← All articles

June 28, 2026 · 9 min read

What is ATR (Average True Range) in Trading

Most traders set their stops at a round number below entry or a fixed 0.5% distance - without knowing whether that distance is too tight or too wide for the current market. When a stop sits inside the market's normal candle-to-candle noise, it is not risk management. It is an invitation to get knocked out of a good trade before it has a chance to move.

ATR gives you the market's own measure of normal movement. A stop inside that range is one the market will trigger through ordinary noise alone, with no structural reason to close the position. A stop beyond it gives the trade the room it needs to develop.

What ATR actually measures

ATR stands for Average True Range. It was developed by J. Welles Wilder and introduced in his 1978 book New Concepts in Technical Trading Systems. The core idea is simple: measure the full range of each candle, then average it.

The "true" part accounts for overnight gaps. For each candle, the true range is the largest of:

ATR is then the 14-period average of these true ranges. On a 15-minute chart, that is the average movement across the last 14 fifteen-minute candles.

Three things ATR is not: it is not directional, it carries no buy or sell signal, and it does not predict the next move. It only describes the recent pace of movement. A rising ATR means candles are expanding - the market is more active and volatile. A falling ATR means candles are compressing - the market is quieter and consolidating.

What ATR measures on a series of candles Six candles of varying size are shown. On the central candle, a bracket marks the high-to-low span as the true range. A shaded horizontal band spans the full chart, labeled as the ATR, representing the average of recent true ranges. ATR: the average of recent true ranges Not direction - just how much the market has been moving ATR (14) avg true range true range
The true range of each candle is its full span - high to low, adjusted for any gap. ATR is the average of recent true ranges, giving you a single number that represents how much this market normally moves per candle.

The problem with ATR across different pairs

Raw ATR in price terms is useless for cross-pair comparison.

If BTCUSDT shows an ATR of $400 on the 15-minute chart and SOLUSDT shows an ATR of $1.20, you cannot directly compare them. BTC is at $65,000 and SOL is at $160. A $400 move on BTC is 0.6% of price; a $1.20 move on SOL is 0.75%. SOLUSDT is actually the more volatile pair - but the raw numbers suggest the opposite.

If you apply a uniform rule like "stop at $200 below entry," you would be giving BTC 0.3% clearance (inside its normal noise) while giving SOL 125% clearance (absurdly wide). Neither is calibrated to how much that pair actually moves.

Cross-pair scalping requires a common unit. That unit is NATR.

What NATR is and why it matters

NATR (Normalized ATR) converts ATR to a percentage of the current price:

NATR = (ATR / price) x 100

Now BTCUSDT's 0.6% NATR and SOLUSDT's 0.75% NATR are directly comparable. You can set your stop logic in NATR terms and apply it consistently across every pair in the market.

ATR vs NATR: why normalization matters for stop placement Two side-by-side panels compare BTCUSDT and SOLUSDT. On BTCUSDT, a 0.5% stop clears the ATR noise band and is labeled safe. On SOLUSDT, the same 0.5% stop sits inside the ATR noise band and is labeled too tight. Same stop %, different result A 0.5% stop works on BTC but sits inside SOL's normal noise BTCUSDT ATR $400 = 0.6% NATR ATR noise band (0.6%) normal candle movement stop at 0.5% - has room ▲ stop clears the noise band entry SOLUSDT ATR $1.20 = 0.75% NATR ATR noise band (0.75%) normal candle movement stop at 0.5% - too tight ▼ stop sits inside the noise band entry
The same 0.5% stop sits below the noise band on BTCUSDT (0.6% NATR) and inside it on SOLUSDT (0.75% NATR). NATR makes the comparison visible - and prevents the same rule from working on one pair while failing on another.

Using ATR for stop placement

The practical rule: your stop distance must be at least 1x NATR from your entry.

If 15-minute NATR is 0.4%, a stop 0.2% below entry will be triggered by the ordinary back-and-forth of the market before the trade has had any time to develop. The position closes at a loss not because the thesis was wrong, but because normal candle noise swept the stop. That is not risk management - that is paying to find out what a candle looks like.

A stop of 0.4-0.5% in that scenario gives the trade the room it needs. Still tight, still structurally defined, but no longer inside the noise.

The structure-first rule still applies. The correct order is:

  1. Identify the structural level that makes the trade thesis wrong (the HL that was broken in a CHoCH, the resistance that should now hold in a breakout).
  2. Place the stop just beyond that level.
  3. Check whether that distance is at least 1x NATR.

If the structural stop is too close - inside 1x NATR - the setup is not tradeable on that pair at that moment. Not because the structure is wrong, but because the market's normal noise makes a clean stop geometrically impossible. Skip it and wait for the next one. See how to set a stop loss in trading for the full stop-placement framework.

One more use: ATR tells you when volatility is spiking. If NATR jumps from 0.4% to 1.2% in an hour, any stop you sized for normal conditions is now three times too tight. Either re-size the position so the wider stop represents the same dollar risk, or step aside until volatility normalizes. See Crypto Risk Management for Scalpers for how this fits into a complete risk framework.

Using ATR for target setting

ATR anchors your targets as much as your stops.

If 15-minute NATR is 0.4%, a 0.4% TP1 is achievable in a single candle - the move is within the normal true range. A 3% target requires roughly seven full ATR cycles to complete, which typically means hours of follow-through or a strong session-level momentum event.

Practical ATR-based targets for scalping:

Always check whether TP1 is blocked by a structural obstacle - a volume profile POC, a prior swing, an unfilled gap. An ATR target that walks straight into a wall will stall. The reward-to-risk math only works when the target is actually reachable.

How NextScalp uses NATR

NextScalp uses NATR as a volatility context signal inside its scoring gate - and it affects whether a setup ships a full Trade Plan or gets downgraded to informational.

A structural setup - a CHoCH, a breakout, a range-edge fade - is evaluated differently depending on the current NATR environment. When volatility is elevated, the same structural entry requires a wider stop to clear the noise, which compresses the reward-to-risk ratio to the first target. If the geometry no longer supports a clean plan at a real R:R, the signal is marked informational rather than shipping fabricated levels.

This is part of the scoring gate's job: not just asking "is the structure valid?" but also "is the risk geometry tradeable at this moment, on this pair, in this volatility regime?" See how NextScalp scores every signal for the full breakdown. For the complete picture of how ATR fits alongside leverage, position sizing, and daily loss limits, see Crypto Risk Management for Scalpers.

Want volatility-aware signal scoring and pre-calculated stops across all Binance USDⓈ-M perpetuals? Try NextScalp free for 7 days.

Try NextScalp free for 7 days