What Is Average True Range (ATR) in Futures Trading?

Average true range is an average of the true range of recent bars, and the true range of a bar is the largest of three distances: its own high to low, its high to the previous bar's close, and its low to that same previous close. Sierra Chart publishes both as separate studies, the second feeding the first.

What that gives you is a figure in the price units of the contract — index points, ticks, handles — describing how far the market has been travelling per bar. Multiply it by what a point is worth and it becomes dollars per contract, which is the form that can be compared with a loss limit. Below is how the number is built, how it differs from the thing the CFTC calls volatility, and what an average day's travel costs against a fixed drawdown.

What is average true range (ATR) in futures trading?

ATR is an average of the true range, taken over a chosen number of bars. Sierra Chart's Average True Range study documents it as exactly that — the printed formula is a simple moving average of that true range series over the input length — and exposes two inputs, the moving average length and the moving average type. Quantower describes its own ATR as a volatility-based indicator covering the range of bars averaged over a period, and names 14 as the period most commonly used for its smoothing.

The reading inherits the timeframe of the chart it is drawn on. An ATR on five-minute bars describes five-minute travel; the same study on daily bars describes daily travel. Sierra Chart is explicit that averaging the true range of daily bars requires a historical daily chart set to one day per bar, with the result pushed onto an intraday chart through its study overlay if you want to see it there. That detail catches people out: with the simple average, a 14-period ATR on five-minute bars covers roughly the last 70 minutes of movement, which is a different quantity from the one most traders mean when they say "the ATR is 60 points".

The output is a distance, averaged over the length you choose. Quantower files it among its volatility indicators and calls it volatility-based; Sierra Chart documents it in an alphabetical reference of studies, with no such label attached.

How is true range different from a bar's high-low range?

True range adds the gap between bars to the measurement. Sierra Chart's True Range study defines it as the maximum of three quantities: high minus low, the absolute distance from the high to the prior bar's close, and the absolute distance from the low to that prior close. On the very first bar of the series there is no prior close to reach back to, so the formula substitutes that bar's own close into the second and third measurements.

The arithmetic matters whenever a session opens away from where the last one finished. Take a prior close of 5,000.00 and a following bar that trades between 5,030.00 and 5,042.00. The visible height of that bar is 12 points. Measured from the prior close, the high is 42 points away and the low is 30 points away, so the true range is 42 — three and a half times the bar's own range.

That is the case true range exists to catch. High minus low alone scores that bar at 12 points and files it as a quiet one, because the 30-point jump from the previous close is invisible to it. Wherever a bar starts away from where the previous one finished, that gap is part of the distance price covered, and of the three measurements only the two that reach back to the prior close count it.

Does ATR always use Wilder's smoothing?

No — the averaging method is a setting. Sierra Chart prints its ATR formula as a simple moving average applied to the true range series, and its moving average type input offers six alternatives to that. Wilders is one of them, alongside a weighted, exponential, smoothed, linear regression or simple skip-zeros average.

Sierra Chart documents the Wilders average separately as a recursion: each value is the previous value plus one-nth of the difference between the current input and that previous value, seeded on the first bar with the input itself. Quantower's page states that its first ATR value comes from the arithmetic mean.

Those are two different weightings of the same true range series, and on anything but a series that never changes they produce different numbers. A one-nth recursion never fully discards an old observation: at a length of 14 the weight on a bar forty bars back is still around 0.004, shrinking every bar without ever reaching zero. A 14-bar simple mean gives each bar inside the window a weight of one fourteenth and each bar outside it exactly zero. So "ATR 14" identifies a length, not a value — the same bars at the same length print a different figure under a different averaging type, and a reading copied from somebody else's screenshot is comparable to yours only when both settings match.

Is ATR the same thing as volatility?

They are different statistics measured in different units. Under Volatility, the CFTC glossary names one statistic and one operand: the standard deviation of a contract's returns, annualized, as the measure of how fast its price changes, scoped to a futures contract, a security, or an instrument an option is written on. Its Historical Volatility entry takes that same annualized standard deviation over a stated number of past trading days, and its Implied Volatility entry is the figure backed out of option prices through an options pricing model.

So the glossary's volatility is built from returns and annualized, which leaves it with no price units attached. ATR is computed from price distances, covers the length you set, and comes out in the contract's own points. Quantower documents a separate standard deviation indicator for the dispersion statistic, describing it as the spread of price around its mean and the root of the variance, with the observation count and the price as its operands.

If you want ATR in comparable percentage terms, Sierra Chart publishes a normalized version: 100 times the ATR divided by the current close, which turns points into a percentage of price. That study is documented as working properly only on a historical daily chart. The two forms answer different questions — how many dollars a contract travels, against how far it travels relative to its own level.

How do researchers measure volatility on index futures?

By summing squared intraday returns. In a 2015 Federal Reserve Board working paper on news announcements and market jumps, Xin Huang measures a day's volatility with realized variance — that day's squared intraday returns added together — introducing it as "the now familiar realized variance" and noting that it consistently estimates the day's quadratic variation as the sampling interval shrinks toward zero. The jump part of the day is then separated by subtracting a realized bipower variation from that total.

The measurement choices in that paper are instructive for anyone reading a volatility number off a screen. Huang samples at five minutes, and states the reason as a trade-off: fine enough for the asymptotics to hold, coarse enough to limit the microstructure noise that contaminates very high-frequency data. The measured figure therefore depends on the interval chosen, exactly as an ATR reading depends on bar size and length.

Keep the paper's scope in mind. Its futures data is five-minute returns on two contracts: the S&P 500 index future at CME, sampled from 3 January 1994 to 30 September 2014 at 94 returns a day for 495,944 in total, and the 30-year Treasury bond future at CBOT, whose sample reaches further back, to November 1988. Nothing in the paper mentions ATR, and its stated subject is how volatility and jumps in those two markets respond to scheduled macroeconomic releases. Its finding on timing is that jumps turned up significantly more often on announcement days than on days with no scheduled release, in those two contracts over those samples, with the bond market responding more than the equity one.

Why does a range measure survive a back-adjusted continuous chart?

Because back-adjustment shifts a stretch of bars by one amount, and a shift of that kind leaves distances inside the stretch alone. Sierra Chart's continuous futures documentation describes the mechanism: on a back-adjusted continuous chart, the prices of earlier contract months are moved up or down by the price gap to the following month, while the most recent month is left unshifted.

Add the same amount to a bar's high, low and close and high-minus-low does not change, and neither does the distance from that bar's high to the previous bar's close inside the same adjusted stretch. The price level does change. Sierra Chart also warns that with back-adjustment switched on, the prices shown for a given past date keep changing as later adjustments accumulate, so a percentage-of-price figure computed over several years of a back-adjusted series is dividing by levels that were edited after the fact.

That reasoning covers bars inside a single adjusted stretch, and it is not extended here to the bar where one contract month hands over to the next. Our explainer on continuous contracts covers how the stitched series is built and which rollover method produces it.

What is an ATR reading worth per contract?

Multiply the reading by the value of one point. Our instruments page lists each contract's tick size and tick value, and the two combine with the contract unit on every row: the E-mini S&P 500 ticks in 0.25 index points at $12.50 a tick, which puts one index point at $50, and the Micro S&P 500 ticks in the same 0.25 at $1.25, which puts one point at $5. Those rows were checked against the exchange's contract specifications on 31 August 2026, and specs change, so the page carries the current figures.

Run an illustrative figure through it. Suppose a daily ATR on the index contract reads 60 points — a round number for arithmetic, not a market forecast. That is 60 × $50, or $3,000 per E-mini contract, as an average day's true range. The micro version of the same contract gives 60 × $5, or $300.

The conversion says nothing about whether either contract is worth holding; it gives the cash size of an average day's travel before anything is held. Tick size and tick value covers the conversion in more detail, including why the identity of unit times tick size is worth checking on any row you rely on.

What does the number mean against a funded account's drawdown?

It tells you how many average days of movement the account's remaining room can absorb. A simulated account here carries a loss line a fixed distance below its highest end-of-day balance, and the drawdown article holds the current amounts: a $50K account carries $2,000 on all three of our programs, as of that article's 8 September 2026 update. The line is live through the session, so touching it ends the account.

Hold the 60-point illustration and divide. One micro contract's average day of true range is $300, so $2,000 of room covers between six and seven such days of movement, in either direction, since range carries no sign. The published ceiling on that account size is four minis or forty micros, per the contract limits article, with ten micros counting as one mini. Four minis at $3,000 of average daily range each is $12,000 of movement passing through the position every average day, against $2,000 of room. Forty micros is the same $12,000.

A ceiling is a permission, and the arithmetic above is the reason it is not a plan. Futures position sizing works through the other half of the calculation, where a stop distance in ticks divides into the remaining room.

Does ATR tell you which way price is going?

Quantower's documentation says no: it states that the indicator gives no reading on trend direction or on reversal, and measures volatility and its structure only. That is the vendor's own statement about its own tool, in its own documentation.

The same Quantower page then lists two interpretations under its useful-cases heading: a divergence, where a new price high or low is not matched by a new ATR high, which it calls a good approach for finding weakness in an existing trend; and ATR rising with price, which it reads as confirmation of trend strength. Both are offered as interpretation, and the page carries no test behind either one. None of the four Sierra Chart study pages cited here offers an interpretation at all: each gives the formula, what the study's inputs are, and any chart type it requires.

Treat the reading as a measurement, which is how the Federal Reserve paper above uses its own volatility measures. A volatility measurement answers how far a contract tends to move, which is one input to position size, and the question of direction sits outside it.

FAQ

What ATR period do most futures traders use?

Quantower's documentation puts the commonest smoothing period at 14, and Sierra Chart lists the length as an input, leaving the number to whoever adds the study. Both treat the length and the averaging method as settings. In a simple average every bar in the window carries a weight of one over the length, so a shorter length puts more weight on each recent bar and a longer one spreads it thinner. Because both are settings, a quoted "ATR" figure means little without the bar size, the length and the average type that produced it.

Is ATR expressed as a percentage?

No. ATR is in the price units of the contract, so an ATR of 60 on an index future means 60 index points. Sierra Chart publishes the percentage form as a separate Normalized Average True Range study, calculated as 100 times the ATR divided by the current closing price and documented as working properly only on a historical daily chart. The points form is the one that converts into dollars per contract.

Can you use ATR on an intraday futures chart?

Yes. Sierra Chart's page says the study works on a chart of any bar timeframe, so it reads the true range of whatever bars the chart holds. The daily figure is the part that needs care: Sierra Chart documents that averaging the true range of daily bars requires a historical daily chart at one day per bar, with the result transferred onto an intraday chart through its overlay study. A 14-period ATR on five-minute bars is not a daily figure.

Does a higher ATR mean more risk on a funded account?

It means a larger dollar range per contract against a drawdown line that does not widen. The line sits a fixed distance below the highest end-of-day balance and never moves down, so if the average daily range of a contract doubles while the room stays the same, an unchanged contract count passes twice the dollar movement through the account. Contract count is the part a trader controls, and the published ceiling does not adjust for volatility.