Anyone do any backtesting on "time-based" exits (ie. stops)?

Time-delayed entries are much more important than time-based exists. As it turns out, the expected gain from time-based exits is zero for very large samples of intraday and swing trades but that is not the case with time-delayed entries since there is a built-in bias in that case.
 
Quote from intradaybill:

Time-delayed entries are much more important than time-based exists. As it turns out, the expected gain from time-based exits is zero for very large samples of intraday and swing trades but that is not the case with time-delayed entries since there is a built-in bias in that case.
More...

care to elaborate on time-delayed entries? And where did you see/find the data for expected value of time-based exits?
 
Quote from bigb:

care to elaborate on time-delayed entries? And where did you see/find the data for expected value of time-based exits?
More...

The idea behind time-delayed entries is managing pullbacks and eventually getting filled at a better price. This does not mean each and every trade should be delayed because some trades work best when filled right away when you get your signal. The trick is in determining when to delay and by how much.

I did a lot of testing and statistical analysis in the past and I found out that in the case when profit and stop-loss exits are used the incremental gain from establishing a time-based exit has an expected value of zero. This is because in a large sample, and at the limit, the ratio of the sum of the incremental gains divided by the some of the incremental losses due to the time-based exits approaches 1.

On the contrary, the increase in equity performance due to time-delayed entries can be substantial. I mean quite substantial.
 
You must log in or register to reply here.
Back
Top