Trailing relative to the spread

[Fairman]

Trailing relative to the spread

A fixed trailing stop of 3 pips behaves very differently when the spread is 0.2 than when it's 1.5. In the second case, a fifth of the stop distance is simply the spread, and a brief widening can tag it.

A spread-aware trail keeps the distance in proportion to the current cost. For instance, the stop trails at a base distance plus twice the current spread. When the spread widens, the stop gives more room. When it narrows, the stop tightens.

The idea is to avoid being stopped out by cost fluctuation rather than by actual price movement. It's easy to implement in code and makes a measurable difference during rollover and news.

The effect can be tested. Run the strategy with a fixed trail and with a spread-adjusted one. Compare the number of stop-outs that occurred during spread spikes and the final result.

Even a simple adjustment, such as freezing the trail while the spread is above twice its median, removes a source of random losses.

During spread spikes, does a fixed trailing stop get hit more often in the records?