quant research // summarizing the paperLOADING
Andrew Y. Chen, Ivo Welch · 2026-07-07
The paper re-tests roughly 200 published stock-market anomalies (long-short equity strategies). Their returns shrink dramatically over time and when excluding tiny stocks: from a median 48 basis points per month before 2006 down to just 7 bp when using both post-2005 data and non-micro-cap stocks. Even small allowances for luck or trading costs wipe out that remaining 7 bp.
Why it matters: Practitioners should be skeptical that published academic anomalies generate meaningful returns for realistic, tradable portfolios of larger stocks in recent years. Much of the historical 'edge' appears concentrated in micro-cap stocks and pre-2006 periods, which are hard or costly to exploit at scale.
⚠ This is a historical backtest study focused on U.S. non-micro equities; it evaluates published anomalies rather than proving no strategy can work.
This paper examines about 200 published long-short anomaly equity portfolios (Chen and Zimmermann, 2022). Over the period through 2005 (December 2005 and earlier) and across all stocks, their median zero-investment return was an impressive 48 bp per month. Using only post-2005 years (January 2006 onward) reduces this to 19 bp. Using only "non-micro" top-3,000 stocks in the top 90% of market capitalization reduces this to 26 bp. Using only post-2005 and non-micro stocks reduces this to 7 bp. Even modest allowances for luck or transaction costs would have eliminated even these 7 bp. The evidence strongly suggests that published academic anomalies have been useless to non-micro-cap portfolio managers in the 21st century. Public stock markets were very efficient.
Go deeper: a full research-committee breakdown of this paper, its assumptions and failure modes, and how its method would apply to a specific ticker or your watchlist. See StockTools AI →
AI summary generated from the paper’s public abstract via arXiv; it may miss nuance — read the source before relying on it. Thank you to arXiv for its open-access interoperability; StockTools is not affiliated with arXiv, and all rights remain with the authors. Educational only, not financial advice.