The thesis
Dual-momentum is a trend-follower: it holds the strongest-trending names by 60-day return and exits when the trend breaks. Its universe is 24 large-cap U.S. names spanning tech, financials, healthcare, staples, and energy — AAPL, MSFT, NVDA, JPM, UNH, XOM, and the like. The logic is deliberately simple: ride winners, cut losers when momentum fades. That simplicity is a virtue, but as the numbers show, it comes with a distinctive risk profile.
Recent activity: parked in cash
The most striking fact is what isn't happening. Across the six most recent scheduled runs (2026-07-10 through 2026-07-17), the strategy executed zero trades every single time. Cash sits at $10,000 and total equity at $10,000 — the book is fully in cash. Read charitably, this is the trend filter doing its job: nothing in the universe currently clears the momentum threshold, so the strategy declines to force a position. Read skeptically, a live strategy earning nothing is an opportunity cost, and a prolonged flat stretch tells us little about edge.
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%. Those are respectable, mid-tier numbers. Two cautions stand out. First, the win rate is just 28.79% across 136 trades — classic momentum behavior, where a handful of large winners must outweigh many small losers. That works only if the tails keep cooperating. Second, turnover is enormous at 2,638%, meaning frequent churn; fees of $136 are modest here but would scale with size and slippage in the real world.
Validation: not passed
This is where enthusiasm should cool. Our walk-forward validation gate returned passed: false. The encouraging parts: 3 of 4 folds were positive, out-of-sample return was 13.34% with an OOS Sharpe of 2.15, and the probabilistic Sharpe ratio (PSR) is a strong 0.893. But the deflated Sharpe ratio (DSR) is only 0.476 — after adjusting for the 6 trials run during development, the evidence for genuine edge falls below our confidence bar.
The fold detail explains the caution. Folds 1, 3, and 4 were solid (+5.85%, +25.59%, +13.34%), but fold 2 (Jan–Jul 2025) lost 7.31% with a Sharpe of -1.05 and a 17.15% drawdown — worse than the full-period max drawdown. That single regime shows how badly the strategy can behave when trends whipsaw.
Verdict
Dual-momentum is a plausible, well-behaved trend-follower with a genuine backtest but not yet a validated edge. The strengths — positive OOS results, high PSR, disciplined cash stance — are real. The risks are equally real: a sub-threshold DSR, a punishing low win rate, one ugly losing regime, and a live book that has been idle for a week. I would keep it in observation, not capital-at-scale, until it either trades through a full trend cycle or the deflated statistics improve.