ATR vs ATR Percent: Compare Volatility Across Prices

An ATR of 8 looks more dramatic than an ATR of 1. It may describe the quieter instrument. Price units are doing part of the talking: an eight-point range on a 2,000-priced instrument is smaller proportionally than a one-point range on a 20-priced instrument.

ATR percent, also called ATRP, divides ATR by the current closing price and multiplies by 100. It puts the range on a relative scale. That helps when a TradingView automation rule applies one volatility filter to several symbols. It does not make those symbols interchangeable at the execution stage.

For an AlgoWay workflow that routes TradingView alerts to execution, define the volatility test in the strategy first, then keep its percentage reading separate from the symbol, quantity and protective price levels in the order instructions.

Put both instruments on the same scale

Fidelity's ATRP definition uses this calculation:

ATRP = 100 × ATR / close

The following numbers are invented arithmetic examples, not market observations. Assume both ATR values use the same timeframe, lookback length and smoothing method.

A has eight times B's absolute ATR. B has 12.5 times A's ATR relative to its closing price. A hypothetical filter requiring ATRP above 2% would admit B and exclude A. A raw ATR filter above 2 would do the reverse. Neither threshold is a trading recommendation; the example shows why the unit belongs in the rule.

Keep the underlying ATR calculation consistent

True range is the largest of the bar's high minus low, the absolute high minus the previous close, and the absolute low minus the previous close. The previous close lets the calculation capture gaps that a high-low range alone can miss. ATR smooths that series. TradingView documents RMA as its default ATR smoothing, with other smoothing choices available in the indicator settings.

A 14-bar reading on a five-minute chart and a 14-bar reading on a daily chart summarize different observations. Dividing each by price does not repair that mismatch. Compare matching timeframes, lengths and smoothing settings; also record the session and data source, since changing which bars enter the calculation changes the input.

TradingView's ATR% example calculates a 14-period RMA of true range, divides the result by close and multiplies by 100. That is different from dividing every historical true range by its own historical close and then averaging those percentages. The order of operations matters when the denominator changes. Check the formula when comparing indicators with similar names.

A rising percentage need not mean a larger price range

There are two moving parts in ATRP: ATR and the closing price. If ATR stays at 2 while the close falls from 100 to 80, ATRP rises from 2% to 2.5%. The absolute range estimate has not grown. It now represents a larger fraction of the price.

Keep raw ATR and close beside ATRP in a research record. Otherwise a percentage change can be mistaken for an increase in absolute movement. The ratio also needs a sensible denominator: a zero price makes it undefined, and negative or very small prices make ordinary positive-price comparisons unsuitable or unstable.

Neither ATR nor ATRP supplies a bullish or bearish direction. A large reading can accompany a fall, a rally or broad movement without useful progress. Nor is 5% ATRP a forecast that the next bar will move 5%, or a probability that a stop will survive.

Use the percentage as a filter, then return to order units

Suppose the research rule is: evaluate on completed daily bars, allow entries only when 14-bar ATRP is above 2%, and require a separate directional entry condition. Write down whether “above” means strictly greater than 2 or greater than or equal to 2. Those are different tests at the boundary.

Also distinguish a state from an event. “ATRP is above 2%” may remain true for several bars. “ATRP crosses above 2%” identifies a transition. Decide whether the strategy can enter again while the filter remains true, rather than letting the filter accidentally become a repeated entry instruction.

For instrument B, a hypothetical stop distance of twice ATR is 2 price units. Its ATRP reading of 5 is not a five-unit stop and is not an order quantity. With the same close and ATR, twice ATR is also 10% of price, but that equivalence depends on using the same reference values. See the worked ATR stop-distance and position-sizing example for converting the distance into monetary exposure.

Two instruments with equal ATRP can still have different spreads, liquidity, contract values and quantity increments. Before using a shared filter in automation, check a small sample of allowed and rejected signals by hand: record close, ATR, the calculated percentage, the decision time and the resulting instruction. A percentage makes the comparison readable. The order still needs its own units.