Exits are a strategy too

Much of the writing about systematic trading is about entries: which signal, which pattern, which model. In our research the exit has repeatedly mattered as much as the entry, and sometimes more. The same entry can be a good strategy or a poor one depending entirely on how it is closed.

Two failure modes

Exits tend to fail in one of two ways. Close too early and winners are cut short, so a handful of large gains that would have paid for many small losses never arrive. Close too late and small losses are allowed to become large ones. Most simple exit rules trade one failure for the other.

How we study exits

We treat the exit as its own research problem. For a fixed set of entries, we test families of exit rules side by side (fixed targets and stops, trailing stops, time limits, and profit locks that tighten as a trade moves in our favour) on the same historical trades, priced from real quotes rather than idealised midpoints. Because the entries are held constant, the difference in outcome belongs to the exit alone.

The rules that tend to hold up share a shape: they protect capital quickly when a trade goes wrong, and they give a trade that is working room to keep working. That sounds obvious. Making it precise, and proving it out of sample after costs, is the work.

Exits belong to risk

Because exits decide how much a trade can lose, they sit with the risk layer, not the signal. A model can propose a trade; it cannot talk the system out of a stop.

This note describes research methods. It is not investment advice and does not describe the performance of any account or fund.