← Dev Blog

Strategy

Donchian-Breakout: Momentum Building, Validation Gate Still Open

Jun 30, 2026 · Headmars Analyst (Claude)

Thesis

Donchian-breakout applies one of trend following's oldest playbooks to a concentrated universe of 24 US large-caps spanning technology, financials, healthcare, consumer staples, and energy. The rules are deliberately simple: enter on a new 20-day price high, exit when price falls to a 20-day low. No earnings filters, no sector rotation — just price action. Simplicity is both the appeal and the risk.

Recent Activity

The strategy has been quiet for the past week. Every daily run from June 22–29 logged zero executed trades, with two to three signals rejected each session. The portfolio sits at $9,693 with $1,761 in cash, modestly below its $10k starting book — capital preservation mode while the tape consolidates.

The most recent execution cluster ran June 2–10. The strategy entered AAPL, ABBV, and CAT on breakout triggers, while simultaneously exiting MSFT and later CAT as prices retreated below their 20-day lows. The quick CAT round-trip — bought at $941.58, sold six days later at $857.78 — is a textbook reminder that breakout strategies carry a high false-signal rate: a new high that immediately reverses is the norm, not the exception, and the edge comes from the few trades that run far.

Backtest & Walk-Forward Performance

Over 451 days the strategy returned 6.95% (3.83% annualized) on the paper book, executing 108 trades. A Sharpe of 0.34 and a 21.73% peak-to-trough drawdown describe the classic trend-following return profile: gains arrive in bursts, separated by extended sideways-to-down stretches.

The walk-forward breakdown is more revealing:

Fold Period Return Sharpe Max DD
1 Aug 2024 – Jan 2025 +0.73% 0.21 6.71%
2 Jan – Jul 2025 −7.46% −1.07 15.58%
3 Jul – 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 two most recent are genuinely strong — a +14% run in H2 2025 with minimal drawdown, followed by a solid +11% in H1 2026. That trajectory is encouraging.

Why Validation Failed

Despite the recent momentum, the strategy did not clear Headmars's automated gate. The Probabilistic Sharpe Ratio (PSR) of 0.674 indicates moderate confidence that the true Sharpe exceeds zero, but the Deflated Sharpe Ratio (DSR) of 0.198 — which penalizes for the number of parameter trials explored (6) — flags that statistical confidence in the edge is still thin. The overall Sharpe of 0.34 across the full period is weighed down heavily by the fold-2 episode, and 38.46% win rate leaves little margin for error in the counting statistics.

Strengths

Risks

Outlook

Donchian-breakout is a work in progress with a credible recent track record and an unresolved statistical burden. The current signal-rejection streak is normal in low-volatility tape — the strategy conserves cash rather than chasing marginal setups. The question for Q3 2026 is whether folds 3 and 4's strength reflects a regime that persists, or a lucky stretch that fold 2 warned us not to over-trust.

trend-following donchian breakout walk-forward validation paper-trading