ATR-Based Stops & Bracket Orders Explained

The same 16-tick stop that works on a quiet Tuesday is noise on CPI morning. Volatility-scaled brackets exist to solve exactly that problem — here's how they work.

Brackets first: what they are

A bracket order attaches two exits to your entry the moment it fills: a stop-loss below (for a long) and a profit target above, linked as an OCO pair — one-cancels-other. Fill the target, the stop cancels; get stopped, the target cancels. On platforms like NinjaTrader 8 this is the mechanism behind ATM strategies.

The point isn't sophistication — it's that the seconds after an entry are when you're most exposed and least rational. Brackets make the worst case defined before the trade exists.

What ATR measures

Average True Range is the moving average (commonly 14 periods) of the true range — each bar's high-to-low span, expanded to include any gap from the prior close. In plain terms: how far does this market typically travel per bar right now? ES might run a 3-point ATR on a quiet afternoon and 8+ on a data morning. ATR says nothing about direction; it measures the current size of the noise.

The problem with fixed-tick stops

A fixed stop encodes an assumption that volatility is constant. It isn't. The consequences show up on both ends:

Traders compensate by manually widening stops on wild days, which reintroduces discretion (and hesitation) at the worst moment.

ATR brackets: scaling both sides

An ATR bracket sets the stop and target as multiples of current ATR — for example stop at 1.5×ATR, target at 2×ATR. Because ATR updates continuously, the bracket auto-sizes: wider when the market is fast, tighter when it's slow, with no manual adjustment.

The subtle benefit is on the target side. If only the stop scales, your reward-to-risk ratio drifts with volatility, and backtested expectancy stops meaning anything. Scaling both sides keeps the ratio constant across regimes — the difference between a strategy that behaves in December the way it did in March, and one that doesn't.

One caveat: ATR multiples deserve per-instrument tuning. 1.5×ATR on ES and on crude oil are different animals in dollar terms, which leads to the third bracket family.

Dollar-based brackets

A $-PnL bracket places the stop and target where your position's P&L would hit a chosen dollar amount — risk $150, target $250 — computed from your live position size and recomputed when you scale in or out. This is the natural bracket for prop-firm traders, whose real constraints are denominated in dollars: daily loss limits and trailing drawdown, not ticks. Defining risk in the account's own units makes "how many of these losses until I breach?" trivially answerable.

Choosing between the three

Bracket typeSized inFits when
Fixed ticksTicks/pointsA structure-based trade where the invalidation level IS a price, not a distance
ATR multipleVolatility unitsSystematic entries traded across changing regimes; keeping R:R constant
Dollar P&LAccount currencyProp/drawdown-constrained accounts; multi-instrument traders normalizing risk
In our tools: CustomDOM offers all three as one-click presets on the DOM — fixed ticks, ATR multiples (1.0×/1.5×/2.0×), and $-PnL brackets recomputed as you scale — plus ATR-shaded ladder coloring so you can see extension in volatility units at a glance. 14-day free trial. As always in futures trading, test any bracket scheme in sim before trading it live.

Frequently asked questions

A stop-loss and profit target attached to your entry automatically as an OCO (one-cancels-other) pair: when one side fills, the other cancels. Brackets remove unprotected seconds after entry and give every trade a defined worst case.
Fixed-tick stops assume constant volatility. ATR stops scale with current conditions — mechanically wider on volatile days, tighter on quiet ones — keeping the probability of a noise stop-out roughly constant across regimes.
If the stop scales and the target doesn’t, your reward-to-risk ratio drifts with volatility. Scaling both sides (e.g., 1.5×ATR stop, 2×ATR target) holds the ratio constant, which keeps a strategy’s expectancy math stable across regimes.