5 ATR Stop-Loss Strategies for Risk Control

Average True Range (ATR) can help express a stop distance in units of recent price volatility. It does not predict direction, guarantee a fill or establish the best exit. The rule also needs a reference price, an update schedule and a position-sizing method.

This guide compares five ATR-based approaches: a fixed initial stop, a trailing stop, a Chandelier Exit, a percentage-normalized stop and a regime-adjusted multiplier. Use LuxAlgo’s AI trading and charting platform to inspect the setup on native charts and ask Quant to implement explicit rules for comparison.

Understand ATR Before Setting a Stop

True Range is the largest of: the current high minus low, the absolute difference between the high and previous close, and the absolute difference between the low and previous close. ATR smooths those True Range values across multiple bars. The largest of the three distances is one bar’s True Range, not the ATR itself.

A 14-period ATR uses 14 bars: daily bars on a daily chart, hourly bars on an hourly chart. Specify the smoothing method as well as the length. TradingView’s ATR documentation describes RMA as its default, with other smoothing options available. Different settings can produce different stop levels.

ATR is expressed in the instrument’s price units. It measures recent movement, including gaps, but it is not a maximum expected move or a probability boundary. See the LuxAlgo Library’s ATR reference for the indicator’s definition.

1. Basic ATR Stop-Loss

A fixed initial ATR stop takes a volatility reading when the trade is established and holds that distance constant. With entry price E, entry-time ATR A and multiplier k:

• Long stop: E − k × A.
• Short stop: E + k × A.

For a hypothetical $100 entry, ATR of $2 and multiplier of 2, the long stop is $96 and the short stop is $104. Freeze the ATR value used for this initial calculation if the intent is a static stop. Recalculating it each bar creates a different rule.

Choose an ATR length and multiplier as test parameters, not universal recommendations. A shorter lookback generally responds faster, while a longer one smooths more history. Neither is automatically better for a volatile or trending market. Check the stop against the setup’s invalidation logic, costs and trade size.

2. ATR Trailing Stop

A trailing variant can use the highest price reached since entry for a long position, or the lowest for a short. To create a stop that only tightens, calculate a candidate level and compare it with the previous stop:

• Long candidate: highest high since entry − k × current ATR.
• Long stop: the greater of the previous stop and the candidate.
• Short candidate: lowest low since entry + k × current ATR.
• Short stop: the smaller of the previous stop and the candidate.

Initialize the stop when the trade begins. For example, a $200 long entry with ATR $5 and multiplier 3 starts at $185. If the highest price becomes $215 while ATR stays $5, the candidate becomes $200. If ATR later increases enough to lower the candidate, the tightening-only rule retains the previous stop instead of widening it.

Specify when updates occur. A rule calculated after a bar closes should not be treated as if that new stop had been active earlier in the same bar. Historical high/low data may not reveal the order in which prices occurred. Inspect the backtest’s execution assumptions rather than assuming every apparent touch could have been traded.

A trailing stop can surrender some unrealized gains or exit before a trend resumes. It does not lock in a guaranteed profit, and after a tightening-only adjustment its distance from price need not equal the latest ATR multiple exactly.

3. ATR Chandelier Exit

The Chandelier Exit uses a rolling lookback extreme, which differs from the highest or lowest price since your entry. StockCharts documents a default of 22 periods and a multiplier of 3:

• Long raw line: highest high over 22 bars − 3 × ATR(22).
• Short raw line: lowest low over 22 bars + 3 × ATR(22).

These are configurable defaults, not optimal settings for every instrument. The raw long line can fall when ATR rises or an older high leaves the lookback; the short line can also move in both directions. If using it as a tightening-only order, add an explicit previous-stop comparison like the trailing rule above.

Decide whether an exit is triggered by an intrabar price crossing, a closing-price condition or another specified event. Those are different strategies. Chandelier levels provide an exit framework; they do not supply a complete entry strategy or guarantee greater precision than other ATR stops.

4. ATR Percentage Stop

To compare volatility across instruments with different prices, express ATR as a percentage of a specified reference price. The usual Average True Range Percent (ATRP) calculation is:

ATRP = 100 × ATR ÷ closing price.

If ATR is $2 and the reference close is $100, ATRP is 2%. A multiplier of 2 gives a distance of 4% of that same $100 reference price, or $4. At the same reference price, converting back to money produces the same distance as 2 × ATR. This is a normalized way to express the rule, not a separate source of trading advantage.

Do not confuse ATRP with taking a fraction of ATR. Multiplying a 50-pip ATR by 20% gives a 10-pip distance, or 0.2 ATR. It does not calculate volatility as a percentage of price, and there is no general reason that 0.2 ATR leaves enough room for ordinary movement.

Keep the ATR interval, reference price and conversion consistent. If the percentage is calculated from one close but applied to a different entry price, the money distance changes. Record which approach the strategy uses before comparing results.

5. Market Volatility ATR Stop

A regime-adjusted approach changes the multiplier under a predefined volatility condition. This is a custom strategy rule, not one standardized indicator with universally best settings. Define whether “market volatility” means the traded instrument’s own history or a separate benchmark.

For an illustrative hypothesis, compare the latest completed bar’s ATR with a specified historical ATR baseline and use one multiplier above a threshold and another below it. Set the baseline length, threshold, multipliers and update timing before evaluating results. If a benchmark is used, align timestamps and avoid using values that were unavailable at the decision time.

ATR already tends to expand after larger ranges. Increasing the multiplier at the same time can widen the distance further. Recalculate planned position size for new trades, and define whether existing stops can ever loosen. Do not widen a live stop simply because a recent loss feels uncomfortable.

Test this added complexity against a fixed-multiplier baseline. A wider stop may reduce stop-outs while increasing average loss or giving back more profit. Evaluate the complete distribution of outcomes rather than only win rate.

Size Positions and Check Execution

A wider stop does not preserve the same dollar risk unless position size changes. For a simple cash-stock example, a $200 planned loss budget and $4 entry-to-stop distance imply 50 shares before costs and execution allowances. For contracts, include tick value or contract multiplier and any currency conversion.

A stop is not a guaranteed loss ceiling. The SEC’s stop-order bulletin explains that triggered stop orders can fill away from the stop price, while stop-limit orders can remain unfilled. Gaps, thin liquidity and trading halts can make actual results differ from the planned distance.

Test ATR Exits with LuxAlgo

Open the relevant instrument and interval on LuxAlgo’s native charts. Inspect recent price movement and the ATR calculation you intend to use. An indicator plotted on a chart describes a level; it does not automatically place a broker stop.

Ask Quant, LuxAlgo’s coding agent, to implement the selected exit with explicit entry, ATR length, smoothing, multiplier, update timing and sizing rules. For a fair comparison, retain the same entry conditions while changing one exit method at a time.

1. Review Code: confirm that the implementation matches the intended rule, including initialization and whether stops can loosen.
2. Run and inspect trades: check sample entries, stop changes and exits against the chart.
3. Check assumptions: use Quant’s strategy controls to review capital, sizing, commissions and slippage.
4. Compare a baseline: use the same instrument, interval and period; examine net result, drawdown, trade count and average loss alongside win rate.
5. Validate separately: test selected settings on data not used to choose them and monitor forward behavior.

A successful compilation does not validate strategy logic or future performance. If the exit permits larger dollar losses because size was left unchanged, that is also an exposure change, not just a better stop rule.

Use the native Journal to compare recorded fills and actual exits with the plan. Check each tool’s ATR implementation and settings rather than assuming the line behaves identically across tools.

FAQs

What is the best ATR stop-loss setting?

There is no universal best multiplier. Test a defined range against the same entry and sizing rules, with costs and separate validation data. A wider stop may reduce stop-outs but increase loss size or profit giveback; a higher win rate alone does not establish improvement.

What is the best ATR period for a stop-loss?

The appropriate length depends on the strategy, interval and desired responsiveness. A 14-period ATR is a common starting point, but it means 14 chart bars rather than always 14 days. Specify smoothing and test alternatives without repeatedly choosing the best-looking historical result.

How do you calculate the ATR stop?

First calculate True Range using the high-low distance and the absolute distances from the previous close, then smooth those values to obtain ATR. For a basic long stop, subtract multiplier × ATR from entry; for a short stop, add it. At a $128.60 long entry, ATR $3.20 and multiplier 1, the planned stop is $125.40 before execution differences.