The 200-day moving average: invested above, out below. In 1992 the paper found different returns on each side on the Dow 1897-1986; in 1999 and 2012 others showed the effect survives neither the count of rules tried nor fees.
Here it is a witness: the 12/10/2026 record predicts failure after publication. If the Lab said "survives", the judge would need examining.
| Market | Unseen period | Verdict | Rule / buy-and-hold | Max drawdown rule / buy-and-hold | Proof |
|---|---|---|---|---|---|
| US stocks (total market) | 1993-01 → 2026-08 (33.7 y) | No proof of an edge yet | +1,170.5% / +3,262.3% | 36.9% / 54.6% | SAP-cfd1492ddc37 |
What the theory brings, measured. Negative witness, as planned by the record: over 33.7 unseen years (since January 1993), the 200-day average earned +1,170% vs +3,262% for buy-and-hold, with a smaller maximum drawdown (36.9% vs 54.6%). The Lab does not say "survives": it behaves like the 1999 and 2012 literature that took apart the 1992 result. It is a control of the judge: the world's most famous rule does not pass, and the judge says so.
Rule written and preregistered on 12/10/2026, before any proof, with its source's parameters; one trial per market, nothing tuned; judged in certified mode by Lab 270, which cuts the unseen period itself: everything after the source's publication. "Rule / buy-and-hold": gain of the rule and of buy-and-hold on the same bars, 10 bp fee per order. Proofs SAP-cfd1492ddc37. Gaps with the source: Ken French US total market (total return) instead of the Dow Jones. Data: Ken French Data Library, FRED (Board of Governors). Research on past data, not advice; no future return is promised.
Brock, W., Lakonishok, J., LeBaron, B. (1992), Journal of Finance 47(5), 1731-1764, DOI 10.1111/j.1540-6261.1992.tb04681.x; Sullivan, R., Timmermann, A., White, H. (1999), JF 54(5), 1647-1691, DOI 10.1111/0022-1082.00163; Bajgrowicz, P., Scaillet, O. (2012), JFE 106(3), 473-491, DOI 10.1016/j.jfineco.2012.06.001.