What Is the Average True Range (ATR) Indicator? Formula and How to Use It

Search Fintrev

Search our guides, reviews and glossary.

The average true range (ATR) is a technical indicator that measures volatility: how far an asset’s price typically moves during one period, including any gap from the previous close. J. Welles Wilder Jr. introduced it in his 1978 book New Concepts in Technical Trading Systems. ATR is expressed in the asset’s own price units (dollars for a stock, pips or price points for a currency pair), and it says nothing about direction. A rising ATR means bars are getting bigger. It does not mean price is going up.

What Is the Average True Range (ATR)?

Wilder designed ATR for commodity futures. Those markets often opened well above or below the prior close and sometimes hit daily price limits, so a bar could show a small high-to-low range even after a large move. A limit-up day, for example, might trade in a narrow band at the limit price. Measuring volatility with the simple high minus low would understate what holders of the contract had just experienced.

True range fixes this by reaching back to the previous close. If the market gapped, the gap counts as part of that period’s movement. ATR then averages those true ranges so one bar does not dominate the reading.

The indicator applies to any market with open, high, low and close data: stocks, indices, forex, futures and crypto. It also works on any chart timeframe. A 14-period ATR on a daily chart averages 14 daily bars, and on a 15-minute chart it averages 14 fifteen-minute bars. The number always means “typical movement per bar on this timeframe.”

The same book introduced the Relative Strength Index (RSI), the Average Directional Index (ADX) and the Parabolic SAR, which is why ATR shows up inside several of Wilder’s other calculations.

The True Range and ATR Formulas

True range

For each period:

TR = max[(H - L), |H - Cp|, |L - Cp|]

where:

• H = current period’s high
• L = current period’s low
• Cp = previous period’s close
• |x| = absolute value, so a downside gap counts the same as an upside gap

When there is no gap, the high-to-low range is usually the largest of the three and true range equals the ordinary range. The other two terms take over when price opens away from the prior close and does not trade back to it.

First ATR value

With no prior ATR to build on, the first value is a simple average of the first n true ranges:

ATR(first) = (TR1 + TR2 + … + TRn) / n

The summation notation sometimes shown for this, (1/n) ΣTRi, means the same thing: add each period’s true range from the first to the nth, then divide by n.

Every later ATR value

After the first value, Wilder used a recursive smoothing formula:

ATR(current) = [ATR(previous) × (n - 1) + TR(current)] / n

Each new bar gets a weight of 1/n and the existing average keeps the rest. This is mathematically equivalent to an exponential moving average with a smoothing factor of 1/n (often called Wilder’s smoothing or RMA). It reacts more slowly than a standard exponential moving average of the same length, which uses 2/(n + 1). Some charting platforms let you switch ATR’s smoothing to a simple, exponential or weighted average, so two charts set to “ATR 14” can show slightly different values.

How to Calculate ATR: A Worked Example

The example below uses a five-period ATR so every step stays visible. The prices are hypothetical, for a currency pair quoted to four decimals, with 1 pip = 0.0001. The close before Day 1 is 1.1000.

All distances are in pips.

Step 1: find each true range. On four of the five days, the high-to-low range is the largest term. Day 3 is different: the pair opened well above the prior close of 1.1010 and never traded back down to it. The bar itself spans only 30 pips, but the distance from the prior close to the high is 90 pips, so the true range is 90.

Step 2: compute the first ATR. Add the five true ranges and divide by five:

(70 + 60 + 90 + 70 + 60) / 5 = 350 / 5 = 70 pips

For comparison, a plain average of the high-low ranges is (70 + 60 + 30 + 70 + 60) / 5 = 58 pips. The 12-pip difference is entirely the Day 3 gap, which is the movement Wilder wanted to capture.

Step 3: update with the smoothing formula. Suppose Day 6 trades with a high of 1.1095 and a low of 1.1000, after the Day 5 close of 1.1070. The three terms are 95, 25 and 70 pips, so TR = 95.

ATR = (70 × 4 + 95) / 5 = (280 + 95) / 5 = 75 pips

A single wide bar lifted ATR by 5 pips, not by the full 25-pip difference between that bar and the old average. That damping is the point of the smoothing. It also means a volatility shock fades out of ATR gradually rather than dropping out all at once after n bars, as it would with a simple moving average.

With the standard 14-period setting, the steps are identical: average the first 14 true ranges, then apply ATR = (prior ATR × 13 + current TR) / 14 to each new bar. Because the recursive formula carries forward the starting value, ATR calculated from a short price history can differ slightly from one calculated from years of data. The difference shrinks as more bars are included.

How to Read ATR Values

An ATR reading is the average true range per bar over the lookback period. If a stock’s 14-day ATR is $1.20, its daily true range has averaged about $1.20 over the last two to three weeks of trading, with recent days weighted slightly more.

Three readings are more useful than the raw number:

• ATR against its own history. An asset’s ATR rising from $1.20 to $2.00 tells you its daily swings have widened substantially. A fall back toward a long-run level suggests the market has calmed. Absolute levels only mean something relative to that asset’s typical behavior.
• ATR relative to price (ATR%). ATR% = ATR / closing price × 100. A $50 stock with a $1.25 ATR has an ATR% of 2.5%, the same relative volatility as a $200 stock with a $5.00 ATR. Stock screeners often filter on ATR% (for example, “ATR% below 2.5%” to find quieter names, or above some level for active day-trading candidates). The threshold is a user choice, not a standard.
• Today’s move against ATR. Intraday traders compare the distance price has traveled today with the daily ATR. If a pair with a 70-pip daily ATR has already moved 90 pips from its low, the session has exceeded its recent average range. That is information about how stretched the day is, not a reliable reversal signal, since trending days can run to two or three times the average.

High and low ATR are not good or bad in themselves. A high ATR means larger potential gains and losses per bar, wider stops and smaller position sizes for the same risk. A low ATR means tighter conditions, which some traders treat as a setup for a later expansion in volatility.

How Traders Use ATR

ATR does not tell you when to buy or sell on its own. Its main role is to convert “how volatile is this market?” into a number you can plug into risk decisions, regardless of how the entry was chosen.

Setting stop-loss distances

A fixed stop such as “20 pips” or “$1” ignores the market’s current behavior. In a quiet market it may be generous. In a volatile one, ordinary noise can hit it. An ATR-based stop scales with conditions:

Long stop = entry price - (multiple × ATR) Short stop = entry price + (multiple × ATR)

Hypothetical example: you buy at 1.1070 with a 70-pip ATR and a 2× multiple. The stop goes 140 pips below entry, at 1.0930. The multiple is a judgment call. Smaller multiples get stopped out more often by normal fluctuation, and larger ones accept a bigger loss when the trade fails.

Trailing stops and the chandelier exit

The chandelier exit, developed by Chuck LeBeau, hangs a trailing stop a fixed number of ATRs below the highest high reached since entry (or above the lowest low for a short). As price makes new highs, the stop ratchets up. If volatility expands, the stop distance widens with it. A frequently cited configuration is a 22-period ATR with a 3× multiple, though traders adjust both.

Position sizing

ATR also connects stop distance to position size, so each trade risks a similar amount of money regardless of how volatile the instrument is:

Position size = account risk in currency / (stop distance in price units × value per price unit)

Hypothetical example, with the assumptions stated: a $10,000 account risking 1% ($100) per trade; a 2× ATR stop on EUR/USD with ATR at 70 pips, so a 140-pip stop; and a pip value of $10 per standard lot. One standard lot would lose $1,400 at the stop, so the position is $100 / $1,400 ≈ 0.07 lots. If ATR doubled to 140 pips, the same rule would cut the size roughly in half. A position size calculator handles the pip-value conversion for pairs where the account currency is not the quote currency.

This volatility-based sizing is the approach the Turtle Traders of the 1980s used. Their “N” was essentially a 20-day ATR, and position size was set so a 1N move equaled a fixed fraction of account equity.

Volatility breakouts

One breakout method adds a multiple of ATR to the most recent close and treats a move beyond that level as a signal. The logic: a move larger than a normal day’s range suggests something has changed. These signals occur less often than with tighter thresholds, and ATR does not indicate which side the breakout will happen on, so the rule is usually applied in both directions or combined with a trend filter.

Setting realistic targets

Daily ATR offers a sanity check on profit targets. A target 200 pips away on a pair averaging 70 pips a day implies several days of favorable movement. That is not impossible, but it is a different trade from an intraday target inside one ATR.

Inside other indicators

ATR is a building block elsewhere. The modern version of Keltner Channels draws bands at a multiple of ATR around an exponential moving average. The Supertrend indicator flips between support and resistance lines placed at ATR multiples. Wilder’s directional movement system (the basis of ADX) normalizes directional movement by smoothed true range.

Choosing the ATR Period and Multiple

Fourteen periods is Wilder’s default and the setting most platforms load. Shorter lookbacks (5 to 10) react faster to new volatility and produce more signals in systems that use ATR thresholds. Longer lookbacks (20 or more) give a smoother reading that changes slowly and generates fewer signals.

For day trading, the usual approach is to calculate ATR on the timeframe being traded (a 14-period ATR on 5-minute or 15-minute bars) and use daily ATR as a separate reference for how much room the whole session typically offers. For swing trading, daily ATR is the common basis for stops and sizing, since positions are held across several daily bars.

Multiples for stops typically fall somewhere between 1× and 3× ATR, but there is no correct value. The right multiple depends on the strategy’s holding period and how much noise it can tolerate. Testing a rule on historical data or in a paper trading simulator shows how often a given multiple would have been hit.

Limitations of ATR

• No direction. ATR rises in sharp selloffs and in sharp rallies. A jump in ATR after a large move against the prevailing trend can be misread as confirming the old trend when it may mark a turning point.
• Lagging. It is built from past bars and smoothed, so it reports volatility that has already happened. After a sudden shock, ATR takes several bars to catch up. After the shock passes, it takes several more to come back down.
• Interpretation is subjective. No ATR level signals a reversal or tells you a market is “too volatile.” Readings only make sense against the same asset’s history.
• Not comparable across assets or timeframes in raw form. A $3 ATR on a $30 stock and on a $300 stock describe very different markets. Use ATR% for comparisons, and do not compare a daily ATR with an hourly one.
• Sensitive to how bars are built. In forex, the daily close depends on the time zone a broker uses to cut daily candles, so daily highs, lows and closes, and therefore ATR, can differ between platforms. Markets that trade around the clock, such as crypto, have few gaps, so true range is usually close to the plain high-low range.
• Not a probability. An ATR of 70 pips does not mean price will move 70 pips today, or that it is unlikely to move 150. It is an average, and individual bars regularly exceed it.

ATR vs. Similar Measures

The most frequent confusion is between ATR and trend indicators. A market can have a high ATR while going nowhere (wide bars that chop back and forth) or a low ATR while trending steadily. Pairing ATR with a trend or momentum tool, such as a moving average or MACD, separates “how much is it moving” from “which way.”

FAQs

The Bottom Line

ATR matters most to traders who need stop distances and position sizes that match current market conditions. It turns volatility into a price-unit number that fits into risk rules, scaling stops wider and positions smaller when bars grow. Because it lags, ignores direction and varies with how bars are built, it works best judged against the same asset's history and paired with a trend tool.