An exit rule states, in advance, the condition that closes a position: a profit, a loss, a date, a price signal, or an expiry you let run. It has to be checkable by somebody who wasn’t there.
Most traders have an exit. Far fewer have an exit rule.
| What it looks like | Can it be tested? | |
|---|---|---|
| A rule | Close at 50% of the credit received, or at 21 days, whichever comes first | Yes |
| A habit | “I usually take profits about halfway” | No — “about” and “usually” have no value |
| A reflex | Close it when the position starts to hurt | No — the condition is a feeling |
That fourth one gets skipped constantly. A profit target and a time exit that both trigger on the same afternoon need a stated winner, or two people testing “the same” rule get different numbers.
The most expensive words in an exit rule are “a percentage” — because nobody says of what.
No. A stop loss is one of seven exits. An exit rule is the umbrella term for whichever of them you have written down — a profit target and a time exit are exit rules too.
Usually yes. Most premium-selling strategies carry a profit target and a time or loss condition together, with a stated order for when both fire at once.
It can be discretionary in life. It cannot be discretionary in a test. Anything that needs your judgement on the day can’t be measured, so it can’t be compared against the alternative.
That is the question the whole cluster exists to answer, and it is not answerable in the abstract — it depends on the structure, the tenor and the underlying. What is clear is that changing it changes the result more than most traders expect.
No test on this site has published a result yet, so this page doesn’t claim one.
The first run that turns on this page is the exit ladder: one set of entries, closed seven different ways, scored on expectancy rather than win rate. This entry will link to it the day it publishes.
