Trent has been working on a mid term trend system and questioning on an appropriate entry.
I suggested he test a raw entry for it’s edge by using an exit after n-days. If an entry has an edge then a win rate above 50% would identify that and the only way to test the entry edge is without any stops and a pure exit after n-days.
Here’s what he’s come back with and associated question:
Quote:
As you suggested I just exited after a set number of days to work out the winning percentage of the entry. With both of these entry techniques with the best parameters in winning percentage is about 58%.
What would you suggest would be a reasonable number? I have applied a simple trailing stop exit to each of these and once that is applied the winning percentage goes down a long way.
Note the highlighted comment. We need to think differently.
The prevailing wisdom is that a trailing stop be used, yet clearly the addition of the trailing stop in this instance severely dilutes the strategy effectiveness.
The answer?
Don’t use a trailing stop.
Instead:
(1) Exit the position after n-days.
(2) Ensure both the entry breakout length and the n-days exit are reasonably robust, i.e. add some variance to both
(3) Lastly, work out the maximum adverse excursion and see if an initial stop helps reduce drawdown
If the W/L ratio is above 1.2 or so then using a 58% entry win rate will provide a solid positive expectancy outcome.
