What Is a Breakout Strategy? History, Mechanics and Honest Limits

Q: Who invented breakout trading?

No single person. Jesse Livermore traded breaks of his pivotal points by judgement in the early twentieth century; Richard Donchian made the idea fully mechanical with his 4-week channel rule around the mid-century; and the Turtle experiment of Richard Dennis and William Eckhardt in the 1980s famously taught Donchian-style breakout rules to novices, showing the method could be written down and followed.

A breakout strategy buys when price pushes beyond a defined ceiling, or sells when it drops through a defined floor, on the premise that a market escaping its recent range tends to keep travelling for a while. It is one of the oldest styles in trading and one of the most testable, because both "the range" and "the break" can be measured exactly: a boundary is a number, and a break is a price crossing it. This page traces the idea from the bucket shops of the 1890s to the fully mechanical channel systems of the 1950s and the famous Turtle experiment of the 1980s, then takes the mechanics apart piece by piece: ranges, channels, compression, session breaks, false breaks, stops and targets, and what a rules engine actually checks when it trades one.

Who created it: Livermore, Donchian and the Turtles
What a breakout actually is
Donchian channels: the first mechanical breakout
Range compression and the volatility cycle
Opening range and session breakouts
False breaks: the defining failure mode
Stop placement and measured moves
How the engine mechanises it
Strengths and failure modes
Deeper reading
Glossary
FAQ
In the Systems Library

Who created it: Livermore, Donchian and the Turtles

Nobody invented the breakout in a single stroke, but three chapters of trading history did most of the work of turning "buy strength through a level" from a hunch into a rulebook.

Jesse Livermore and the pivotal point

Jesse Livermore, photographed for Financial World magazine in 1923. Public domain.

Jesse Livermore (1877 to 1940) began as a teenage quotation-board boy in Boston, then made his first money betting on price movements in the so-called bucket shops, storefront operations where customers wagered on ticker prices without any shares changing hands. He was eventually banned from most of them for winning too often, and moved on to the real exchanges of New York, where he made and lost several fortunes across four decades of speculation.

Livermore's method, as far as it can be reconstructed, was an early breakout discipline. He waited for what he called pivotal points: prices at which a stock had repeatedly stalled, so that a decisive move through them signalled the line of least resistance had been established. He insisted on waiting for the market to confirm before committing, added to positions only as the move proved itself, and treated a failure at the pivotal point as a reason to get out, not to argue. His thinking reached the public in two books: Reminiscences of a Stock Operator (1923), the lightly fictionalised biography written by journalist Edwin Lefevre and still one of the most read trading books a century later, and his own How to Trade in Stocks (1940), published shortly before his death, which set out the pivotal-point method explicitly. Livermore traded by judgement rather than by a fixed formula, so he is the breakout's grandfather rather than its engineer; the engineering came next.

Richard Donchian and the channel rule

Richard Donchian (1905 to 1993), a Yale-educated son of Armenian immigrants, is widely called the father of trend following, and he is the person who made the breakout fully mechanical. In 1949 he launched Futures, Inc., generally credited as one of the first publicly held commodity funds, and for decades he published a weekly newsletter on technical methods in the futures markets.

His signature contribution is disarmingly simple: the 4-week rule, later generalised as the Donchian channel. Buy when price exceeds the highest high of the previous four weeks (roughly 20 trading days); sell and reverse when price falls below the lowest low of the previous four weeks. There is no judgement anywhere in the rule. No chart patterns, no opinions about value, no forecast; just a lookback window, a high, a low, and a crossing. That property, that the entire strategy can be written down and tested on historical data by anyone, is what makes Donchian the pivot point of this story. Almost every mechanical breakout system since is a variation on his channel.

The Turtle experiment

The most famous field test of Donchian-style breakouts was run by Richard Dennis, a Chicago futures trader who had turned a small stake into a large fortune in the trading pits, and his partner William Eckhardt, a mathematically minded trader and close friend. In the early 1980s the two argued about whether great traders are born or trained. To settle it, Dennis advertised for novices, and between 1983 and 1988 he taught two small classes of recruits, nicknamed the Turtles, a complete mechanical system built around channel breakouts: enter on breaks of the recent N-day high or low, size positions by volatility, cut losses at a fixed multiple of that volatility, and follow the rules without deviation.

The recruits traded Dennis's capital under those rules, and the experiment became one of the best known stories in trading, largely through later accounts by participant Curtis Faith and by author Michael Covel, who tracked down and published the rules and interviewed many of the original class. The detail worth keeping is not any single number from the story, but the design lesson: the system's edge, whatever it was, lived in the rules and in the discipline of following them through long losing stretches, not in anyone's feel for the market. Breakout trading, more than most styles, is a test of whether you can keep taking small losses while waiting for the occasional large move that the rules exist to catch.

What a breakout actually is

Markets spend much of their time in ranges: stretches where price oscillates between a rough ceiling and a rough floor. The ceiling behaves as resistance, a zone where selling has repeatedly absorbed buying; the floor behaves as support, the mirror image. A breakout is the moment price passes through one of these boundaries and, crucially, stays there.

That second clause matters. Technicians distinguish between rejection, where price pokes above resistance and is promptly pushed back inside the range, and acceptance, where price trades beyond the old boundary and the market carries on doing business at the new, higher (or lower) prices. Only acceptance is a breakout in the useful sense. The logic behind trading it is straightforward supply and demand: a resistance level exists because sellers were willing to act there. When price moves through it and holds, that supply has been consumed, traders positioned short against the level are forced to cover, and traders who sold their holdings early are pulled back in, all of which adds fuel in the direction of the break.

Everything else on this page is elaboration on three engineering questions: how do you define the boundary (channels, sessions, patterns), how do you decide the break is real (confirmation, compression context), and what do you do when it is not (stops, and the false-break problem). Readers coming from trend following will recognise the family resemblance: a breakout entry is very often the front door into a trend-following position, while momentum trading asks a related but distinct question about the strength of a move already underway.

Donchian channels: the first mechanical breakout

A Donchian channel draws two lines on a chart: the highest high of the last N bars, and the lowest low of the last N bars. The area between them is, by construction, exactly where the market has traded over that window; the lines are the range made literal. The classic setting is N = 20, echoing Donchian's four weeks of daily bars, but the construction works on any timeframe and any lookback.

The entry rule is the crossing: go long when price exceeds the previous upper band, go short when it falls below the previous lower band. Note the word "previous": the band is computed on the bars before the current one, so the current bar can actually cross it; a band that includes the current bar can never be exceeded by definition.

Complete systems typically add a second, shorter channel for the exit. A common shape, and the one the Turtle rules used, is to enter on a break of the longer channel (say the 20-bar high) and exit when price crosses the opposite band of a shorter channel (say the 10-bar low). The asymmetry is deliberate: the long entry channel makes the system slow to get excited, while the shorter exit channel makes it quicker to concede when the move stalls. The gap between the two channels is where the strategy's character lives; widen the entry and you trade rarely but on bigger moves, tighten the exit and you keep more of each move but get shaken out more often.

Donchian channel with price oscillating between the 20-bar high and low, then closing above the upper band for an entry.

Range compression and the volatility cycle

Volatility is cyclical. Quiet markets, where bars are small and ranges tight, tend to be followed by loud ones, and vice versa; expansion follows contraction the way a spring releases after being coiled. Breakout traders exploit this by adding a compression precondition: do not just trade any break of any level, but prefer breaks that emerge from an unusually quiet stretch, because that is where the odds of genuine expansion are most favourable.

The classic compression flags are the narrow-range bars popularised by Toby Crabel's work on short-term price patterns. An NR4 bar has the smallest high-to-low range of the last four bars; an NR7 is the narrowest of the last seven. Both are pure observations; they say the market has gone quiet, nothing more. The trade comes from pairing the observation with a trigger: the prior bar was an NR4, and the current bar breaks above its high. Compression, then release, in one two-bar rule. Related squeeze logic works the same way with indicators instead of bars: when a volatility band (such as a Bollinger Band) contracts inside a wider average-range band (such as a Keltner channel), the market is coiled, and the eventual escape from the squeeze is traded as the breakout.

The honest framing is important here. Compression identifies the coil, not the direction of the release, and not even that a release is imminent; quiet markets can stay quiet. That is why compression conditions are almost always preconditions attached to a separate directional trigger, rather than signals in their own right.

Shrinking bar ranges ending in a gold NR7 bar, with a later green expansion bar breaking the NR7 high trigger line.

Opening range and session breakouts

Some ranges are defined by the clock rather than by a lookback window. An opening range breakout marks the high and low of the first stretch of a trading day, perhaps the first 30 or 60 minutes, and trades the escape from that box, on the reasoning that the early auction establishes the day's initial balance and a decisive departure from it often sets the day's direction.

Currency markets, which trade around the clock, have their own version. During Asian hours, with London and New York asleep, many pairs drift in a narrow overnight range. When European liquidity arrives at the London open, order flow surges, and the escape from the Asian range is one of the most traded session patterns in retail FX: mark the overnight high and low, then buy a break above the high (or sell a break below the low) in the first hours of the London session.

One structural honesty note. The "session" part of a session breakout is purely a clock filter: it restricts trading to certain hours and reads no price at all. On its own it decides nothing; it is always paired with a real price condition, the break itself. And session logic only makes sense on charts fast enough to see sessions; on four-hour bars and slower, a single bar spans most of a session, so session rules either lift entirely or fall back to plain channel breaks. How much any of this matters depends heavily on the chart speed you trade, a topic covered properly in trading timeframes explained. Readers interested in the market-structure story behind quiet accumulation ranges and their resolution will also find a much older treatment of the same shape in the Wyckoff method.

Price contained in a shaded Asian-session range, then breaking the overnight high just after the London open marker.

False breaks: the defining failure mode

The breakout's one great enemy has a dozen names: false break, fakeout, failed break, stop run, bull trap. The shape is always the same. Price pushes through the boundary, everyone watching the level acts, and then the move dies and collapses back inside the range, leaving the breakout traders holding positions at the worst prices of the day.

Why is this so common? Because obvious levels concentrate orders. Just above a well-watched high sit two clusters: the stop-loss orders of traders who are short against the level, and the entry orders of breakout traders. A modest push into that pocket triggers both, producing a burst of buying that needs no genuine new demand behind it. If real buyers do not follow through, the burst exhausts itself and the market falls back, often accelerating as the trapped breakout buyers bail out. This stop-run reversal is so reliable a pattern that some traders run the mirror-image strategy, fading breaks rather than following them.

Mechanical systems fight false breaks with confirmation, and every form of it trades the same currency: a later entry price in exchange for a truer signal.

Close beyond, not touch beyond. Requiring the bar to close outside the level, rather than merely trade through it intraday, filters the spike-and-collapse pattern at the cost of a worse fill.
A margin, scaled to volatility. Requiring price to exceed the level by some multiple of average true range means a drifting nudge does not qualify; only a push that is unusual for the current volatility does.
Participation. Requiring above-average volume (or, in futures-style logic, strong directional order flow) on the breakout bar demands evidence that real business, not just resting stops, is driving the move.
The retest entry. The most patient version: let the break happen without you, then enter when price returns to the broken level and holds it, old resistance behaving as new support. You miss the breaks that never look back, in exchange for a better price and a cleaner invalidation point on the ones that do.

None of these removes false breaks. They shift the mix: fewer, later, truer entries versus more, earlier, noisier ones. Where a system sits on that spectrum is a design choice, not a solved problem.

Side-by-side panels contrasting a false break that pokes above resistance and collapses with a true break that closes above, retests and resumes.

Stop placement and measured moves

A breakout trade comes with a built-in invalidation point, which is one of the style's quiet virtues. If the premise is "the market has accepted prices beyond the old boundary", then price returning decisively back inside the range refutes the premise, and the trade should end. The natural stop therefore sits back inside the broken range: below the breakout level for a long, often below the most recent minor low, or at a fixed multiple of average true range beneath the entry so that the stop breathes with current volatility rather than sitting at a fixed distance in all conditions. Stops placed too tight, just under the level itself, live exactly in the pocket where post-breakout retests trade, and get collected by the very noise the trader should expect.

Targets are less standardised, and the two main schools disagree on principle. The measured move school projects the height of the broken range: if a market oscillated in a 100-point box and breaks upward, the first objective is roughly 100 points above the breakout, on the logic that the energy stored in the range approximates the energy of the release. The trend-following school, Donchian's and the Turtles' school, sets no target at all: the whole point of entering on strength is that occasionally the market keeps going far beyond anything a range projection would suggest, and those outliers are what pay for the many small failed attempts. A measured target caps the loser-paying outliers; an open-ended trailing exit endures more giveback at the end of each move. Neither answer is free; a system simply has to pick one and be tested as a whole.

How the engine mechanises it

Inside our engine, every idea on this page exists as an exact, testable condition, with its assumptions written down. A few examples of breakout conditions used by systems in the library, paraphrased in plain English:

Channel breakout. The close crosses above the previous 20-bar Donchian upper band, price at a fresh 20-bar high. This is the classic channel rule with no interpretation anywhere in it, just a number being exceeded.
Narrow-range breakout. The prior bar was the narrowest of the last four (an NR4), and the current close breaks above its high. Compression, then release, in one two-bar rule; the related NR4 and NR7 flags are also available on their own as pure quiet-market observations that trigger nothing by themselves.
London breakout. Between 07:00 and 10:00 broker time, the close breaks above the recent Asian-hour highs. On four-hour charts and slower there is no session to speak of, so the clock gate is lifted and the rule falls back to a plain 20-bar channel break on an up bar. The same lifting applies to session filters generally: on H4 and above a clock window would do nothing useful, so it stands aside rather than pretending to filter.
Volatility breakout (Keltner). The close exceeds a 20-period average plus two average-true-range units, so only a push that is unusually strong relative to current volatility qualifies, not any drift past a fixed line.
Confirmed structural break. The bar breaks above the prior confirmed swing high on an up close, with its range larger than one ATR and volume above 1.2 times its 20-bar average, a break of structure that demands both size and participation before it counts.

Two honesty notes carried directly from the engine's own documentation. First, "volume" on spot forex charts is tick volume, a count of price updates, not true traded size; it is a serviceable proxy for participation, not a measurement of it. Related order-flow style readings are likewise built from bar ranges and volume rather than a true bid and ask split, and are labelled as proxies for that reason. Second, condition names are labels, not verdicts: a flag called "rising volatility" describes expanding ranges, and one labelled as a high-volatility read may in fact describe contraction, so every condition is defined by what it computes, not by what its name suggests. And nothing here predicts. A breakout condition observes that a boundary has been crossed under stated circumstances; whether the market continues is what testing on unseen data exists to estimate, never something a rule can promise.

Strengths and failure modes

Strengths. Breakout logic is objective end to end, which makes it unusually easy to state, automate and test honestly; there is no discretionary step where hindsight can hide. It is self-confirming in design: it only ever enters when the market is already doing the thing it wants to profit from, so it can never miss a large move entirely, and by construction it participates in every major trend, since every major trend begins by breaking out of something. Its invalidation is built in, giving each trade a natural, structurally meaningful stop. And it needs no forecast; it reacts to what has happened rather than opining about what will.

Failure modes. The mirror image of each strength. Most individual breakouts fail; a majority of signals in most tested variants lose money, and the style is only viable if the occasional extended move outweighs the many small losses, which makes the equity curve lumpy and psychologically hard to sit through. In choppy, range-bound conditions the strategy is systematically wrong-footed: it buys every upper boundary and sells every lower one, precisely the opposite of what a range rewards, and the resulting whipsaw, a string of small losses plus spread and slippage on every attempt, is the style's characteristic cost. Slippage bites harder here than in most styles, because breakouts by definition enter during fast, one-sided moments when fills are worst. Above all, the approach is regime-dependent: it earns in expansion regimes, when volatility is cycling from quiet to loud and moves extend, and it bleeds in contraction and chop, and no filter reliably announces in advance which regime comes next. Compression preconditions, confirmation rules and regime filters tilt the mix; none of them changes the underlying dependence, and a backtest of any of it is a record of the past, not a promise about the future.

Deeper reading

Jesse Livermore, Wikipedia: the life behind the pivotal-point method, bucket shops to Wall Street.
Richard Donchian, Wikipedia: the father of trend following and originator of the channel breakout.
Richard Dennis, Wikipedia: the trader behind the Turtle experiment with William Eckhardt.
Reminiscences of a Stock Operator, Wikipedia: background on Edwin Lefevre's 1923 classic.
Reminiscences of a Stock Operator at the Internet Archive: a library scan of the book in the archive's catalogue.
Breakout (technical analysis), Wikipedia: the concise formal definition of breaks of support and resistance.
Price channels (Donchian channels), StockCharts ChartSchool: construction and use of N-period high-low channels, with charts.

Glossary

Breakout: A move in which price passes through a defined boundary, such as a range high, and holds beyond it rather than returning inside.
Range: A stretch of trading bounded by a rough ceiling and floor, within which price oscillates without establishing a trend.
Support and resistance: Price zones where buying (support) or selling (resistance) has repeatedly halted movement, forming the boundaries a breakout must cross.
Acceptance: The market continuing to trade beyond a broken level, treating the new prices as fair; the hallmark of a genuine breakout, as opposed to rejection.
Donchian channel: A pair of lines at the highest high and lowest low of the last N bars; the mechanical definition of the recent range, introduced by Richard Donchian.
4-week rule: Donchian's original system: buy a break of the prior four-week high, sell and reverse on a break of the prior four-week low.
NR4 / NR7: A bar whose high-to-low range is the narrowest of the last four (NR4) or seven (NR7) bars; a flag of range compression, not a directional signal.
Squeeze: A volatility-contraction condition, typically a tighter volatility band trading inside a wider average-range band, marking a coiled market awaiting release.
Opening range breakout: A trade taken when price escapes the high-low box built in the first portion of a session, such as the first hour of the day.
London breakout: The FX version of the session breakout: trading the escape from the quiet overnight Asian range as European liquidity arrives at the London open.
False break (fakeout): A push through a boundary that fails to hold and collapses back inside the range, trapping breakout entries; often driven by stop orders clustered at the level.
Stop run: A quick thrust into a pocket of resting stop orders that triggers them and then reverses, the mechanism behind many false breaks.
Confirmation: Any extra requirement, a close beyond the level, a volatility-scaled margin, above-average volume, that a break must meet before it is traded; later entries in exchange for truer ones.
Retest: Price returning to a broken level after the break; if the old ceiling now holds as a floor, the retest offers a second, cleaner entry.
Measured move: A target projected by adding the height of the broken range to the breakout point, on the logic that the stored range roughly sizes the release.
Whipsaw: A rapid sequence of losing signals in choppy conditions, where each break fails and reverses; the characteristic cost of breakout trading in ranges.

FAQ

What is a breakout strategy in simple terms?

It is a rule that buys when price pushes above a defined ceiling, or sells when it drops below a defined floor, on the premise that a market escaping its recent range often continues in the direction of the escape. The ceiling and floor can be a Donchian channel, a session range, or a chart level, but the trade is always the crossing itself.

Who invented breakout trading?

No single person. Jesse Livermore traded breaks of his "pivotal points" by judgement in the early twentieth century; Richard Donchian made the idea fully mechanical with his 4-week channel rule around the mid-century; and the Turtle experiment of Richard Dennis and William Eckhardt in the 1980s famously taught Donchian-style breakout rules to novices, showing the method could be written down and followed.

What is the difference between a breakout and a fakeout?

A breakout is price moving through a level and holding beyond it; a fakeout is a break that fails and returns inside the range, often after triggering the stop orders clustered at the level. Systems reduce fakeouts by requiring closes beyond the level, volatility-scaled margins, volume confirmation or retest entries, but no filter removes them entirely.

Do breakout strategies work in ranging markets?

Ranging markets are the hardest environment for breakouts, because most breaks fail there and the strategy buys every ceiling and sells every floor, the opposite of what a range rewards. That is why many breakout systems add compression or regime filters, and why traders often pair the style with mean-reversion approaches that prefer ranges.

What is a London breakout?

A session breakout specific to currency markets: mark the quiet range built overnight during Asian hours, then trade the escape from it in the first hours after the London open, when European liquidity arrives. The session window itself is only a clock filter, so the rule is always paired with an actual price break of the overnight range.

In the Systems Library

Systems in the Orion RFX Systems Library are tagged by style, including breakout and volatility-breakout, and each ships with a plain-English playbook of its exact rules, validated on data it never trained on. See how automated trading systems work for how a rulebook like the ones on this page becomes a running system, then compare breakouts with their close cousins, trend following and momentum trading, and read how timeframes change what a session breakout even means.