Anomaly Replication Lab
  • The zoo
  • Screener
  • Multiple testing
  • Correlations
  • Composite
  • Methodology
  • Guides
  • Resources
Sign in
Anomaly Replication LabSign in
  • The zoo
  • Screener
  • Multiple testing
  • Correlations
  • Composite
  • Methodology
  • Guides
  • Resources

Educational replication, not advice. Long-short returns are decile spreads or published factor series from the Kenneth R. French Data Library and AQR, value-weighted, US equities, and gross of transaction and shorting costs — which for high-turnover signals like short-term reversal is the whole story. Reproducing a premium in-sample is not evidence it will persist. Past performance does not predict future results.

© 2026 Austin Levenberg. All rights reserved. This site’s code, design, written analysis, and computed results are not licensed for reuse or redistribution — terms of use · privacy.

Data: Kenneth R. French Data Library, AQR Capital Management public data sets.

Part of austinlev.dev — four research tools, and a note on where this is headed.

← the zoo
ValueFaded

Value (earnings-to-price)

Low P/E stocks outperform high P/E stocks.

Basu · Journal of Finance · published 1977 · original sample 1957–1971 (reported ≈ +6.7%/yr, t ≈ 2.7)

How our version differs from the paper. Basu sorted NYSE stocks into quintiles and equal-weighted them; we use value-weighted deciles from all listed stocks, which is milder. The decay figure compares our replication in-sample to our replication after publication, so it isn’t affected by this — but the level gap between “reported” and “our replication” partly is.
Our replication
+3.3%
1951–2026, per year
Full-sample t
2.4
Sharpe 0.30
Post-publication
+2.0%
since 1978, t = 0.9
Decay
51%
of the premium is gone

The whole history

Growth of $1 in the long-short strategy, gross of trading and shorting costs. The dashed line marks the year the paper was published — everything to its right is out-of-sample.

$1 in the long-short strategy, gross of costs. Grey = the sample the paper had; blue = everything since.

Before and after publication

The comparison the site is built around.

WindowReturn/yrVolSharpet-statMax DD
Paper's sample (1957–1971)+3.5%11.4%0.361.2-30%
After publication (1978–2026)+0.8%15.3%0.130.9-71%
Full sample (1951–2026)+3.3%14.8%0.302.4-71%

Against the factor models

Is there a risk-adjusted alpha left once you control for the standard factors? Alpha is annualized; t-stats use Newey–West standard errors.

ModelAlpha/yrAlpha tR²Biggest loadings
CAPM (market only)+4.2%1.90.01Mkt-RF -0.12
Fama–French 3-factor-1.0%−0.80.57HML +1.09, SMB +0.25, Mkt-RF -0.01
Fama–French 5-factor-2.4%−1.80.59HML +1.04, RMW +0.33, SMB +0.33
FF5 + momentum-2.5%−2.00.59HML +1.04, RMW +0.33, SMB +0.33

Rolling performance

Trailing 3-year Sharpe ratio. Long stretches below zero are common.

By decade

1950s+10.0%/yrSharpe 0.91
1960s+3.7%/yrSharpe 0.31
1970s+12.6%/yrSharpe 0.74
1980s+3.2%/yrSharpe 0.22
1990s-0.3%/yrSharpe -0.02
2000s+14.0%/yrSharpe 0.86
2010s-5.2%/yrSharpe -0.47
2020s-5.4%/yrSharpe -0.24

The risk explanation

Same distress-risk logic as book-to-market; E/P just measures cheapness with the income statement instead of the balance sheet.

The behavioral explanation

Naive investors equate a good company with a good stock and overpay for earnings growth that doesn't materialise.

Why it might have decayed

Highly correlated with book-to-market value, so it shares that factor's lost decade. Negative-earnings firms have to be dropped, which makes the sort noisy.

Long-short = top-decile minus bottom-decile of stocks univariately sorted on the characteristic. Value-weighted, US equities, monthly, gross of transaction and shorting costs. The original-paper figures are approximate and annualized for comparison.