Best ATR Period for Trading: What Really Works
If you are spending hours tweaking your ATR period trying to find the magic number, I need to stop you right now. You are optimizing the wrong thing.
ATR is the very essence of breakout trading. In the Mr. Breakouts Formula – point of initiation plus space plus time plus filters – ATR is how we calculate the “space” component. It defines the breakout bands, the distance price needs to travel before we consider it a real breakout. So naturally, traders assume that nailing down the perfect ATR period is critical.
I ran the numbers on 18 years of data. And the results will change how you think about this parameter entirely.
What Is the ATR Indicator and Why It Matters
The Average True Range (ATR) is one of the most widely used volatility indicators in trading. Developed by J. Welles Wilder, ATR measures how much a market moves on average over a given number of bars.
In breakout trading, ATR plays a central role. It is the tool we use to define how wide our breakout bands are – how much “space” price needs to cover before triggering an entry. It also feeds into stop losses, position sizing, and volatility filters.
Because ATR is embedded in so many parts of a breakout system, traders naturally assume that finding the “perfect” ATR period is the key to unlocking better performance. But when you actually test it, the data tells a very different story.
The ATR Period Trap: Why Traders Over-Optimize
Here is what typically happens. A trader builds a breakout strategy that uses ATR for the breakout channel width. Then they start optimizing step by step – testing ATR period 5, then 10, then 14, then 20. They find one that looks great in backtesting, lock it in, and feel confident.
This is what I call “overfitting square two.” You have moved past the beginner mistakes, but now you are curve-fitting at a more subtle level. The “best” ATR period you found is the best for that specific historical data, not for the future.
This step-by-step optimization approach is one of the most common mistakes I see traders make, and it is something I cover extensively when teaching people how to build breakout strategies from scratch.
My ATR Period Test: Visual Proof on 18 Years of Data
To settle this once and for all, I ran a visual test on E-mini S&P 500 hourly data spanning 18 years. I took a working breakout strategy and changed only the ATR period while keeping everything else identical.
The strategy setup was straightforward: a short entry between 9 and 10 AM, end-of-day close at 4 PM as the point of initiation, a multiplier of 3, and a one-bar exit. This is based on a strategy I built in a previous video.
When I changed the ATR period from 10 to 20 to 40, here is what happened to the breakout bands on the chart: almost nothing. The bands barely moved. It was subtle – very subtle – almost no difference at all.
And the equity curves? Nearly identical across all three ATR period settings. The performance, the drawdowns, the shape of the curve – all essentially the same.
The Real Contrast: What Actually Changes Everything
Now here is where it gets interesting. While the ATR period barely registers, other parameters in the formula produce completely different results when you change them.
Using the same strategy, when I changed the moving average period or the time parameters, everything shifted dramatically – completely different entries, completely different trades. And when I changed the multiplier from 3 to 5, or down to 0.25? A completely different strategy. Not subtle adjustments – a fundamentally different set of results.
| Parameter Changed | Impact on Strategy |
|---|---|
| ATR Period (10 → 20 → 40) | Barely noticeable – bands and equity curves nearly identical |
| Moving Average Period | Dramatic change – completely different entries |
| Multiplier (3 → 5 or 0.25) | Completely different strategy and results |
| Time Parameters | Dramatic change – different trade timing entirely |
This is a huge insight. If your strategy only works with one specific ATR period, that is a red flag – not a feature. It means your strategy is likely curve-fitted. But if changing the multiplier or time window destroys performance, that is expected because those parameters actually define what the strategy does.
What You Should Focus on Instead
If ATR period is not the lever you should be pulling, what is? Based on my testing, here are the things that actually move the needle in breakout trading:
1. The multiplier. This defines how wide your breakout bands are relative to volatility. Changing a multiplier from 3 to 5 gives you a completely different strategy. This is where real edge lives – not in whether ATR looks back 14 bars or 20.
2. Time parameters and point of initiation. When your breakout window opens and closes matters enormously. The 9-10 AM entry window with a 4 PM close produces fundamentally different results than other time combinations.
3. Filters. Adding the right filters to your strategy – like trend filters or volatility regime filters – can filter up to 80% of bad trades. This has a far bigger impact than any ATR tweak. And a note here: when ATR is used inside filters rather than for the point of initiation, the period can be a little bit more sensitive. We need to be more careful with filters.
4. Robustness testing. Instead of finding the “perfect” ATR setting, focus on making sure your strategy works across a range of settings. This is the foundation of knowing whether a strategy is real.
A Smarter Way to Handle ATR Period: The Fixed Set Approach
So if you should not optimize ATR period step by step, how should you handle it? Here is the approach we use at my hedge fund: fixed sets instead of optimization.
Rather than testing every ATR period from 5 to 50 looking for the best one, I use a set of three values and treat them as switches:
- 15 – close to the standard default of 14, captures short-term volatility
- 25 – a medium-term lookback for broader volatility context
- 40 – a longer-term measure for smoother volatility readings
I test my strategy with each of these three values. If the strategy works well across all three, I have something robust. If it only works with one of them, I know I am likely overfitting.
This is the opposite of step-by-step optimization. You are not searching for the magic number. You are testing whether the strategy is resilient across a reasonable range of volatility lookbacks. And because you are only testing three values – not fifty – you massively reduce the risk of curve-fitting.
How This Applies to Building Real Strategies
This ATR insight is part of a bigger principle I follow when building breakout strategies: never over-optimize individual parameters. Especially parameters that do not actually change results much.
A good breakout strategy should be robust across a range of reasonable settings. If changing the ATR period by ten bars dramatically alters your results, the strategy is fragile – and fragile strategies break in live trading. But if changing the multiplier or time window by a meaningful amount changes results, that is normal and expected.
This is the difference between strategies that work in backtesting and strategies that work with real money. Spend your optimization budget on the parameters that matter – multiplier, time, filters – and use fixed sets for the ones that do not, like ATR period. It is exactly why I built the testing framework inside BreakoutOS – to make it easy to validate strategies properly before risking capital.
Key Takeaways
- ATR calculates the “space” component of the Mr. Breakouts Formula – it is central to breakout trading
- But the ATR period itself has minimal impact – on 18 years of E-mini S&P data, changing from 10 to 20 to 40 barely moved the breakout bands or equity curves
- The multiplier, time parameters, and moving average period are what actually change your strategy’s behavior dramatically
- Stop optimizing step by step – use a fixed set of three values (15, 25, 40) and treat them as switches
- ATR period is slightly more sensitive when used in filters than in the point of initiation – be more careful there
- If your strategy only works with one specific ATR period, it is likely curve-fitted
Stop tweaking the parameter that does not matter. Start testing the ones that do.
Want to find the optimal ATR settings for your breakout strategies? BreakoutOS includes the Component Optimizer and Breakout Filter Inputs Calibrator – tools that show you the precise input ranges that matter. No more guessing. Explore them at breakoutos.com