Why is the normal distribution not normal? - page 20

New comment
 
I wonder if everyone analyses the time series of prices over time...
 
getch >>:
Маскимально возможный доход на любом отрезке времени будет получен при проходе по ВСЕМ локальным экстремумам, расстояние между которыми не меньше Spread пунктов. Времени здесь нет.

If the TS does not contain time, then the yield is maximized per transaction anyway. But it is important for us - humans - to earn a lot per month, not per transaction, which in the limit may take a lifetime and not end during it!

P.S. I don't analyse price series over time. I maximise returns by finding the maximum as the average number of pips per unit time.

Can you feel the difference?

 

So there is a notion of an optimal transaction that is in no way time-independent.

You do analyse price series with time in mind, convincing yourself otherwise.

 

Well, then we're talking about the same thing.

Hallelujah!

 
No, completely different approaches.
 
Neutron >> :

If the TS does not contain time, then the return is maximised per transaction anyway. But it is important for us - humans - to earn a lot per month, not per transaction, which in the limit may take our whole life time and not end during it!

Not per transaction, but per number of transactions.

 
There may be different approaches. This does not say what is right and what is wrong. The important thing is that the goal is the same, so that the solution is optimal and does not depend on the way in which the problem is solved.
 
getch >> :

Not per transaction, but per number of transactions.

We give the spread from the transaction.

 

With such conceptual differences, even the goals are different.

All the same, in the EA, the optimal approach is many times small.

 
Yes, we give the spread from the transaction, and the maximum profit will be under the condition I wrote above. The advisor shows this.
1...131415161718192021222324252627...47
New comment