The thesis
Dual-momentum is about as classic as systematic trading gets: hold the strongest-trending names by 60-day return, and exit when the trend breaks. It runs on a 24-name universe of large-cap U.S. equities — mega-cap tech (AAPL, MSFT, NVDA), financials (JPM, V, MA), staples (PG, KO, WMT) and a few cyclicals. The logic is intentionally simple, which is both its appeal and its liability: simple rules resist overfitting, but they also offer no cushion when the whole universe stalls.
Recent activity
Here the picture is quiet — arguably too quiet. The last six scheduled runs (2026-07-20 through 2026-07-27) all report 0 executed, 0 rejected, with the book holding $10,000 in cash and $10,000 total. In other words, the strategy is fully flat, holding no positions at all. That is not a malfunction; it is the trend-break exit doing its job when nothing in the universe clears the momentum bar. Still, a week of pure cash means zero participation in whatever the market has done since, and it underscores that dual-momentum spends real stretches on the sidelines.
Backtest and validation
The backtest is genuinely attractive at the headline level: 23.5% total return, a 12.52% CAGR over 451 days, final equity of $12,349.63, and a Sharpe of 0.95 against a contained 15.67% max drawdown. The catch is the shape of those returns. Win rate is just 28.79% across 136 trades — this is a strategy that loses most of the time and relies on a handful of large winners, so its equity curve will feel lumpy. Turnover is a striking 2,638%, meaning frequent rotation and, over time, non-trivial fee drag ($136 in fees against ~$2,350 of profit).
The walk-forward validation is where honesty matters most: it does not pass. Across four folds, three were positive — including a standout Fold 3 (+25.59%, Sharpe 3.32) and a solid Fold 4 (+13.34%, Sharpe 2.15) — but Fold 2 lost -7.31% with a -1.05 Sharpe and a 17.15% drawdown. Out-of-sample return lands at 13.34% with a strong 2.15 OOS Sharpe, and the Probabilistic Sharpe Ratio is high at 0.89. But the Deflated Sharpe Ratio — which discounts for the 6 trials run — sits at just 0.476, below the confidence bar. That deflation is exactly why the gate reads passed: false.
Verdict
Dual-momentum has real merit: positive out-of-sample results, disciplined risk (it goes to cash rather than force trades), and a transparent thesis. But the DSR failure and the ugly Fold 2 are warnings that the 23.5% headline may owe something to trial selection. The low win rate and high turnover demand patience and cheap execution. My read: promising and worth keeping live in paper, but not yet a strategy I'd trust with meaningful capital until it clears validation cleanly.