Overlap studies
Parabolic SAR
Trend-following stop-and-reversal level derived from recent highs and lows.
- Engine
- TA-Lib
- Function
SAR- Input series
- high, low
- Outputs
- 1
How it works
Parabolic SAR recursively moves a stop level toward the trend's extreme point using an acceleration factor that increases up to the configured maximum and reverses sides when price crosses the stop.
Use case
Use Parabolic SAR as a trend-direction, trailing-stop, and reversal-level feature, with caution in sideways markets.
Parameters
| Parameter | Type | Default | Bounds or options | Description |
|---|---|---|---|---|
| Acceleration | number | 0.02 | 0…1.7976931348623157e+308 | Acceleration-factor step applied as the trend extends. |
| Maximum Acceleration | number | 0.2 | 0…1.7976931348623157e+308 | Maximum acceleration factor allowed during a trend. |
Outputs
| Output | Type | Description |
|---|---|---|
| SAR | number | Parabolic stop-and-reversal level for the bar. |
Reading the result
Parabolic SAR produces a trailing price-level series whose internal state accelerates as a directional move extends and resets when the model reverses direction. The output is best interpreted relative to price and the inferred trend side.
Because the calculation is path-dependent, an earlier repaired or revised bar can affect later values even when the recent bars themselves did not change.
Common mistakes
- Treating the output as a guaranteed stop level or execution instruction.
- Using aggressive acceleration settings without evaluating frequent reversals.
- Comparing only the latest value from datasets with different starting history.
- Applying crossover logic without accounting for gaps and price discontinuities.
DataCat workflow notes
For research, derive distance from close and side-of-price as explicit downstream features. Keep the raw price-level output and quality state so transformations remain reproducible.