Direct answer
ATR’s main limitation is that it describes recent volatility, not the future path of price. Because it is computed from past ranges, ATR can become stale when market conditions change. It can also be misunderstood as a standalone trading or risk signal, even though it does not incorporate costs, execution quality, or the trader’s specific constraints.
Mechanism or definition
ATR stands for Average True Range. In plain terms, it summarizes how much prices have been moving over a chosen lookback period. The “true range” component reflects the size of movement that matters between consecutive periods (for example, using the current high and low and how those relate to the previous close). Averaging those values produces a single number that represents recent variability.
Key point: ATR is an estimate of typical movement magnitude based on a fixed method and a fixed window. Changing the lookback length changes the estimate, which means ATR is not a universal constant for a market.
Evidence or example
Consider a market that experiences mostly steady movements for weeks, then enters a sudden news-driven phase. ATR may rise quickly because the recent true ranges become larger. If you interpret ATR as “the market will continue moving like this,” you are making an assumption about persistence that ATR itself does not verify. After the event passes, volatility can compress, and ATR may fall even though any prior conclusion is no longer valid.
Another common example: two traders use ATR with different lookback lengths. One sees a higher ATR and expects wider typical movement; the other sees a lower ATR and expects tighter movement. The difference is not a mystery—each trader is using a different window, so the estimate is different by design.
Limitations and risks
-
No prediction of direction: ATR is about movement size, not whether price will rise or fall. Treating ATR as a directional indicator is a conceptual failure mode.
-
Stale volatility estimates: Because ATR is based on recent ranges, it can lag when volatility regimes shift (for example, from calm to turbulent and back). A lagging estimate can be less useful exactly when decisions need to adapt.
-
Sensitivity to assumptions and parameter choices: Lookback length, the way “true range” is calculated by a platform, and how data is sampled can change the ATR values. That makes comparisons across feeds or platforms uncertain.
-
Costs and execution are not included: ATR does not measure spreads, commissions, slippage, or order-fill behavior. Even if ATR suggests a certain movement magnitude, real outcomes can differ when costs and execution quality matter.
-
Historical relationships may not persist: If someone observes that “ATR tends to widen before X,” that is not a guarantee of recurrence. Relationships based on past data can fail when market structure or participant behavior changes.
Verification or next question
To independently verify ATR’s relevance, you can test its behavior under assumptions you control: check how ATR changes after major events, compare ATR using multiple lookback lengths, and examine how outcomes differ when you include realistic costs and execution assumptions. A useful next question is: Which ATR window length best matches the time scale of the decisions you are trying to evaluate?