Where to put a stop loss: structure, buffers and when to move it
Where to put a stop loss: just past the level that proves your trade idea wrong, plus a small buffer. How to find that level, size it and manage it.
Key takeaways
- Put the stop just past the structure your trade is built on, plus a 0.2 ATR buffer, not at a round number or a random percentage.
- Skip trades where the stop is closer than 0.25 ATR or further than 4 ATR, instead of nudging the stop to fit.
- Move the stop toward profit, never away from it: to entry after target 1, and close stale trades after 32 candles.
In this article
Put your stop loss just past the price that proves your trade idea wrong, such as below the swing low or support zone the trade depends on, plus a small buffer for noise. Then size the position from that stop, not the other way round.
That one sentence covers most of it, but each part hides a decision. Which level proves you wrong? How big a buffer? What if the stop ends up very close or very far? And what do you do with it once the trade is moving? This guide works through each question with the rules the setups on this site use.
What is a stop loss for?
A stop loss is not there to limit how much you lose in dollars. Position size does that. The stop marks the price where your reason for being in the trade no longer holds. If you bought because support held, the stop goes where support has clearly failed. If you bought a breakout, it goes where the breakout has clearly failed.
Seen that way, a stop is a statement about the chart: “if price trades here, I was wrong.” That makes it easy to accept when it is hit. You were not unlucky. The market showed you the idea did not work, and you lost the amount you planned to lose.
Once the stop is set, the dollar risk is controlled by the position size. That part is covered in how much to risk per trade.
Structure-based stops vs percentage stops
There are two common ways to choose a stop.
A percentage stop sits a fixed distance from entry, say 2% or 5%, on every trade. It is easy to apply, but it ignores the chart. Sometimes 2% lands in the middle of normal noise. Sometimes it lands right on the low other traders are watching.
A structure-based stop sits just past the level the trade depends on: the swing low under a pullback, the bottom of a support zone, the wick that swept a low. The distance changes from trade to trade, and the position size changes with it.
Structure-based stops
- Tied to the reason for the trade, so a hit means the idea really failed.
- Adapt to quiet and wild markets on their own.
- Easy to review in a journal: you can see if the level was a good one.
Percentage stops
- Ignore the chart, so they often sit inside the structure.
- Too tight on volatile coins and too loose on quiet ones.
- Get hit on retests that the trade idea allowed for.
Example. BTC sweeps a low at $61,200 on the 1-hour chart and closes back up at $62,000. A 1% stop would sit at $61,380, which is above the low that was just swept. A simple retest of that low would stop you out even though the idea, that the sweep marked the bottom, is still intact. The structure-based stop goes below $61,200.
How big should the buffer be?
Placing the stop exactly at the low is not enough. Obvious lows, round numbers and the previous day's low attract other traders' stops, and price often pokes just past them before turning. Those pokes are called sweeps, and whole setups are built around them, such as the SMC sweep and change of character and the previous day sweep. If those setups can profit from a sweep, a stop sitting right at the low is on the other side of that trade.
So the stop gets a buffer. The setups on this site use 0.2 ATR. ATR (average true range) is the typical size of a candle on that chart, so the buffer grows on volatile coins and shrinks on quiet ones.
Worked example, continued. On the 1-hour chart the ATR is $400. The buffer is
0.2 × $400 = $80.
Stop = $61,200 − $80 = $61,120. Entry is $62,000, so the stop is $880 away, which is 2.2 ATR.
Risking 1% of a $10,000 account ($100): $100 ÷ $880 = 0.1136 BTC, or about $7,045 of exposure.
Round-trip fees at 0.1% each way are about $14, which is 0.14R. That is under the 0.3R cap, so the trade passes.
Round numbers deserve extra care. If a low sits at $60,000, many stops will be at $59,990 or $59,900. Go past the cluster, not into it. The buffer usually does that. If it does not, the stop probably needs to sit a little lower, and the position gets a little smaller.
When is a stop too tight or too wide?
A structure-based stop can still be a bad stop if it is very close or very far. The setups here only accept a stop between 0.25 and 4 ATR from the entry. With an ATR of $400 that means between $100 and $1,600 away.
Too tight
A stop closer than a quarter of a normal candle sits inside ordinary movement. It gets hit even when the idea was right. Tight stops also mean large positions, and fees are charged on the position, so a tight stop can lose a big slice of R to fees alone.
Too wide
A stop more than 4 ATR away is a different trade. It needs a long time to play out and a much smaller position to keep the same dollar risk. The targets also have to be far away to pay at least 1.5R, which makes them less likely to be reached in time.
When a stop falls outside the band, skip the trade. Do not nudge the stop inside the band to make it fit. A stop moved for convenience no longer marks where the idea is wrong.
How and when should you move a stop?
Moving a stop is fine in one direction only: toward profit. The rules here use a simple plan.
To entry after target 1
Target 1 must be at least 1.5R from entry. When price reaches it, half the position comes off and the stop on the rest moves to the entry price. The remaining half can no longer turn into a full loss. The details are in managing a setup.
Worked example, continued. R is $880. Target 1 is a swing high at $63,500, which is $1,500
away, or 1.70R. Target 2 is the previous day's high at $64,400, which is $2,400 away, or 2.73R.
Half the position (0.0568 BTC) comes off at target 1 for about $85, which is 0.85R. The stop moves to $62,000.
If the rest is stopped at entry, the trade ends at about +$85 before fees.
If the rest reaches target 2, it adds about $136, for a total of about +$222, or 2.2R before fees.
Trailing stops
A trailing stop follows price as the trade moves: for a long, under each new higher swing low, or under a moving average. It can keep you in a strong trend far longer than a fixed target would. The cost is that trails give back part of the move when they are hit, and a trail that is too close behaves like a too-tight stop. Slower trend systems, such as trend following, often exit on a moving average instead of a target. Whatever you use, decide it before the trade.
Time stops
Price is not the only reason to get out. A setup is a bet that something happens soon. The setups here close whatever is still open after 32 candles: 32 hours on the 1-hour chart, about 5 days on the 4-hour chart. If neither the stop nor a target is reached by then, the reason for the trade has faded.
Never widen a stop
Moving a stop further away when price comes near it is the most expensive habit in trading. It turns a planned 1R loss into 2R or 3R, and it usually happens exactly when the idea is failing. If you feel the urge to widen a stop, take that as the sign the trade is over.
Common mistakes
- Stopping exactly at the obvious low. That is where everyone else's stops are, so a routine sweep takes you out before the move you were right about.
- Choosing the stop from the position size. Picking how much to buy first and then fitting a stop to it puts the stop somewhere that means nothing on the chart.
- Using the same percentage on every coin. A 2% stop is wide on a quiet large coin and inside the noise on a volatile small one.
- Widening the stop as price approaches it. This turns small, planned losses into the large ones that do real damage to an account.
- Moving to entry too early. Jumping the stop to entry before target 1 often gets you stopped on a normal pullback just before the trade works.
- Not placing the stop at all. A stop that only lives in your head will not be there when price gaps through it while you sleep.
What to check
- What proves the trade wrong? Name the level: the swept low, the bottom of the zone, the pullback low. If you cannot name it, you do not yet have a trade.
- Is the stop past the obvious spot? Add the 0.2 ATR buffer below the low, and make sure the stop is not sitting just under a round number where stops cluster.
- How far is the stop in ATR? Divide the distance from entry to stop by the ATR. Between 0.25 and 4 is workable. Outside that band, skip the trade.
- Did you size from the stop? Divide your dollar risk by the stop distance. The stop sets the size, not the other way round.
- Do fees fit? Round-trip fees divided by your dollar risk should be 0.3R or less. Tight stops on short timeframes often fail this.
- Is target 1 at least 1.5R away? If the stop is fine but the target is too close, the trade does not pay enough for its risk.
- Do you know how the stop will move? Write down when it goes to entry, whether you will trail, and the candle on which the trade closes on time.
Stops only work as part of a full routine. See what to check before any trade, and for the levels stops are built around, support and resistance. Nothing here is financial advice. Crypto is volatile, and a stop can fill worse than its price in a fast market.
