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 operates over a 24-name large-cap US universe spanning tech (AAPL, MSFT, NVDA), financials (JPM, V, MA), healthcare (JNJ, UNH), staples (PG, KO, WMT) and energy (XOM, CVX). The design is deliberately mechanical — rank, hold the leaders, cut the laggards — which keeps it readable and auditable. It went live after a reviewer approval on 2026-05-31 flagged as a built-in hand-written strategy at risk level 0.
Recent activity
The story of the last week is inactivity. The scheduled runs on 2026-10-29, 2026-10-30 and 2026-11-02 each executed zero trades and rejected zero, leaving the book flat at $10,000 cash and $10,000 total. In other words, the strategy currently sees no names clearing its trend threshold and is parked defensively on the sidelines. That is a feature, not a bug, for a trend follower — but it also means no live P&L signal is accumulating to confirm or refute the backtest.
Backtest performance
Over 451 days the backtest returned 23.5% (final equity $12,349.63), a 12.52% CAGR, with a Sharpe of 0.95 and a 15.67% max drawdown. Two numbers deserve scrutiny. The win rate is just 28.79% across 136 trades — characteristic of momentum, where a minority of large winners carries a majority of small losers, but it demands discipline to sit through the losing majority. Turnover is a striking 2,638%, implying the book churns many times over; fees totalled $136 (one per trade) with no FX cost, so costs stayed modest here, but that churn is a sensitivity to watch if spreads or slippage rise.
Validation
This is where caution is warranted: validation did not pass. Across four walk-forward folds, three were positive. Fold 3 was excellent (+25.59%, Sharpe 3.32) and fold 4 solid (+13.34%, Sharpe 2.15, the out-of-sample segment). But fold 2 lost 7.31% with a 17.15% drawdown and a negative Sharpe of -1.05 — a reminder that a regime without clear trends punishes this approach hard.
The statistical verdict is the real flag. The Probabilistic Sharpe Ratio is a healthy 0.893, but the Deflated Sharpe Ratio — which penalises for the 6 trials run — falls to 0.476, below the threshold that would mark the result as unlikely to be luck. In plain terms: once you account for how many variants were tested, the edge is no longer statistically convincing.
Verdict
Dual-momentum has an honest, legible thesis and genuinely strong stretches (folds 3 and 4). Against that sit a failed validation, a punishing fold-2 regime, a low win rate that tests conviction, and heavy turnover. Its current all-cash stance is consistent with a disciplined trend follower waiting for a signal — but until live results or a cleaner validation accumulate, treat the 23.5% headline as promising rather than proven.