Backtest evidence

Is there a Parabolic SAR step and maximum that works across every period?

Parabolic SAR Settings by Timeframe: 36 Tests

Finding

No setting led every period. On hourly USDJPY, step 0.05 with a maximum of 0.10 made +2,062.7 pips over 643 trades in 2025 and +2,969.8 pips in 2024, while 0.03/0.10 fell from +3,910.5 pips in 2024 to +981.4 in 2025. Settings positive in both 2024 and 2025 numbered 23 of 36 on hourly bars, 11 of 36 on 15-minute and 4 of 36 on four-hour. The 0.02/0.20 default was profitable for the full year but lost 1,032.5 pips in the first half of 2025 and made 1,847.7 in the second, so the halves do not agree.

Key results

Positive in both years
23 of 36 hourly, 11 of 36 on 15-minute, 4 of 36 on four-hour
Best hourly setting, step 0.05 / maximum 0.10
643 trades and +2,062.7 pips in 2025, +2,969.8 pips in 2024
The 0.02/0.20 default by half-year on hourly bars
−1,032.5 pips in the first half of 2025, +1,847.7 in the second

Scope

  • USDJPY, 2025-01-01 to 2025-12-31, with 2024 under the same rules and 2025 split into halves
  • 15-minute, hourly and four-hour bars
  • 6 step values (0.005 to 0.05) x 6 maximum values (0.05 to 0.50) = 36 settings; 432 runs across three timeframes and four windows
  • Flat 0.3 pip spread, 0.1 lots, zero slippage

Method

Measured with Formiq's backtester. A long fills at the close of the bar where SAR flips from above price to below, a short where it flips the other way, and the only exit is the opposite flip. Orders fill at the close of the flip bar, not at the plotted SAR value. No stop or target in the main sweep.

Limitations

  • One pair and two full years: USDJPY in 2024 and 2025.
  • Parabolic SAR implementations can differ in their initial direction, order of extreme updates and prior-bar clamps. These numbers use Formiq's implementation.
  • Orders fill at the close of the flip bar, not at the plotted SAR value.
  • Spread is flat at 0.3 pips and slippage is zero in the main sweep.
  • The 36-setting grid brackets the common default; it does not test every possible decimal.
  • The four-hour default made 130 to 138 trades a year, so its win rate and profit factor come from fewer trades than the hourly and 15-minute versions.