How to Use Parabolic SAR (Stop and Reverse) for Day Trading

Parabolic SAR (stop and reverse) is a trailing stop indicator that prints a dot below price in an uptrend and above price in a downtrend, and flips sides the moment price crosses it. For day trading, it is used in two ways: as a mechanical trailing stop that tightens as a move accelerates, and as a trend filter that tells you which direction you are allowed to trade. It works well in trending intraday sessions and badly in chop, so the whole craft is knowing which session you are in.

This guide covers Wilder’s formula, the acceleration factor settings that matter intraday, running SAR as a trailing stop, why it fails in ranges, pairing it with ADX, a worked 5 minute example, and a frank note on the evidence, which is thinner than most tutorials admit.

What Parabolic SAR measures

J. Welles Wilder introduced the indicator in his 1978 book New Concepts in Technical Trading Systems, the same book that gave traders RSI, ATR and ADX. SAR stands for stop and reverse: Wilder designed it as an always in the market system where the stop for a long position is also the entry for a short one. The “parabolic” part refers to the shape the dots trace as the stop accelerates toward price during a strong move.

Each bar, the indicator computes tomorrow’s SAR from today’s SAR, the extreme point (EP) of the current trend, and an acceleration factor (AF):

SAR(next) = SAR(current) + AF * (EP - SAR(current))

EP  = highest high since the uptrend began (lowest low in a downtrend)
AF  = starts at 0.02, increases by 0.02 each time EP makes a new extreme,
      capped at 0.20

Rules:
  Uptrend:   SAR(next) may not be above the low of the last two bars
  Downtrend: SAR(next) may not be below the high of the last two bars
  Reversal:  if price touches SAR, flip direction, set SAR = old EP,
             reset AF to 0.02

Three consequences follow. The dot only ever tightens. It moves faster after every new high, which is why it hugs a strong trend. And after a reversal the AF resets, so the new dot starts far from price and moves slowly, which is when the indicator is most vulnerable to a quick counter move.

Acceleration factor settings for intraday charts

The defaults, 0.02 step and 0.20 maximum, came from daily commodity charts in the 1970s. They still work on intraday timeframes because the formula is unit free; the step is a fraction of the distance to EP, not a price amount. What changes is how much noise a 1 minute chart carries, and noise is what triggers premature flips.

Change one number at a time and judge the result on flips per session, not on a single winning day. Fewer flips without noticeably later exits is an improvement; one bar earlier on the best trade of the week is usually curve fitting.

Using SAR as a trailing stop

This is the lowest risk way to use the indicator and the one with the most logic behind it. You take entries from something else, a breakout, a pullback to VWAP, a candlestick reversal such as those covered in our candlestick patterns primer, and SAR manages the exit.

1. Enter long on your own signal while the dot is already below price. If the dot is still above price, the trend by SAR’s definition has not turned; wait for the flip or skip the trade.
2. Place the initial stop at the current dot. On a 5 minute chart of a liquid stock that is often a reasonable distance; if it is too wide for your risk per trade, reduce size rather than tightening the stop by hand.
3. At the close of every bar, move the stop to the new dot. Never move it back.
4. Exit when price trades through the dot. Whether you actually reverse is a separate decision; most day traders do not.

Why SAR whipsaws in ranges, and the ADX filter

Because SAR is always in the market and always tightening, a sideways session produces a flip every few bars, each one a small loss plus commissions and slippage. Eight or ten losers in a row on a choppy day is the biggest reason people abandon the indicator.

Wilder’s own answer was ADX, which measures trend strength without regard to direction. The common rule of thumb pairs a 14 period ADX with SAR:

• ADX below 20: no trend. Ignore SAR flips, or turn the indicator off entirely for the session.
• ADX between 20 and 25: developing trend. Take SAR in the direction of the higher timeframe only.
• ADX above 25 and rising: trending. SAR flips in the direction of the trend are tradeable, counter trend flips are exits only.
• ADX above 40 and turning down: the move is mature. Expect the next SAR flip to be a real reversal rather than noise, and tighten anyway.

Time of day is the other practical filter. The first fifteen minutes and the lunch hour produce most intraday whipsaws in US equities, so many SAR users only honor flips from roughly 9:45 to 11:30 Eastern and in the last ninety minutes.

A worked example on a 5 minute chart

Take a liquid large cap stock that gaps up on earnings and trends for the morning. Bar by bar, without invented prices:

1. Bar 1 to 3 after the open: price rips higher, SAR flips below price on bar 2. AF is 0.02, so the dot sits well below the lows and barely moves. You do not chase; you wait for a pullback.
2. Bar 5: price pulls back toward VWAP and prints a bullish engulfing candle. Dot is still below price, ADX has climbed above 25. You enter long with the stop at the dot.
3. Bars 6 to 12: each new high bumps AF by 0.02. By bar 12 the AF is at its 0.20 cap and the dot is only a few ticks under the recent lows. Your stop has been ratcheted up every bar.
4. Bar 14: a wide range down bar trades through the dot. You are stopped out near the highs of the move. SAR flips above price with a fresh 0.02 AF. You do not short, because ADX is still high and the higher timeframe is up.
5. Bars 15 to 20: price grinds sideways, SAR flips twice more. ADX rolls over below 25. You stop trading the stock for the session.

The indicator earned its keep during the trend and produced only noise afterward; the filters, not the dots, made the difference. Volume confirmation helps too, and our OBV for scalping guide covers a complementary approach.

Computing SAR yourself

Coding it once removes any mystery about the two bar rule and the reset. A minimal Python version:

def parabolic_sar(highs, lows, step=0.02, max_af=0.20):
    n = len(highs)
    sar = [None] * n
    up = highs[1] > highs[0]
    ep = highs[0] if up else lows[0]
    sar[0] = lows[0] if up else highs[0]
    af = step
    for i in range(1, n):
        prev = sar[i - 1]
        cur = prev + af * (ep - prev)
        if up:
            cur = min(cur, lows[i - 1], lows[i - 2] if i > 1 else lows[i - 1])
            if lows[i] < cur:
                up, cur, ep, af = False, ep, lows[i], step
            elif highs[i] > ep:
                ep, af = highs[i], min(af + step, max_af)
        else:
            cur = max(cur, highs[i - 1], highs[i - 2] if i > 1 else highs[i - 1])
            if highs[i] > cur:
                up, cur, ep, af = True, ep, highs[i], step
            elif lows[i] < ep:
                ep, af = lows[i], min(af + step, max_af)
        sar[i] = cur
    return sar

Run it against your platform’s output for a day of bars; small differences on the first few values are normal because platforms seed the initial trend differently.

What the evidence actually says

Be honest with yourself about this part. Academic tests of simple technical rules on liquid markets generally find little or no edge after transaction costs, and Parabolic SAR is a simple rule. Backtests that show SAR reversal systems winning tend to be on trending periods, on instruments picked after the fact, or without realistic slippage. There is no published, replicated study we are aware of showing a raw SAR system beating buy and hold on intraday US equities after costs.

That does not make the indicator useless. Used as a trailing stop it is a disciplined way to take profits and limit losses. Used as a standalone signal generator, it is a coin flip with a commission attached. Day trading itself is a high risk activity; the SEC’s investor guidance on day trading is short and worth reading before you risk real capital.

Troubleshooting

The dot flips every two or three bars. You are in a range or on too short a timeframe. Check ADX; if it is under 20, stop. If ADX is fine, step up from 1 minute to 5 minute bars, or lower the step to 0.01.

SAR exits me right before the big move. The AF reached its cap and the dot was hugging price during a normal pullback. Lower the maximum from 0.20 to 0.10 so the trail stays looser late in a move, or accept the exit and re enter on the next pullback with the dot back under price.

The indicator gives back most of the profit at the end of a trend. That is by design: SAR waits for price to cross the dot. Combine it with a partial profit target, for example selling half at a fixed multiple of your initial risk and trailing the rest with SAR.

Frequently asked questions

What are the best Parabolic SAR settings for day trading?

Start with Wilder’s defaults, a 0.02 step and 0.20 maximum, on 5 or 15 minute bars. Drop to 0.01 and 0.10 on 1 to 3 minute charts or volatile names to reduce flips. There is no universally best setting; tune for fewer whipsaws on your specific instrument rather than for the biggest single win.

Is Parabolic SAR a leading or lagging indicator?

Lagging. It is computed from past highs and lows and can only tighten toward price, so it confirms a trend after it has started and signals a reversal after price has already moved against you. Its value is in mechanical stop placement, not in predicting turns.

Can I use Parabolic SAR alone?

You can, but a raw SAR reversal system loses steadily in ranging markets because every flip costs a small loss plus fees. Pair it with a trend strength filter such as ADX, a higher timeframe bias, or a time of day rule, and use it mainly for exits rather than entries.

Does Parabolic SAR work for scalping?

Only with a looser setting and a strong trend filter. On 1 minute charts the default parameters flip constantly. Scalpers who use it typically set the step to 0.01, cap the AF at 0.10, and only take flips that agree with the 5 or 15 minute direction, using it purely as a trailing stop.

Is Parabolic SAR better than a moving average stop?

They serve different purposes. A moving average stop keeps a constant lag and lets you hold through pullbacks. SAR accelerates as the move matures, so it exits earlier on strong runs and protects more open profit. Many traders use a moving average as the trend filter and SAR as the stop.

The bottom line

Parabolic SAR is a trailing stop with a formula, and it is excellent at that one job in a trending session. Treat the dot as your stop, ratchet it every bar, and take entries from a separate signal while ADX confirms the market is actually moving. Turn it off, or at least stop honoring flips, when ADX is low or the clock says lunch.

Do not expect it to be a system on its own, and do not trust a backtest that says otherwise unless it includes costs and a choppy period. Everything above is educational and not investment advice; test it on paper first and size positions so a string of whipsaws is an annoyance, not a disaster.