Thread Closed : New approach to

"Using another coefficient could be an even better solution. This has been suggested in several other threads too."



If enough suggestions along these lines are made, maybe beneficial change will take place. I hope you don’t mind me making these suggestions. They were not directed at you.



FWIW, I had an English prof back in college circa the 1970’s. He thought every idea for a story had already been used before. Indeed, nearly every movie you see today is a variation on something Shakespeare wrote. I’m sure every post I’ve made is redundent to some extent.



I think C2 is great, but its tragic flaw is that it boosts unrealistic systems. If I hammer this point too often I apologize, but I can see C2 being even greater…perhaps a starting point for CTAs as well as THE marketplace for trading advice. It will not happen though if it’s not tightened up. Pros coming in and seeing a SR’s of 1000 will just write the site off as BS.

"I hope you don’t mind me making these suggestions. "



I don’t mind. On the contrary. I merely tried to explain what is happening: Probably not a computational error but a (in my view) bad choice of the time periods. But as I remember it most people supported this choice for various reasons. Perhaps they will be convinced by the example you gave.



Another way to solve this would be to display the SR only for much older systems. Then again, many vendors will be against it because it means that their system won’t get a SR for a very long time.



Note that if the SR is based on monthly periods, and the SR is displayed in May for a system that started in March, then the SR is based on only three values of the system: the return in March, the return in April and the return in May. I’m not suprised that some extreme values occur in such cases. But who wants to wait 30 months?

"Note that if the SR is based on monthly periods, and the SR is displayed in May for a system that started in March, then the SR is based on only three values of the system: the return in March, the return in April and the return in May. I’m not suprised that some extreme values occur in such cases. But who wants to wait 30 months?"



You said that already, and I got it. Let ME repeat: Net Profit / Max DD ratio

is going to look good (or not) right from the get-go and would be one of the best ways to rank systems. The data is already in the report on C2. Just have the ratio as one of the selectable ranking criteria.