Time-series momentum: each month, be long a market if its last-12-month return (above the risk-free rate) is positive, short otherwise, sized to the same volatility for every market.
A positive average premium across a broad basket of futures (58 in the paper), especially in extreme markets.
| Market | Unseen period | Verdict | Rule / buy-and-hold | Max drawdown rule / buy-and-hold | Proof |
|---|---|---|---|---|---|
| Asia-Pacific ex Japan | 2012-06 → 2026-08 (14.2 y) | No proof of an edge (a loss on the unseen period) | -38.0% / +145.0% | 61.9% / 29.3% | SAP-ed106b457ad4 |
| Europe | 2012-06 → 2026-08 (14.2 y) | No proof of an edge yet | +10.3% / +260.3% | 51.6% / 31.1% | SAP-3bb2d1e84c17 |
| Japan | 2012-06 → 2026-08 (14.2 y) | No proof of an edge yet | +62.4% / +266.7% | 44.4% / 28.9% | SAP-9bbdecc89388 |
| North America | 2012-06 → 2026-08 (14.2 y) | No proof of an edge yet | +93.2% / +561.4% | 56.8% / 24.1% | SAP-239198fd301e |
| EUR/USD | 2012-06 → 2026-10 (14.3 y) | No proof of an edge (a loss on the unseen period; Deflated Sharpe below 50%) | -12.6% / -9.5% | 26.4% / 30.9% | SAP-1375252ac1b9 |
| GBP/USD | 2012-06 → 2026-10 (14.3 y) | No proof of an edge (a loss on the unseen period; Deflated Sharpe below 50%) | -29.6% / -14.1% | 31.1% / 37.6% | SAP-49ae5f418e37 |
| USD/JPY | 2012-06 → 2026-10 (14.3 y) | No proof of an edge yet | +58.6% / +101.4% | 21.7% / 20.3% | SAP-7fa3f9e03da3 |
What the theory brings, measured. Over 14.2 unseen years (since June 2012), the rule played market by market, without leverage, stays below buy-and-hold in 4 regions out of 4, with deeper drawdowns than the market. On currencies, two markets out of three lose. What the theory really promises concerns a basket of dozens of futures, where some trends offset others' reversals: a single market does not keep that edge. The measured lesson: time-series momentum is a portfolio property, not a signal to play alone.
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-ed106b457ad4, SAP-3bb2d1e84c17, SAP-9bbdecc89388, SAP-239198fd301e, SAP-1375252ac1b9, SAP-49ae5f418e37, SAP-7fa3f9e03da3. Gaps with the source: No leverage (exposure capped at 100%); one market per proof, while the promise concerns the basket; spot currencies, without the rate differential; gold band left out (monthly averages). Data: Ken French Data Library, FRED (Board of Governors). Research on past data, not advice; no future return is promised.
Moskowitz, T. J., Ooi, Y. H., Pedersen, L. H. (2012), "Time series momentum", Journal of Financial Economics 104(2), 228-250, DOI 10.1016/j.jfineco.2011.11.003; Hurst, Ooi, Pedersen (2017), JPM 44(1), 15-29, DOI 10.3905/jpm.2017.44.1.015.