The thesis
Dual-momentum is a straightforward trend-follower: hold the strongest names by 60-day return from a 24-stock universe of large-cap US equities (AAPL, MSFT, NVDA, JPM, JNJ, XOM and peers), and exit when the trend breaks. There is no bet on mean reversion here — the strategy assumes winners keep winning until they don't, and it accepts frequent small losses in exchange for occasionally riding a big move.
Backtest performance
Over 451 days the strategy returned 23.5%, growing a $10,000 book to $12,349.63, for a CAGR of 12.52%. The Sharpe ratio is a modest 0.95 and the maximum drawdown was 15.67%.
The headline number that jumps out is the 28.79% win rate across 136 trades. On its own that reads badly, but it is exactly the signature of a working trend-follower: most positions are cut for small losses when the trend fails, and the aggregate return is carried by a minority of large winners. What deserves more scrutiny is turnover of 2,638% — the book is churning heavily, and while fees came in at only $136 in this test, that level of trading is fragile to slippage and any real-world cost that the backtest understates.
Validation — the caveat that matters
Walk-forward validation did not pass. Across four out-of-sample folds, three were positive and one was sharply negative: fold 2 (Jan–Jul 2025) lost 7.31% with a -1.05 Sharpe and a 17.15% drawdown on 60 trades — a choppy, trendless regime that whipsawed the strategy. Folds 3 and 4 were strong (25.59% and 13.34%), so the edge is real but regime-dependent.
The deflated Sharpe ratio (DSR 0.476) is the sticking point. With a probabilistic Sharpe of 0.893 the strategy looks decent in isolation, but after adjusting for the six trials run in its development, the DSR sits below the bar we require. In plain terms: some of that backtest shine may be selection, not skill. The strategy stays live but flagged, not blessed.
Recent activity
Recent comments tell a quiet story. Every scheduled run from 2026-08-28 through 2026-09-04 reports 0 executed, 0 rejected, cash $10,000, total $10,000. The strategy is fully in cash and has been for over a week — no name in the universe is clearing its trend filter. That is by design, not a bug: a trend-follower with nothing trending should hold cash rather than force positions. But it also means the strategy is currently contributing no live signal, and we have no fresh out-of-sample data points to update our confidence.
Verdict
Dual-momentum is a defensible, well-behaved implementation of a well-understood idea. Its strengths — disciplined exits, cash when trends are absent, a genuine winner in favourable regimes — are real. Its risks are equally clear: a failed validation gate, a nasty trendless-market fold, punishing turnover, and a low win rate that demands patience from anyone watching the equity curve. We keep it live, on a short leash, and wait for the trends to return.