The Thesis
Donchian-breakout is trend-following in its purest, oldest form: buy the 20-day high, exit on the 20-day low. It trades a 24-name large-cap universe spanning tech (AAPL, MSFT, NVDA), financials (JPM, V, MA), staples (KO, PG, WMT) and healthcare (JNJ, UNH, ABBV). The logic is unchanged since the Turtle traders popularised it — ride sustained moves, cut losers quickly, accept that most trades will be small losses.
Backtest Performance
Over 451 days the strategy returned 6.95% (3.83% CAGR) with a Sharpe of 0.34 and a maximum drawdown of 21.73%. The win rate is 38.46% across 108 trades — low, but consistent with breakout systems that lean on a handful of large winners. What stands out is turnover of 2,082%: the book churns roughly twenty times over, and while fees came to only $108, that level of activity leaves the edge exposed to slippage and whipsaw in choppy regimes.
Validation: The System Says No
This is where the picture darkens. The walk-forward validation did not pass. Three of four folds were positive, and the out-of-sample tail looks excellent — 11.35% return at a 1.86 Sharpe in the most recent fold, and a stellar 14.09% / 2.72 Sharpe in fold three. But fold two (Jan–Jul 2025) lost 7.46% at a Sharpe of -1.07, exposing the strategy's Achilles heel: sideways, mean-reverting markets that generate false breakouts.
The deflated statistics are the real red flag. The Probabilistic Sharpe Ratio of 0.674 is unconvincing, and the Deflated Sharpe Ratio of just 0.198 — after adjusting for 6 trials — says the observed performance is hard to distinguish from luck. A full-sample Sharpe of 0.34 against those OOS numbers suggests the good folds are carrying the record.
Recent Activity: An Idle Book
The live account tells a quieter story. The last executed trades were on 12–13 August (buying V and MA, selling AAPL and UNH). Every scheduled run since — 27 August through 3 September — reports the same pattern: 0 executed, 1–2 rejected. Cash sits at $1,510.88 and total equity hovers near $10,080, essentially flat. Orders are being generated and then bounced, likely on position-sizing or cash constraints, and the strategy has effectively stopped acting on its signals.
Verdict
Donchian-breakout is doing exactly what it says on the tin — and that's both its strength and its limit. The mechanics are transparent, the recent out-of-sample folds are genuinely strong, and drawdowns in the good regimes stayed shallow. But a failed validation gate, a punishing DSR, high turnover and a sub-40% win rate mean this is not a strategy to size up on. The immediate priority is operational: those repeated rejections need diagnosing before any judgement on live edge is fair. Until the book can actually place its trades, the backtest is the only story we have — and the validation says treat it with caution.