Donchian channels, explained

What a Donchian channel is

A Donchian channel draws two lines over a price chart:

• The upper band: the highest high of the last N periods.
• The lower band: the lowest low of the last N periods.

That is the whole indicator. No smoothing, no weighting, no formula beyond a running maximum and minimum. A 20-day Donchian channel simply answers: what is the highest price of the last 20 trading days, and the lowest?

The signal is just as plain. When price trades above the upper band, it has done something it could not do at any point in the last N days, and a breakout trader takes that as evidence a trend may be starting. A break below the lower band is the mirror-image signal to sell or go short.

Where it came from

Richard Donchian (1905–1993) is often called the father of trend following. In 1949 he launched Futures, Inc., generally considered the first publicly available managed futures fund, and for decades he published a newsletter built on systematic rules. His most famous was the weekly rule: buy when price exceeds the high of the past four calendar weeks, sell when it breaks the four-week low. Simple as it sounds, the four-week rule became one of the most studied systems in futures research and kept working, in various forms, for decades.

Dennis and Eckhardt's system was a direct descendant. What the Turtle rules call a 20-day breakout is a Donchian channel with N set to 20.

How the Turtles used it

The Turtles ran two channel lengths side by side:

• 20 days (System 1): a faster signal that catches trends early but gets whipsawed more, filtered by skipping the signal after a winning breakout.
• 55 days (System 2): a slower signal that only fires on major moves, taken without exception.

They also used shorter channels to exit: System 1 positions were closed on a 10-day break against the position, System 2 on a 20-day break. The channel handled both doors, in and out.

Why breakouts work (when they work)

The logic is unglamorous: every large trend must pass through a breakout. A market cannot rise 80% without first making a new 20-day high, so a breakout trader is guaranteed a seat in every big move. The cost of that guarantee is false signals, and there are many. Most breakouts stall and reverse, which is why the channel never worked as a standalone system and the Turtles never traded it as one. It needed volatility-based position sizing to keep the failed breakouts small, and wide exits to let the real trends pay for them.

Using Donchian channels today

The indicator ships with virtually every charting platform, usually under "Donchian channel" or "price channel." Three practical notes from the modern literature:

• Parameters are not magic. 20 and 55 were choices, not laws. Research since suggests breakout systems are robust across a wide band of lengths, and slower has generally aged better than faster.
• The channel is an entry technique, not a system. Sizing, stops, and exits decide whether it makes money.
• Expect a low win rate. Breakout trend following often wins on 30 to 40 percent of trades. The math works through asymmetry, not accuracy.

For the deepest treatment of channel breakouts inside a full portfolio, Andreas Clenow's Following the Trend rebuilds and stress-tests the approach with modern data.