Donchian Channel Breakout Trading

20 days is roughly one trading month and was used by the Turtles in their System 1. It provides a good balance between signal frequency and quality. Shorter periods give more signals but more false breakouts; longer periods give fewer but higher conviction signals.

System 1 uses 20-day entry / 10-day exit with a filter (skip if previous breakout was winner). System 2 uses 55-day entry / 20-day exit with no filter. System 2 catches bigger, longer-term trends but trades less frequently.

False breakouts are inherent to breakout trading. The Turtle filter (skip after winner) helps. Also use proper position sizing so false breakouts don't hurt too much. Expect 60-65% of breakouts to fail; the 35-40% that work should be large winners.

Close is more reliable. An intraday breakout that doesn't close beyond the channel is often a false signal. Wait for the daily close to confirm. This filters some false signals at the cost of slightly later entry.

Typically 35-45% win rate. This is low but expected for trend-following. The key is that winners are much larger than losers (profit factor 2-3). Many small losses are offset by fewer large gains.

Track each breakout signal and its outcome. If a long breakout was profitable, skip the next long breakout signal. If it was a loss, take the next signal. The logic: trends often start after failed breakouts, so winners often follow losers.

The Turtles added 1 unit for each 0.5× ATR move in favor, up to 4 units. This is aggressive. A more conservative approach: add after significant new breakout (e.g., 20-day high while already long from earlier breakout).

If price gaps beyond the channel, you can: 1) Enter at the gap open (immediate), 2) Wait for close to confirm, 3) Use limit order at channel level hoping for pullback. Gaps are tricky; many traders reduce size on gap entries.

The Turtles used the same parameters (20/10 or 55/20) across all markets, relying on ATR-based sizing to normalize. You can customize, but ensure you're not curve-fitting. One benefit of standard parameters is simplicity and avoiding over-optimization.

Follow Turtle-style limits: max 4 units per market, max 6 in correlated markets, max 10 one direction, max 20% total heat. Exit oldest units first when scaling out. Regularly review correlation exposure.

Use walk-forward optimization: optimize on past data, test on subsequent out-of-sample, roll forward. Use multiple markets to ensure parameters aren't market-specific. Prefer parameters near the middle of good performance ranges rather than at extremes.

ATR is used for: 1) Position sizing (equalizes volatility risk), 2) Stop placement (2× ATR), 3) Pyramid levels (0.5× ATR increments). ATR normalizes the system across different volatility regimes and markets.

Donchian struggles in ranging regimes. Options: 1) Accept drawdowns during ranges, 2) Add regime filter (ADX, volatility), 3) Use adaptive periods. The Turtles accepted range drawdowns as cost of catching trends; they sized to survive drawdowns.

Yes. Some traders use Donchian breakouts for trend following and mean reversion (fade breakouts) in ranging conditions. Use ADX or channel width to identify regime. In low ADX, fade breakouts to middle line. In high ADX, follow breakouts.

Monitor correlations dynamically; they change over time. Group highly correlated markets and apply group limits. During crises, correlations spike. Consider reducing overall exposure during high correlation periods, or ensure diversification across uncorrelated asset classes.