Direct answer
To assess Parabolic SAR, collect (1) the price series needed for the indicator, (2) the parameter settings used in the SAR formula, (3) the timeframe and price source so calculations are consistent, and (4) a set of verification checks that reveal whether the output is internally consistent and usable. Because indicator behavior changes with market regime, execution conditions, and data quality, you should also note limitations and potential failure modes before treating SAR output as informative.
Mechanism or definition
Parabolic SAR (often written “Parabolic SAR”) is a trend-following indicator that generates a sequence of dots around price and switches its placement when the underlying trend direction changes. “Assessing” it therefore means checking whether the SAR dots and switch points are calculated from the expected inputs and whether their behavior matches the assumptions behind the method.
The minimum data inputs are:
- A historical time series with enough bar resolution to support your timeframe (commonly OHLC: open, high, low, close). Without at least high/low values, you cannot reproduce the indicator’s incremental step-and-acceleration logic.
- The indicator parameters used by the calculation. In practice this includes values such as the starting acceleration factor, the acceleration increment, and a maximum acceleration cap (the common “three-parameter” setup). Even if you use software defaults, you need to record the exact parameter values so assessments are repeatable.
- The timeframe definition (for example, 1-minute vs 1-hour bars). SAR is computed over ordered bars; changing timeframe changes which sequence of highs and lows the indicator “walks” through.
- The price source and data treatment (for example, whether you use raw broker bars, adjusted data, or resampled bars; and how you handle missing or outlier bars). Data provenance matters because SAR is sensitive to high/low extremes.
A practical assessment also requires separating stable mechanics from variable conditions:
- Stable mechanics: SAR’s reliance on ordered highs/lows, its parameter-driven step/acceleration, and the concept of dot placement switching.
- Variable conditions: market volatility regime, trading costs and execution quality, and how your data provider constructs bars.
Evidence or example
Here is a concrete, self-check oriented example of what to collect and verify (no live data assumed):
- Lock the calculation setup
- Choose a timeframe and ensure your OHLC bars match that timeframe.
- Record the three SAR parameters (start acceleration, acceleration step, and maximum acceleration). If your platform has defaults, note them.
- Reproduce the SAR output consistency
- Confirm that SAR dots update at each bar according to the same formula and parameter set.
- Check flip points: when direction switches, the SAR placement should jump from one side of price to the other in a way that matches the indicator’s definition.
- Inspect edge cases
- Look at how SAR behaves near the beginning of your dataset. Many implementations need a “seed” period; results near the start can differ because the prior trend context is limited.
- Identify bars with extreme highs or lows. Since SAR is influenced by highs/lows and trend acceleration, one anomalous bar can affect subsequent dots.
- Compare across internally consistent data treatments
- If you use resampling (e.g., build 1-hour bars from 5-minute bars), verify that the highs and lows remain consistent with your method.
- If adjusted price data is involved, document what was adjusted and whether highs/lows were altered by that adjustment.
What counts as “evidence” here is not predicted performance; it is whether you can independently verify that the SAR output follows from the recorded data and settings.
Limitations and risks
At least one material limitation is that SAR can be sensitive to whipsaws in choppy conditions. Because SAR accelerates during persistent moves, short-term reversals can cause frequent switch behavior even if the broader context is not trending.
Common failure modes to account for:
- Data quality and provenance problems: inconsistent bar construction, missing bars, or mismatched timeframe definitions can produce different SAR dots.
- Parameter mismatch: two assessments using different acceleration parameters are not directly comparable.
- Regime dependence: relationships observed in one market environment may not hold in another, so historical similarity does not guarantee future behavior.
- Practical trading realism (even though you should not treat SAR as a standalone signal): execution costs, slippage, and latency can change whether the “switch points” translate into achievable results.