Are Value and Momentum Still Everywhere?

Thirteen years after publication, Jose Ordonez of Alpha Architect re-examined "Value and Momentum Everywhere" by Asness, Moskowitz, and Pedersen, which documented positive average returns to value and momentum across individual stocks, equity indexes, bonds, currencies, and commodities. Using AQR's public datasets and Kenneth French's Data Library, the analysis covers July 2013 to July 2026. All returns are hypothetical paper portfolios that exclude fees, transaction costs, and taxes. The "everywhere" portfolio is shown only through January 2025, when AQR's non-stock series end.

The "everywhere" long/short portfolio averaged 5.58% annually with 4.48% volatility before publication, a 1.25 Sharpe ratio. Afterward, it averaged 0.48% with 2.36% volatility, a 0.20 Sharpe. Across global stock selection, pre-publication returns averaged 6.34% a year (1.18 Sharpe), falling to 2.11% (0.44 Sharpe) afterward. In the U.S., pre-publication returns averaged 5.94% (0.84 Sharpe), compared with 1.22% (0.21 Sharpe) afterward. In each long/short set, value lagged sharply, with the U.S. value leg down 39.99% cumulatively, while momentum offset much of the damage. The author notes that the weakness was most pronounced in the 2018-2020 "quant winter."

In the long-only U.S. analysis, a 50/50 value and momentum portfolio's CAPM alpha dropped from 3.03% (t=5.55) before publication to -0.66% (t=-0.44) after. The combination compounded at 12.50% against the market's 14.24%, with momentum at 15.42% and value at 9.16%. The combination's maximum drawdown relative to the market was 27.21%, from a July 2014 peak to an October 2025 trough, compared with 8.47% in the pre-publication sample.

The author states that the data add to the evidence of post-publication decay but does not answer whether positive factor premia can be expected going forward. Three arguments are offered. First, the post-2020 global stock selection Sharpe of 0.71 is closer to the pre-publication level, though the author cautions that the start date was chosen and the average return is not statistically significant (t=1.40). Second, the paper's core findings still hold for combinations: value and momentum remain negatively correlated within all eight asset classes (-0.43 to -0.67 after publication), and average pairwise cross-asset correlations rose to 0.27 for value and 0.33 for momentum from 0.16 and 0.23. However, seven of 16 individual strategies had negative average returns after publication, and the long/short combination results are not statistically significant given 139 and 157 monthly observations. Third, the author suggests value's underperformance may reflect a risk that investors are paid to bear over time, a hypothesis the post does not test, with the U.S. valuation gap between cheap and expensive stocks widening to near-record levels.