1992issue C121-9
Turtle trading as one testable Donchian breakout procedure
Turtle trading taught basic trend-following and Donchian breakout rules after about two weeks of classroom work, then moved trainees from single-lot practice to full speed. The harder skill was following those rules while losing on about 65 to 70 percent of trades.
- The method used basic trend-following and channel-breakout rules rather than complicated algorithms. The harder skill was following those rules while losing on about 65 to 70 percent of trades.
- The breakout system bought a new high over X days and sold a new low over X days, initiating on 50-day extremes and liquidating on 20-day opposite extremes.
- The same lookback was applied across markets rather than fitted market by market. Position size was worked backward from a tested profit-and-loss profile, with no more than 2 percent of bankroll on any single trade and no more than 10 percent on total open exposure.
- If a required hold lasted three or four months and volatility felt uncomfortable, the prescribed adjustment was smaller size, not cutting the trade early.
Classroom rules before live accounts
Classroom instruction lasted about two weeks and covered trend-following rules, simple chart patterns, channel breakouts, and money-management concepts before trainees received live accounts.
Trainees first practiced single-lot mechanics for about two weeks, then were told to trade full speed. The method used basic trend-following and channel-breakout rules rather than complicated algorithms.
One Donchian breakout lookback
Entries used a horizontal-channel Donchian breakout: a new high over X days was a buy and a new low over X days was a sell, with 50-day extremes used to initiate and 20-day opposite extremes used to liquidate.
The same lookback concept was applied across markets rather than market-by-market optimization, on the view that past best-fit parameters tend to change. A longer breakout window was treated as more significant than a shorter one, for example a 50-day new high versus a 20-day new high.
Risk caps that stay with the hold
Ruin was defined as a 50 percent account drawdown, with risk-of-ruin targets illustrated as 1 or 2 percent for an institutional trader and 5 percent for an individual trader.
Position sizing worked backward from a tested profit-and-loss profile, with a hard stop of no more than 2 percent of bankroll on any single trade and no more than 10 percent of bankroll on total open exposure.
Even a signal with everything going for it was framed as having only about a 52 to 53 percent chance of being right, so the edge was treated as a long-run frequency, not a forecast of the next move.
If a required hold lasted three or four months and volatility felt uncomfortable, the prescribed adjustment was smaller size, not cutting the trade early.
Following the procedure while most trades lose
The harder skill was following those rules while losing on about 65 to 70 percent of trades.
All readings on this track · 6 readings
- 1992Turtle trading as one testable Donchian breakout procedure
- 2001Hourly Donchian and Turtle rules locked after one walk-forward window
- 2001Critiquing Turtle trend following as a drawdown budget
- 2010Evaluate Turtle-style bands as one mechanical procedure
- 2012Quiet-tape cycle clocks for index futures
- 2020Constructing Turtle-style dual Donchian trend-bands