Volatility & bandsAverage True Range · ATR
The average size of a bar's true range — the benchmark measure of volatility magnitude.
Works in most conditionsEngine-computed on a fixed sample series
What it is
The Average True Range, created by J. Welles Wilder, is the benchmark measure of volatility magnitude — how much a market moves, expressed in the instrument's own price units. It answers a single question cleanly: how big is a typical bar's move right now, so I can size stops, targets and positions to the market's actual turbulence? Crucially, ATR is directionless — it tells you the size of moves, never their sign, so a high ATR accompanies both violent rallies and violent sell-offs. Its great innovation over a plain high-minus-low range is the concept of true range, which accounts for overnight gaps that a simple range would miss. ATR is one of the most widely used building blocks in all of technical analysis, feeding stops, position sizing and countless other indicators.
How it's calculated
The foundation is the true range for each bar, defined as the greatest of three quantities: the current high minus the current low, the absolute value of the current high minus the previous close, and the absolute value of the current low minus the previous close. Including the previous close is what lets true range capture gaps — if a market opens far above yesterday's close, the high-to-previous-close distance captures that jump where a plain range would not. The Average True Range is then Wilder's smoothed moving average of the true range, conventionally over 14 periods. Wilder's smoothing weights recent bars while retaining a long memory, so ATR responds to new volatility without jumping around on a single bar. The output is in price units, so an ATR of $2 means the instrument typically moves about $2 per bar.
Reading it, step by step
A high ATR means large, fast bars and an agitated market; a low ATR means quiet, small bars and a calm one. The number must be read in the instrument's own price units and in context — an ATR of $2 is huge on a $20 stock and trivial on a $2,000 one — so it is a relative, per-instrument gauge, not a cross-market one. The trend of ATR carries information: a rising ATR flags expanding volatility, which often accompanies breakouts, news and climactic moves, while a falling ATR signals a market settling into quiet, sometimes ahead of a coiling breakout. Because ATR is directionless, a spike in it tells you a big move is happening but not whether it is up or down, so it must never be read as bullish or bearish. You read ATR to gauge how much room to give a trade, not which way to take it.
Best timeframes and settings
The default 14-period ATR works on every timeframe, since it simply measures the true range of whatever bars it is fed, from 1-minute scalping charts to weekly position charts. A shorter ATR period reacts faster to volatility changes and is useful for scalping where you want the stop to adjust quickly, while a longer period gives a smoother, slower-moving volatility read better suited to position trades. The responsiveness-versus-noise trade-off is straightforward: fewer periods make ATR twitchier and more current, more periods make it steadier and laggier. Because ATR feeds stops and sizing, the choice of period indirectly sets how quickly your risk parameters adapt. Most traders keep 14 and let the timeframe do the rest, matching the ATR to their holding period.
When and where to use it
ATR's dominant use is risk control, and it is relevant in every market regime and asset class because every instrument has a volatility that needs measuring. Use it to place stops a multiple of ATR beyond price so they sit outside normal noise, to size positions so that a one-ATR move equals a fixed dollar risk, and to set targets scaled to how much the instrument actually moves. It automatically gives volatile names wider stops and smaller size and quiet names tighter stops and larger size, which normalizes risk across a portfolio. It is not a directional or timing tool, so do not expect entries from it. Avoid comparing raw ATR across different instruments; for that, a normalized version expressed as a percentage of price is required.
Strategies that use it
The volatility-stop strategy places the initial stop at a multiple of ATR — say 2 or 3 — beyond the entry, so the stop is beyond the market's ordinary noise and only triggers on a genuine adverse move; this is the foundation of ATR trailing stops and the Chandelier Exit. The position-sizing strategy sets share or contract size so that a one-ATR adverse move equals a fixed fraction of the account, automatically shrinking size in volatile names and expanding it in calm ones to keep risk constant. The volatility-breakout strategy watches for ATR to compress to a multi-period low and then trades the expansion when a bar's true range surges, anticipating that contracted volatility precedes a directional break. In each, ATR supplies the volatility yardstick that turns a raw entry into a properly sized, properly stopped trade.
Combining it with other indicators
ATR is a companion to directional tools rather than a standalone, so it pairs with anything that supplies a trade direction and needs a stop or size. Trend and breakout systems use ATR to place volatility-scaled stops and targets around their entries. Keltner Channels are literally built on ATR, using it to set band width around a moving average, so ATR underlies that whole indicator. ADX complements it neatly: ADX says whether a trend is present and ATR says how much that market moves, together informing both whether to trade and how much room to give. Bollinger Bands, built on standard deviation, offer an alternative volatility read that can be cross-checked against ATR. Wherever a strategy needs to convert a signal into sized, stopped risk, ATR is the usual tool.
Where it fails
ATR's most important limitation is that it is directionless, and the classic mistake is reading a spike as bullish or bearish when it signals only that a big move — up or down — is occurring. Because it is expressed in absolute price units, it cannot be compared across instruments or across large changes in an instrument's own price over time; normalized ATR exists precisely for that. A single climactic bar can lift ATR sharply and leave stops unusually wide for a while, which may be appropriate or may over-widen risk depending on context. ATR also lags shifts in volatility because it is a smoothed average, so it confirms a change in turbulence rather than predicting it. The defenses are to treat ATR strictly as a magnitude gauge, to use a normalized version for cross-comparison, and to remember that its readings describe volatility, not direction or timing.
A worked example
A trader wants to buy a stock at $100 and risk a fixed $500 on the trade, and the 14-period ATR reads $2.50. Choosing a 2-ATR stop, the stop goes at $100 − 2 × $2.50 = $95, which sits beyond the stock's normal daily noise so ordinary wiggles will not trigger it. With $5 of risk per share and a $500 risk budget, the trader buys 100 shares, and the ATR-based sizing means a genuinely volatile move of two ATRs against them costs exactly the planned $500. A month later the same stock's ATR has risen to $4 as volatility expanded around an earnings run; a fresh trade with the same $500 budget and 2-ATR stop now uses an $8 stop and only about 62 shares, automatically shrinking size to hold risk constant. The example shows ATR doing its real job — not calling direction, but scaling the stop and the size to the market's current volatility so that risk stays uniform.