CustomTradingPlatform .com
Log in Get started free
Home / Guides & research / Volatility

Guide · Volatility

ATR vs Fixed Stop Losses

The short answer

A fixed stop changes your risk every time volatility changes. An ATR stop keeps the risk fixed and moves the size instead.

The choice

A fixed stop is a constant distance — 20 pips, $1.50, 0.5% — applied to every trade. An ATR stop is a multiple of recent average true range, so it widens when the market is volatile and tightens when it is quiet.

fixed : stop = 20 pips, always
ATR : stop = 2 × ATR(14), whatever that is today

The difference matters because volatility is not constant. A 20-pip stop on EURUSD is generous in a quiet August session and negligible around a rate decision.

What each one gets right

  1. ATR normalises across instruments and regimes. The same rule can be applied to gold and to EURUSD and mean something comparable in both — the stop is always the same distance in units of what the market is currently doing.
  2. Fixed stops are predictable. You know the risk before the trade and the arithmetic never surprises you. For a discretionary trader taking a handful of positions, that is worth something real.
  3. ATR stops adapt without you noticing. In a volatility expansion, a fixed stop is hit by noise while an ATR stop is not. That is the single strongest argument for ATR.
  4. Fixed stops cannot be gamed by their own input. ATR is computed from the same price series being traded, so a volatility spike widens the stop at exactly the moment the position is most at risk.

Position size makes them equivalent in risk

The comparison is often framed as "which stop risks less", and that framing is wrong. With correct position sizing, both risk the same amount — the stop distance is an input to the size, not to the risk.

fixed : 20 pips → smaller stop → larger position → same $ risk
ATR : 55 pips → wider stop → smaller position → same $ risk

So the real question is not risk but hit rate: which stop is touched by noise more often. That is an empirical question about the instrument and timeframe, not a matter of preference.

Where this goes wrong

Whichever you choose, choose it before the trade and apply it to every trade. A stop selected after entry, sized to the loss you are willing to look at rather than to the market, is not a stop — it is a hope with a price attached.

A reasonable default

For systematic testing across multiple instruments, ATR is the better default, because it is the only one that makes results comparable between markets with different typical ranges. Every formation rule on this site uses ATR for exactly that reason — thresholds like `minBodyAtr` and `minGapAtr` exist so that a rule written once means the same thing on gold and on EURUSD.

For discretionary trading on a single instrument you know well, a fixed stop is defensible and easier to apply consistently, which usually matters more than the marginal improvement.

The third option nobody mentions

Percentage stops sit between the two: a stop placed a fixed percentage from entry, so it scales with price but not with volatility.

How each behaves when conditions change
Stop typePrice doublesVolatility doubles
Fixed distanceEffectively halvesUnchanged, so hit far more often
PercentageScales correctlyUnchanged, so hit far more often
ATR multipleScales correctlyWidens with it

Percentage stops solve the instrument-price problem and not the volatility problem, which is why they are common in equities — where prices vary by orders of magnitude — and rare in FX, where they do not.

What to check before choosing

The question is empirical and specific to your instrument and horizon: which stop is touched by noise more often, for the same dollar risk.

  1. Measure the noise. Compare the typical adverse excursion on winning trades against each candidate stop distance. A stop inside that range is being hit by trades that would have worked.
  2. Check across regimes. Split the sample into high- and low-volatility periods. A fixed stop that performs well overall often does so by working in one regime and failing in the other.
  3. Price the difference. A wider stop needs a smaller position for the same risk, so it is not free — it trades a lower stop-out rate for a smaller position in the trades that do work.

What both stops share

Whichever you choose, two properties matter more than the choice itself, and they apply equally to both.

  1. The stop must be placed before entry. A stop chosen after a position moves against you is a decision made under exactly the conditions that impair it.
  2. The stop determines the size, not the risk. Wider stop, smaller position, same dollar loss. Anyone whose risk changes with their stop distance has the causation backwards.

What this depends on

MethodThe rule is stated precisely enough to run, and the same rule produces every figure quoted.
MeasuredWhere a number comes from our own testing it carries its sample size and its data vintage.
LimitsCosts, slippage and non-USD accounts change the arithmetic. See methodology.