Thesis
Dual-momentum is a trend-following strategy with a simple mandate: hold the strongest-trending names by 60-day return, and exit when the trend breaks. It fishes in a pond of 24 large-cap US names spanning tech, financials, healthcare, staples, and energy — AAPL, MSFT, NVDA, JPM, V, JNJ, XOM, and peers. The design bets that recent relative strength persists, and that a mechanical exit on trend break caps the damage when it doesn't.
Recent Activity
Here the live story diverges sharply from the backtest. Across the last six scheduled runs (2026-07-30 through 2026-08-06), the strategy executed zero trades each day — no entries, no rejections. Cash and total equity have held flat at $10,000, meaning the book is fully idle and uninvested. No name in the universe is currently clearing the trend filter, so the strategy is doing exactly what a disciplined trend-follower should when there is nothing to hold: sitting on its hands. That is defensible behavior, but it also means recent performance contributes nothing, and the strategy is earning no return while capital waits.
Backtest Performance
Over 451 days the backtest returned 23.5% (CAGR 12.52%), ending at $12,349.63 with a Sharpe of 0.95 and a max drawdown of 15.67%. Two numbers deserve scrutiny. First, the win rate is just 28.79% across 136 trades — typical for trend-following, where a few large winners must pay for many small losers, but it demands strict exit discipline to work. Second, turnover is a remarkable 2,638%, implying heavy churn; at $1 per trade the fee drag is modest here, but at scale or with slippage this churn is a real cost.
Validation
The walk-forward validation did not pass, and this is the headline risk. Of four folds, three were positive, and out-of-sample return was a healthy 13.34% with an OOS Sharpe of 2.15. The Probabilistic Sharpe Ratio (0.893) is encouraging. But the Deflated Sharpe Ratio is only 0.476 across six trials — below the confidence bar — which flags that the strong result may partly reflect selection across configurations rather than genuine edge.
The fold detail explains the caution. Fold 2 (Jan–Jul 2025) lost 7.31% with a 17.15% drawdown — deeper than the full-period max drawdown, a sign the strategy can bleed badly in choppy, trendless regimes. Folds 3 and 4 were excellent (Sharpe 3.32 and 2.15), but that dispersion — one ugly period offset by two strong ones — is exactly what regime-sensitive momentum looks like.
Verdict
Dual-momentum has a coherent thesis and genuine out-of-sample returns, but it carries a failed validation, a low win rate, and clear regime risk. The current all-cash live posture is honest rather than alarming — no signal, no position. I would treat this as a promising-but-unproven strategy: keep it on paper, watch how it behaves when a trend finally triggers, and weight the failed DSR heavily before trusting it with real capital.