Where does a stop logically belong?

A stop has exactly one job: to mark the price at which the reason for the trade no longer exists. Everything else — how much you can afford to lose, how much you want to make, how many shares you'd like — is a different question, answered later.

Two honest ways to locate one

• Structural. Place it beyond the chart feature that defined the setup: below the swing low the bounce was based on, outside the range the breakout escaped, past the trendline the trade was leaning on. If price gets there, the picture that justified the entry is gone. The exit is information arriving.
• Volatility-based. Place it k × ATR away, so ordinary movement can't reach it. Useful when the structure is fuzzy or the instrument is noisy.

Most practitioners use both as a cross-check: find the structural level, then confirm it sits outside about one ATR of normal noise. The structural half is the marked pivots below — nothing about them refers to an account balance.

The anti-pattern

The stop derived from the wallet: "I can only stand to lose $200, so I'll put my stop $1 away." That places the exit where your account balance says, on a chart that has no idea your account exists. The market will hit it at exactly the rate a $1 distance implies — which on most instruments is often.

The correct sequence is the reverse, and it's worth memorising:

Chart → stop → size.

The chart decides the stop. The stop decides the size (Unit 1's formula). If the resulting size is uncomfortably small, the trade is simply smaller — not the stop tighter.

Worked example

A breakout above a $98–$100 range, entering at $100.00. The most recent swing low is $96.20. ATR(14) is $1.50. Risk budget: $500.

The second row risks the same $500 on paper, and will be stopped out by an ordinary Tuesday. It also carries 6.9× the exposure, which means slippage and gaps hit it 6.9× as hard. It isn't a tighter risk; it's a bigger bet with a shorter fuse.

Two rules that follow

• Match the stop to the timeframe the signal came from. A setup read off a 5-minute chart has 5-minute structure; a daily setup has daily structure. Using a 5-minute stop on a daily thesis makes noise exits far more likely, because ordinary daily movement is many multiples of five-minute structure.
• Never widen a stop after entry. The moment you move it, the risk you computed stops being the risk you have, and every formula in Unit 1 silently becomes wrong. Moving a stop toward the entry as a trade works is a different action with different consequences — that's the next-but-one lesson.

Try it now

1. Take the last marked swing low on the schematic above as your structural level and the final close as a hypothetical entry, and write the distance down in dollars. The chart chose that number; you did not.
2. Now on real prices. Mark the most recent clear swing low in the year below, then take Apple's latest 14-day ATR from the table under the chart, and express the distance from a hypothetical entry to that level in ATR units. Is it more than 1?

3. Compute the share count for a hypothetical $40,000 account at 1% risk using that structural distance — then again using an arbitrary $0.50 stop. Compare the two exposures and note which one your wallet chose rather than the chart.