Parabolic SAR Trading: Stop and Reverse Signals Guide
Parabolic SAR is a price-based stop-and-reverse indicator. Its dots mark the state of a recursive calculation; they do not establish that a market has a strong trend or that the next trade will profit. A useful SAR strategy must specify when a signal is known, whether it closes or reverses a position, and how orders are filled.
What the Dots Mean
Wilder's Parabolic SAR typically appears below price in its rising state and above price in its falling state. A change of side indicates a reversal in that model. It can alternate repeatedly in a narrow range, so reading the direction is different from measuring trend strength.
TradingView's SAR documentation describes three inputs: Start, Increment, and Maximum acceleration. Its defaults are 0.02, 0.02, and 0.20. The initial value and increment are separate controls even when their values match.
Calculation and the Missing Rules in a One-Line Formula
The candidate update is:
Next SAR = prior SAR + prior AF × (prior EP − prior SAR)
EP tracks the current run's extreme. AF increases when a new extreme is made, up to its cap. The calculation also includes reversal/reset logic and constraints against the preceding two bars' lows or highs. The recurrence alone is not the full indicator.
A reversal tests price penetration using the bar's range; it does not inherently require a closing-price cross. A strategy may choose to wait for the close, but that is an additional trading rule.
As an arithmetic illustration, prior SAR 100, EP 110, and AF 0.02 yield a candidate of 100.20. The final plotted value can differ after constraints or a reversal. This example shows one update, not a forecast or a simulated trade.
Choose Settings as a Research Question
Pineify's current SAR component exposes the same three inputs and generates TradingView's built-in ta.sar calculation with one SAR output. Its drawing uses cross markers. It does not infer the most profitable settings for a particular ticker.
| Input | What changing it asks | What to measure |
|---|---|---|
| Start | Should the new run begin with a different acceleration? | Early exits and initial stop distance |
| Increment | Should acceleration change faster after new extremes? | Holding duration, turnover, and whipsaws |
| Maximum | Should acceleration be allowed a different upper bound? | Late-trend exits and execution costs |
There is no verified universal “crypto setting” or requirement that shorter charts use higher acceleration. Record the default as a baseline, select a small set of alternatives before examining results, and keep an untouched validation period. Report all variants, including those that lose money after costs.
A parameter that produces the highest historical return on one stock is not evidence that it generalizes to another stock, exchange, or interval. Check nearby settings as well: an isolated peak deserves investigation before further optimization.
Three Distinct SAR Strategies
1. Closed-Bar State Change
Define a long entry when a completed bar changes from the SAR-above state to SAR-below. Specify whether a later opposite change closes the long or also opens a short. These produce different exposure and transaction costs.
If the strategy calculates after the bar closes, do not assume it bought at an earlier intrabar cross. TradingView's strategy execution documentation explains that the default broker emulator normally fills newly created market orders on the next available tick, commonly the following bar's open in historical tests. Execution settings can change that behavior.
Inspect a sample of entries, reversals, and gaps in the trade list. A plausible chart marker is not enough to verify the order sequence.
2. Higher-Timeframe Direction Filter
One testable variant permits lower-timeframe long signals only when the last completed higher-timeframe bar has the chosen bullish SAR state. Fix the higher timeframe and state definition before testing.
Using the still-forming daily SAR on an intraday chart can produce signals that change before the daily close. TradingView's repainting guidance explains why current and confirmed higher-timeframe values behave differently. A historical test must not apply the final daily state to an earlier intraday decision.
Compare the filtered rule with the unfiltered baseline. Count the trades removed and the opportunities missed, as well as net returns and drawdowns. Fewer entries alone do not prove better accuracy.
3. SAR as an Exit for Another Entry Rule
Keep the original entry rule fixed and use SAR only to update the protective exit. Specify whether the order updates after each completed bar or during realtime calculation. These are different strategies.
For a long position, a proposed stop must remain below the applicable market price when submitted. If the model has already reversed and places SAR above price, do not blindly use that dot as a new long protective stop. Define an exit action for that state. Apply the corresponding logic to short positions.
A trailing stop can reduce the amount given back during some trends, but does not automatically lock in net profit. Gaps, slippage, commissions, and stop-order mechanics can produce a worse result than the plotted level.
Combining SAR With Other Indicators
A moving-average condition, an RSI condition, or a MACD condition can be written as an additional filter. None establishes a higher win rate merely by agreeing with SAR. Several indicators computed from the same price series can repeat similar information.
Define timing precisely. “RSI was below 30 sometime recently” needs a lookback and an invalidation rule. “MACD confirms” needs a specific event: a MACD-line/signal-line cross is different from the MACD line crossing zero. A histogram zero cross corresponds to the former when the histogram is MACD minus its signal.
Change one filter at a time while retaining the same entry timing, sizing, exits, and cost model. Otherwise you cannot tell which change caused the difference.
A Practical Verification Sequence
- Add the built-in SAR to a standard candlestick chart and record its three inputs, instrument, interval, and session.
- Mark several completed-bar state changes, including a sideways period and a gap.
- Write separate entry, exit, reversal, sizing, and stop-update rules.
- Build the strategy and compare its calculated SAR and order timestamps with the reference chart.
- Inspect the trade list, costs, drawdown, exposure, and losing sequences before optimizing.
- Evaluate the fixed rule on later unseen data and in a simulated forward session. Keep discrepancies in the log.
Pineify's Visual Editor or AI Coding Agent can help express those rules in Pine Script. Compile and review the resulting script, then run the strategy in TradingView. Syntax checking does not demonstrate correct fills, profitable parameters, or automatic live order placement. A plotted SAR indicator by itself submits no protective broker order.
Questions and Answers
Does a SAR flip require the bar to close beyond the dot?
No. The indicator's reversal logic uses price penetration. Waiting for a completed-bar state change is a separate strategy choice, and it changes when an order can be submitted.
What settings should I start with?
Use the documented defaults as a comparison baseline. Keep Start, Increment, and Maximum separate, and test any alternatives with the same execution assumptions. No asset-specific optimum is established here.
Does adding RSI or MACD remove false signals?
No. A filter changes the set of eligible trades. Measure its effect after costs and account for delayed or missed entries; agreement between indicators is not a calibrated success probability.
Is a SAR dot an active stop-loss?
No. It is a calculated level. A strategy or trader must submit and update the appropriate order. Live execution can differ from that level.
Why can SAR lose repeatedly in a range?
Alternating price moves can repeatedly change the SAR state. A reversal system can then trade back and forth, incurring losses and costs. A trend filter can be tested, but cannot guarantee that this stops happening.