Strategy Overview
The donchian-breakout strategy applies one of systematic trading's oldest rules: enter long when a stock closes at a 20-day high; exit when it closes at a 20-day low. Applied across a 24-stock universe of large-cap US equities — spanning technology, financials, healthcare, consumer staples, and energy — the system is designed to capture sustained directional moves while cutting positions that fail to hold their breakout levels. The logic is intentionally simple: no filters, no earnings-season overlays, no macro regime gates.
Backtest Summary
Over 451 days of paper trading, the strategy produced a total return of +6.95%, equivalent to a 3.83% CAGR. The aggregate Sharpe ratio of 0.34 and a win rate of just 38.46% across 108 trades are textbook trend-following numbers: many small losers, offset by fewer but larger winners. The maximum drawdown of 21.73% is the sharpest risk marker — a figure that would test the patience of most systematic traders during adverse regimes. Turnover was high at 2,083%, though with flat $1 per-trade fees across all 108 executions, friction remained manageable.
Fold Analysis: A Tale of Two Regimes
The four-fold walk-forward validation tells a nuanced story.
| Fold | Period | Return | Sharpe | Max DD | Trades |
|---|---|---|---|---|---|
| 1 | Aug 2024 – Jan 2025 | +0.73% | 0.21 | 6.71% | 30 |
| 2 | Jan 2025 – Jul 2025 | -7.46% | -1.07 | 15.58% | 38 |
| 3 | Jul 2025 – Dec 2025 | +14.09% | 2.72 | 3.14% | 20 |
| 4 | Dec 2025 – May 2026 | +11.35% | 1.86 | 6.01% | 26 |
Three of four folds were profitable, and the trajectory is notable: folds 3 and 4 delivered Sharpe ratios of 2.72 and 1.86 respectively, with drawdowns contained well below 7%. The out-of-sample fold — fold 4, the most recent held-out period — returned +11.35% with a Sharpe of 1.86, the system's strongest risk-adjusted reading.
Fold 2 is the cautionary counterpoint. The -7.46% loss with a -1.07 Sharpe and a 15.58% drawdown most likely reflects the choppy, mean-reverting conditions that chronically punish breakout rules — a known Achilles heel for pure Donchian systems.
Validation Verdict: Gate Not Cleared
Despite the improving recent folds, the strategy did not pass validation. The Probabilistic Sharpe Ratio (PSR) of 0.674 implies only a 67.4% probability that the true edge exceeds zero — suggestive but not definitive. More critically, the Deflated Sharpe Ratio (DSR) of 0.198 penalises the six parameter trials that preceded this configuration, reducing confidence to 19.8%. The multiple-testing adjustment is working as intended: an observed Sharpe of 0.34 across a modest 108-trade sample simply does not clear the statistical bar when alternatives have been explored.
The strategy remains live in paper-trading mode, accumulating the evidence needed to revisit the gate.
Recent Activity
The past week has been quiet. Daily scheduled runs from July 2–9 each logged zero executed trades, with 1–3 signals rejected per session — the system is generating candidates but none clearing its entry criteria. The paper portfolio stands at roughly $9,884 against a $1,761 cash reserve, broadly flat week-over-week.
June was more active. The system rotated through MSFT, CAT, UNH, KO, ABBV, and AAPL. Two round-trips stand out: MSFT entered at $450.24 and exited at $404.21, while CAT entered at $941.58 and was stopped out at $857.78 — both losses entirely consistent with the 38% win-rate profile. The system is behaving as designed; the question is whether the winners, when they arrive, will be large enough to justify the drawdown.
Bottom Line
Donchian-breakout is structurally sound and shows improving recent performance. The final two folds hint at a strategy that may be finding its footing in more trending conditions. However, the validation gate exists precisely to guard against recency bias: a DSR of 0.198 across six trials does not constitute a deployable edge. The next 60–90 days of live paper performance — particularly through any renewed market volatility — will be the most informative data the system has yet generated.