Definition-first verification
ATR (Average True Range) is a volatility metric built from “true range” for each bar, then averaged over a lookback period. True range uses the current bar’s relationship to the previous bar’s close, so it reflects both intrabar movement and gap-like jumps. Because ATR is defined mathematically, you can verify information about ATR by confirming: (1) the exact definition, (2) the calculation inputs (OHLC), and (3) the averaging method over the chosen period.
Mechanism and reproducible calculation steps
To verify ATR independently, use a small, fully specified dataset (for example, 15 consecutive bars) with known Open, High, Low, and Close values. Then reproduce ATR using these steps:
-
Compute true range for each bar (except the first, which needs a previous close):
- Let H = current high, L = current low, Cprev = previous close.
- Compute TR = max(H − L, |H − Cprev|, |L − Cprev|).
-
Choose assumptions before averaging:
- Fix the time frame (e.g., 1-hour bars) and the lookback length (commonly called the period).
- Confirm whether your reference uses a simple moving average or a smoothed/recursive averaging approach. Different providers implement the averaging step differently.
-
Compute ATR for each position you can verify:
- If using a simple average: ATR(period) = average(TR over the last period bars).
- If using smoothing: apply the provider’s documented recurrence on each new bar, using the initial value specified by that method.
-
Round-trip check:
- Compare your recomputed ATR values to the claimed ATR output for the same instrument, timeframe, and period.
- If results differ, treat it as evidence that at least one assumption (bar construction, averaging method, or data cleaning) does not match.
Evidence and example verification approach
A practical verification workflow is to start with one source of truth for raw OHLC (the same feed/provider you will compute from), then compare:
- Whether the other party’s ATR explanation matches the same true range definition.
- Whether their stated lookback period matches the one you used.
- Whether their ATR output aligns when you recompute it from the same OHLC.
For example, take a short segment where a gap-like move is visible between a previous close and the next bar’s high or low. In those cases, true range should reflect more than just (High − Low). If an ATR description claims to use true range but your recomputation does not pick up the same effect on those bars, the claimed method or the input data is inconsistent.
Limitations and failure modes to account for
Even with correct formulas, ATR can be hard to verify across sources because implementations and data handling vary. Material limitations include:
- Averaging method differences: some references use a simple average, others use smoothing; the same TR series can produce different ATR paths.
- Bar/timeframe mismatches: converting to a different timeframe, using different session cutoffs, or applying time zone adjustments changes OHLC bars.
- Data cleaning and corporate actions: missing bars, adjusted prices, or different handling of holidays can change OHLC inputs.
- Rounding conventions: intermediate rounding or display rounding can create small mismatches.
- Interpretation limits: ATR measures volatility magnitude relative to recent history; it does not, by itself, predict future direction or guarantee outcomes.
Verification checklist and next question
To verify information about ATR, you can ask four concrete questions:
- What true range formula is used?
- What OHLC source and timeframe are assumed?
- Is the ATR averaging a simple average or a specific smoothing method?
- Do the reported ATR values match a recomputation using the same inputs and assumptions?
If a provider cannot specify these items clearly, treat their ATR “value” claims as not fully verifiable. Next, you can verify related volatility concepts by checking whether they use different definitions of range or normalization, since small definitional changes can produce meaningfully different outputs.