Partial exits and the arithmetic of taking half
Posted: Sat Oct 31, 2026 6:42 am
Closing half a position at a first target and letting the rest run is a popular approach. It reduces stress and locks in some gain. The question is whether it improves expectancy.
The arithmetic depends on the distribution of outcomes. If the first target is hit often and the runner rarely reaches the second, partials smooth results but may reduce the total. If the runner frequently extends, partials cost potential profit.
Costs also matter. Splitting an order into two exits on a scalp means paying commission on two closings, and spread cost is the same either way.
A backtest comparing full exit at the first target, full exit at the second, and a half-and-half split reveals which fits the data. Differences are often small, making psychological comfort the deciding factor.
Whichever is chosen, consistency matters more than the specific split.
Logging each partial as its own line in the journal makes the real contribution of the runner easy to see after a few hundred trades.
Does taking partials improve results, or mainly make the trading feel easier?
The arithmetic depends on the distribution of outcomes. If the first target is hit often and the runner rarely reaches the second, partials smooth results but may reduce the total. If the runner frequently extends, partials cost potential profit.
Costs also matter. Splitting an order into two exits on a scalp means paying commission on two closings, and spread cost is the same either way.
A backtest comparing full exit at the first target, full exit at the second, and a half-and-half split reveals which fits the data. Differences are often small, making psychological comfort the deciding factor.
Whichever is chosen, consistency matters more than the specific split.
Logging each partial as its own line in the journal makes the real contribution of the runner easy to see after a few hundred trades.
Does taking partials improve results, or mainly make the trading feel easier?