Stock Market Anomalies: A 2026 Cross-Market List
A catalog of documented equity return anomalies — size, value, momentum, low-volatility, short-term reversal, accruals and more — with academic attribution, plus how SurgeFlow lets you inspect factor
Category: Factor Research
What a market anomaly is
Stock market anomalies are documented cross-sectional return patterns that standard asset-pricing models leave unexplained — including size, value, momentum, low-volatility, short-term reversal, profitability, investment and accruals. Each has academic attribution and debated robustness. SurgeFlow lets you inspect how factor proxies for several of these behave across four markets using live data.
More precisely, an anomaly is a regularity in returns that the prevailing benchmark model — historically the CAPM, later the Fama-French three- and five-factor models — does not account for. Whether a given anomaly reflects compensation for risk, behavioural mispricing, or simply data-mining and survivorship is genuinely contested in the literature. Many documented effects have weakened out-of-sample or after publication (McLean and Pontiff, 2016). This page is a reference catalogue with attribution, not a claim that any of these patterns will persist or that any of them is investment advice.
A 2026 catalogue of documented anomalies
The most-cited cross-sectional anomalies, with their canonical academic attribution. Robustness varies by market, period and definition; treat every entry as contested rather than settled. - Size — small-cap stocks have historically shown higher average returns than large-caps; Banz (1981), formalised as SMB in Fama and French (1993). The premium has been unstable and partly attributed to liquidity and microstructure. - Value — high book-to-market (cheap) stocks vs. low (expensive); Fama and French (1992, 1993), the HML factor. Underperformed for much of the 2010s ('value's lost decade'). - Momentum — past 3-to-12-month winners continue to outperform past losers; Jegadeesh and Titman (1993), Carhart (1997) as WML/MOM. Subject to occasional sharp crashes (Daniel and Moskowitz, 2016). - Low volatility / low beta — low-risk stocks earn returns not commensurately lower than high-risk stocks; Black, Jensen and Scholes (1972); Frazzini and Pedersen (2014), 'betting against beta'. - Short-term reversal — one-month losers tend to rebound; Jegadeesh (1990), Lehmann (1990). Largely a liquidity/microstructure effect. - Profitability / quality — robustly profitable firms outearn unprofitable ones; Novy-Marx (2013), the RMW factor in Fama and French (2015). - Investment — conservative (low-asset-growth) firms outperform aggressive ones; Aharoni, Grundy and Zeng (2013); the CMA factor in Fama and French (2015). - Accruals — firms with high accruals relative to cash earnings subsequently underperform; Sloan (1996). - Liquidity — less-liquid stocks have historically carried a premium; Amihud (2002); Pastor and Stambaugh (2003). - Post-earnings-announcement drift — prices continue drifting in the direction of an earnings surprise; Ball and Brown (1968); Bernard and Thomas (1989).
How factor proxies are constructed and tested
SurgeFlow does not assert that any of the above premia exist; it builds factor proxies and exposes the machinery so you can inspect their behaviour yourself. The factor system implements Fama-French-style long-short proxies — SMB (size), HML (value), RMW (profitability), CMA (investment), WML (momentum) and a liquidity factor (LIQ) — and computes daily factor returns behind a publication gate, with an effective-direction resolver so a sleeve's sign is checked against convention rather than assumed.
Estimation and testing follow standard methods: cross-sectional risk-premium estimation via Fama-MacBeth, joint model tests via the Gibbons-Ross-Shanken (1989) GRS F-test, and multicollinearity screening via VIF. An important honesty caveat applies to the current Factor Backtest V2 evidence. The Phase-1 run is research_backtest_insample — in-sample factor attribution, not live out-of-sample rolling-signal evidence. GRS and the other diagnostics are reported as labels, not used as sleeve-exclusion gates ('diagnostics are labels, not portfolio gates'). Data-quality gates — point-in-time alignment, survivorship handling and minimum-observation thresholds — apply to the data, not to portfolio selection. The backtest composite ranks sleeves by a composite of risk-adjusted return (Sharpe) and Information Ratio across equal-weight and risk-parity variants. The US composite winner is the WML+RMW Top-30 risk-parity sleeve, ranked #1 by both Sharpe and Information Ratio over roughly ten years of in-sample history. Read that as a qualitative in-sample ranking, not an out-of-sample return forecast.
What SurgeFlow data this uses
This catalogue is built on SurgeFlow's open multi-market equity data science layer — research data, not investment advice — spanning four markets: US, China, Japan and Hong Kong. Pricing and fundamentals come from FMP (US), Tushare (China, Hong Kong), and J-Quants and EDINET (Japan). The factor proxies (SMB, HML, RMW, CMA, WML, LIQ), daily factor returns, the publication gate, the effective-direction resolver, Fama-MacBeth estimates, GRS tests and VIF diagnostics are all queryable, alongside the Factor Backtest V2 sleeve rankings. Institutional-ownership context lives in the Whales tab, which combines SEC EDGAR 13F filings in the US with EDINET, HKEX and CSRC disclosures elsewhere — so it covers global ownership, not just 13F. You can pull any of this into Google Sheets today via the public CSV/JSON snapshots with =IMPORTDATA(...) or via Apps Script with a free API key; a Google Workspace Marketplace add-on is in review, not published.
Inspect the proxies yourself
The credible way to use an anomaly catalogue is to interrogate it, not to trust it. Create a free (beta) API key at /account and query factor proxies across the four markets: the free public API (beta) allows 2,000 requests per day and exposes /api/v1/markets/{market}/realtime, /screen and /hotlist, plus /api/v1/summary. The OpenAPI spec is at /api/public/openapi.json and an AI-readable index at /llms.txt. From there you can compare how a size or momentum proxy behaves in, say, Japan versus the US, read the factor methodology on /factors, and review the in-sample backtest framing on /backtesting. None of this is investment advice, no premia are asserted, and the in-sample caveats above travel with the data.
Sources and further reading
Primary, academic, and SurgeFlow data references for the claims above: - Banz, R. (1981). The relationship between return and market value of common stocks. Journal of Financial Economics. - Fama, E. and French, K. (1992, 1993). The cross-section of expected stock returns; Common risk factors in the returns on stocks and bonds. - Fama, E. and French, K. (2015). A five-factor asset pricing model. Journal of Financial Economics. - Jegadeesh, N. and Titman, S. (1993). Returns to buying winners and selling losers. Journal of Finance. - Carhart, M. (1997). On persistence in mutual fund performance. Journal of Finance. - Novy-Marx, R. (2013). The other side of value: the gross profitability premium. Journal of Financial Economics. - Sloan, R. (1996). Do stock prices fully reflect information in accruals and cash flows? The Accounting Review. - Frazzini, A. and Pedersen, L. (2014). Betting against beta. Journal of Financial Economics. - Amihud, Y. (2002). Illiquidity and stock returns. Journal of Financial Markets. - Gibbons, M., Ross, S. and Shanken, J. (1989). A test of the efficiency of a given portfolio. Econometrica. - McLean, R. and Pontiff, J. (2016). Does academic research destroy stock return predictability? Journal of Finance.