The thesis
Dual-momentum is a classic trend-follower: hold the strongest trending names by 60-day return and exit when the trend breaks. It runs over a 24-name large-cap universe spanning tech, financials, healthcare, staples, and energy (AAPL, MSFT, NVDA, JPM, V, JNJ, XOM, and so on). The logic is deliberately simple — let winners run, cut on weakness — which historically produces a low hit rate offset by a few large winners. The strategy is currently marked live.
Recent activity: parked in cash
The most striking feature of the last week is inactivity. Every scheduled run from 8 to 15 September executed 0 trades and rejected 0 candidates, leaving the book flat at $10,000 cash and $10,000 total equity. No name in the universe is clearing the entry threshold, so the strategy is sitting fully in cash rather than forcing marginal positions. That is arguably correct behaviour for a trend-follower in a directionless tape — but it also means the live track record is currently generating no signal to evaluate.
Backtest performance
Over 451 days the backtest returned 23.5% (final equity $12,349.63, ~12.5% CAGR) with a Sharpe of 0.95 and a max drawdown of 15.67%. The win rate is only 28.79% across 136 trades — textbook momentum, where most trades are small losers and the returns come from a minority of large winners. Turnover is heavy at 2,638%, though total fees were a modest $136 and FX cost was zero. A sub-1.0 Sharpe with a 16% drawdown is respectable but not exceptional.
Validation: it did not pass
Here is the honest part. Walk-forward validation across four out-of-sample folds returned passed: false. Three of four folds were positive, and the aggregate OOS return of 13.34% with an OOS Sharpe of 2.15 looks encouraging — but the deflated Sharpe ratio (DSR 0.476) falls short of the bar once you account for the six trials tested. The probabilistic Sharpe ratio (0.893) is healthier, yet the gate correctly weighs the deflation penalty.
The fold detail explains the risk. Fold 2 (Jan–Jul 2025) lost 7.31% with a 17.15% drawdown and a Sharpe of -1.05 — a full regime where the strategy bled. Folds 1, 3, and 4 were strong (Sharpe 1.23, 3.32, 2.15), but that dispersion is the point: performance is highly regime-dependent, clustered in trending markets and punished in choppy ones.
Verdict
Strengths: a coherent, low-cost thesis, a solid headline backtest, and disciplined risk behaviour (it goes to cash rather than over-trade). Risks: it failed formal validation, the returns lean on a handful of winners, and one full fold showed the strategy can lose money for months. The current all-cash posture underlines that this only works when trends exist. Treat it as a promising but unvalidated candidate — worth monitoring live, not yet worth conviction.