Track 2 · Lesson 4 · 6 min
Breakouts and channels
Buying new highs: Donchian's four-week rule, the Turtles, Darvas boxes and the 52-week high.
What a breakout is
A breakout is a close beyond the edge of a recent range: above the highest price of the last 20 days, say, or the last year. The idea is simple. A stock that escapes its range has found buyers willing to pay more than anyone has recently, and that is often how a new trend starts.
Breakout plays buy those new highs and then need a way out, because many breakouts fail and fall back. The exit is usually the opposite edge of a shorter channel, or a trailing stop.
Channels
A channel boxes the price in between two lines. The oldest is the Donchian channel: the highest high and lowest low of the last N days. Richard Donchian, who started one of the first public managed-futures funds in 1949, is credited with the "four-week rule": buy when price beats the highs of the last four weeks, sell when it breaks the lows.
Other channels measure distance from an average instead. A Keltner channel sits a couple of average true ranges (ATRs) above and below a 20-day EMA; Bollinger Bands sit two standard deviations away from a 20-day average. A close above the top line is the breakout.
Nvidia with the four-week rule, last two years
The violet line is the highest high of the prior 20 days, the amber line the lowest low. A close above violet buys; a close below amber sells.
The Turtles
In 1983, commodity traders Richard Dennis and William Eckhardt settled a bet about whether trading could be taught. They recruited about two dozen novices through newspaper ads, taught them a breakout system in two weeks and gave them money to trade. Their rules, published by former Turtles in 2003, had two versions:
- System 1: buy a 20-day breakout and sell a 10-day low, but skip the next breakout if the last one would have been a winner (then wait for a 55-day breakout instead).
- System 2: buy every 55-day breakout and sell a 20-day low.
Both used a stop two ATRs (the Turtles called it 2N) below the entry, sized every position by volatility and also sold short. Be The Puck tests one long position at a time.
New highs: Darvas and the 52-week high
Nicolas Darvas, a professional dancer, described in 1960 how he bought stocks breaking out of "boxes": a top that held for a few days, then a floor that held under it. When price broke the top, he bought; if it fell through the floor, he sold.
Academics found something similar. Thomas George and Chuan-Yang Hwang showed in 2004 that stocks trading close to their 52-week high tended to keep outperforming. Their explanation: traders anchor on the old high and hesitate to bid past it, so good news takes longer to show up in the price.
Failed breakouts
Most breakouts don't run. Across Be The Puck's tests, breakout plays were right on about 32% of their calls, against 30% for trend plays. They make it back because a failed breakout is usually cut quickly while a real one can run for months. Some traders even trade the failures: Linda Bradford Raschke's "Turtle Soup" setup, from her 1995 book with Larry Connors, bets against breakouts to new 20-day extremes that quickly reverse.
Every breakout play on the S&P 500
On SPY since 2005, after trading costs.
| Play | Checks | Switches / yr | Time in | Annual return | Worst drawdown | Calls right |
|---|---|---|---|---|---|---|
| Four-Week RuleRichard Donchian | Daily | 6.2 | 64% | +6.9% | −26.9% | 41% |
| Turtle System 1Richard Dennis & William Eckhardt | Daily | 8.2 | 47% | +3.8% | −17.5% | 30% |
| Turtle System 2Richard Dennis & William Eckhardt | Daily | 5.6 | 49% | +3.6% | −18.2% | 32% |
| Clenow TrendAndreas F. Clenow | Daily | 5.6 | 49% | +4.3% | −15.5% | 32% |
| Darvas BoxNicolas Darvas | Daily | 5.8 | 59% | +4.7% | −30.5% | 34% |
| Keltner ChannelChester W. Keltner & Linda Bradford Raschke | Daily | 8.0 | 33% | +2.5% | −13.5% | 29% |
| Bollinger BreakoutJohn Bollinger | Daily | 9.5 | 33% | +0.7% | −23.9% | 26% |
| Chandelier ExitChuck LeBeau | Daily | 8.2 | 61% | +5.6% | −26.7% | 36% |
| Wilder Volatility SystemJ. Welles Wilder Jr. | Daily | 8.4 | 67% | +5.9% | −17.8% | 38% |
| P&F Double TopA.W. Cohen & Thomas J. Dorsey | Daily | 1.0 | 89% | +10.6% | −37.3% | 55% |
| Chande Kroll StopTushar S. Chande & Stanley Kroll | Daily | 16.9 | 66% | +4.8% | −35.8% | 26% |
| Acceleration BandsPrice Headley | Daily | 12.0 | 12% | +2.0% | −10.1% | 28% |
| TTM SqueezeJohn F. Carter | Daily | 5.7 | 11% | +1.6% | −17.1% | 30% |
| NR7 BreakoutToby Crabel | Daily | 36.2 | 52% | −0.3% | −54.3% | 31% |
| Bollinger SqueezeJohn Bollinger | Daily | 2.6 | 7% | +0.9% | −17.8% | 16% |
| Buy & holdbenchmark | Never | 0 | 100% | +11.0% | −55.2% | – |
Check yourself