The Thesis
Dual-momentum is one of the more durable ideas in quantitative equity research: rank a fixed universe by 60-day return, hold the strongest trending names, and exit when the trend breaks. The Headmars implementation runs this logic daily across 24 large-cap U.S. equities spanning tech, financials, healthcare, consumer, energy, and industrials — a diversified but high-quality universe designed to give the momentum signal room to breathe without straying into micro-cap noise.
Backtest Overview
Over 451 days of paper trading the strategy compounded to $12,349 from a $10,000 start, a total return of 23.5% and an annualised CAGR of 12.5%. The Sharpe ratio sits at 0.95 — respectable for a single-factor strategy, though not exceptional. The max drawdown of 15.7% is moderate; combined with a win rate of just 28.8% across 136 trades, the strategy is clearly a typical trend-follower: most individual trades lose small, and a handful of winners carry the return.
Turnover is notable at 2,638% annualised, which translates to roughly seven full portfolio rotations per year. At $1 per trade in fees (totalling $136 over the period), transaction costs are negligible in paper-trading terms — but that assumption deserves scrutiny before any live deployment at scale.
Fold-by-Fold Breakdown
The walk-forward validation split the history into four equal windows, and the dispersion tells the real story:
| Fold | Period | Return | Sharpe | Max DD |
|---|---|---|---|---|
| 1 | Aug 2024 – Jan 2025 | +5.85% | 1.23 | 3.9% |
| 2 | Jan 2025 – Jul 2025 | −7.31% | −1.05 | 17.2% |
| 3 | Jul 2025 – Dec 2025 | +25.59% | 3.32 | 4.0% |
| 4 | Dec 2025 – May 2026 | +13.34% | 2.15 | 7.4% |
Fold 3 is the standout — a Sharpe of 3.32 and a 25.6% return with minimal drawdown suggests the strategy caught a clean trending market. Fold 2 is the concern: a −7.3% return and a 17.2% drawdown in a single five-month window indicates the model can chop badly when momentum signals are noisy or mean-reverting conditions dominate.
Validation Status: Not Passed
The platform's automated validation gate returned failed, driven primarily by a Deflated Sharpe Ratio (DSR) of 0.476 against a trial count of 6. The PSR of 0.893 is acceptable in isolation, but once adjusted for the number of parameter configurations tested, the probability that the observed Sharpe is a genuine edge — rather than the best outcome of a search process — drops materially. Three of four folds are positive, and the out-of-sample Sharpe of 2.15 (Fold 4) is encouraging, but the DSR flags that the overall result may carry more luck than the headline numbers suggest.
Current Activity
As of June 18, 2026, the strategy has been flat for at least six consecutive scheduled runs, holding 100% cash at $10,000. No trades have been executed or rejected since at least June 10. This means no names in the universe are currently clearing the momentum threshold — a posture that, given recent market conditions, may be entirely rational. Patience is a feature of trend-following, not a bug.
Strengths and Risks
Strengths: Clean conceptual thesis with a long academic track record. The out-of-sample Fold 4 performance (+13.3%, Sharpe 2.15) suggests the model retains edge in trending regimes. Low fees and a liquid universe limit implementation drag.
Risks: The failed DSR gate is not a cosmetic issue — with only 451 days of history and 6 trials, the statistical foundation is thin. Fold 2's drawdown shows the strategy has no defensive mechanism in choppy or reversing markets. High turnover could erode returns meaningfully at larger capital sizes.
Bottom Line
Dual-momentum earns its place as a live strategy on Headmars, but it warrants cautious position sizing until it accumulates a longer out-of-sample track record. The current all-cash stance is a clean signal the model is working as designed — waiting for confirmation before committing capital.