.jpg)
.jpg)
A Donchian channel draws two lines you could calculate on the back of a napkin: the highest high and the lowest low of the last N price bars. That is the whole indicator — and it is the engine behind one of the most studied trend-following systems in market history.
This guide explains Donchian channels from the arithmetic up: how the three lines are built, which period to use, exactly how a breakout is traded, and why the same tool that made the Turtle Traders famous still whipsaws traders in sideways markets. If you want to put these mechanics to work on real charts, a structured technical analysis course will take you further than any single article can.
- A Donchian channel = highest high and lowest low over N bars, with a midline halfway between.
- The classic setting is 20 periods (Richard Donchian's four-week rule); 55 is the slower, higher-conviction cousin.
- A close beyond the upper or lower band is a breakout signal — the core of the Turtle Traders' system.
- Unlike Keltner (ATR) or Bollinger (standard deviation), Donchian bands track raw price extremes, not an average.
- Breakout systems win on a minority of trades — they survive on a few large winners, so risk control matters more than the signal.
What is a Donchian channel?
A Donchian channel is a volatility-and-trend indicator made of three lines: an upper band equal to the highest high over a set number of bars, a lower band equal to the lowest low over the same window, and a middle line sitting exactly halfway between them. Price riding the upper band signals strength; price pinned to the lower band signals weakness.
It was developed by Richard Donchian, the fund manager widely called the "Father of Trend Following." His four-week breakout rule — buy a new four-week high, sell a new four-week low — is the direct ancestor of the modern 20-period channel. (Source: Wikipedia, "Richard Donchian," 2025.)
How a Donchian channel is calculated
The math is deliberately simple, which is part of why it has survived for decades. For a chosen lookback of N bars:
- Upper band = the highest high of the last N bars.
- Lower band = the lowest low of the last N bars.
- Middle line = (upper band + lower band) ÷ 2.
Work a quick example. Say a stock's highest high over the last 20 sessions is $182.40 and its lowest low is $168.20. The upper band sits at $182.40, the lower band at $168.20, and the midline at (182.40 + 168.20) ÷ 2 = $175.30. A daily close above $182.40 prints a fresh 20-day high — a long breakout. (Figures are illustrative, to show the arithmetic, not a forecast.)
Notice what the bands do not use: no moving average, no standard deviation, no smoothing. They step up only when a genuinely higher high prints, and they stay flat the rest of the time. That staircase shape is the signature of the indicator.
The middle line is more useful than most beginners realize. In an uptrend, pullbacks often stall near the midline before price pushes to a new high, so it acts as a rough dynamic support level. In a downtrend, the midline becomes resistance. Read the three lines together and you get a compact picture: direction from where price sits, and the width of recent volatility from the gap between the bands.
What is the best Donchian channel period?
The honest answer: there is no universally "best" period, only trade-offs between speed and reliability. Two settings dominate because the Turtle Traders used them.
The 20-period channel is the classic default — roughly one month of trading, matching Donchian's four-week rule. It reacts faster, catches trends earlier, and produces more signals, including more false ones. The 55-period channel is slower and higher-conviction: fewer signals, later entries, but a much stronger filter against noise. (Source: Wikipedia, "Donchian channel," 2025.)
A common structure uses both: a fast channel for entries and a slower one as a trend filter, taking longs only when the slow channel confirms an uptrend. If you already combine signals this way, the same logic appears in other trend-following indicators like ADX and Supertrend. Shorter periods (10–15) suit active intraday traders; longer ones (55+) suit position traders riding multi-week moves.
Match the period to your timeframe, not to a backtest
The temptation is to optimize the number until it looks perfect on past data. Resist it. A setting that fit last year's chart will not fit next year's, and over-tuning simply memorizes noise. Pick a period that matches how long you can realistically hold a position — days, weeks, or months — and keep it consistent. A trader who flips between 15 and 55 every time a trade goes wrong has no system at all, just a moving target.
How to trade a Donchian channel breakout
A Donchian channel breakout strategy is rules-based by design — the appeal is that it removes discretion. Here is the mechanical sequence for a classic 20-day long setup.
What to do with this: paper-trade the sequence before risking capital, and journal every signal. Many traders also confirm the move with a volume indicator like OBV, so a breakout backed by rising participation carries more weight than one on thin volume.
The Turtle Traders: the system that made Donchian channels famous
In 1983, veteran trader Richard Dennis made a bet with his partner William Eckhardt: that trading could be taught to complete beginners. He recruited a group of trainees — the "Turtles" — and handed them a rules-based system built on Donchian-style channel breakouts. One trainee, Curtis Faith, later published the rules in Way of the Turtle (2007).
The system ran on two timeframes. System 1 was the fast, 20-day engine; System 2 was the slow, 55-day one. The exits used a shorter opposite channel, so a trade closed only after the trend had clearly turned.
| Rule | System 1 (short-term) | System 2 (long-term) |
|---|---|---|
| Entry breakout | 20-day high / low | 55-day high / low |
| Exit (opposite channel) | 10-day low / high | 20-day low / high |
| Skip filter | Skip entry if the previous signal won | No skip — take every signal |
| Initial stop | ~2 volatility units (N) | ~2 volatility units (N) |
Source: TurtleTrader.com, "Original Turtle Trading Rules," 2025; Curtis Faith, "Way of the Turtle," 2007. Exit parameters vary slightly between accounts.
What made the experiment famous was the result: several of the trainees went on to run money professionally, and the story became the standard case study for the idea that disciplined, rules-based execution beats gut instinct. The parameters mattered less than the discipline to follow them without exception, trade after trade, through the losing streaks.
The lesson for a modern trader is not to copy the parameters blindly — markets have changed since 1983 — but to copy the structure: a clear entry trigger, a volatility-based stop, and a mechanical exit that takes the decision out of your hands at the worst possible emotional moment.
Donchian vs Keltner vs Bollinger: what's the difference?
All three are "channel" indicators, so beginners mix them up. The real difference is what sets the band width — and that changes what each one is good for.
| Indicator | Bands are set by | Middle line | Best at |
|---|---|---|---|
| Donchian | Raw price extremes (highest high / lowest low) | Midpoint of the two bands | Pure breakout discipline |
| Keltner | ATR multiple around an EMA | Exponential moving average | ATR-normalized trend riding |
| Bollinger | Standard deviation around an SMA | Simple moving average | Volatility squeezes and reversals |
Source: StockCharts ChartSchool, 2025; FXOpen research, 2025.
The key distinction: Donchian is the only one of the three whose edges ignore an average entirely and lock onto absolute highs and lows. That makes it blunt but honest — the band moves only when the market genuinely makes a new extreme. For the ATR-based alternative, see how the ATR-based Keltner channel compares; choose Donchian when you want a mechanical breakout, Keltner when you want a smoother trend envelope, and Bollinger when you are hunting squeezes.
Do Donchian channels actually work?
Here's the catch: a Donchian channel is excellent at one job and poor at another. In a trending market, it keeps you on the right side of a sustained move for weeks. In a range-bound market, it generates repeated false breakouts — price pokes above the band, you buy, and it immediately falls back inside. That is the dreaded whipsaw.
Be realistic about hit rate. Channel-breakout trend systems typically win on a minority of their trades; one widely-cited backtest referenced on Wikipedia's Donchian entry (default 20-period setting) reported a win rate near 35% — meaning most individual signals lost money even where the overall system was positive. (Source: Wikipedia, "Donchian channel," 2025; methodology not independently verified.)
That is not a flaw to be fixed; it is the nature of trend following. The system makes money because the few winners are far larger than the many small, quickly-cut losers. If you cannot stomach a string of six or eight losing breakouts before the one that pays for them all, this is the wrong tool for your temperament — and no setting will change that.
Most traders who abandon Donchian channels do so after making one of a handful of avoidable errors:
- Trading every breakout in a range. Without a trend filter, a sideways market will hand you loss after loss as price ping-pongs between the bands.
- Moving the stop further away. Widening a stop to avoid being closed out turns a small, planned loss into a large, unplanned one — the exact opposite of what the system needs.
- Taking profits early. Cutting a winner at a round number throws away the oversized trend that funds the whole approach.
- Over-optimizing the period. A number curve-fit to history rarely survives live markets; a robust setting beats a perfect one.
Frequently asked questions
Trading involves substantial risk of loss and is not suitable for every investor. This article is educational content, not investment advice.
.jpg)
.jpg)
.jpg)
.jpg)
.jpg)
.jpg)
.jpg)
.jpg)
.jpg)
.jpg)
.jpg)
.jpg)
.jpg)
.jpg)
.jpg)
.jpg)
