What the Strategy Does
Donchian breakout is one of the oldest systematic trading ideas in the book: enter a position when a stock closes at a 20-day high, exit when it closes at a 20-day low. No sentiment, no earnings calls, no macro narrative — just price structure.
The live universe spans 24 large-caps across six sectors (tech, financials, healthcare, consumer staples, consumer discretionary, energy, and industrials). The breadth is deliberate: trend-following works best when it can rotate freely across whatever is actually breaking out, rather than being pinned to a single theme.
Backtest Snapshot
Over 451 trading days the strategy compounded to a +6.95% total return on a notional $10,000 base (final equity $10,695), implying a CAGR of roughly 3.83%. That is modest in absolute terms, but the style characteristics are textbook trend-following:
- Win rate of 38.5% — most trades lose a little; the strategy profits when winners run.
- Max drawdown of 21.7% — meaningful, and the primary source of the weak Sharpe.
- Sharpe of 0.34 — positive but thin, reflecting the lumpy, momentum-driven return profile.
- Turnover of ~2,083%/yr — very high. At $1 per trade in fees (108 trades, $108 total), friction is contained at current paper-trading scale, but real-money slippage in this regime deserves scrutiny before any live allocation.
Walk-Forward Validation: A Tale of Two Halves
The four-fold walk-forward tells a more nuanced story than the headline numbers suggest.
| Fold | Period | Return | Sharpe | Max DD |
|---|---|---|---|---|
| 1 | Aug 2024 – Jan 2025 | +0.73% | 0.21 | 6.7% |
| 2 | Jan – Jul 2025 | −7.46% | −1.07 | 15.6% |
| 3 | Jul – Dec 2025 | +14.09% | 2.72 | 3.1% |
| 4 (OOS) | Dec 2025 – May 2026 | +11.35% | 1.86 | 6.0% |
Three of four folds are positive, and the out-of-sample fold (Fold 4) is the best-performing period on both return and risk-adjusted basis — an encouraging sign that the edge has not been fully arbitraged away. Fold 2, however, was a bruising -7.5% in six months, and that volatility drags down the full-sample Sharpe to 0.34.
Despite the strong OOS result, the strategy did not pass Headmars's formal validation gate. The Probabilistic Sharpe Ratio (PSR) of 0.674 and Deflated Sharpe Ratio (DSR) of 0.198 reflect the core problem: with only six trials and high return variance across folds, there is insufficient statistical evidence that the Sharpe is reliably above zero after adjusting for multiple testing. Validation failure is not a verdict on the strategy's potential — it is the system working as intended, demanding more evidence before promotion.
Recent Activity: Patience Mode
The live paper account (current equity ~$9,488) has been quiet since mid-June. Every scheduled run from June 19 through June 25 executed zero trades, rejecting two to three signals per session. This is not a bug — it reflects the strategy waiting for genuine 20-day-high breakouts that do not currently exist in the universe. Discipline here is a feature, not an oversight.
The last burst of activity (late May through early June) showed the rotation the strategy was designed for: entering KO and UNH on defensive breakouts, exiting MSFT at a loss and CAT at a quick profit, and adding AAPL and ABBV on momentum. Cross-sector rotation in a single week is par for the course when multiple names hit channel highs simultaneously.
Strengths and Risks
What works: The OOS fold is the strongest in the entire test window — Sharpe 1.86, drawdown under 6% — which suggests the strategy found a gear heading into 2026. Three of four folds are positive. The logic is transparent and robust to look-ahead bias.
What to watch: The 21.7% max drawdown is the clearest risk: a real-money account running this strategy would need to stomach a prolonged underwater period. Turnover is high enough that execution quality (spreads, order routing) would materially affect live returns. And the DSR of 0.20 is a genuine caution flag — the edge is real but narrow, and the statistical case for it needs more out-of-sample reps before it earns a larger mandate.
Bottom line: Donchian breakout is a disciplined, well-specified strategy in a recognized style class. The OOS performance is genuinely promising. The validation framework is doing its job by demanding patience — and so, for now, is the strategy itself.