Where to place your stop loss, and why not elsewhere
Updated on
The short answer
A stop answers a question, not an amount: where does my idea become wrong? It therefore goes on the far side of the level that would invalidate your reading, and beyond the instrument’s ordinary range so that you are not taken out by breathing. The most widespread mistake is doing the opposite: deciding you want to risk fifty euros, then placing the stop at the matching distance. It is the size that adjusts to the amount, never the stop.
A stop is not chosen by amount
This is the inversion that costs most, and it looks perfectly logical: I want to risk fifty euros, so I put my stop where I lose fifty euros.
The problem is that the market does not know your budget. A stop placed at that distance almost always falls inside the instrument’s normal movement, and you are taken out for nothing, often just before your idea plays out.
The correct order is the reverse: the distance comes from the chart, and it is the size that then adjusts to bring the loss back to fifty euros. The full calculation is on our page about risk per trade, and its consequences from one instrument to another on the one about position size by instrument.
The two markers that work
They combine, and the stop goes beyond both, not beyond one of them.
The first is structural. Your idea rests on something precise: a low that held, a high that broke, an area price left from. If that level gives way, the reason for the trade is gone. So the stop goes on the far side of that level, and nowhere else.
The second is the ordinary range. Every instrument and every timeframe has a usual breathing, measurable with any average-range indicator. A stop placed inside that breathing will be hit by noise, without anything having happened.
When the two contradict each other, meaning the structural level is closer than ordinary noise, it is not the stop that should be brought in: it is the entry that is badly chosen, or the timeframe that does not fit.
Why never exactly on the level
Two reasons, one human and one mechanical.
The first is that the level is visible to everyone. That is precisely where stops pile up, so precisely where price often goes to take a wick before turning. Leaving a margin beyond the level costs a little distance and avoids a great many pointless exits.
The second is that a stop is not triggered on the price you see. A buy is closed at the selling price, so its stop is measured against the bid; a sell is closed at the buying price, so against the ask. On a sell, the stop is therefore reached one spread earlier than the chart suggests, which surprises everyone the first time. The detail of the two prices is on our page about order types.
The places not to put it
Three locations are systematically bad, for the same reason: everybody thinks of them.
Just below a round number, where orders visibly pile up. Just beyond the previous day’s extreme or the previous session’s, one of the most watched markers in the market. And at a round distance from the entry price, twenty points exactly and so on, which has no relation to what the instrument does.
A fourth case is not a location but a moment: a tight stop placed just before an economic release protects nothing, since price jumps from one level to another without passing through the prices in between. That is explained on our page about the economic calendar.
Once placed, it is never moved away
Moving a stop further away mid-trade is the decision that turns a small loss into a large one, and it is almost always the same story: price approaches, you convince yourself the level will hold, you shift it a few points, then again.
Bringing it closer is another matter. Bringing your stop back to the entry price once the trade is under way, or trailing it behind price, are legitimate exit decisions, provided they are decided before entering and not in the emotion of the moment.
Either way, note it down. A trade whose stop has moved is no longer the one you planned, and counting it as such distorts any review, as our page about the trading journal explains.
What AlphaGPT does
Every plan offered carries a stop, and it is from the distance between the entry and that stop that the size is computed on your real equity. The two go together: if you decide to place your stop elsewhere, the size no longer holds.
The spread is counted in that distance, which avoids sizing on a theoretical gap more favourable than reality.
What the tool does not do: it does not move your stops, does not bring them to break-even automatically and does not trail price for you. Those actions happen in your platform, and they remain decisions that are yours.
Trading carries a risk of capital loss.