Parabolic SAR: The Trailing Stop Indicator

How Parabolic SAR Works

The parabolic SAR, short for Stop and Reverse (SAR), is a trailing stop and reversal system developed by J. Welles Wilder in 1978. Wilder is the same name behind the Relative Strength Index, the Average True Range (ATR), and the Average Directional Index (ADX). The parabolic SAR is the fourth indicator from that set and the one most explicitly designed to tell a trader where to place a stop loss.

The indicator plots a series of dots either above or below price. When the dots sit below the candles, the system is in an uptrend and each dot is acting as a rising stop-loss level. When the dots sit above the candles, the system is in a downtrend and each dot is a falling stop. The dots never plot on price itself. They are always on one side or the other, which is why the indicator is binary at any given moment: long or short, never neutral.

The "parabolic" name describes the shape the dots form on the chart. As a trend extends and price keeps making new extremes in the trend direction, the dots curve toward price faster and faster, tracing a parabolic arc. That acceleration is the whole point of the system. The longer a trend runs, the tighter the stop gets, until eventually price intersects the dots and the system flips to the opposite side.

The acceleration factor

The math behind the curve is driven by a single variable Wilder called the Acceleration Factor (AF). The AF starts at 0.02. Every time price prints a new extreme in the direction of the current trend, the AF increases by a step of 0.02, up to a cap of 0.20. On the next bar, the SAR moves toward price by the AF multiplied by the distance between price and the prior SAR.

The formula reads simply: next SAR equals prior SAR plus AF times (extreme point minus prior SAR). In an uptrend the extreme point is the highest high since the trend began. In a downtrend it is the lowest low. As new highs (or lows) keep printing, the AF ratchets higher, the dots move closer to price faster, and the trailing stop tightens. If price stops making new extremes, the AF stops ratcheting and the dots advance at a constant rate, eventually catching up to price.

The TSLA chart above shows the SAR dots tracking below price through uptrends and flipping above price on each reversal. In the long bull runs the spacing between price and dots compresses as the AF ratchets up, until a pullback finally intersects the dots and the system reverses. In choppier stretches the dots flip back and forth on small swings, which is the characteristic whipsaw behavior the indicator is known for.

Reading the Dots

The visual grammar of parabolic SAR is the simplest of any indicator Wilder published. A dot below a candle is a long signal in effect: a trader following the system is holding long, and the dot marks where the stop sits. A dot above a candle is a short signal in effect: stop above, trade is short. When price touches or crosses a dot, the trade closes and the system immediately reverses to the other side. That is the literal meaning of Stop and Reverse — there is no flat state.

The flip event

A flip is the most important event the indicator produces. On the bar where price intersects the SAR, the dot moves from below price to above (or vice versa) and the position direction is inverted. The new SAR starts at the most extreme point of the previous trend (the highest high if the prior trend was up, the lowest low if it was down), and the AF resets to 0.02. The cycle begins again.

What this means in practice: every flip is simultaneously an exit from one trade and an entry into the opposite trade. A pure mechanical implementation of the system is always in the market. A discretionary implementation treats the flip as an exit signal only and waits for a confirming setup before taking the reverse direction. Most experienced traders use the second interpretation, because the mechanical reverse-on-every-flip rule is what generates the worst drawdowns in sideways markets.

Dot spacing tells you trend age

Early in a fresh trend, the dots sit far from price. The AF is low and the SAR is still anchored near the prior trend's extreme. As the trend matures and the AF ratchets up, the spacing compresses. Dots that started ten percent away from price end up one or two percent away by the time the trend is fully extended. Reading the spacing is a quick way to gauge trend age. Wide dots mean fresh trend with room to run. Tight dots mean mature trend approaching its mechanical exit.

NVDA is a useful study because its trends are typically long and steep, exactly the regime the parabolic SAR was designed for. In the extended bull runs visible on the chart, the SAR dots track below price for months, with the spacing tightening progressively as the AF ratchets up. The flip out comes when momentum exhausts and price finally pulls back into the rising dots.

Using SAR as a Trailing Stop

The most defensible use of parabolic SAR is as a trailing stop underneath an existing long (or above an existing short) that was entered for some other reason. The indicator was not built to generate entries. It was built to manage exits. Used that way, it solves a problem most discretionary traders struggle with: when to take profits on a winning trade.

The trailing-stop workflow

A typical use looks like this. A trader takes a long position on some independent signal — a breakout, a moving-average crossover, a fundamental catalyst, whatever the strategy calls for. From the entry bar forward, the parabolic SAR is plotted on the chart. The stop is set wherever the current SAR dot sits. Each bar, the dot advances; the trader adjusts the stop upward to match. The position stays open until price touches the dot, at which point the stop executes mechanically.

The advantage over a fixed-percentage trailing stop is automatic adaptation. A fixed five percent trailing stop treats every market the same. A SAR stop in a fast trend ratchets up aggressively because new highs keep raising the AF. A SAR stop in a slow trend advances gently because few new extremes are printing. The indicator's own logic does the volatility-aware tightening that a trader would otherwise have to apply by hand.

Acceleration captures parabolic moves

The reason the system uses an accelerating factor rather than a fixed distance is exactly to capture the late stage of a parabolic move. In the final weeks of a strong trend, price often moves vertically. A static trailing stop sitting twenty percent below price gives back too much when the move finally reverses. A SAR stop, with its AF maxed out near 0.20, sits one or two percent below price and exits the trade within a bar or two of the top.

The trade-off is the inverse case: in a long but gentle uptrend, the accelerating SAR will sometimes get tagged on a routine pullback before the trend has actually ended. A trader using SAR as a trailing stop accepts that some winners will be cut short in exchange for locking in the gains on the truly parabolic ones. That tradeoff matches the philosophy of most trend-following systems: sacrifice average winners to catch outsized ones.

Sizing the initial stop

One detail worth flagging: on the bar of entry, the parabolic SAR may already be plotted very close to price if the trend that produced the entry is already extended. A long entered after a ten-day rally will have the SAR dot a few percent below the entry price. That stop distance may be tighter than the trader's risk plan calls for. The honest fix is to size the position so that the dot distance equals the trader's intended dollar risk, rather than overriding the indicator's stop with something wider.

SAR Failures in Sideways Markets

The parabolic SAR has one large, predictable failure mode: it does not work in range-bound markets. The system assumes a trend exists. When price chops sideways within a horizontal range for weeks, the dots flip from below to above and back, week after week, generating a continuous sequence of stop-outs and reverses. Every flip costs the trader the spread plus slippage plus whatever range was traveled before the next reversal. The cumulative bleed is substantial.

Why the system breaks in chop

The mechanism is straightforward. In a sideways range, price keeps oscillating between the same support and resistance levels. The SAR's reset-on-flip logic starts a new trend cycle from each recent extreme, but the AF never has time to ratchet up before price reverses again. The dots end up tracking close to price on both sides, which means even small swings inside the range trigger flips. A two percent intraday pullback in a five percent range is enough to convert a long to a short, which then gets flipped back to long on the next bounce.

In strongly trending periods the same mechanism is what makes the system valuable. The AF ratchets up because new extremes keep appearing, the stop tightens at the right moment, and the eventual flip locks in most of the move. The behavior is the same in both regimes. What changes is whether the regime rewards or punishes the behavior.

Filtering with ADX

The standard fix is to gate the system on a trend-strength filter. Wilder paired parabolic SAR with his own Average Directional Index (ADX) for exactly this reason. The rule is simple: only trade SAR signals when ADX is above 25. Above that threshold, the market is trending strongly enough that SAR's flip-and-reverse logic captures real moves. Below 25, the market is either ranging or transitioning, and SAR signals should be ignored.

The combination is one of the cleaner indicator pairings in the Wilder family. ADX identifies regime; SAR manages execution within the trending regime. Many trend-following systems still encode this pairing forty-eight years after the indicators were first published, because the failure mode of SAR alone is so reliable that the ADX gate is the obvious patch.

On SPY, range-bound months are easy to identify by the SAR pattern: the dots flip from one side of price to the other every few bars, generating a continuous sequence of small whipsaws. Those are precisely the months where a pure SAR system bleeds. A trader running SAR with an ADX filter sits out those periods entirely and reactivates when ADX reclaims 25.

Other failure modes

Two other regimes punish the parabolic SAR. Choppy reversals after earnings or macro events generate large overnight gaps that flip the system on the open, often at the worst possible price. And extremely volatile single-day spikes inside an otherwise trending stock can clip the SAR and reverse the system right before the original trend resumes. Both cases share a common feature: an abrupt break of structure that does not actually mark the end of the prevailing trend.

Common Mistakes

Trading every flip as a reversal entry

The most expensive mistake. The indicator's name says Stop and Reverse, and a naive reading suggests every flip is both an exit and a new entry. In trending markets that interpretation occasionally works. In aggregate across all market regimes, the mechanical reverse rule produces negative expected value, because the bulk of flips happen in non-trending conditions where the opposite direction is also wrong. Use the flip as an exit. Look elsewhere for the next entry.

Running SAR without a trend filter

A close second. Parabolic SAR alone, applied to any liquid market with no regime filter, generates roughly as many losing whipsaw sequences as winning trend captures. Net result is often negative after costs. Adding the ADX-greater-than-25 filter transforms the performance because it removes the regime in which the system structurally cannot work. Skipping the filter and then concluding the indicator is broken is a category error.

Using SAR on very short intraday timeframes

The default 0.02 / 0.20 parameters are tuned for daily charts. Applied to one-minute or five-minute bars without adjustment, the AF ratchets too aggressively relative to the noise, and flips occur every few bars. Day traders who want to use SAR intraday typically lower the step to 0.01 and cap the AF at 0.10, which slows the system enough to be usable on five- and fifteen-minute charts. Even then, intraday SAR is for trend-day setups, not for chop.

Ignoring overnight gaps

The SAR dot is calculated from the prior bar's close. A gap open that crosses the dot triggers the stop at the open price, which may be substantially worse than the dot itself. Traders who hold overnight should treat SAR as a guide rather than a precise execution level. The mechanical exit will fill somewhere on the gap; how far from the dot depends entirely on the size of the gap. Risk planning should assume the worst-case fill, not the dot level.

Over-tuning the parameters

Backtesting SAR with step 0.03 and maximum 0.25 on a single ticker over a single five-year window and finding better numbers does not mean the new parameters generalize. The 0.02 / 0.20 defaults survived decades of out-of-sample testing across futures, equities, and currencies precisely because they hold up across regimes. Custom parameters that beat the defaults on one dataset almost always underperform on the next. Risen surfaces parabolic SAR alongside ADX, ATR, and the rest of the Wilder family on every U.S. ticker chart, with sensible defaults preserved.

Keep Learning