Take profits fill exactly, stops do not
Limit orders fill at their price or better. Stop orders fill at the next available price, which in a fast market can be well beyond the stop level. That means the average win is capped by the target, while the average loss is stretched by slippage.
A strategy planned on a 1:1 reward-to-risk using exact fills becomes something worse in practice. Two-pip target, two-pip stop on paper. In reality the target brings exactly two pips and the stop brings two plus whatever slippage the market adds. A scalper doing this sixty times a day has paid for the asymmetry many times over.
Backtests that assume both sides fill precisely overstate performance. Good backtests add a slippage model to stops specifically and not to limits.
The fix is not to avoid stops, since protection matters. The fix is accounting honestly for the extra cost and sizing the stop distance to include expected slippage.
A rough way to estimate it: take the last hundred stop-outs, measure the difference between the stop price and the actual fill, and average it. That number belongs in every backtest.
What was the average slippage on stop-outs compared with the average on take profits over the last few months?
Take profits fill exactly, stops do not
Take profits fill exactly, stops do not
It’s Fairman 
-
FXS Prop Desk
- Posts: 1689
- Joined: Wed Oct 07, 2026 9:52 am
Re: Take profits fill exactly, stops do not
Because limit exits fill at the price and stop exits fill at the next available one, the realised reward-to-risk ends up worse than planned. Recomputing expectancy with the average stop slippage from actual fills, rather than the planned stop, gives a more realistic picture of the strategy.
FXS Official Team | The Idea Lab & Strategy