Donchian channel and Turtle breakout
BreakoutIntermediate
A Donchian channel is the highest high and lowest low over a lookback window. A break of the upper channel is a long breakout, and a break of the lower channel is a short. The Turtle trading experiment turned that idea into a full system, with volatility-based size and a shorter channel for exits.
Not investment advice. Nothing here is a recommendation to buy or sell anything. Backtests and historical studies are not guarantees, and a rule that worked in one sample can fail in the next. Trading can lose money. Links are for education, not endorsements. Do your own research.
- Rules
- A basic rule is long on a new N-day high and short on a new N-day low, with N often 20. The published Turtle system used a 20-day breakout and a slower 55-day breakout, risked a fraction of equity per ATR unit, and exited on a break of a shorter channel, such as 10 days. One version skipped a 20-day signal if the previous breakout had been a winner.
- Where AI fits
- The useful automation is the portfolio version: many markets, one risk budget, and position size tied to ATR. AI can code that and test whether a regime filter should block breakouts when markets are highly correlated. Tuning the channel length on a single chart is the version most likely to fool you.
- Markets
- A basket of futures across commodities, rates, currencies, and equity indexes. That diversification is part of the original design.
- Timeframes
- Daily bars. Weekly for a slower trend sleeve.
- Risks
- Most breakouts fail, so the result depends on a few large trends and on cutting the losers. One market can spend years in a drawdown. The rules are public, and a backtest that ignores the portfolio and the position-sizing rules is not the strategy.