Is there proof of profit from using Autotrading? - page 24

New comment
 
khorosh:
You can remove the bad bars in the quotes.


I don't think that's a good idea. You will get gaps, the quality of quotes in the tester will be low, if the EA uses dates, the results may be lefty.

It is better to make a CSV file with the sections where you cannot trade and read it in the tester.

One more thing. OK, we have taught the robot to work profitably in the absence of hazards. And then we release it, all delicate and subtle, into the real, brutal world. I think the result is predictable.

 
Alexey Volchanskiy:


I don't think it's a good idea. You will get gaps, the quote quality will be low in the tester, if the EA uses dates, the results may be lefty.

It is better to make a CSV file with the sections where you cannot trade and read it in the tester.

One more thing. OK, we have taught the robot to work profitably in the absence of hazards. And then we release it, all delicate and subtle, into the real, brutal world. I think the result is predictable.

I think that in a crash we should not expect miracles - just that the optimizer will not have slippages
 
Younga:
I think in a catastrophe, you shouldn't expect miracles - the optimizer just won't get stuck.

Oh, I agree with the optimizer, of course.
 

Better yet, create a library file and enter the dates into it so that they don't interfere with optimisation.

 
I realise this is too weird a question - what about the trades opened by the tester before the crash? What are the options?
 
Younga:
I understand this strange question - what should we do with trades opened by the tester before the crash? What are the options?


Well, there aren't many options.

  1. Close before disaster strikes
  2. Close regardless of disaster by TP and SL
  3. Close regardless of the catastrophe by internal algorithm
  4. At the time of the crash, do nothing; set TP and SL to a huge margin so they won't trigger.
1...1718192021222324
New comment