Thesis
Donchian breakout is a textbook trend-following rule: buy when price touches a new 20-day high, exit when it falls to a 20-day low. The universe spans 24 large-cap U.S. names across tech, financials, healthcare, consumer, energy, and industrials — liquid, well-covered stocks where breakouts tend to carry informational weight.
The logic is durable: price making a 20-session high reflects sustained buying pressure, and the symmetric exit enforces discipline without discretion. No forecasting, no earnings calls, no macro calls — just price structure.
Backtest Summary
Over 451 days, the strategy compounded to +6.95% total return (3.83% CAGR) on a $10,000 starting book, finishing at $10,695. That headline number is adequate but not exciting.
The deeper numbers tell a mixed story:
| Metric | Value |
|---|---|
| Sharpe ratio | 0.34 |
| Max drawdown | 21.73% |
| Win rate | 38.46% |
| Trades (108 total) | ~$1/trade in fees |
| Turnover | 2,082% |
A Sharpe of 0.34 is below the threshold most institutional frameworks require, and a 21.7% peak-to-trough drawdown is substantial for a strategy that's only returning ~7% total. The win rate of 38% is consistent with trend-following — these systems make money by letting winners run, not by being right often — but it demands that average winners meaningfully exceed average losers.
Walk-Forward Validation
The strategy ran a 4-fold walk-forward backtest, and the results are genuinely interesting:
| 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 positive, and the final out-of-sample fold — the one that matters most — returned +11.35% with a Sharpe of 1.86. That's a strong signal. Fold 2's −7.46% was the weak link, likely coinciding with the choppy, mean-reverting conditions of early-to-mid 2025 that are notoriously hostile to breakout strategies.
Despite the encouraging recent folds, the validation gate returned failed. The Probabilistic Sharpe Ratio sits at 0.674 and the Deflated Sharpe Ratio at just 0.198 — after adjusting for the number of trials (6) and backtest length, the statistical confidence that the true Sharpe is positive remains below the required threshold. This is an honest result: the system shows promise but lacks sufficient track record to clear the overfitting bar.
Recent Activity
The strategy has gone quiet. Every scheduled run from June 15–22 recorded zero executions, with 2–3 signals rejected each day. Cash has sat at $1,760.89 throughout, with total equity drifting between $9,258 and $9,448 — suggesting open positions are treading water.
The last executed trades (early June) were a mixed bag: a CAT round-trip at a loss (bought at $941.58, sold at $857.78), offset by entries in KO, UNH, ABBV, and AAPL. The signal drought since then implies prices in the universe haven't been making clean 20-day highs — a sign of a consolidating or directionless tape.
Strengths and Risks
Strengths: The out-of-sample trajectory is improving. Folds 3 and 4 show the strategy can perform well in trending markets, and the rule-based structure eliminates execution bias entirely.
Risks: The max drawdown of 21.7% is the most pressing concern — it's nearly three times the total return. The strategy is also high-turnover (2,082% annualized), which amplifies fee drag at scale. And the failed PSR/DSR validation means we should treat the backtest as exploratory, not predictive, until more live data accumulates.
Donchian breakout is worth watching through the next trending phase. If it can string together a third consecutive positive fold, the statistical case for graduation becomes materially stronger.