Thesis and Approach
The donchian-breakout strategy applies one of the oldest rules in systematic trend following: enter long when price clears a 20-day high, exit when it falls to a 20-day low. Running across a 24-stock universe of large-cap U.S. equities — spanning technology, financials, healthcare, consumer staples, energy, and industrials — the strategy aims to ride sustained directional moves while cutting losers mechanically, with no discretionary override.
Backtest Overview
Over 451 days, the strategy returned 6.95% in total (CAGR 3.83%) with a Sharpe ratio of 0.34. The more sobering figure is the 21.73% maximum drawdown: at its worst, the portfolio shed more than a fifth of its value before recovering. With a win rate of just 38.46% across 108 trades, donchian-breakout follows the classic trend-follower profile — most trades lose small, and a handful of extended winners are expected to carry cumulative performance. Turnover of 2,083% reflects how aggressively the 20-day bands cycle positions within a concentrated universe.
Cross-Validation: A Tale of Two Halves
The four-fold walk-forward validation paints a more nuanced picture. The first two folds (August 2024 – July 2025) were difficult: Fold 1 barely broke even at +0.73% with a flat Sharpe of 0.21, while Fold 2 posted a -7.46% return and a Sharpe of -1.07 — the worst stretch in the test window, consistent with a choppy, range-bound market where breakout signals fail repeatedly.
The second half reverses that narrative sharply. Fold 3 (July – December 2025) delivered +14.09% with a Sharpe of 2.72, and Fold 4 — the designated out-of-sample period — matched with +11.35% and a Sharpe of 1.86. Three of four folds finished positive, and the trajectory is clearly improving.
Despite this momentum, the strategy did not pass formal validation. A Probabilistic Sharpe Ratio of 0.674 and a Deflated Sharpe Ratio of 0.198 — estimated over only six trials — fall well short of the thresholds needed to rule out overfitting. With the full-period Sharpe at 0.34, there is insufficient statistical confidence that the observed edge is real and durable rather than lucky timing.
Recent Activity
The past week has been quiet. No trades executed on June 11, 12, 15, or 16; a handful of signals were rejected each day. Portfolio value slipped from $9,657 on June 9 to $9,363 by June 16, suggesting open positions drifted lower as no new breakout triggers fired.
The most recent executions (June 9–10) rotated toward defensives: 23 shares of KO at $83.69 and 3 shares of UNH at $412.62 were added, while MSFT was sold at $404.21 and CAT was stopped out at $857.78 — a realized loss versus the June 4 entry at $941.58. The shift away from tech and industrials and into consumer staples and healthcare may reflect the strategy responding to narrowing high-momentum breadth in those sectors.
Outlook
donchian-breakout is live and generating signals, with genuine promise in its two most recent folds. The risks are real, however: a sub-40% win rate demands patience through losing streaks, the peak drawdown of 21.73% is steep for a single-strategy allocation, and formal statistical validation has not been cleared. The OOS results are encouraging early evidence — not proof of a durable edge. This strategy is worth monitoring closely, but capital commitment should remain sized accordingly until the PSR and DSR cross acceptance thresholds.