What ATR multiplier should I use for a volatility stop?

There is no correct universal k. Wilder suggested a constant near 3 in the original system, and published variants mostly run between 2 and 3.5 on daily data. Smaller multiples exit faster but whipsaw more; larger ones give back more open profit. Match the multiple to your holding period and test it on the instrument, treating defaults as starting points.

Is Supertrend the same as a volatility stop?

Same core mathematics: an ATR multiple offset from a reference price with a ratchet. Supertrend fixes the reference at the bar midpoint, plots continuously, and flips bias when price crosses, so it behaves as a stop-and-reverse regime line. It is best read as one packaged implementation of the volatility-stop idea rather than a separate concept.

Should a volatility stop trigger on a close or an intrabar touch?

It is a real design choice. Intrabar triggering matches how a live stop order fills but gets tagged by wicks and news spikes; close-based triggering ignores wicks at the cost of exiting later, sometimes well beyond the line. Backtest whichever version you will actually execute, because the two can produce genuinely different results on the same settings.

Why was my volatility stop hit just before the move resumed?

Because k × ATR bounds typical noise, not extremes: one wide-range bar or a fast sweep can pierce any statistical offset, and volatility often expands exactly where trends pause. If it happens repeatedly, the multiplier or lookback may be too tight for the timeframe, or entries may be arriving late in the leg. No distance eliminates it.

Does a volatility stop widen when ATR expands?

The freshly computed level does move farther from price when ATR expands, but ratcheted implementations refuse to loosen: they keep the tightest stop already reached. Only the initial stop, set at entry, fully reflects entry-time ATR. If an implementation lets the trailing level slip backward in fast markets, that is a design flaw worth checking.

Concept

Volatility Stop

Volatility Stop, also known as k×ATR, is a Risk, Sizing & Exits concept. The Library holds 1 implementation, a working definition you can pull into Quant.

Top Volatility Stop indicator

The top custom implementation, built on the original standard Volatility Stop formula.

1 total

Want to trade Volatility Stop? The implementation below is one prompt away from a backtested strategy in Quant.

Statistical Trailing Stop

Indicator

What is a Volatility Stop?

A volatility stop is a stop-loss placed a multiple of current volatility away from price, in practice almost always k times the average true range. Instead of a fixed dollar, tick, or percent distance, the offset breathes with the market: it sits farther away when ranges expand and tightens when they contract. In the trailed form the level recalculates as new bars print and ratchets in the trade's direction, never loosening, until price crosses it and the position exits.

The idea traces to Welles Wilder's 1978 volatility system, which trailed a stop a constant near 3 times a short average true range away from the trade's most favorable close. Later variants, one family within the broader trailing method taxonomy, differ mainly in the reference they offset from: the latest close, the bar midpoint (Supertrend), or the highest point since entry (the chandelier stop). Published multipliers commonly fall between 2 and 3.5, but no constant is canonical.

It matters because sizing the stop to volatility separates 'the trade is wrong' from 'the market is noisy.' A stop inside the instrument's normal bar-to-bar range will be hit regardless of direction; a k × ATR offset places the exit outside typical fluctuation for the chosen k. The same quantity also standardizes risk across instruments, which is why it often sets both the initial stop distance and, from there, the position size.

How to calculate a Volatility Stop

All variants share one construction; only the reference point, lookback, and multiplier differ.

1Compute the average true range over a chosen lookback. Published variants use windows from about 7 bars (Wilder's original) to 22 (a common chandelier setting); shorter windows react faster and jitter more.
2Multiply the ATR by a factor k to get the offset. A larger k means fewer whipsaw exits and more giveback; a smaller k the reverse.
3Subtract the offset from the reference price for longs, or add it for shorts. The reference is the main design choice: the latest close, the bar midpoint, or the highest high (or close) reached since entry.
4Ratchet and exit. Each bar, keep the tighter of the new level and the prior stop, and exit when price crosses it; decide in advance whether a close beyond the line or any intrabar touch triggers.

How it's calculated

A trailing stop placed k ATRs away from price that only ratchets in the trade's favor and flips direction when price crosses it.

\operatorname{TR}_t = \max(H_t - L_t, \lvert H_t - C_{t-1} \rvert, \lvert L_t - C_{t-1} \rvert)

\operatorname{ATR}_t = \frac{\operatorname{ATR}_{t-1} \times (n - 1) + \operatorname{TR}_t}{n}

\text{Uptrend: } \operatorname{Stop}_t = \max(\operatorname{Stop}_{t-1}, C_t - k \times \operatorname{ATR}_t)

\text{Downtrend: } \operatorname{Stop}_t = \min(\operatorname{Stop}_{t-1}, C_t + k \times \operatorname{ATR}_t)

\text{Flip to downtrend when } C_t \text{ crosses below } \operatorname{Stop}_t\text{; flip to uptrend when } C_t \text{ crosses above } \operatorname{Stop}_t

\text{On a flip, restart the stop at } C_t - k \times \operatorname{ATR}_t \text{ (new uptrend) or } C_t + k \times \operatorname{ATR}_t \text{ (new downtrend)}

TR_t: true range of bar t

H_t: high of bar t

L_t: low of bar t

C_t: close of bar t

C_(t-1): close of the prior bar

ATR_t: Wilder-smoothed average true range at bar t

ATR_(t-1): prior bar's ATR (the first value is seeded with the simple average of the first n true ranges)

n: ATR period (commonly 14; Wilder used 7)

k: ATR multiplier (commonly 2 to 3)

Stop_t: volatility stop level at bar t

Stop_(t-1): prior bar's stop level

t: bar index

Wilder's original volatility system measured k × ATR from the extreme close reached during the trade (the SIC) rather than ratcheting bar by bar; the result is nearly identical.

Chandelier Exit anchors the same offset to the highest high or lowest low of the last 22 bars with k = 3, and SuperTrend measures it from the bar midpoint (H_t + L_t) / 2.

How traders use it

As the exit engine of trend-following systems: after a breakout or crossover entry, the position simply rides until the k × ATR line is crossed, replacing any fixed target.
As the initial stop and the sizing input: placing the first stop k × ATR away keeps risk proportional to current conditions, and dividing per-trade risk by that distance yields the position size used in volatility-targeted sizing.
As a regime filter: plotted continuously with a flip rule, the ratcheting line becomes a trend classifier, and which side price sits on gates other signals (the ATR trailing regime reading).
As a way to de-cluster stops: a statistical offset is not anchored to visible swing points, so it often lands away from the obvious levels where resting stops pool (stop placement vs liquidity pools).

Volatility Stop vs related concepts

Chandelier Stop: A specific volatility stop: k × ATR trailed from the highest point reached since entry. The generic concept leaves the reference open; the chandelier pins it to the trade's extreme.

Supertrend: The same k × ATR offset measured from the bar midpoint, ratcheted, and drawn as a continuous line that flips bias on a cross: a volatility stop packaged as a regime indicator.

ATR Bands: Envelopes plotted k × ATR on both sides of price or an average for context. They track both directions at once and do not ratchet, so they frame volatility rather than enforce an exit.

Structure Stop: Anchored to a chart feature, typically the swing that would invalidate the idea, rather than a statistical distance. Volatility stops track noise width; structure stops track where the thesis fails.

Concept family

Risk, Sizing & Exits

37 concepts mapped · 37 in the Library

Volatility Stop FAQ

Turn Volatility Stop into a trading strategy.

Take the implementation from this page into Quant, then build on it, backtest it on real data, and keep refining it in conversation.

Previous conceptTWAP/VWAP/POV Execution Next conceptVolatility-targeted Sizing