On the unequal probability of a price move up or down - page 16

New comment
 
The "weight" of currencies is different, so there will always be asymmetry
 
Younga:
When calculating the coefficient the depth required is.... How much is it?
I don't choose the depth, I set the identity of the increments. So the correlation is one at the time interval at which it is set. On what interval do I set it? On the duration of the trade plus a little to the left. You won't need more than three days if the duration of the transaction is within a couple of days. If interested, I can repeat the trick in real time for the public.
 
On the length of the trade plus a little to the left... at the moment of opening how is it calculated ? and to repeat the experience - I would love to watch...
 
Younga:
On the length of the trade plus a little to the left... at the moment of opening how is it calculated ? and to repeat the experience - I would love to watch...
Didn't want to start tonight as I'm away on business tomorrow morning. Won't be able to be at the computer. I will start the repeat focus tomorrow closer to evening.
 
Mikhael1983:
I'll start a repeat focus tomorrow evening.


Better tell me how to construct the reference lines q .

 

I got held up. My apologies. However, let's get started.

Current market situation (showing 2 days in tf M5):


 

On one chart, it looks like this:

So, we buy on a volume of 8 EURUSD, sell on a volume of 3 GBPUSD:


 
No ... well that's not interesting ... correlation from when to recalculate, why 3 and 8 have survived the correlation ... That way you can't even point out the errors (maybe there are some...)
 
Younga:
No ... well that's not interesting ... correlation from when to recalculate, why 3 and 8 have survived the correlation ... That way you can't even point out the mistakes (maybe there are some...).
I could have suggested a different ratio, but having thought about it, I decided to continue with the same ratio of volatilities EDq and PDq, for now it is possible. The correlation is calculated (I finally understood the question) on the time interval at which the charts are shown (in this case 288*2=576 samples in TF M5). However, it does not matter, the correlation coefficient will be 1 at any shorter time interval. Why is it not interesting if you can repeat the deal and get the profit, guaranteed by me? )
 
Mikhael1983:
I could have suggested a different ratio, but after thinking about it, I decided to continue with the same ratio for the sake of uniformity, for now it is possible. The correlation is calculated (I finally understood the question) on the time interval at which the charts are shown (in this case 288*2=576 samples in TF M5). However, it does not matter, the correlation coefficient will be 1 at any shorter time interval. Why is it not interesting if you can repeat the deal and get the profit, guaranteed by me? )

I'm interested in opposing you having a description of the system - you've described it all, for observation... I will watch further. In the meantime - and why the correlation will remain (=1)on the future plot?

1...91011121314151617181920212223...184
New comment
Reason: