What the Strategy Does
Dual-momentum is one of the oldest quant ideas in the book — and for good reason. The agent scores every name in its 24-stock universe (large-cap equities across tech, financials, healthcare, consumer, and energy) on 60-day trailing return, holds the strongest trending names, and exits when momentum rolls over. No forecasting, no macro overlay: pure price-relative strength with a hard trend-break exit.
The universe skews toward liquid mega-caps (AAPL, NVDA, MSFT, JPM, V, MA, among others), which is a deliberate choice — thin names punish momentum strategies with slippage and mean-reversion noise.
Backtest Summary
Over 451 days of simulation, the strategy returned +23.5% on a $10,000 paper account, ending at $12,349.63 — a CAGR of roughly 12.5%. The Sharpe ratio came in at 0.95, which is respectable for an equity-only strategy. Maximum drawdown reached 15.67%, and total turnover was a high 2,638% — the agent traded actively, logging 136 round-trips.
The win rate of 28.8% is the number that jumps out. Fewer than three in ten trades closed profitable. That's not unusual for a trend-follower (you cut losses fast and let winners run), but it does mean the P&L is heavily concentrated in a small number of large winners. Those trades carry the whole book.
Walk-Forward Validation: A Yellow Flag
The four-fold walk-forward tells a more nuanced story:
| Fold | Period | Return | Sharpe | Max DD |
|---|---|---|---|---|
| 1 | Aug 2024 – Jan 2025 | +5.85% | 1.23 | 3.92% |
| 2 | Jan 2025 – Jul 2025 | −7.31% | −1.05 | 17.15% |
| 3 | Jul 2025 – Dec 2025 | +25.59% | 3.32 | 4.04% |
| 4 | Dec 2025 – May 2026 | +13.34% | 2.15 | 7.43% |
Three of four folds are positive, and the out-of-sample (OOS) return on the most recent fold is a strong +13.34% with an OOS Sharpe of 2.15. But the validation gate still shows failed — and the reason sits in the probabilistic metrics.
The Probabilistic Sharpe Ratio (PSR) is 0.893 and the Deflated Sharpe Ratio (DSR) is 0.476. The DSR in particular accounts for the number of trials run against the data (6 in this case) and adjusts for multiple-testing bias. A DSR of 0.48 means there is a meaningful probability that the observed Sharpe is an artifact of the search process rather than a genuine edge. With only 6 trials that number isn't alarming, but it's below the threshold the platform requires before a strategy earns a passing grade.
Fold 2's −7.31% return (with a 17.15% drawdown) is also worth watching. That period roughly corresponds to the market turbulence of early-to-mid 2025; this strategy got hit hard during that regime, suggesting it underperforms in choppy, trendless markets — which is exactly what theory predicts for a pure momentum system.
Live Status
The agent deployed on 2026-05-31 with a $10,000 paper allocation and has logged two scheduled runs since (June 1 and June 2) with no trades executed. The portfolio is sitting fully in cash. This may reflect a weak momentum signal across the universe at the current juncture — the strategy is waiting for a clear trend leader to emerge before entering.
Outlook
Strengths: Solid aggregate return, a strong recent fold, and a theoretically sound, well-documented thesis. The 24-name liquid universe reduces execution risk.
Risks: High turnover, concentration in a few winning trades, regime sensitivity (see Fold 2), and a DSR that doesn't yet clear the validation bar. The strategy would benefit from a longer track record and possibly a volatility filter to reduce exposure during trendless regimes.
The platform's decision to keep this agent in paper-trading mode pending a validation pass is the right call. Dual-momentum is promising — it just needs more out-of-sample evidence before it earns larger capital allocation.