Donchian Channel Strategy: The Original Turtle System
In 1983, Richard Dennis and William Eckhardt recruited 21 people off the street, taught them a simple breakout strategy, and gave them each $1 million to trade. They became known as the Turtles — and they made over $175 million in 4 years. Their core strategy? Donchian Channels.
What Are Donchian Channels?
Created by Richard Donchian in the 1940s, Donchian Channels plot the highest high and lowest low over a specified period. The channel has three lines: the upper (highest high), lower (lowest low), and middle (average of the two).
The default period is 20 — the same period the Turtles used. The upper line marks the highest price in the last 20 periods. The lower line marks the lowest. When price breaks above the upper line, it signals a new 20-period high — a potential uptrend. When it breaks below the lower line, it signals a new 20-period low — a potential downtrend.
The Original Turtle Rules
The Turtle System 1 (shorter-term) used 20-day Donchian Channels:
- Entry (long): Buy when price breaks above the 20-day high
- Entry (short): Sell when price breaks below the 20-day low
- Stop: 0.5 × ATR(20) from entry price (ATR-based stop, not fixed)
- Exit (long): Close when price breaks below the 10-day low
- Exit (short): Close when price breaks above the 10-day high
The system uses a 10-day exit channel (shorter than the 20-day entry channel). This asymmetry — slow to enter, quick to exit — lets profits run while cutting losses faster.
Position Sizing (The Turtles' Secret)
The real secret of the Turtles was not the breakout signal — it was position sizing. They risked exactly 2% of their account per trade, calculated using ATR:
Position size = (Account × 2%) / (ATR × pip value). If account = $100,000, ATR = 50 pips, pip value = $10/lot: position = ($100,000 × 0.02) / (50 × $10) = 4 lots.
This normalised risk across all instruments. A volatile instrument with a large ATR got a smaller position. A quiet instrument with a small ATR got a larger position. The dollar risk was always the same: 2% of the account.
Adapting for CFD Trading
The Turtle system was designed for commodity futures, but it adapts well to CFDs:
- Best instruments: Index CFDs (S&P 500, DAX 40), gold (XAU/USD), and major forex pairs. These trend well — the key requirement for breakout systems.
- Avoid: Range-bound instruments and exotic pairs. Breakout strategies lose money in ranging markets.
- Timeframe: Daily charts work best — this matches the original Turtle design. 4H charts can work but produce more false breakouts.
- Spread impact: CFD spreads are wider than futures. For a 20-day breakout, the spread is a small percentage of the move — acceptable. For shorter timeframes, spread becomes a bigger drag.
Why Most Traders Fail with This Strategy
The Donchian strategy has a win rate of only 35-45%. Most trades lose. The strategy is profitable because the winners are 3-5× larger than the losers — trend following math.
Most traders cannot handle a 60% loss rate. After 5 consecutive losses (which happens regularly), they abandon the system. Then the one big trend comes that would have recovered all losses and made a year's profit — and they missed it.
The Turtles succeeded because they followed the rules mechanically. Dennis and Eckhardt designed the system to be rule-based precisely because they knew emotional traders would fail.
Adding a Trend Filter
To reduce false breakouts, add a trend filter:
- ADX > 25: Only take breakouts when ADX indicates a trending market. Filters out 40-50% of false breakouts.
- 200 SMA alignment: Only take long breakouts when price is above the 200 SMA. Only short breakouts when below.
- Volume confirmation: Breakout candle should have above-average volume. Low-volume breakouts fail 60%+ of the time.
These filters reduce the number of trades but improve the win rate from 35% to 45-50% — making the strategy more psychologically manageable.
Disclaimer: CFDs are complex instruments and come with a high risk of losing money rapidly due to leverage. 70-80% of retail investor accounts lose money when trading CFDs. Backtesting does not guarantee future results. Always consider whether you can afford the potential loss of your capital.