All articles
Strategy · Technical Analysis

What the ATR Filter Actually Does — And Why Most ORB Traders Skip It

8 min read ATR · Opening Range · Risk

Most traders who use an Opening Range Breakout strategy stop at the basics: wait for price to break outside the range, enter in the breakout direction, set a stop, and target a reward. That is the framework. And it works — but not on every day.

The part most people skip is asking a simple question before the entry: is this opening range a meaningful one?

That question is what the ATR filter answers. It is one of the quietest quality gates in rules-based trading, and skipping it is one of the most common reasons ORB traders experience sessions that feel completely different from the setups they practised on.

What ATR actually measures

ATR stands for Average True Range. Despite sounding technical, the concept is straightforward: it measures the average size of price movement over a set number of recent candles — including any gaps between sessions.

Think of it as a barometer for how active a market is right now. A higher ATR means price has been moving larger distances per candle. A lower ATR means price has been moving smaller distances. It is not predicting direction — it is measuring volatility.

Plain English

If the average candle on Gold over the last 14 periods has moved $12, the ATR is 12. If recent candles have been quieter and averaging $6 moves, the ATR is 6. The system uses a 14-period RMA (a smoothed average) read from the M1 chart at exactly 13:45 UTC, when the Opening Range closes.

What the Opening Range is — and why its size matters

The Opening Range (OR) is the high and low formed during the first 15 minutes of the New York session, from 13:30 to 13:45 UTC. Those two levels become the boundaries for the trade.

A breakout above the OR high is a long signal. A break below the OR low is a short signal. The stop sits at the opposite boundary. That means the OR width is also your risk — 1R.

This is why the size of the OR matters. If the OR is unusually small, the trade has almost no room to breathe before hitting the stop. If it's unusually large, you're risking a much bigger amount than normal — and often because something abnormal happened in those 15 minutes, like a news spike or a flash move.

Neither extreme produces reliable trades. The ATR filter screens them both out.

How the filter works

At 13:45 UTC, once the OR is complete, you check one thing: is the OR width between 0.5× and 8.0× the current ATR?

The check — two boundaries
Lower bound (too tight): OR width ≥ 0.5 × ATR
Upper bound (too wide): OR width ≤ 8.0 × ATR
Result: Both must pass. One failure = no trade.
< 0.5×
Too tight
Range too small — no meaningful boundary to trade
0.5–8.0×
Valid range
Normal volatility — breakout has structural meaning
> 8.0×
Too wide
Abnormal session — likely news spike or extreme event

Why "too tight" is a problem

When the OR is very small — less than half the typical M1 candle movement — a breakout above or below it is essentially meaningless. Price drifts outside that boundary constantly as part of normal market noise.

You're not breaking out of a range. You're just getting a wiggle. The trade has no structural foundation, and your stop is so close to your entry that a single tick of noise can stop you out before any real move develops.

The lower bound filters these sessions cleanly. If the market didn't build a proper 15-minute range, there is nothing meaningful to break out of.

Why "too wide" is equally dangerous

The upper bound catches the opposite problem: an OR that is far larger than normal.

When Gold forms an unusually large 15-minute range, it is usually because something specific happened — a high-impact economic release, an unexpected news event, or a sharp liquidity move. These sessions look dramatic. They can look like incredible opportunities.

They aren't. The volatility in those sessions is reactive and unpredictable. The stop distance (the OR width) is enormous, meaning your risk per trade has ballooned well beyond what the system is designed around. The edge disappears.

A note on the upper bound

The upper limit is 8.0× — not a tighter number. This is deliberate. At current Gold prices in the $3,000–$5,000 range, a normal 15-minute OR routinely spans 3–6 times a single M1 candle ATR. Tighter upper limits (like 3.0×) were calibrated when Gold was trading near $2,000. Using them today would incorrectly reject the majority of perfectly valid sessions.

The filter is designed to catch genuine extremes — not to restrict normal, healthy volatility.

The two types of sessions this eliminates

Compressed sessions
What they look like: Price barely moves in the first 15 minutes. The OR is a tiny band. The candles inside it are small and directionless.

The breakout, if it comes, is often a false one. There was no range worth breaking.
Explosion sessions
What they look like: A major data release or news event fires in the 13:30–13:45 window. Price shoots 40+ points in one direction, forming a massive OR.

Your stop is 40+ points. Your risk has multiplied. The trade is uncontrolled.
Result
No structural edge. Frequent stop-outs from normal noise before any real move has space to develop.
Result
No position sizing consistency. The system's R-based risk model breaks down when the OR width is 5× its normal size.

How to read it in practice

At 13:45 UTC — not before, not after — you note the ATR reading and the OR width, then run the check.

  1. 1 Note the ATR from your M1 chartRead the ATR (14, RMA) indicator at 13:45 UTC exactly. This is your baseline — a number representing average recent candle movement.
  2. 2 Calculate the OR widthOR High minus OR Low. Simple subtraction.
  3. 3 Check the lower boundIs OR width ≥ 0.5 × ATR? If no — no trade. If yes — continue.
  4. 4 Check the upper boundIs OR width ≤ 8.0 × ATR? If no — no trade. If yes — the ATR filter passes. Proceed to the rest of the checklist.

This takes about 15 seconds. It happens at the exact moment the OR closes. It is not a gut feel check — the numbers either pass or they don't.

What this filter is not

The ATR filter does not predict direction. It does not tell you whether the breakout will succeed. It does not replace the bias filters, the A+ scorecard, or the London profile classification. It is a single gate at a specific point in the session workflow — checking that the structural conditions for a valid trade exist before anything else happens.

If the OR doesn't pass this gate, none of the other analysis matters. You log the session as a no-trade, note the ATR reading, and move on. That is a legitimate outcome. Knowing when not to trade is exactly as valuable as knowing when to trade.

The broader principle

Every professional trading system has filters that screen out sessions where the edge doesn't apply. This is one of them. The ATR filter exists because the Opening Range Breakout works in certain conditions — and not in others.

Compressed ORs produce noise-stopped entries. Extreme ORs produce mis-sized risk. Neither is the market environment the system is built for. The filter keeps you trading only the sessions where the conditions actually match what has been tested.

That is what separates a rules-based system from guessing. Not every trade taken — but every trade taken only when the rules say it is valid.

Free Resource
The 4 London Profiles — Free Guide

Learn how to classify London's behaviour before the NY session opens — the foundational filter that determines whether a session is worth trading at all.

Download Free Guide →