The thesis
Dual-momentum is a deliberately simple idea: hold the strongest trending names by 60-day return, and exit when the trend breaks. It runs across a 24-name large-cap universe spanning tech, financials, healthcare, staples and energy — AAPL, MSFT, NVDA, JPM, V, UNH, XOM and the like. There is no attempt to be clever about entries; the strategy just tries to own what is working and cut what isn't.
That design shows up clearly in the numbers. The win rate is only 28.79% across 136 trades, yet the strategy is net positive. That asymmetry — mostly small losses, occasionally large winners — is the signature of a trend-following system doing its job rather than a broken one.
Recent activity
Right now the strategy is doing nothing, and that is worth stating plainly. The last six scheduled runs (20–27 October) each executed 0 trades and rejected 0, leaving the book at $10,000 in cash and $10,000 total. No name in the universe is currently clearing the momentum filter cleanly enough to hold, so dual-momentum is fully defensive. That is the strategy behaving as designed in a trendless tape — but it also means recent live performance is simply flat.
Backtest and validation
Over 451 days the backtest returned 23.5% (final equity $12,349.63), a CAGR of 12.52%, a Sharpe of 0.95, and a max drawdown of 15.67%. Turnover is high at 2,638%, with $136 in fees — churn is a real cost here, not a rounding error.
The honest headline, though, is that walk-forward validation did not pass. Across four folds, three were positive, but fold 2 (Jan–Jul 2025) lost 7.31% with a 17.15% drawdown — the ugly regime this approach must survive. Folds 3 and 4 were strong (Sharpe 3.32 and 2.15), which flatters the aggregate.
The statistical gatekeepers tell the cautionary story. The Probabilistic Sharpe Ratio is a healthy 0.893, but the Deflated Sharpe Ratio is 0.476 — below the significance bar once you penalise for the 6 trials that went into tuning this. In plain terms: the result could still be selection luck, and the framework is right to withhold its stamp.
The verdict
Dual-momentum is a coherent, low-maintenance strategy with a believable mechanism and a trend-follower's honest risk shape. The strengths are real: positive aggregate return, controlled drawdown in three of four folds, and a disciplined willingness to hold cash — as it is doing now — rather than force trades.
The risks are equally real: a failed validation, a deflated Sharpe that doesn't survive multiple-testing scrutiny, one genuinely painful losing fold, and punishing turnover. We're keeping it live but unproven. It stays a candidate, not a conviction, until out-of-sample data earns the edge the backtest only suggests.