Where Should You Put a Stop Loss?
A stop goes at the price that proves the read wrong, which is usually below the swing low the trade rests on or a multiple of ATR from entry. A tighter stop is hit far more often, so the placement decides how frequently you are stopped rather than how much you risk.
The stop loss page covers what a stop is. This one is about the choice nobody makes explicitly: exactly where it goes, and what each answer costs.
How it works
The structural answer: below the level the read rests on.
If the reason for the trade is a sequence of higher lows, then below the last higher low that reason has stopped being true — which is the reversals page’s first condition, used as a stop.
The volatility answer: a multiple of ATR from entry.
Both are defensible and they are different tools. The structural stop asks “where is my read wrong?” The ATR stop asks “how far does this market normally travel?”
Using whichever is closer is choosing the answer you prefer, which is the failure this page exists to name.
What tightness actually costs
Measured across 25 evenly spaced entries on the shared history:
| Stop | Stopped out | Reached target |
|---|---|---|
| 0.5 ATR | 21 of 25 | 4 |
| 1 ATR | 19 of 25 | 6 |
| 2 ATR | 7 of 25 | 16 |
A tight stop is hit three times as often. That is not a subtlety — it is the dominant effect, and it is entirely mechanical: a smaller box around price is entered by ordinary noise.
The entries here were spaced evenly rather than chosen, so no skill is being measured. The table is about the geometry of stop placement, not about a strategy.
And a wider stop is not free. Each of those seven losses is four times the size of one of the twenty-one. The choice moves the loss between frequency and size; it does not remove it.
Below the wick
A wick is a price that already traded and was rejected. Putting the stop above it means putting it somewhere price has demonstrably been able to reach.
Below the wick is the more defensible placement, and it is wider — which is the trade the table above measures.
The obvious level problem
The level you can see is the level everyone can see, so a pool of sell orders sits just underneath it. That is exactly what the liquidity page says price travels toward.
The practical response is a small buffer beyond the obvious price, sized in ATR rather than guessed — not far enough to change the trade, far enough that a sweep of the level does not automatically take you.
The honest caveat: everybody knows this, so the buffer is also somewhat crowded. There is no placement that nobody else thought of.
The stop that moves
Two ways a stop changes after placement, one legitimate and one not, and they are easy to confuse.
A trailing stop moves in one direction only — up behind a long, never down. It converts an open profit into a floor, and every mechanism on this site that does it automatically (Supertrend, Parabolic SAR, a moving average) shares that one-way property.
Moving a stop away from price is the other thing entirely. It looks like patience and it is the removal of the only defined limit on the loss.
The test that separates them is simple: does the move make the worst case smaller or larger? A trail can only shrink it. Anything that enlarges it is not a stop adjustment, it is abandoning the stop.
And there is a third case worth naming: moving to break-even. That shrinks the worst case, so it passes the test — but it also tightens the stop, and the table above measures what tightening does. Break-even stops convert some winners into scratches, which is a real cost paid for a real comfort.
A worked example
Find the level first, before thinking about size. Below the last higher low, below its wick.
Check the distance against ATR. If it is four times a typical bar, this is a wide setup and the position must be small.
Add a buffer if the level is very obvious, a fraction of ATR.
Then size from that distance — the risk per trade page — and if the resulting position is too small to be worth taking, that is the answer for today.
Never move it once placed. Moving a stop away from price converts a defined loss into an undefined one, which is the only genuinely unrecoverable mistake on this page.
The original data
Across our study of 24,971 trading videos, 29 cover stop placement specifically. The median one gets 2,934 views, 86% never pass 50,000, and the median length is 8.7 minutes.
29 videos is one of the smallest fields measured here, and the corpus carries description text for only 4 of them — far too few to say anything about how the topic is written.
What is worth noting is the contrast. The broader stop loss topic carries 139 descriptions, of which 87 mention being wrong — the highest rate in this glossary. The narrow question of where the stop actually goes attracts almost nobody.
When it fails
Stopped out, then right
This is the outcome that makes people move stops, and it is unavoidable. 21 of 25 at half an ATR — most of those were not the market proving the read wrong, they were the market breathing.
The response is a wider stop and a smaller position, not no stop.
The structure is too far away
Sometimes the honest level is far enough away that the position becomes negligible. That is information: the setup is not available at a size worth trading today.
Squeezing the stop closer to justify a bigger position is the exact inversion of the method.
You used whichever was nearer
Two methods, and picking the tighter one each time is not a method. Choose one, write it down, and apply it to the trades you do not like as well.
You picked the low afterwards
Several lows qualify, and the one that looks obviously correct is obvious because you can see what happened next.
Related
Stop loss is what a stop is and why it exists at all.
ATR is the measurement that turns “wide” and “tight” into numbers.
And risk per trade is what converts the distance into a position size, which is where the stop actually becomes risk.
The rule I hold to is that the stop goes at the level, and then the position size adapts to that distance - never the other way round. What I used to do was decide how much I wanted to trade and then find a stop that fitted, which is choosing the answer first. If the level is too far away for a sensible size, the trade is not available today. That is a real answer and it took me a long time to accept it.
— Michael Whitman, from this video
This page is educational, not financial advice. Test every idea on your own charts before risking money.