TRIED & TRADED
← Field guide · exits

Exit 2 — the stop loss

  • A stop loss closes the trade once the loss reaches a stated size.
  • That size is a percentage of what you staked — the credit you took in, or the debit you paid. Never of what you could lose.

What it is

  • Your stake is what you put up to open the trade. The credit received on a credit trade; the debit paid on a debit one.
  • The stop is a percentage of that stake. 100% means you have lost exactly what you staked.
  • You close when the loss reaches it — not when the price reaches it. Those are two different exits and the difference matters.
  • It caps the size of the worst trade, which is the entire argument for it.

Every stop on this site is a percentage of what you staked. A percentage with no “of what” is not a rule, it is a guess.

The same spread, in numbers

Sell the 95 put, buy the 90 put, 45 days out, stock at $100. You take in $93 — that is the stake. The spread is $5 wide, so the most you can lose is $407: the width, less the credit you keep.

Stop, as % of stakeYou loseBuy the spread back atThat is this much of max loss
100% of stake$93$1.8623%
200% of stake$186$2.7946%
300% of stake$279$3.7269%
No stop at all$407$5.00100%
  • Read the first two columns together and it is simple: 200% of a $93 stake is a $186 loss.
  • The fourth column is the same rung on the other scale. That $186 is 46% of the most this trade could ever cost you.
  • The two scales are miles apart on a credit trade, and that gap is the whole reason the wording has to be precise.
  • Here, the max loss is more than four times the stake — $407 against $93. So even a 300% stop is still inside the risk: there is road past it.

Arithmetic from the specification above, at 25% volatility and a 4% rate. Not a measured result.

Why it is measured against what you staked

There are two ways the industry writes this, and only one of them works everywhere. Here is why we chose the one we did.

The same rule, on a credit trade and a debit trade

Credit tradeDebit trade
Your stake isthe credit you receivedthe debit you paid
100% of stake meansyou took in $93 and it now costs $186 to close — so you are down $93you have lost the lot
Max loss isthe width less the credit — several times your stakethe debit. The same number as your stake
Is a 200% stop possible?Yes. There is more road past itNo. You cannot lose more than you paid
  • On a debit trade the two scales are the same thing. What you staked is the maximum loss, so 100% of stake and 100% of max loss are one number.
  • On a credit trade they are nowhere near. On the spread above the max loss is more than four times what you staked.
  • So a percentage quoted without its base means one thing on one trade and something four times different on the other.
  • Anything above 100% on a debit trade is meaningless — there is nothing left to lose.

Why not just use percent of max loss?

  • It reads more naturally, and on a defined-risk spread it is perfectly clear.
  • But it does not exist on every trade. A naked put or a strangle has no maximum loss to take a percentage of.
  • And where it does exist, it is not always known. On a calendar the ceiling is a modelled number, because one leg is still alive when the other expires.
  • What you staked is known exactly, at entry, on every structure there is. A modelled number should never be the thing that decides a real loss.

One scale that works on every trade beats a friendlier one that breaks on half of them.

You will see it written the other way

  • The common phrasing is a multiple of the credit — “a 2x stop”, “2x credit received”.
  • It is ambiguous. It can mean the loss reaches twice the credit, or the spread’s price reaches twice the credit. Those are two different exits, one twice the size of the other.
  • Published explainers get this wrong regularly, sometimes inside a single paragraph.
  • So we do not use the phrase. Every stop here is stated as a percentage of the stake, and the stake is named.

One thing the scale does not fix

  • The stop is set off the credit. What you can lose is set by the width.
  • Widen the spread and the credit barely moves; the maximum loss moves a lot.
  • So the same stated stop protects a different fraction of the risk on a $5 spread than on a $10 one.
  • A stop is not portable between specifications, however clearly it is worded.

Where it bites

A stop can be hit by the spread, not the stock

  • Option prices move on volatility as well as on the underlying.
  • A jump in implied volatility can widen a spread enough to trigger the stop while the stock has barely moved.
  • You are then out of a position whose thesis never broke.

And the mid price may not be a price

  • Stops are usually tested against the mid of the bid and ask.
  • On a wide market, the mid is a number nobody was willing to trade at.
  • A stop that only ever triggered on a fictional mid did not happen.
Why a stop can lower the win rate and still be right

Adding a stop turns some trades that would have recovered into realised losses. The win rate falls. Whether the strategy is better off depends on what those cut trades would have cost had they not recovered — which is an expectancy question, and a measurable one.

Common questions

Is a stop loss a real order, or a rule?

Either. A resting stop order on a spread can fill badly in a fast market, so many traders run it as a rule checked at a set time instead. The two behave differently and a test has to say which one it modelled.

Does a defined-risk spread need a stop at all?

It is a genuine argument. The maximum loss is already known and capped, so the stop is buying a smaller worst case at the cost of cutting trades that would have come back. That is exactly the sort of trade-off worth measuring.

Can you have a 200% stop on a debit spread?

No. Your stake and your maximum loss are the same number on a debit trade, so 100% is the end of the road — you cannot lose more than you paid. Any rung above 100% only means something on a credit trade.

What stop do most people use?

Two to three times the credit is the range you see quoted most often. That is convention, not evidence.

Can a stop be set as a share of the account instead?

Yes, and it is a cleaner way to think about sizing. It is a different rule though — it depends on how many contracts you traded, so two traders running the same strategy would exit at different times.

How this shows up in our tests

No stop-loss result is published yet.

The queued run is a credit spread traded with and without a stop, over the same entries, scored on expectancy and on the worst drawdown. Both numbers matter: the stop is supposed to buy the second at the cost of the first.

The hub

The seven ways
to exit

The umbrella

What is an
exit rule?

Related exit

The technical
exit