Strategy Overview
Donchian-breakout applies one of the oldest rules in systematic trading: enter when price touches a new 20-day high, exit when it falls to a 20-day low. The logic is purely trend-following — ride names already in motion and step aside when they lose it. The universe spans 24 large-cap U.S. equities across technology, healthcare, financials, consumer staples, and energy, giving the strategy broad sector exposure without venturing into illiquid territory.
Backtest Snapshot
Over 451 days, the strategy returned +6.95% (3.83% CAGR), ending at $10,695 on a $10,000 base.
| Metric | Value |
|---|---|
| Total Return | +6.95% |
| CAGR | 3.83% |
| Sharpe Ratio | 0.34 |
| Max Drawdown | −21.73% |
| Win Rate | 38.46% |
| Trades | 108 |
The Sharpe of 0.34 reflects an inconsistent return stream rather than a structurally negative one. The more pressing concern is the 21.73% max drawdown relative to the total return earned — the strategy can spend extended periods meaningfully underwater. A 38.46% win rate is normal for trend-following; the model depends on winners running far enough to more than offset frequent small losses. Whether that payoff asymmetry is genuinely present is what the walk-forward analysis tries to answer.
Walk-Forward Validation
The strategy ran through four out-of-sample folds, and the cross-period picture is sharply divided.
| Fold | Period | Return | Sharpe | Max DD |
|---|---|---|---|---|
| 1 | Aug 2024 – Jan 2025 | +0.73% | 0.21 | −6.71% |
| 2 | Jan 2025 – Jul 2025 | −7.46% | −1.07 | −15.58% |
| 3 | Jul 2025 – Dec 2025 | +14.09% | 2.72 | −3.14% |
| 4 | Dec 2025 – May 2026 | +11.35% | 1.86 | −6.01% |
Three of four folds are profitable, and the most recent out-of-sample period — Fold 4 — delivered +11.35% at a Sharpe of 1.86. That is genuinely strong. Fold 2, however, was a −7.46% period with a Sharpe of −1.07, illustrating the strategy's vulnerability to choppy, range-bound markets where 20-day breakouts repeatedly fail and stop out.
Despite the encouraging OOS result, the strategy did not pass the Headmars validation gate. The Probabilistic Sharpe Ratio (PSR) sits at 0.674, meaning roughly a 67% probability the true Sharpe exceeds zero after accounting for return path noise. More critically, the Deflated Sharpe Ratio (DSR) drops to 0.198 once adjusted for six trial variants — only a 20% probability the edge survives the multiple-testing penalty. The recent outperformance may owe something to a favorable trending regime rather than a structural, repeatable advantage.
Recent Activity
The last six scheduled runs (June 8–15) have been largely quiet. Only four trades executed across the period: a KO buy and CAT sale on June 10, and a UNH buy and MSFT sale on June 9. Several signals were rejected — likely by position-sizing or risk controls — while cash has held at $1,760.89 for four consecutive sessions. Portfolio value has drifted from $9,715 on June 8 to $9,369 on June 15, reflecting mark-to-market erosion rather than new trading losses.
The recent rotation pattern is worth noting: entries into defensive and healthcare names (KO, UNH, ABBV) alongside exits from technology (MSFT) and industrials (CAT) suggest the strategy is gravitating toward names currently holding 20-day highs in the prevailing environment.
Balanced Assessment
Donchian-breakout is a structurally sound approach on a quality large-cap universe, and its two most recent folds are the best it has posted. The concern is statistical: with six trials and a DSR below 0.20, the confidence band around the true edge is wide. A second live regime resembling Fold 2 — a choppy six months that erased 7.5% — would test conviction hard. Practitioners comfortable with low win rates, episodic deep drawdowns, and a multi-year horizon will find the thesis credible. But the validation gate is right to hold the bar higher before endorsing full deployment.