Where the stop goes, and what it decides

The stop comes first

Most people size a position by deciding how much they want to own, then place a stop wherever it feels comfortable. That is backwards, and it is why position sizes drift with mood rather than with risk.

The correct order is fixed:

1. Find the price at which the trade idea is wrong. That is the stop.
2. Distance from entry to stop is 1R, expressed per share.
3. Decide the account risk for this trade — commonly 0.5% to 1%.
4. Divide. That gives share count. You do not get to choose it.

1R per share = entry − stop Risk budget = account × risk % Shares = risk budget ÷ 1R per share

Worked: a $30,000 account risking 1% is a $300 budget. Entry $48.00 with a structural stop at $45.60 gives 1R of $2.40 per share, so 125 shares, a $6,000 position. Move that stop to $47.40 and 1R is $0.60, so the same $300 budget buys 500 shares and a $24,000 position. Same risk, four times the exposure, purely because the structure was tighter.

This is also the trap. A tight stop lets you carry a large position, and a large position on a stop that was tight for cosmetic rather than structural reasons is how a normal loss becomes several. Tighten the stop only when the chart genuinely offers a tighter invalidation, never to justify more shares.

Structure-based stops

Place the stop just beyond the price level that would prove the idea wrong: below the low of the base, below the anchored VWAP that generated the signal, below the consolidation floor, beyond the pattern's defining swing.

The advantage is that the exit means something. When a structural stop fills, the reason you entered has been invalidated, so the exit carries information rather than just recording a bad hour. Signal and invalidation should be defined by the same object wherever possible.

Volatility-based stops (ATR)

Average true range measures how far an instrument typically moves in a period, including gaps. An ATR stop scales the distance to the instrument's own noise level instead of to a fixed percentage.

Stop = entry − (k × ATR14) common k values: 1.5 to 3.0 for swing timeframes

Worked: entry $48.00, ATR14 of $1.20, k of 2 gives a stop at $45.60 and 1R of $2.40. If that stock's ATR later rises to $2.00, the same k puts the stop at $44.00 and the same risk budget buys fewer shares automatically. That auto-adjustment is the point: the position shrinks as the instrument gets noisier, without you having to remember to shrink it.

Percentage stops

A flat 7% or 8% stop is the simplest rule and the weakest one, because it ignores both structure and volatility. A 7% stop on a low-volatility utility is enormous; on a leveraged ETF or a small-cap biotech it is inside a normal Tuesday. It has exactly one virtue: it is impossible to rationalise away, which for a beginner is worth something.

In practice the useful combination is structure first, with an ATR check: if the structural stop is under 1 ATR away, it is inside noise and will be hit at random. Widen it to at least 1 ATR and take fewer shares, or skip the trade.

Where not to put a stop

• On a round number. $50.00, $100.00, and $250.00 attract resting orders from everyone else. Sit a little beyond, not exactly on it.
• Exactly at the obvious swing low. Every chart reader sees the same low. Price routinely trades a few cents through it and reverses. A small buffer beyond the obvious level costs a little R and avoids a lot of undercut-and-rally.
• Inside the noise. Anything under roughly 1 ATR of the entry will be hit by random movement regardless of whether the idea was right.
• At a dollar figure you can tolerate losing. The market does not know your comfort level. If the structural stop is bigger than your tolerance, cut share count, not the stop.
• Nowhere. A mental stop is a stop you will renegotiate at the worst possible moment.

Never move a stop down

Moving a stop further away converts a defined 1R loss into an undefined one, and it always happens under pressure, which is exactly when judgement is worst. It is the single most destructive habit available to a retail trader, because one widened stop can erase the gains from a dozen disciplined trades.

Moving a stop up in a long, to breakeven or behind a trailing structure, is a different action entirely and is fine, provided the rule was written before the trade. The asymmetry is deliberate: risk may only be reduced, never increased, after entry.

Log a moved stop mistake tag every single time it happens. After 50 trades that tag usually explains most of the losses beyond −1R, and that is a rule change rather than a feeling. Journal fields that surface it →

Educational content, not financial advice. No live profit-and-loss figures are claimed anywhere on this site; backtest and walk-forward results are always labelled as such. Full terms: /terms.html

Tools referenced in this guide

• Trader's risk toolkit — position sizer that turns entry, stop, and account risk into a share count.
• TradeLog — logs planned 1R and flags moved-stop trades so the pattern is countable.
• R-multiple & expectancy guide — why every result should be measured in R rather than dollars.