For two years I treated the Parabolic SAR like a magic trend arrow. Dots below price: buy. Dots above price: sell. Simple. And for two years, ranging markets ate me alive — I'd buy the dot flip at the top of a range, get chopped for −30 pips, flip short at the bottom, get chopped again. My 2023 journal shows 41 SAR-based trades with a win rate of 34% and a net loss of $2,180. The indicator wasn't broken. My understanding of it was.
Everything changed when I actually learned the math behind the dots. The SAR isn't a signal — it's an accelerating trailing stop that assumes a trend exists. Once I started using it only in confirmed trends, and only for exits and trailing rather than entries, those same dots became one of my most reliable tools. My 2024 SAR trades: 38 trades, 61% win rate, +$4,940. Same indicator. Different brain. Here's what I wish someone had shown me in 2022, including the formula, the settings I actually tested, and how I size every SAR trade through the DarkWolf MT5 Signal Calculator.
How the Parabolic SAR Actually Calculates Its Dots
The SAR (Stop and Reverse) was built by J. Welles Wilder — the same mind behind RSI and ATR. Each new dot is computed from the previous dot, an acceleration factor, and the extreme price of the current trend:
| Variable | Meaning | Example value |
|---|---|---|
| SAR old | The previous period's SAR dot value | 1.0820 |
| AF | Acceleration factor — starts at step, grows each new extreme, capped at maximum | 0.02 → 0.04 → … → 0.20 |
| EP | Extreme point — highest high in an uptrend, lowest low in a downtrend | 1.0895 |
Work it once by hand and the indicator demystifies itself. Say SAR old = 1.0820, AF = 0.02, EP = 1.0895: SAR new = 1.0820 + 0.02 × (1.0895 − 1.0820) = 1.0820 + 0.00015 = 1.08215. The dot creeps toward price, accelerating as the trend extends. (Wilder's original parameters and more background are covered in Investopedia's technical indicator library.) That's the key insight: the SAR assumes the trend continues and chases it faster over time. In a real trend, beautiful. In a range, it whipsaws — flipping dots every few candles and generating the chop that cost me $2,180.
Dialling SAR Signals Into the Signal Calculator
I don't take SAR dot-flips as entries anymore. I enter on trend confirmation (a pullback to the 20 EMA in the trend direction), then use the SAR dot as my trailing stop. The calculator turns the SAR distance into a position size:
| Input Field | What it means | Sample value |
|---|---|---|
| Currency pair | The trending pair you're trading | EUR/USD |
| Account balance | Current equity | 20,000 |
| Risk percentage | Risk per SAR-trailed trade | 1.0 |
| Stop loss (pips) | Entry price minus current SAR dot distance | 35 |
| Entry price | Your pullback entry level | 1.0920 |
The practical detail: measure the stop as the distance from your entry to the current SAR dot, then add a 5-pip buffer. SAR dots get tagged by wicks constantly; the buffer keeps you in trends that are still valid. My DarkWolf MT5 Signal Calculator run for the example above gives 0.57 lots at $200 risk — and I never adjust size mid-trade as the SAR trails closer. The shrinking stop is profit protection, not an excuse to add.
Trading SAR Reversals Without Getting Chopped
Here's my current rule set, the one that flipped my SAR results from −$2,180 to +$4,940. Rule one: the SAR is only active when ADX (14) reads above 25. Below 25, the market is ranging, and SAR signals are random noise — I ignore every dot. This single filter eliminated about 60% of my losing SAR trades.
Rule two: entries come from price action, not dot flips. I wait for the trend direction (SAR dots + price above the 50 EMA), then enter on a pullback to the 20 EMA with a bullish engulfing or pin bar. The dot flip alone is never an entry trigger for me anymore.
Rule three: the SAR dot is my trailing stop, updated at each candle close. I don't move it intra-candle — that's how wicks steal your position. When price closes beyond the dot, I'm out, no second-guessing. Rule four: no SAR trades 30 minutes before or after red-flag news. A CPI print will flip dots and stop you out in both directions within a minute. I learned that watching a $340 winner become a $180 loser in 40 seconds.
Why SAR Lags — and When That Matters
Every SAR criticism boils down to one fact: it's a lagging indicator built from past prices. The dots trail the market by design. In a strong trend the lag is a feature — it keeps you in. At trend exhaustion the lag is a tax — you give back the last portion of the move.
How big is the tax? I measured it across my 38 winning 2024 SAR trades: the SAR exit captured on average 71% of the total trend move, giving back 29% at the end. Compare that to my old fixed-target exits, which captured 52% because I kept taking profit too early. The lag costs you the tail; impatience costs you the middle. I'll pay the lag tax every time.
The lag is fatal in exactly one situation: the V-reversal. When a trend ends with a violent snapback instead of a rounded top, the SAR dot sits far from price and the reversal hits your trailing stop for a full loss before any dot flip warns you. My defense: if a single candle closes beyond the 20 EMA against my position, I exit manually instead of waiting for the SAR. That rule saved me $1,100 on a USD/CHF reversal in October 2024. For trend-strength confirmation I pair the SAR with ADX trend-strength analysis — the combination filters out most V-reversal traps.
Three SAR Trades on a $100,000 Account
November 2024, funded $100K account, three textbook SAR-trailed trades. Full ledger:
| Trade | Setup | SAR exit | P&L |
|---|---|---|---|
| 1. USD/JPY long | ADX 31, pullback to 20 EMA at 154.80, SAR dot 154.45 | Trailed 9 days, exited 157.90 on dot flip | +$3,100 (1.0 lot, $1,000 risk) |
| 2. EUR/GBP short | ADX 28, bearish engulfing at 0.8310, SAR dot 0.8345 | Trailed 6 days, exited 0.8225 | +$2,250 (1.5 lots, $750 risk) |
| 3. AUD/USD long | ADX 26, pin bar at 0.6520, SAR dot 0.6485 | Wicked out day 2, −35 pips | −$700 (1.0 lot, full stop) |
Net: +$4,650 on $2,450 risked. Trade 3 is the honest part — the SAR stop got tagged by a wick on day two, and price then rallied 120 pips without me. That's the lag tax in reverse: sometimes the stop is too tight at the start of a trend. I accept it. A system that wins 2 out of 3 and banks +$4,650 doesn't need to be perfect.
I Tested Three SAR Settings Through the Calculator
Default SAR settings are 0.02 step, 0.20 maximum. I backtested alternatives on EUR/USD H4, 2022–2024, then ran the winners through the calculator at live size:
| Settings (step/max) | Backtest result | Calculator sizing ($20k, 1%) | Verdict |
|---|---|---|---|
| 0.02 / 0.20 (default) | +18.4%, 54% win rate, 96 trades | 35-pip avg stop → 0.57 lots | My live setting — balanced |
| 0.01 / 0.10 (slow) | +11.2%, 61% win rate, 71 trades | 52-pip avg stop → 0.38 lots | Fewer, wider stops — good for swing accounts |
| 0.03 / 0.30 (fast) | +4.1%, 47% win rate, 138 trades | 24-pip avg stop → 0.83 lots | Overtrades — costs eat the edge |
The fast setting was the surprise: more trades, worse results. The tighter SAR flips constantly in minor pullbacks, and the extra 42 trades added roughly $800 in spread costs that the backtest's gross profit couldn't cover. Defaults win for intraday; slow wins for swing. The FCA's research on retail trading costs backs up what my spreadsheet showed — frequency is a cost center most traders ignore.
Parabolic SAR Trend Reversal Trading: FAQs
What are the best Parabolic SAR settings?
The default 0.02 step and 0.20 maximum work well for most intraday and swing trading — my EUR/USD H4 backtest showed +18.4% with defaults. Slower settings (0.01/0.10) suit swing traders who want fewer whipsaws; faster settings (0.03/0.30) overtrade and bleed costs. Don't optimize the settings to fit past data; pick one and trade it.
Does Parabolic SAR work in ranging markets?
No, and this is where most traders lose money with it. The SAR assumes a trend exists, so in ranges it flips dots constantly and generates false signals. I only trade SAR setups when ADX reads above 25. Below that, the indicator is decoration. My ranging-market SAR losses totaled $2,180 before I added this filter.
Should I use Parabolic SAR for entries or exits?
Exits, in my experience. Dot-flip entries get you in late and chopped in ranges. I enter on price-action pullbacks in the SAR-confirmed trend direction, then use the SAR dot as a trailing stop. This division of labor — price action for entries, SAR for exits — is what flipped my results positive.
How does Parabolic SAR compare to a moving average trailing stop?
The SAR accelerates as the trend extends, so it trails tighter in mature trends and looser early on. A moving average trails at a constant distance. I prefer the SAR's acceleration in strong trends because it locks in more profit near exhaustion. In choppy trends, though, the MA is more forgiving — the SAR will stop you out on normal pullbacks.
What timeframe works best for Parabolic SAR?
H4 and daily for swing trading, H1 for intraday — never below M15 in my experience. Lower timeframes multiply the whipsaws faster than they multiply opportunity. My live SAR trading happens on H4; I use H1 only to time entries inside H4 trends. The indicator's lag becomes unmanageable on scalping timeframes.
Can Parabolic SAR be used for scalping?
I don't recommend it. The SAR's lag means scalping timeframes generate constant false flips, and the spread costs on 100+ trades a month destroy the edge. My fast-setting backtest (0.03/0.30) proved this: +4.1% gross became nearly breakeven after costs. Scalpers are better served by order flow or price action.
Why does my Parabolic SAR keep getting me stopped out?
Three usual causes: trading it in ranging markets (check ADX), placing stops exactly at the dot with no buffer (add 5 pips), or using it on too low a timeframe. In my journal, 70% of premature SAR stop-outs were wick tags, not real reversals. The buffer and the ADX filter fixed most of them.
Does Parabolic SAR repaint?
The current, still-forming candle's dot can shift as price moves — that's normal calculation, not repainting. Once a candle closes, its SAR dot is final and never changes. Only ever make decisions on closed candles. If your platform shows dots dancing on the live candle, ignore them until the close.