Crypto news and analysis
Beginner · Trading

How stop losses work

Learn how stop orders support predefined exits, why trigger and fill prices differ, and how volatility, gaps, and sizing affect actual loss.

12 min read3-question quizUp to 135 XP

A sell stop at $49 cannot create buyers at $49 if adverse news moves the best bid directly from $50 to $45. It is a planned exit mechanism, not a guaranteed price or insurance policy. Its value comes from connecting thesis invalidation, position size, and an executable instruction before stress distorts judgment.

Stops solve only part of downside control. A venue may fail, a market may gap, liquidity may vanish, or a stop-limit may remain unfilled. Traders should combine the order with conservative sizing and a contingency plan. The aim is to make a foreseeable error survivable, not to claim that every loss can be held to an exact number.

What you will learn

  • Place invalidation according to thesis and market structure rather than arbitrary loss preference
  • Distinguish stop-market execution risk from stop-limit non-fill risk
  • Calculate how stop distance changes position size and portfolio exposure

Invalidation comes before the stop price

An invalidation level marks evidence that the original setup no longer deserves capital. For a range breakout, it might be sustained acceptance back inside the range; for a trend setup, it might be a structural low breaking. A stop order is one method of acting on that evidence, but the analytical reason should exist before the order level is selected.

Setting a stop solely at the dollar loss a trader emotionally dislikes can place it inside ordinary market noise. Instead, estimate plausible volatility around the setup, identify where the thesis fails, and then reduce quantity until the resulting loss fits the budget. If the required quantity is impractically small, skipping the trade is a valid risk decision.

Trigger price is not fill price

A stop remains dormant until its chosen reference reaches the trigger. The venue then creates the configured order. A stop-market seeks available bids or offers and may slip; a stop-limit refuses prices outside its boundary and may not execute. These are different failure modes, so the trader must choose which one the scenario can tolerate.

Gaps make the distinction visible. If adverse news moves the best available bid from $50 to $45 without trades in between, a sell stop at $49 cannot manufacture buyers at $49. A stop-market may execute near $45 or lower, while a stop-limit at $48 may sit unfilled. Position sizing must account for this residual tail risk.

Stop placement must respect market behavior

Obvious recent highs and lows attract attention because many participants observe them, but no rule says a stop near those levels will be targeted or protected. What matters is whether normal volatility can reach the area without disproving the thesis. Measures such as recent range or average true range describe past movement; they do not forecast a maximum future move.

Liquidity matters as much as chart structure. A stop cluster in a shallow market can activate many stop-market orders at once, worsening fills. Review depth, spread, trading interruptions, and the position's size relative to normal volume. For a large exposure, reducing size earlier under predefined conditions may depend less on one order executing during stress.

Operate stops as part of a complete plan

After a partial entry or exit, check that the stop quantity still matches the open position. Use reduce-only where appropriate and understood. Avoid canceling a protective order merely because price approaches it. If new evidence justifies a change, document that evidence; discomfort with realizing a loss is not new market information.

Prepare a fallback for a failed trigger, rejected order, or unavailable venue. That may include alerts, independent connectivity, smaller exposure, or a written emergency procedure. Holding assets on several venues can diversify one operational dependency but introduces more custody and account-security risk, so contingency design always involves trade-offs rather than perfect protection.

Reality check

Common misconceptions

A stop loss guarantees the maximum amount that can be lost.

A stop is an instruction activated at a trigger. Gaps, slippage, unavailable liquidity, and system failures can produce a much larger realized loss.

A wider stop makes a trade safer.

A wider stop reduces sensitivity to ordinary movement only if quantity falls accordingly. With unchanged quantity, it simply increases planned capital at risk.

Moving a stop farther away gives a thesis time to work.

If the change only avoids recognizing loss, it abandons the original risk contract. Any adjustment needs new evidence and a recalculated portfolio loss.

Before you act

Risks and limitations

  • Adverse gaps can bypass the trigger level and force execution far beyond the planned price.
  • Stop-limit instructions may activate but remain unfilled as the market continues moving against the position.
  • Ordinary volatility can repeatedly trigger poorly placed exits, accumulating spread, fees, and small losses.
  • Canceling, enlarging, or moving stops under emotional pressure can convert a bounded plan into open-ended exposure.

Key takeaways

  1. Derive a stop from thesis invalidation, then derive quantity from the loss budget.
  2. Treat triggers and executions as separate stages with separate failure modes.
  3. Expect actual loss to differ from planned loss during gaps or thin liquidity.
  4. Review stop quantities and reduce-only settings after every partial fill.
  5. Document evidence for any risk-plan change made while a position is open.

Primary and further reading

Knowledge check

Test your understanding

Score at least 2 out of 3 to complete this lesson. Explanations appear after you submit.

1. Where should a risk-first trader begin when selecting a stop?
2. What is the main additional risk of a stop-limit compared with a stop-market?
3. How should position quantity change when valid stop distance increases and loss budget stays fixed?