The Thesis
Dual-momentum keeps it simple: hold the strongest trending names by 60-day return, and exit when the trend breaks. It runs live over a 24-name large-cap universe spanning tech, financials, healthcare, staples, and energy — AAPL, MSFT, NVDA, JPM, UNH, XOM and their peers. This is textbook trend-following: let winners run, cut positions when momentum fades. It is a well-studied factor, and the strategy's design reflects that discipline rather than a novel bet.
Recent Activity
The more striking story is what hasn't happened. Across six consecutive scheduled runs from July 6 to July 13, the strategy executed zero trades and rejected zero candidates, holding a flat $10,000 in cash the entire time. No name in the universe cleared its entry criteria, so the book stayed empty. That is defensible behavior for a trend model in a choppy, directionless tape — sitting out is a position — but a week of full cash also means zero participation in any upside, and it underscores that the live track record is still thin.
Backtest Performance
Over 451 days, the backtest returned 23.5% (12.52% CAGR), lifting a $10,000 book to $12,349.63 on a 0.95 Sharpe with a 15.67% max drawdown. The profile is unmistakably momentum: a 28.79% win rate across 136 trades means the strategy is wrong roughly seven times in ten, yet still profits because the winners are large enough to carry the losers. The cost side deserves attention — turnover ran a hefty 2,638%, and while fees totaled only $136 here, that churn is a real drag in live conditions and a source of slippage the backtest may understate.
Validation — The Sticking Point
This is where enthusiasm should cool. The strategy failed its validation gate. Walk-forward testing across four folds shows three positive and one badly negative: fold 2 (Jan–Jul 2025) lost 7.31% with a −1.05 Sharpe and a 17.15% drawdown, while folds 3 and 4 delivered strong 25.59% and 13.34% runs. Out-of-sample return held up at 13.34% with a 2.15 OOS Sharpe, and the Probabilistic Sharpe Ratio is a healthy 0.893. But the Deflated Sharpe Ratio — which penalizes for the 6 trials tested — sits at just 0.476, below a confidence threshold. In plain terms: after accounting for how many variants were tried, we cannot be confident the edge is real rather than lucky.
Verdict
Dual-momentum is a coherent, disciplined strategy with a genuine backtest edge and sensible risk controls. But the low win rate, high turnover, one ugly fold, and a failing deflated-Sharpe check all argue for caution. Treat it as promising-but-unproven: watch whether it re-engages the market and whether live results resemble folds 3–4 or fold 2 before sizing it up.