What the numbers mean
locked in = (exit − entry) × units × share closed
worst case = locked in + (new stop − entry) × units remaining
Close half a 100,000-unit position at 1.0900 from an entry of 1.0850 and $250 is banked. Move the stop on the rest to the entry price and the worst outcome for the whole trade becomes +$250 rather than −$500.
That is the appeal, and it is real: the trade can no longer lose.
The half nobody calculates
Scaling out does two things at once, and only one of them is visible on the screen.
It removes downside from a trade that is currently winning. It also removes upside from every trade that would have run, because half the position is no longer there when the target arrives.
The losers, meanwhile, are unaffected. A trade that goes straight to the stop was never partially exited — it loses the full amount. So the distribution changes in exactly one direction: winners get smaller, losers stay the same size.
The last row on this calculator prices that. On a trade that reaches 2R, scaling half at 1R gives up a quarter of the result. Do it on every trade and the strategy's expectancy falls by roughly the same proportion, permanently.
When it is worth paying
The exchange is legitimate when it buys something you actually need:
- It buys the ability to hold. If banking a partial is what lets you leave the rest alone to the target, the expectancy lost is cheaper than the expectancy lost by closing everything early out of discomfort.
- It buys a smoother curve. Lower variance is worth real money to anyone whose behaviour degrades in a drawdown, and to anyone trading a funded account with a hard daily limit.
It is not worth paying when the partial is simply relief — taken because the position being green is uncomfortable, at a level chosen by the feeling rather than by the plan.
The specific mistake
Scaling out before 1R — before the trade has earned back what it risked.
At that point the partial is too small to protect anything meaningful and large enough to remove most of the trade's ability to pay for its losers. A method with a genuine edge gets traded into a negative one this way, without any individual decision looking wrong.
The calculator warns below 1R for that reason. If a partial belongs in the plan, it belongs at a level decided in advance — and grading the outcome afterwards in R-multiples is what shows whether the habit is paying for itself.