A stop-loss is the most misunderstood order in retail trading. Beginners place it at a round dollar amount they would not mind losing, conspiracy theories treat it as bait for “stop hunters”, and a surprising number of traders skip it entirely in favour of a promise to themselves. This sheet is the working procedure: where the stop belongs (just beyond the level that proves your idea wrong), how much room it needs (enough that normal noise does not take it), why the in-your-head version fails exactly when it matters, and what even a well-placed stop cannot promise you.
The stop is your invalidation point made executable
Every trade worth taking has a reason — a level held, a trend resumed, a range floor bought. If the reason can be stated, it can be falsified: there is some price at which the reason is simply wrong. The stop-loss is that statement turned into an order. It is not where losing starts to hurt; it is where the idea stops being true. The two questions sound similar and produce completely different stops.
This is why the popular method — “I'll risk $50, so my stop goes 10 pips away” — is backwards. The market does not know your budget. If the level that invalidates your idea sits 40 pips away and your stop sits at 10, ordinary noise will take you out while the idea is still intact, over and over. The dollar budget is honoured somewhere else entirely: position size. The chart sets the distance; the size makes the distance affordable.
The procedure, start to finish
- State the trade idea in one sentence, including the level that would prove it wrong — for a long position, usually the structure low the entry depends on.
- Place the stop just beyond that level, not on it. Add a buffer for the spread and for ordinary noise around an obvious price.
- Measure the distance from entry to stop, in pips.
- Derive the position size from that distance, so the loss at the stop equals your planned risk — 1% of equity in this survey's examples.
- Enter the stop as a real order on the platform, attached to the position, at the same moment the trade is opened.
- Leave it alone. Moving a stop is allowed in one direction only: toward less risk, never toward giving a losing trade more room.
Worked through on the survey's anchor: you buy EUR/USD at 1.0850 because the price has held a floor at 1.0815 twice. The idea is wrong if that floor breaks, so the stop goes a buffer below it — say 1.0808, making the distance 42 pips. On a $10,000 account risking 1% ($100), the sizing formula prints $100 ÷ (42 × $10) = 0.23 lots. If the stop is hit, the loss is about $97. If the price instead travels the same 42 pips in your favour, the gain is the same $97 before costs — that symmetry is the whole point of sizing from the stop.
Structure: beyond the level that proves you wrong
Why “just beyond” rather than “exactly on”? Because obvious levels are where everyone's orders cluster — buyers defending the floor, stops from earlier longs resting beneath it. Prices probe such levels, overshoot by a few pips, and come back, frequently enough that a stop placed on the line itself donates money to ordinary churn. The buffer does not need to be heroic: a handful of pips on a major pair, more on noisier instruments. What it must do is move your exit from “the level was touched” to “the level actually failed”.
The same logic chooses between candidate levels. A stop under the most recent minor dip is cheap but fragile; a stop under the structure the whole idea rests on is wider but meaningful. Prefer the level whose break would genuinely change your mind, and handle the wider distance the only correct way — by reducing size, not by abandoning the level that matters for one that flatters the size you wanted to trade.
Volatility: room to breathe without donating
Structure says where the idea fails; volatility says how much the price wobbles on its way to anywhere. A pair that covers 80 pips on a typical day will routinely cross 20 pips of pure noise, which means a 15-pip stop on such a pair is not a risk decision — it is a donation schedule. Before fixing the distance, look at the instrument's recent daily range. If your structural stop sits inside the band the pair crosses on an average quiet day, either the entry needs to be closer to the level or the trade is not there.
Volatility also changes: the same pair needs more room in a news week than in a holiday lull, so a fixed pip number carried from trade to trade is wrong in both directions at different times. A trailing stop — one that follows a profitable position at a set distance — inherits the same rule: trail it tighter than the noise band and ordinary backfill will close a position whose idea is still working.
The mental-stop autopsy
A “mental stop” is a stop that exists as a promise: I will close manually if it reaches 1.0808. The mechanism of failure is not subtle, and almost everyone who trades long enough performs this autopsy on themselves once. The promise is made by a calm person before the trade. It must be kept by a stressed person during it — watching a losing position, feeling the loss as a verdict on themselves, with a brain that is exceptionally good at generating one more reason to wait. “It's about to bounce.” “The level hasn't really broken.” “I'll close on the next candle.” The price that was supposed to end the trade arrives, and instead of an exit it starts a negotiation.
The platform order has none of this machinery. It does not feel the loss, does not remember that the last two stop-outs reversed, does not hope. The honest comparison is never between mental stops and hard stops in the abstract; it is between your discipline at its worst moment and software. Take the software. And if you find yourself preferring the mental version because hard stops “keep getting hit”, read that message correctly: the stops are too close — a placement problem, covered in the two sections above — not an argument for removing the safety equipment.
Stop-hunting: what is real, what is paranoia
The fear that someone can see your stop and will trade to take it deserves a straight answer. What is real: stops cluster at obvious places — under round numbers, under widely watched swing lows — and large participants know it. A pool of resting stops is a pool of liquidity, and prices do get pushed into such pools, trigger the orders, and reverse. That is a property of order flow in every market, and it is the practical reason this sheet says beyond the level, not on it.
What is paranoia: the idea that a regulated broker is watching your individual 0.23-lot stop and moving a multi-trillion-dollar market to collect it. Retail stops at a single broker are rounding errors against the flow in a major pair. Protection against the real effect costs a few pips of buffer and slightly less obvious levels. The imagined effect has a popular but ruinous “protection” — trading without stops — which exchanges a few pips of occasional hunting for the open-ended losses the rest of this sheet exists to prevent.
The honest caveat: gaps, slippage, and what a stop cannot promise
When the price touches your stop, the platform sends a market order — and a market order fills at the best available price, not at a promised one. In normal conditions on a major pair the difference is a pip or less. Around major news, or in thin hours, the fill can land further away: that difference is slippage. The harder case is a gap — the price jumping across your stop level without ever trading at it, most familiar over the weekend close. A stop at 1.0808 in front of a Monday open at 1.0780 fills near 1.0780: in the worked example above, the planned $97 loss becomes roughly $160. The plan said 1%; the market delivered 1.6%.
The conclusions are practical, not dramatic. The stop bounds your intended loss, and it does so reliably in ordinary conditions — that alone makes it non-negotiable. For the extraordinary conditions: be cautious about holding leveraged positions through weekends and scheduled news, and size every position so that even a fill well past the stop is survivable. The routine case and the bad case both argue for the same two habits this sheet began with — real orders, modest size.
Stop distance in, size out — the wider-stop-smaller-size trade-off, live.
Place the order, attach the stop, watch one get hit. Better to meet that feeling where it costs nothing.