The publication rule, the result of applying it to library v0.1 on data through 2026-09-30, the sample data against the production data, and the cross-validation against public benchmarks. The full record is the PDF below.
Status vocabulary. PASS: meets all criteria of the publication rule in force and is eligible for publication as a return factor; MONITOR: does not meet the rule but shows either a Fama-MacBeth t-statistic above 1.0 or a positive holdout information ratio; computed and reported as a risk descriptor, not presented as a source of return; FAIL: does not meet the rule on any count; computed and reported as a risk descriptor, not presented as a source of return.
Planned for library v0.2. All t-statistics Newey-West adjusted (lag 6); adds (iv) a holdout-period joint Fama-MacBeth t-statistic > 1.0.
| Family | Fama-MacBeth t-statistic > 2.0 | Holdout-period information ratio > 0 | Decile spread > 0 net of 20 bps per side | Turnover / mo | Status |
|---|---|---|---|---|---|
| Momentum Total return from month-end t-12 to month-end t-1 (skips the most recent month) | +0.72 · not met | +0.78 · met | -0.02 · not met | 55% | MONITOR |
| Low volatility Negative of annualized standard deviation of daily returns over trailing 252 trading days | -3.20 · not met | -1.22 · not met | -0.28 · not met | 17% | FAIL |
| Size (small minus big) Negative log market capitalization (current shares x month-end price; sample-issue approximation) | +2.75 · met | +0.24 · met | +0.14 · met | 11% | PASS |
| Short-horizon reversal Negative of trailing 21-trading-day total return | -0.12 · not met | -0.76 · not met | -0.09 · not met | 172% | FAIL |
Construction sample 2016–2022, holdout 2023-01-01 onward; 128 monthly cross-sections; 482 stocks. Each criterion cell prints the observed value against the threshold in its column heading. Fama-MacBeth t-statistics are uncorrected for autocorrelation. The “Decile spread > 0 net of 20 bps per side” column is the annualized D10 − D1 return as a decimal fraction (-0.02 = -1.6%) after a linear haircut of 20 basis points per side on measured turnover.
Library v0.1. The validation record is reissued with each monthly issue and on every library change.
Hypothetical results. Backtested and simulated figures are hypothetical: computed from historical data under stated assumptions, gross of transaction costs unless stated, not the results of actual trading or of any account, and not achievable by any investor. Not prepared for use in the recipient's advertising. Disclosures.
Data note. Source: publicly available daily prices, split- and dividend-adjusted close and volume. The universe is the current S&P 500 constituent list, so companies that left the index over the period are absent and the figures carry survivorship bias. Market capitalization uses one current share count per name, applied to every month-end price, not the count as reported on each date. Figures illustrate the methodology and format.
| Data input | Sample issues | From commercial launch |
|---|---|---|
| Coverage universe | 482 stocks (503 current S&P 500 constituents less 21) + 28 ETFs and volatility indices | Screened U.S. equities plus liquid ETFs, point-in-time |
| Index membership | Current constituents | Point-in-time membership; delisted and demoted names retained |
| Share counts | Current shares outstanding applied to historical prices | As-reported share counts on each date |
| Fundamentals | None; the library uses price and volume only | None |
| Corporate actions | Split and dividend adjustment embedded in the adjusted close | Splits, dividends, mergers, delistings, symbol changes |
| Macro data | VIX and VIX3M; Treasury, credit and gold ETFs as proxies | Rates, term structure, credit spreads, volatility indices |
| In-house spread | Benchmark | Months | Period | Correlation | Beta to benchmark | Sign agreement |
|---|---|---|---|---|---|---|
| Momentum | French UMD | 127 | 2016-02 to 2026-08 | 0.900 | 1.33 | 87% |
| Size (small minus big) | French SMB | 127 | 2016-02 to 2026-08 | 0.597 | 0.80 | 67% |
| Momentum, sector-neutral variant | French UMD | 127 | 2016-02 to 2026-08 | 0.835 | — | — |
The benchmark series are value-weighted 2×3 sorts on the full CRSP universe; the in-house spreads are equal-weighted deciles within the sample universe, so correlation rather than level agreement is the relevant test. The size spread correlates at 0.60 with SMB; the lower figure is expected because SMB spans the full market while the sample universe contains large caps only, so the in-house spread measures size dispersion within the S&P 500 rather than small-versus-large across the market. Sector neutralization removes the component of momentum that is a bet on sector rotation; both variants are carried with their own validation.
Full specification, construction rules, decile results by factor, cost sensitivity and the correlation of monthly spreads are in the validation record (PDF, 12 pp).
The benchmark comparison uses only public series from the Kenneth R. French Data Library and can be reproduced from them: correlation, beta and sign agreement of the in-house momentum and size spreads against UMD and SMB. The comparison, with the series used, is supplied in reproducible form on request.
Enter your firm email to open the sample issues. The Monthly Quantitative Note follows each month; unsubscribe any time.
Firm addresses only. The address is used to send the Note and to reply to enquiries, and for nothing else. Privacy.