932-1942 was solid, dividends reinvested, but below the post-WWII expansion decades, due to the double destabilization of the depression then the war.

The third trough marks the exit from stagflation, July 1982 at 6.64. The context combines the Fed Funds peak under Paul Volcker, at 19.1% as a monthly average in June 1981 per the Federal Reserve, and the economic deterioration produced by forced disinflation: unemployment reached 10.8% in November 1982 per the BLS. The real S&P 500 return over the following ten years was very strong, dividends reinvested, at the start of the 1982-2000 secular bull market. See our note on stagflation for a fuller treatment.

A fourth deep value passage deserves mention without reaching the strict 10 threshold: March 2009 at 13.32. This GFC trough is less low than the previous three, but it is the lowest since March 1986. The real S&P 500 return over the following ten years was comparable, dividends reinvested, to the decade after the 1982 trough. These four empirical troughs produce a statistical regularity: the deep value zone has systematically preceded a decade of solid real returns, which is precisely the CAPE predictive property as documented through the Shiller regression.

Historical Peaks: Three Passages in the Extreme Zone

The first peak in the extreme regime is September 1929 at 32.56. This level was reached at the end of the Roaring Twenties decade, in a context of sustained economic growth, massive technological progress (automobiles, electrification, radio) and broad-based public speculation in equities. The Black Tuesday crash of October 29, 1929 and the subsequent collapse produced a negative real S&P 500 return over 1929-1939, with dividends reinvested per Shiller’s data, an empirical ex post validation of the CAPE’s predictive power.

The second peak is the Nifty Fifty peak of January 1966 at 24.1. This peak is lower than the other two but constitutes an interesting case: it preceded not an immediate crash but fifteen years of stagflation and weak real returns. The real S&P 500 return over 1966-1976 was negative, dividends reinvested, and similar from peak to the 1982 Volcker trough. CAPE 24.1 was not an immediate crash signal; it was a signal of weak decadal forward returns, which validated empirically even without a single sharp correction. Adding the rate backdrop to that decadal reading gives the excess CAPE yield of the S&P 500, which sets the cyclically adjusted earnings yield against the real 10-year Treasury rate.

The third peak is the absolute historical record: December 1999 at 44.20. This level, never reached before or after in the series, was produced by the dot-com bubble: record capitalizations for internet companies without earnings, broad optimism on “new economy” productivity. The Nasdaq Composite lost 78% of its value between 10 March 2000 and 9 October 2002 per Nasdaq. The real S&P 500 return over 1999-2009 was negative, dividends reinvested, even worse than the post-1929 decade, marking the “lost decade” for U.S. equities. Eco3min’s catalogue of historical market crises places them on a common timeline.

The fourth peak opened in November 2021 at 38.58. The CAPE dipped below 30 several times between June 2022 and October 2023 (27.08 in October 2022), then reached 41.12 in August 2026, its highest since September 2000. Unbroken since November 2023, the phase above 30 spans 35 months as of September 2026, shorter than the June 1997 to August 2001 episode (51 consecutive months). The reading of the 2024-2026 regime at CAPE 35-40 covers the competing theses that structure the debate on this persistence.

Intermediate Phases: Normal and Elevated Regimes

Between deep value troughs and extreme peaks, the CAPE has spent most of its history in the “normal” (13-20) and “elevated” (20-30) zones. The historical CAPE median over 1881-2025 sits around 16-17 per Shiller-data calculations, placing this normal zone as the statistical dominant of the ratio.

Normal regime phases cover notably the 1950s-1960s decades (sustained post-WWII economic growth, moderate inflation, CAPE oscillating between 10.5 and 24.1), the 1980s-1990s before the final dot-com phase (CAPE between 6.6 and 25.0 over 1982-1995), and certain post-crisis windows. The real S&P 500 returns over the decades following a CAPE in the normal zone have been intermediate (dividends reinvested, Eco3min calculation on Shiller’s data for a starting CAPE in the normal zone): the standard return zone of U.S. equities, neither the outperformance derived from deep value troughs nor the underperformance derived from extreme peaks.

Elevated regime phases (CAPE 20-30) are historically rarer but documented over extended periods. The mid-to-late 1960s show a CAPE between 17 and 24, mostly above 20, a stable elevated zone before the 1970s degradation. The 2010s post-GFC kept the CAPE between 22 and 30 in 62% of the decade’s months, between a low of 19.7 in July 2010 and a peak of 33.3 in January 2018. The real S&P 500 return over 2010-2020 was solid, dividends reinvested, above the historical median, suggesting that the stable elevated zone can produce solid returns as long as valuation does not tip into the extreme.

This four-regime typology (deep value, normal, elevated, extreme) is not a theoretical convention but a statistical reading of the historical distribution. It structures the reading of peaks and troughs as well as that of intermediate phases, giving each ratio level an empirical comparative reference rather than a mechanical threshold to respect or cross. The distinction with other valuation measures structures the CAPE’s uniqueness as a regime thermometer.

The FASB 142 Break and Pre-Post 2001 Comparability

Any historical reading of the Shiller series meets a structural accounting break that occurred in July 2001: the adoption of FASB Statement 142 by the Financial Accounting Standards Board, which modified the accounting treatment of goodwill (acquisition premium) in U.S. listed companies’ accounts. Before 2001, goodwill was amortized over a maximum of forty years, spreading its accounting impact over very long periods. From July 2001, goodwill ceases to be amortized and is subject to annual impairment tests: any depreciation is immediately and fully booked as a charge.

The practical effect on S&P 500 GAAP earnings has been significant: recession phases now produce massive concentrated depreciations, against smoothed amortizations before. Post-2001 GAAP earnings are therefore structurally more volatile, and tend to be depressed relative to pre-break earnings. For the CAPE, which uses GAAP earnings in the denominator, this means post-2001 values are mechanically higher than they would have been under the pre-FASB 142 methodology. Related analysis: the Eco3min frame for valuing equities.

This break is a central argument of the CAPE critique, developed notably by Jeremy Siegel in “The Shiller CAPE Ratio: A New Look” (Financial Analysts Journal, 2016). The argument: comparing a 2026 CAPE of 40 to a 1929 CAPE of 32.56 overvalues the present relative to the past, because reported earnings since 2002 are depressed by FAS 142, while 1929 earnings were not. Siegel proposes recomputing the ratio on national accounts (NIPA) profits, which carry no such break. Where the current reading falls relative to the 1929 and 1999 extremes is laid out in the macro-regime comparator, which juxtaposes two historical states and the data each carried. A related angle on real equity valuation is available in real equity valuation stripped of fiat money.

Shiller’s data allow the predictive-power argument to be tested. Computed by Eco3min, the correlation between the log of the CAPE and the real return over the following ten years is −0.59 for starting points from 1881 to 2000 and −0.72 for those from 2001 to 2016: the relationship has not weakened since the accounting break, but the sub-period covers only some fifteen years of starting points. The complete methodological critique of the CAPE by Siegel and the Shiller response cover this debate in detail.

Last updated — 3 November 2026