ATR Trailing Stop Strategy

The ATR Trailing Stop Strategy uses Average True Range (ATR) to place a stop level that follows price by a volatility-scaled distance, widening when the market is choppy and tightening when it calms down. Traders use it mainly as an exit and trade-management rule, letting the stop ratchet in the direction of an open position while never moving against it. It is a tool for defining when a trade is "no longer valid" on your terms; it is not predictive and does not guarantee any outcome, and trading always carries the risk of loss.

How it works

Average True Range measures how much an instrument typically moves over a chosen lookback period, so it adapts to changing volatility instead of using a fixed point/pip distance. The strategy multiplies ATR by a factor (commonly around 2 to 3.5) to get a "trailing distance," then keeps a stop line that distance away from a reference price. For a long position the stop sits below the reference and is only allowed to move up (or stay flat) as price advances; for a short position it sits above and only moves down. The key mechanic is the ratchet: once the stop tightens, it never loosens again while the position is open. The reference can be the current close (as in the example below) or the highest high / lowest low since entry — the latter is the "chandelier" variant, which anchors the trail to the extreme rather than the close. When price closes through the trailing line, the strategy treats that as the exit signal. Because the distance is volatility-based, the same multiplier gives a wider stop in turbulent conditions and a narrower one in quiet ranges, which is the intent: it aims to sit outside ordinary noise while still defining a clear invalidation level. It says nothing about whether a trade will work out; it only standardizes how and when you step aside. Want to confirm your version behaves the way you intend? Paste your strategy into the ForexCodes Strategy Validator to get validated Pine Script v6 plus an automatic check for look-ahead bias, repainting, and intent mismatches before you take it live.

Entry rules

This strategy is fundamentally an exit/trade-management overlay, so entries are typically supplied by a separate signal (trend filter, breakout, moving-average cross, or your own discretionary read).
A common pairing: take long entries only when price is above a longer-term moving average (e.g. 200-period) and short entries only when below, so the trailing stop manages trades in the prevailing direction.
On a fresh entry, initialize the trailing stop at the current ATR distance from the reference price (below for longs, above for shorts) and record it as the active stop.
Decide and document your direction bias before entry; do not retro-fit the trailing stop to justify a trade after the fact.
Evaluate ATR and trailing levels on confirmed (closed) bars to avoid acting on an unconfirmed, still-forming candle.

Exit rules

Long exit: close the position when price closes below the long trailing stop line.
Short exit: close the position when price closes above the short trailing stop line.
Each bar, recompute the candidate stop as the reference price minus (ATR x multiplier) for longs, or plus (ATR x multiplier) for shorts.
Ratchet rule: the long stop is updated only to the higher of the previous stop and the new candidate (it never drops); the short stop only to the lower of the previous and new candidate (it never rises).
Optionally combine with a time-based or opposite-signal exit so a position is not held indefinitely in a slow drift that never triggers the trailing line.

Risk notes

ATR scales position-management distance but does not size your position; decide risk per trade separately and never risk more than you can afford to lose.
A wider multiplier means a more distant stop and therefore a larger potential loss per trade if it is hit; a tighter multiplier reduces that distance but increases the chance of being stopped out by noise.
Stops can be jumped over by gaps, news, or thin liquidity, so the actual exit price may be worse than the trailing line (slippage).
Volatility-based stops adapt to recent conditions only; a sudden regime change can leave the stop too wide or too tight relative to what is actually happening.
Nothing here is predictive. Past behaviour of ATR or any backtest is illustrative only and past performance does not indicate future results. Trading involves the risk of loss.

Pine v6 example

//@version=6
strategy("ATR Trailing Stop Strategy (Illustrative)", overlay = true, default_qty_type = strategy.percent_of_equity, default_qty_value = 10)

// --- Inputs ---
atrLen  = input.int(14,  "ATR Length", minval = 1)
atrMult = input.float(3.0, "ATR Multiplier", minval = 0.1, step = 0.1)
maLen   = input.int(200, "Trend MA Length", minval = 1)

// --- Indicators (built-in namespaces) ---
atr = ta.atr(atrLen)
ma  = ta.sma(close, maLen)

// --- Direction bias for entries (example filter, not predictive) ---
longBias  = close > ma
shortBias = close < ma

// --- Trailing stop state (persists across bars) ---
var float longStop  = na
var float shortStop = na

// Candidate stop levels for this bar (close-referenced variant)
longCandidate  = close - atrMult * atr
shortCandidate = close + atrMult * atr

// --- Manage long trailing stop (ratchets up only) ---
if strategy.position_size > 0
    longStop := na(longStop) ? longCandidate : math.max(longStop, longCandidate)
else
    longStop := na

// --- Manage short trailing stop (ratchets down only) ---
if strategy.position_size < 0
    shortStop := na(shortStop) ? shortCandidate : math.min(shortStop, shortCandidate)
else
    shortStop := na

// --- Entries (managed by the trailing stop afterwards) ---
if strategy.position_size == 0 and longBias
    strategy.entry("Long", strategy.long)
    longStop := longCandidate
else if strategy.position_size == 0 and shortBias
    strategy.entry("Short", strategy.short)
    shortStop := shortCandidate

// --- Exits on a confirmed close through the trailing line ---
if strategy.position_size > 0 and not na(longStop) and close < longStop
    strategy.close("Long", comment = "ATR trail hit")
if strategy.position_size < 0 and not na(shortStop) and close > shortStop
    strategy.close("Short", comment = "ATR trail hit")

// --- Plots ---
plot(strategy.position_size > 0 ? longStop  : na, "Long Trail",  color = color.green, style = plot.style_linebr)
plot(strategy.position_size < 0 ? shortStop : na, "Short Trail", color = color.red,   style = plot.style_linebr)
plot(ma, "Trend MA", color = color.new(color.blue, 50))

Pitfalls

Reading the stop off the forming bar: if you act on intrabar values, the trailing line can move and re-trigger within the same candle, which looks fine in hindsight but is not reproducible live. Evaluate on bar close.
Letting the stop loosen: if your code uses the raw candidate instead of math.max/math.min against the prior stop, the 'trailing' stop will drift back toward price and stop trailing. The ratchet is the whole mechanism.
Repainting illusions: trailing levels that reference future-looking constructs (or are plotted with lookahead) can appear to have caught every reversal. Confirm there is no look-ahead bias before trusting any backtest.
Mismatching the reference price: a close-referenced trail and a highest-high (chandelier) trail behave differently. Make sure the code matches the variant you describe, or the documented intent will not match the chart.
Over-tuning the multiplier to a single instrument/timeframe so the curve looks clean, then seeing very different behaviour elsewhere. Treat the multiplier as a robustness question, not an optimisation target.
Forgetting slippage and gaps: a backtest that fills exactly at the trailing line overstates how tightly the stop is honoured in real, illiquid, or news-driven markets.
Treating the trailing stop as an entry signal. It manages risk on an existing position; it does not tell you when an entry exists.

FAQ

Is the ATR trailing stop a complete strategy on its own?

Not really. It is an exit and trade-management rule. You still need an entry method and a direction bias (the example uses a moving-average filter). It standardises how you leave a trade and define invalidation; it does not decide which trades to take, and it makes no claim about outcomes.

What ATR length and multiplier should I use?

There is no universally correct setting. A 14-period ATR with a roughly 2 to 3.5 multiplier is a common starting range, but the right values depend on the instrument, timeframe, and how much noise you are willing to sit through. A larger multiplier gives more room (and a larger potential loss if hit); a smaller one tightens the stop but increases the chance of being shaken out. Test across conditions rather than tuning to one chart.

Why did my trailing stop seem to catch every top and bottom in the backtest?

That is usually a warning sign, not a feature. It often comes from reading values off unconfirmed bars or from look-ahead/repainting in how the level is computed or plotted. A stop that 'knew' the high in advance cannot be reproduced live. This is exactly the kind of intent mismatch and look-ahead bias the ForexCodes Strategy Validator is built to catch.

Want this as a clean, validated Pine script? run it through the Validator → or convert a trading strategy to Pine Script →

Educational & software only — not financial advice, not a recommendation to trade. Backtests are illustrative; past performance does not predict future results. Trading involves substantial risk of loss.