Where to place a stop loss has nothing to do with 2%. It has nothing to do with any number you picked before you looked at the chart. The stop belongs at the price where the reason you took the trade stops being true. Not before it, not well past it. Right there.
I used to size stops the other way. Pick a percentage, apply it to the entry, done in four seconds. It felt disciplined because it was a rule. It wasn’t a rule. It was a number with no relationship to the chart underneath it. Every time I ask myself where to place a stop loss now, I ask a different question than I used to. Not “how much am I willing to lose.” Instead: at what price does the reason for this trade stop being true.
That question sounds small. It isn’t. It changes where the stop goes on almost every trade I take, and it’s the difference between a stop that protects the idea and a stop that just protects a percentage on a spreadsheet.
A percentage stop doesn’t know what the chart is saying
Say you buy a stock at $60 and your rule is a 2% stop. That puts you out at $58.80. Now say the level that actually justified the trade — the floor that held twice before, the reason you clicked buy, the thing that answers where to place a stop loss on this trade — sits at $58.90. Your stop is inside the level’s own noise. A normal wick down to $59.10, completely unremarkable behavior for that exact level, doesn’t touch you. But a wick to $58.85 does, and that wick tells you nothing. The level never failed. You just handed the trade to a number you picked before you’d looked at where the level actually lives.
Flip it around. Same stock, same 2% rule, but this time the real level sits at $59.60. Price closes at $59.30 — a clean close below the floor, the kind that actually says the level gave way. This is the clearest case for why to place a stop loss by the level instead of the percentage: your 2% stop at $58.80 is still open. You’re down another sixty cents on a trade that already told you it was wrong, just waiting for the percentage to catch up to a decision the chart already made.
Both of those are the same mistake wearing different clothes. A percentage doesn’t read the chart. It reads your account size. Those are two different things, and only one of them is on the screen in front of you. Once you accept that, the question of where to place a stop loss stops being a math problem and becomes a reading problem — you’re reading the level, not calculating a percentage.
Where to place a stop loss when the level is the entry
If a level is the reason you’re in the trade, the level is also the thing that has to fail for you to be wrong. So the stop goes just past it — far enough to survive the noise that level normally produces, not one cent further. This is the entire method, and everything else in this article is detail on how to measure that distance instead of guessing it.
That means before I enter anything, I look at how that exact level, or a nearly identical one on the same ticker, has behaved recently. Does price tend to poke half a point below it and snap back? Does it tend to wick a full point through before reversing? That range — the normal noise around that specific level, on that specific stock — is what tells me how far past it my stop needs to sit. Not a formula. A pattern I can point to on the same chart, usually from the last few times that level or one like it got tested — the same pattern that answers where to place a stop loss on the next trade at this level.
A level that’s failed cleanly looks different from a level that’s just breathing. Noise is a wick that reverses inside the same candle or the next one, a close that comes right back above the line. Failure is a close that holds below it, and then the next candle holding there too instead of snapping back. If I’ve watched that level get tested twice already and both times the noise stayed inside forty cents, then forty-five or fifty cents past the level is my room. Not eighty. Not fifteen. That measured gap is where to place a stop loss on this exact setup.
How much room is “enough” versus “too much”
This is the part that trips people up once they’ve accepted that percentage stops answer the wrong question about where to place a stop loss. They swing the other way and give every trade a wide berth “just in case,” which is its own version of not reading the chart. Enough room means: past the wick highs or wick lows this level has already produced, so ordinary noise can’t touch you. Too much room means: past a second, unrelated level that was never part of your reason for the trade. Somewhere between those two extremes is where to place a stop loss correctly.
If a floor has been tested three times and the deepest wick below it was thirty cents, a stop forty cents below the floor has room for noise. A stop a dollar fifty below it doesn’t have “extra safety” — it has slack that costs you money on every loss and doesn’t buy you anything, because a move that far past the level was never going to be a normal test of it. By the time price is a dollar fifty through your level, the level already failed. You’re just paying to find out later than you needed to — which is another way of saying you never actually decided where to place a stop loss, you just delayed the decision.
The same logic works in reverse for a resistance level you’re shorting into, or a support level under a call option’s underlying. The direction changes. The measurement doesn’t. Look at what the level has already shown you, and place the stop at the edge of that, not past a nearby number that happens to look tidy. That edge is where to place a stop loss on any level-based entry, long or short.
The trade where I got the room right
AAPL had a floor at $221 that had already been tested twice in the two weeks before — once with a wick down to $220.55 that closed back at $221.60, once with a wick to $220.70 that closed at $221.90. Both times, the noise stayed inside about fifty cents of the level. That was my sample, and it’s what told me where to place a stop loss before I ever clicked buy.
I bought at $221.80 on the third approach and put my stop at $220.35 — about sixty-five cents below the level, forty-five past it, a little beyond what the two prior tests had shown me but not by much. Price dipped to $220.90 that same afternoon, inside the noise range I’d already seen twice, and closed the day at $222.70. Two sessions later it was at $226.40. The stop never came close, because it wasn’t guessing. It was sitting exactly where the chart’s own history said noise ends and failure begins.
The trade where I learned this the expensive way
Before I traded this way, I had a position in a mid-cap, entry at $34.20, level underneath at $33.80. I put my stop at $34.05 — fifteen cents below entry, twenty-five cents above the actual level. It felt tight and disciplined. It was neither. It was inside the level’s noise, not past it — proof that feeling disciplined and knowing where to place a stop loss are not the same skill.
Price wicked to $33.95 an hour later, took my stop, and closed that same candle at $34.55. By end of day it was at $35.10. Nothing about the level had failed. My stop had never been past it — it was sitting on top of the level’s normal breathing room, and normal breathing room is exactly what took it out. I didn’t lose because I was wrong about the trade. I lost because I placed the stop somewhere the chart was always going to visit on a good day. That trade is the reason I stopped guessing at where to place a stop loss and started measuring it off the level’s own recent behavior instead.
The round-number trap
One more thing worth naming here, briefly, because it compounds the problem of where to place a stop loss even after you’ve started measuring levels correctly. A lot of traders don’t just place stops too tight — they place them on the nearest round number. $220.00 instead of $220.35. $34.00 instead of $34.05. Round numbers are where everyone’s stop clusters, and clusters of stops are exactly the kind of liquidity a level can get swept for before it reverses. That’s a topic on its own — I’ve written about the sweep mechanics elsewhere — but the short version is: if your stop sits at a price a thousand other traders would also pick by habit, you’re not measuring the level anymore. You’re measuring the crowd instead of measuring where to place a stop loss based on what the level itself has shown you.
What this actually gives you
A stop placed this way isn’t smaller because you’re being conservative and it isn’t bigger because you’re giving the trade “room.” It’s the size the chart says it needs to be, no more. Sometimes that’s twenty cents. Sometimes it’s a dollar twenty. The number changes every time because the level changes every time. That’s the whole point — the stop answers to the chart, not to a spreadsheet formula you decided on before you opened it.
Most of what I do now is exactly that measuring, done before the trade exists. By the time I’m in, the stop is already sitting where the level’s own history said it should, and there’s nothing left to decide while price is moving. Where to place a stop loss was never really a risk question. It was a reading question, answered before the trade started, by a version of me that still had the patience to look.
I trade with a blank chart and a few lines in Static, the free daily chart room run by Draw Lines Make Money. If this way of thinking about where to place a stop loss makes sense to you, you can sit in and watch how it’s done live:
Join the free Static chart room →
Disclosure: that’s an affiliate link — I may earn a commission if you join a paid tier later, at no extra cost to you. The free room is free.