A stop loss is the single most important piece of trade management most people get wrong — not because the concept is hard, but because the placement usually isn't thought through.
What it actually is
A stop-loss order is a standing instruction to exit a position automatically once price reaches a predetermined unfavorable level. It exists to enforce a decision you made calmly, in advance, at a moment when you might not be calm — the entire point is that it doesn't require you to be watching, or to trust your own judgment under pressure in real time.
The main types
- Fixed (hard) stop — a static price level set at entry and left alone. Simple, predictable, but doesn't adapt as a position moves favorably.
- Trailing stop — a stop that moves up (for a long position) as price rises, locking in gains while still giving the trade room to breathe, but never moving down. It converts "protect against loss" into "protect an unrealized gain" once the trade is working.
- Volatility-based (ATR) stop — placement determined by the asset's own recent volatility (a multiple of its Average True Range) rather than an arbitrary percentage, so a calm asset gets a tighter stop and a wild one gets more room — the same fixed percentage stop is effectively tighter on a volatile asset and looser on a quiet one, which is usually backwards.
The placement mistakes that quietly ruin a good strategy
Too tight, relative to the asset's own noise. A stop placed inside an asset's normal hourly wiggle gets triggered by noise, not by the thesis actually being wrong — you can be right about the trade and still get stopped out by an ordinary fluctuation.
Round numbers everyone else uses. A stop parked exactly at an obvious round number or a widely-watched technical level sits exactly where a disproportionate number of other stops cluster — a level more likely to get probed and swept before a real move resumes.
No stop at all, "because I'll watch it." The single most common way a manageable loss becomes an account-threatening one. A plan that depends on always being present, always calm, and always willing to act against your own hope is not a real risk plan.
Moving the stop away from the entry after the trade goes against you. Widening a stop because price approached it is the mechanism by which a small, planned loss becomes a large, unplanned one. The stop's value comes specifically from being decided before emotion is involved — moving it under pressure defeats the entire purpose.
The honest tradeoff
Every stop-loss placement trades off two failure modes against each other: too tight and you get stopped out by ordinary noise on trades that would have worked; too loose and a real loss runs further than it should before it's cut. There's no placement that eliminates both — the right balance is a property of the specific asset's volatility and the strategy's own historical behavior, which is exactly why it needs to be backtested, not guessed.
This is why Atradian's live strategies use a dynamic, volatility-scaled stop (via each asset's own ATR) rather than one arbitrary percentage applied everywhere, and why every backtest reports how much of the result came from stop-outs specifically — visible in the trade-exit reasons behind every Strategy Lab result.