How to Set a Stop Loss (and Why Where You Put It Matters More Than the Amount)

The most common stop-loss question is "what percentage should I use?" The more useful question is "at what price is my reason for holding gone?" Here's the difference — and how to place stops that survive contact with a real market.

Illustration coming soon

article-stop-loss-hero.svg

vector illustration: a chart with an uptrending price line, a small X marking the entry, and a clean horizontal stop line below the most recent swing low; a large percentage symbol faded in the background with a line through it, suggesting the "% question" is the wrong frame.

What a stop loss is — and what it isn't

A stop loss is an order that exits a position if price reaches a level you chose in advance. That's it mechanically. The reason it matters is what it forces you to do: decide, while you're calm and uninvested, at what point you will accept that the trade didn't work.

Three things a stop loss is not:

• Not a guarantee. A stop order is filled at the next available price. If the market gaps past your level — overnight, or on a surprise announcement — you may be filled well beyond it. The stop doesn't prevent loss; it defines the loss you were willing to plan for.
• Not a prediction. Setting a stop at $47.50 doesn't mean you expect the stock to fall to $47.50. It means: at that price, your reason for owning the stock no longer holds, so you're leaving by prior arrangement.
• Not a punishment or a gesture. A stop that's placed at a level you'd never actually act on is not a stop; it's decoration. The level has to be one you're willing to honor.

Seen that way, the question "what percentage should I use?" is the wrong starting point. The right starting point is structural: where does this trade stop being right?

The placement principle: stops go where the trade idea breaks

Every trade is a bet on a specific idea — "this pullback in an uptrend will find buyers," "this tight range will resolve upward," "this channel will hold." The stop belongs at the price where that specific idea is objectively invalidated, not at a round number of dollars you'd feel okay losing.

The cleanest way to see this is through the three setups that dominate swing trading (each covered in depth in The Anatomy of a High-Probability Swing Setup):

• The pullback. You're buying a dip in an established uptrend, expecting support to hold. The idea is broken if price closes through the swing low that defined the pullback — so the stop goes just below that low. The chart tells you exactly where you're wrong.
• The coil (consolidation). You're buying a tight range resolving upward. The stop sits inside the range — typically just below the breakout trigger or the range's low. Because the range is narrow, the stop is close, which is the whole appeal.
• The channel. You're buying near the lower line of an established upward channel. The stop goes just below that line: if price is back inside the channel, fine; if it's decisively below it, the channel's gone.

Notice the common thread: the pattern's shape supplies the stop. That's why traders who work from defined structures find stops easy, and traders who trade "this stock just looks strong" find stops impossible — there's no level where the idea breaks, because there was never a defined idea.

Illustration coming soon

article-stop-placement.svg

teaching diagram, three labeled panels in consistent annotation style: (1) "Pullback": uptrend dipping to a rising moving average, entry arrow at the bounce, stop line under the swing low; (2) "Coil": tightening range, entry arrow at the upside break, stop inside the range; (3) "Channel": rising parallel lines, entry arrow at the lower line, stop just below it. Each panel shows the invalidation level the structure implies.

Why percentage stops and round numbers fail

The two most popular ways to choose a stop — a fixed percentage and a round number — share one flaw: neither is derived from the chart, so neither means anything at the price where it sits.

• Fixed percentages ignore context. In a wide, volatile stock, 3% below entry can be ordinary daily noise — you'll be stopped out before the trade has a chance to work. In a tight coil, 3% might be most of the range, so your stop sits so far away the risk is enormous. The same number is too tight in one name and too wide in another.
• Round numbers are where everyone else is. A stop at $50 under a $52 stock is visible to every other participant looking at the same chart. When many orders cluster at an obvious level, that level is where price can behave abruptly — not because of any single trader, but because of the crowd of orders. Your stop should mark the place your idea breaks, not the place everyone's orders pile up.
• Too tight means death by a thousand cuts. Repeatedly stopped by noise, you pay the spread, the slippage, and the emotional tax of watching trades work after you're out. A stop that's too tight doesn't control risk; it just converts it into guaranteed small losses.
• Too wide means risk you never signed up for. A distant stop makes the per-share risk huge, which either blows your risk budget or forces you into positions too small to matter.

The discipline that resolves all four: measure the distance the chart demands, then convert that distance into size — never the other way around.

From stop distance comes size

This is where stop placement stops being about the chart and becomes about survival. The formula that ties it together:

Position size = (account × risk per trade) ÷ (entry − stop)

Suppose you're willing to risk 1% of your account on the trade. If your entry is $50 and your stop is $47.50, the risk is $2.50 a share, so you can take a meaningful position for a defined cost. Move the stop to $45 because it feels safer, and the risk per share doubles — so either the position halves, or your actual risk quietly doubles. The stop distance doesn't just protect you; it sets how large you can responsibly be.

Two companion rules, covered properly in Swing Trading 101:

1. Risk a small, fixed slice per trade — commonly around 1%. A losing streak is guaranteed to happen; the question is whether it's a bruise or a knockout.
2. Only take trades where the realistic reward is a multiple of the risk — 2R or better. A tight, logical stop is what makes that arithmetic possible: the closer the invalidation point, the smaller the risk you're asking the reward to pay for.

This is why professionals obsess over structures with close invalidation points. Tight stops aren't about being stingy with losses — they're leverage on being right.

Trailing stops: protecting what the trade gave you

Once a trade works, the situation that justified the original stop has changed. The swing low you were protecting is now far below price, and the risk that remains is no longer "the entry was wrong" but "the move is ending." A trailing stop ratchets the exit up as the trade progresses, so a gain that exists is a gain that's protected.

Common methods:

• Below successive swing lows — as the stock makes higher lows, the stop follows them up.
• Below a rising moving average — the same long-term average used for trend context can serve as a moving line of invalidation.
• A fixed distance from price — simpler, but it inherits the same context problem as fixed-percentage entries; it works best as a backstop, not as the primary logic.

One rule makes trailing stops safe: the stop only ratchets one way. It moves up with the trade; it never moves down. A trailing stop that gets loosened "just in case" isn't trailing — it's the original stop being slowly abandoned, which is the beginning of most account disasters.

And set expectations honestly: a trailing stop will always leave money on the table. The point isn't to capture the final point of the move; it's to convert paper gains into decisions made in advance. Missing the exact top is the design working as intended.

The honest limits: earnings and gaps

A stop loss manages gradual adverse movement. It does not manage gaps — the overnight jump, or the surprise announcement that opens price far beyond your level. When that happens, your stop fills at the gap price, and the protection you planned is whatever the market happened to offer.

The professional's answer isn't a tighter stop; it's not being in the position when the gap is likely. The classic case is earnings. A flawless chart setup reporting in three days is a different instrument from the same chart reporting in three months — the report can gap the stock through any stop you can draw. The routine fix is a calendar check: before every entry, look up the earnings date and decide in advance whether this position will be held through it. That decision belongs in the same step as placing the stop, because it defines whether the stop is your exit or merely your backstop.

The process: a stop is a decision made early

Everything above is technique; this last part is the part that decides who survives. A stop loss works because it converts a future emotional decision into a present logical one. The exit is decided when there's no loss yet, no hope yet, no "one more day" yet. That's the entire point.

Two habits to protect:

• Never widen a stop to avoid taking the loss. If the level was right at entry, it's still right now — the only thing that changed is that taking the loss became unpleasant. Widening converts a defined, planned loss into an undefined, unplanned one.
• Journal the stops you honored and the ones you didn't. In a few months the journal will show you which placements fit your actual tolerance — not the tolerance you'd like to have. That personal calibration is worth more than any default setting.

Putting this into practice with Relaxfolio

Stop placement needs two inputs: a structure with a defined invalidation point, and a calendar check for events. Both are research questions you can run in plain English across thousands of US-listed stocks and ETFs:

• "Pullbacks" — uptrending stocks near their moving averages, with distance to support shown; the pullback structure hands you its own stop level (below the swing low).
• "Upward channels" — channel candidates ranked by their position in the channel; the lower line is the natural invalidation.
• "Breakout stocks" — fresh range resolutions with relative volume, where the stop sits inside the range you're buying the break of.
• "Upcoming earnings" — the event check that belongs in the same routine as stop placement, so you're never holding a routine trade into a report you didn't plan for.

Open any single name from a screen and you can review its multi-year chart with 50- and 150-day moving averages — the trend context that tells you where the invalidation genuinely sits, rather than where a percentage rule of thumb would put it. Screens can be saved, with updated results delivered by email for ongoing review. The research is organised for you; the stop decision — where the idea breaks, how much risk it deserves, and whether to be in the position at all — remains yours.

A stop-placement checklist

Before you enter, run through six questions:

1. Is my stop at a level where the trade idea is invalidated — not at a level where I'd "feel okay" losing that much?
2. Is the distance a property of the chart — a swing low, a range, a channel line — rather than a fixed percentage or a round number?
3. Is my position size derived from the stop distance and a fixed risk budget (commonly ~1%)?
4. Have I checked the earnings date, and decided in advance whether this position is held through it?
5. Will I commit to never moving the stop down, and to trailing it only in the direction of the trade?
6. Would I still take this trade if the stop were hit tomorrow — with the loss I planned, and nothing more?

Six honest answers will tell you more than any default setting. The amount you risk matters far less than the reasoning behind it: a stop loss is where your process meets the market, and a process that decides in advance is the only kind that survives contact.

Continue the swing series: Swing Trading 101: Capturing the Meat of the Move, The Anatomy of a High-Probability Swing Setup, and Volume Tells the Truth.

---

See how pullbacks and channels surface structures with defined stop levels — ask "pullbacks" or "upward channels" for a structured view. Explore swing research → · Open Relaxfolio →

This article is for educational purposes only and is not investment advice. All investing involves risk, including possible loss of principal.