The thesis
donchian-breakout is a classic trend-following system: buy when price closes at a 20-day high, exit when it closes at a 20-day low. It trades a 24-name large-cap universe spanning tech (AAPL, MSFT, NVDA), financials (JPM, V, MA), staples (PG, KO, WMT, COST) and energy (XOM, CVX). The logic is deliberately simple and hand-written — the reviewer approved it at risk 0 as a built-in strategy. There is nothing clever hiding here, which is both its appeal and its ceiling.
Recent activity
The strategy went to work on 2026-10-29, executing four buys: WMT (20 @ $121.17), HD (7 @ $328.57), ABBV (8 @ $289.91) and MA (3 @ $636.60). Since then it has been quiet in the worst way — every scheduled run from 2026-10-30 through 2026-11-04 executed zero trades while rejecting one or two signals each time. Cash has sat frozen at $959.13 throughout, which strongly suggests those rejections are the account running out of buying power rather than a lack of breakouts. Total equity has drifted between roughly $9,912 and $10,391, so the portfolio is marginally in the green on a $10k base, but it is fully invested and unable to act on new signals.
Backtest and validation
Over 451 days the backtest returned +6.95% (final equity $10,695), a 3.83% CAGR. That headline hides the texture: a 0.34 Sharpe, a 21.73% max drawdown, and a 38.46% win rate across 108 trades. The low hit rate is normal for trend following — the method survives on a few large winners — but pairing it with a sub-0.4 Sharpe and a 20%+ drawdown is an uncomfortable ratio. Turnover of 2,082% also means the strategy churns hard; fees of $108 are modest here but would bite at scale.
Walk-forward validation failed, and the detail is instructive. Three of four folds were positive, with folds 3 and 4 genuinely strong (+14.09% at Sharpe 2.72; +11.35% at Sharpe 1.86). But fold 2 lost 7.46% at a −1.07 Sharpe, and that single regime drags the full-sample Sharpe down to 0.34. The out-of-sample window looks excellent (11.35% return, 1.86 Sharpe), yet the deflated Sharpe ratio of 0.198 — against a PSR of 0.674 across 6 trials — is what sinks the gate. In plain terms: once you account for the number of configurations tried, the evidence that this edge is real rather than lucky is thin.
Verdict
The recent out-of-sample strength is real and worth watching, but the failed DSR gate and the brutal fold-2 behaviour say the same thing the drawdown does — this is a volatile, regime-dependent system. The frozen cash and string of rejected runs are a more immediate problem: a trend-follower that cannot open new positions is just holding four stocks. Keep it live, keep it small, and don't mistake two good folds for a validated edge.