Volatility & bandsDonchian Channels · DC
The highest high and lowest low of the last N bars — the original breakout channel.
Works best in trending marketsEngine-computed on a fixed sample series
What it is
Donchian Channels, created by the pioneering trend trader Richard Donchian, are the original breakout channel. The upper line is simply the highest high over the last N bars, the lower line is the lowest low over the same N bars, and a midline runs through the average of the two. Because the channel is drawn from raw price extremes, a close beyond the upper line is a literal new N-bar high and a close below the lower line a literal new N-bar low. For a beginner, that makes the tool wonderfully concrete: it maps the recent trading range exactly, so the channel edges are the precise prices that define breaking out of that range. It answers the question, has price just done something it has not done in the last N bars? That simplicity is why it became the backbone of some of the most famous trend-following systems ever traded.
How it is calculated
The calculation involves no smoothing or weighting at all. For a chosen lookback of N bars, the upper channel equals the maximum high observed over those N bars, and the lower channel equals the minimum low over the same window. The midline is the arithmetic average of the upper and lower channels. The default lookback is 20 bars, though the famous Turtle system used a 20-day channel for entries and a shorter 10-day channel for exits. As each new bar arrives, the window rolls forward, so the channel expands when a new extreme is set and can contract as old extremes drop out of the lookback. There are no other inputs — the entire indicator is just the rolling highest high and lowest low, which is exactly what makes a break of it unambiguous.
Reading it, step by step
Price touching or closing beyond the upper channel means a new N-bar high has been made, the signature of emerging or continuing upside strength, and touching the lower channel means a new N-bar low. The width of the channel is itself informative: a narrow channel reflects a tight, coiled range that is primed to break, while a wide channel reflects a market that is already moving with force. The midline can act as a rough equilibrium and a trailing reference, but in the classic reading only the outer lines matter for signals. Because the boundaries are raw extremes, a single dramatic bar can redraw them, so you read a breakout as a break of the prior established extreme rather than of a line that itself just jumped. In essence, the channel encodes the recent range and the outer edges are the exact breakout triggers.
Best timeframes and settings
The 20-bar default is the standard starting point, and it works across daily, weekly, and intraday charts wherever breakout trading is applied. Longer lookbacks such as 50 or 55 bars produce wider, slower channels that only signal on major breakouts and suit position trading, while shorter lookbacks such as 10 or even 5 bars produce tight, fast channels with far more frequent but lower-quality signals. The core trade-off is that a shorter channel catches moves earlier but generates many more false breakouts, while a longer channel filters the noise at the cost of later entries and larger stops. The Turtle approach of using a longer channel for entries and a shorter one for exits is a durable template. Choose the lookback to match your holding period: swing and position traders lean longer, active intraday breakout traders lean shorter.
When and where to use it
Donchian Channels are a trend-following breakout tool, so their natural home is trending markets and instruments capable of sustained directional runs, historically commodities and futures but equally equity indices and liquid stocks. They are designed to keep you positioned for large moves, accepting many small false starts as the cost of catching the occasional big trend. In persistently ranging markets the channel produces repeated failed breakouts as price pokes beyond the edge and snaps back, which is the environment where the tool struggles most. Avoid trading every channel touch in choppy conditions, and be prepared for a low win rate even in good conditions, since the profitability comes from the size of the winners rather than their frequency. This is a tool for traders comfortable with being wrong often but right big.
Strategies that use it
The first and most famous strategy is the Turtle breakout: buy when price makes a new 20-day high and exit the long when it makes a new 10-day low, with the mirror for shorts, letting a shorter opposite channel trail the position. The second is a squeeze breakout: watch for the channel to narrow to an unusually tight range, then trade the first decisive close beyond either edge in the direction of the break, since the compression often precedes an expansion. The third is a trend-filtered entry: only take upper-channel breakouts while a longer-term average or the ADX confirms an uptrend, which screens out many of the false breaks that plague the raw system in ranges. Across all three, disciplined position sizing and a willingness to sit through frequent small losses matter far more than signal accuracy, because the edge lives in the tail of large trend captures.
Combining it with other indicators
Donchian Channels pair naturally with tools that confirm whether a breakout has genuine backing. The ADX or a long-term moving average provides trend context, helping you take only the breakouts that align with a real directional bias and skip the range-bound fakes. The Average True Range is an essential companion for stops and sizing, since breakout systems need stops placed beyond normal volatility, and ATR quantifies exactly that. Volume adds conviction, as a channel breakout on expanding volume is more likely to hold than one on thin participation. Comparing Donchian with Keltner or Bollinger Channels can also be instructive, since the different width methods together paint a fuller picture of range and volatility. The recurring theme is to use partners to separate the breakouts worth trading from the noise the raw channel cannot filter on its own.
Where it fails
The channel's raw-extreme construction is its Achilles heel: a single spike bar redraws the boundary, and in choppy markets breakouts fail again and again as price nudges past the edge only to reverse. This produces a characteristically low win rate, and traders who cannot stomach a long string of small losses tend to abandon the system right before the big trend that would have paid for them all. Beginners also misuse it by trading every touch without a trend filter, or by setting stops too tight for a system that is meant to breathe. The fixes are to expect and accept frequent false starts, to filter entries with a trend tool so you fade fewer ranges, to size positions with ATR so no single loss is large, and above all to let winners run since the entire edge is in capturing the occasional outsized move. It is a system that punishes impatience and rewards discipline.
A worked example
Suppose a commodity future has been consolidating and its 20-day Donchian upper channel sits at 152.00, meaning that is the highest high of the last twenty sessions. Price grinds sideways for a week, the channel narrows, and then a strong session closes at 152.60 — a new 20-day high and a valid Turtle breakout. You buy, and using the Average True Range of about 1.50 you set an initial stop a couple of ATRs back and note that your trailing exit will be a new 10-day low. Over the following weeks the market trends higher to 168, and your 10-day low exit line trails up beneath it the whole way. Eventually a pullback prints a fresh 10-day low near 164, and you exit there, banking roughly twelve points on a trade whose initial risk was around three. Along the way you also took two earlier breakout attempts that failed for small losses, and this single large winner more than paid for both — the essence of how a Donchian breakout system makes money.