A stop loss based on market structure is placed at a price level where the chart itself tells you your trade idea is wrong — not at whatever dollar amount your account can tolerate losing. The difference matters more than most traders realize.
Fixed stops — "I'll risk $200 on this trade" — are account-centric, not market-centric. They have nothing to do with how the underlying asset is actually moving. Structure-based stops start from the chart and work backward to position size. That sequence is what separates well-placed stops from ones that get hit on normal volatility before the trade has had a chance to develop.
Market structure refers to the pattern of swing highs and swing lows that define trend direction, and the zones where price has repeatedly reversed or consolidated. These aren't arbitrary lines — they represent levels where a meaningful number of buyers or sellers have previously made decisions. When price revisits those levels, something real tends to happen.
For stop placement, the most relevant structural elements are:
Your stop goes beyond the level that, if broken, invalidates the trade thesis. Not near it — beyond it.
If you're buying into an uptrend or off a support level, your thesis is that buyers are in control. The structural invalidation point is the most recent significant swing low. If price breaks below it, the structure that justified the trade no longer holds.
The practical rule: place the stop a few ticks or a small buffer below that swing low, not directly at it. Price often wicks through clean levels before reversing — a stop sitting exactly at the swing low will get hit on a false break more often than one placed just beyond it. The buffer doesn't need to be large; it just needs to account for normal spread and noise at a key level.
What you're not doing is placing the stop at a round number below your entry because it looks tidy on a risk calculator. If the swing low is $4.30 below your entry and you only wanted to risk $2.00, that's a position sizing problem — reduce the size, not the stop.
The logic inverts for short positions. Your thesis is that sellers are in control and price is likely to move lower. The trade is wrong if price reclaims the most recent swing high — that's the structural level that matters.
A common mistake here is placing stops too tight above a resistance level. Resistance tends to get tested aggressively before it fails or holds, so a stop directly at the prior high is frequently triggered on the initial test. Give it room — the same buffer principle applies.
Not all assets move the same way, and not all market environments are the same. A stop that makes sense for a low-volatility large-cap stock will be absurdly tight for a high-beta small-cap or an options-heavy name during earnings season.
Average True Range (ATR) is useful here — not as the stop itself, but as a sanity check on whether your structural stop is realistic given how the asset normally moves. If the swing low is 0.5 ATR away and you're in a trending, volatile name, expect that level to be tested frequently without meaning much. If the swing low is 2-3 ATR away, you're likely giving the trade appropriate room.
The structural stop sets the location. ATR tells you whether that location makes sense given the asset's typical behavior.
This is where the framework comes together. Once you know where price invalidates your trade idea, you know your risk per share or per contract. From there, you size the position based on how much of your account you're willing to risk on the trade — typically a fixed percentage.
If the structure demands a wider stop than your risk tolerance allows at your preferred position size, you have two options: reduce the size to keep the dollar risk acceptable, or skip the trade. What you don't do is tighten the stop to fit a predetermined size. That approach just guarantees you'll be stopped out before the trade plays out.
Million Candles' Scan Mode identifies structural levels — including swing points and consolidation zones — across charts automatically, which can speed up the process of finding where logical stop placement sits before you've committed to a trade direction.
Structure-based stops aren't static. As a trade develops and new swing points form, the invalidation level changes. In a long trade, each new higher low that forms is a candidate to move your stop up to. You're not trailing by a fixed amount — you're following the structure as it builds.
The rule is simple: only move a stop in the direction of the trade, and only to a new structural level. Moving a stop wider because the trade is going against you is not stop management — it's hope, and it defeats the entire point of having a stop.
Structural stops won't be right every time. Some trades will be stopped out at the swing low and then reverse sharply without you. That's an acceptable outcome — it means the market gave a confusing signal, not that your framework was wrong. The goal is to be out when your thesis is genuinely invalidated, and in when it's holding. Trade sizing, not stop tinkering, is how you manage outcomes over a series of trades. As always, this is educational content meant to explain how traders approach stop placement — it's not a recommendation to make any specific trade or risk any specific amount of capital.