This thread is going to go around in circles so I won't bother, after this one but... Let's all keep an open mind... yeah?
It is a function in TS to evaluate the ability to capture maximum profit per trade. I have found it to be very handy.
If you make 2000 trades per year and that it is not beneficial each year it can be very doubtfull: not only it should be beneficial each year but almost each month since more than 100 hundred trades per month must show something significantly different from 0 : statistical significance grows with the size of sample so if it is not positive (nearly) each month it is already doubtfull. More generally it is more important to look at the distribution of subsamples means (of each month, week,...) rather than the one based on total aggregated datas.