Donchian Channels

Keltner and Bollinger wrap their bands around an average and scale them with clever measures of volatility. The Donchian Channels throw all of that away and ask the most literal question a chart can ask: over the last N bars, what was the highest price, and what was the lowest? Draw a line across the top at the highest high and one across the bottom at the lowest low, and you have the entire indicator. No averages, no standard deviation, no ATR — just the recent extremes.

That plainness is not a weakness; it is the point. The question "are we at a new high?" is one of the oldest and most powerful in trading, and the Donchian channel answers it at a glance. It is also the foundation of one of the most famous experiments in market history — the Turtle Traders — which makes it the perfect place to look honestly at what a breakout (price pushing beyond a prior high or low into new territory) really is, and how often it lets you down.

The idea behind Donchian channels is almost primitive, and that is its strength. A new high means, by definition, that everyone who bought in the recent past is now sitting on a profit and nobody is trapped underwater overhead — there is no recent supply of frustrated sellers waiting to get out at cost. That clean overhead is why a fresh high can sometimes run: the path of least resistance is up. A new low is the mirror image. So a tool that simply marks the highest high and lowest low of a chosen window puts the single most important structural fact — "are we breaking new ground?" — right on the chart.

Richard Donchian, often called the father of trend-following, formalised this in the mid-20th century. Decades later, in the 1980s, the trader Richard Dennis used a Donchian-style breakout rule to settle a bet: he claimed trading could be taught to novices. He recruited a group he nicknamed the Turtles, handed them a mechanical system built largely on entering new N-day highs and exiting on N-day lows, and several of them went on to make substantial returns. The story is legendary, and it is usually mis-told. The Turtles did not win because breakouts are magic. They won because they followed a disciplined rule with strict risk limits through the many losing breakouts to catch the few enormous trends — which is the honest heart of cut the losses, let the profits run.

Two lines and, sometimes, a third:

• The upper channel is the highest high of the last N bars.
• The lower channel is the lowest low of the last N bars.
• The midline (optional) is simply the average of the two.

Because each line tracks a maximum or a minimum, they do not move smoothly — they sit flat, then step. The upper line stays perfectly level until price prints a new high, at which point it jumps up to the new high and flattens again. This gives Donchian channels their distinctive staircase look. The classic signal could not be simpler: price closing above the upper channel is a bullish breakout; closing below the lower channel is a bearish breakdown.

The lookback N is the one real choice. A short window (say 20 bars) gives frequent, early breakouts — more signals, more false ones. A long window (say 55 bars) gives fewer, later, more selective breakouts. The Turtles famously used a shorter entry window and a longer one as a filter. There is no magic N.

Watch a breakout — and remember it is a bet, not a certainty. illustrative

A stock has gone quiet, drifting sideways for weeks inside a tight band. The Donchian channel is dead flat: the upper line pinned to the range top, the lower line to the range bottom. Then one day price closes decisively above the upper channel — a new high the stock has not seen in the whole window. A breakout trader takes the entry, not because the high guarantees more highs, but because a fresh high with clean overhead has a real, if modest, tendency to lead to more. Critically, she sets her exit before she is in: if price falls back below the channel, the breakout has failed and she is out for a small loss.

The candles below show exactly this shape — a flat, boring range on the left, then a clean push to new highs on the right. That transition, from range-bound to a fresh high, is the moment a Donchian breakout is trying to catch.

Here is the part beginners refuse to believe: most breakouts like this one fail. Price pokes above the channel, sucks in buyers, then falls back into the range — a false breakout. The disciplined trader knows this and does not care, because her losses on the failures are small and her rare winners, the ones that turn into months-long trends, are large. The method does not need to be right often. It needs to be right big when it is right, and cheap when it is wrong.

The Donchian channel is honest to a fault: it reports the recent high and low, and nothing more.

It cannot tell a real breakout from a false one. The channel marks that price reached a new extreme; it has no idea whether that extreme will hold. Most will not. A breakout is a probability, not a promise, and the false-breakout rate is high enough that trading without stops is ruinous.

It whipsaws badly in a range. In a choppy, trendless market, new highs and lows are precisely where price reverses. Donchian breakouts fail again and again in such conditions. The tool cannot know a trend is absent; you must.

It lags, like every tool here. The channel is built entirely from past highs and lows, so it confirms new ground after price has already reached it. You enter after the high, never before.

It says nothing about the company. A new high on the chart is a fact about price, not about value. Plenty of new highs belong to companies you would never want to own; the channel cannot see the difference.

The Turtle legend sets a trap of its own, and beginners walk into a few predictable ones.

1. Expecting breakouts to win often. They do not. Breakout systems win a minority of trades and live on a handful of big winners. Judging one on a short losing streak — as almost everyone does — throws away the method just before it might have worked.

2. Trading breakouts in a range. The one market Donchian is worst in. New extremes in a choppy market are reversal points, and the channel will whipsaw you to death.

3. Skipping the stop. The entire logic is "wrong cheaply, right big." Remove the stop and you keep the many small losses but expose yourself to one large one — the exact opposite of the design.

4. Copying the Turtles' numbers, not their discipline. People memorise "20-day high" and forget the strict position sizing and risk limits that actually made the system survive. The lookback was never the edge; the risk control was.