Correlation, allocation in a portfolio. Calculation methods - page 4

New comment
 
CHINGIZ MUSTAFAEV:

on any

Genghis, another option
check out the correlation:
Eurusd and 1/usdchf
One of the best highly correlated bundles.
If no fraction, close to -1, i.e. inverse correlation
 
CHINGIZ MUSTAFAEV:

on any

i.e. on a year or a month there is no sliding period? something is wrong. On large ones it may not be, but on short ones, it is unlikely that there would not be a period with a tangible correlation.

 
Valeriy Yastremskiy:

i.e. on a year or a month there is no sliding period? On large ones it may not be, but on short ones, it is unlikely that there would not be a period with a tangible correlation.

Forget sampling.
 
Renat Akhtyamov:
Genghis, another option.
check out the correlation:
Eurusd and 1/usdchf
One of the best highly correlated bundles.
If without fractions, close to -1, i.e. inverse correlation
Same eggs, only in profile.
 
CHINGIZ MUSTAFAEV:
Forget sampling

how? Where does the data come from and how?

 
Valeriy Yastremskiy:

how? Where does the data come from and how?

First we estimate on a sample of plus or minus infinity (all available history)
Then any period you want
 
CHINGIZ MUSTAFAEV:
First we estimate on a sample of plus or minus infinity (all available history)
Then any period you want

I am asking, is there no period with correlation in the entire history? The fact that there is minimal correlation in the entire history is normal, but the fact that there are no interdependent periods is not very normal) At least the crises should have given the same behaviour.

 
Valeriy Yastremskiy:

I am asking, is there no period with correlation in the whole history? The fact that there is minimal correlation throughout the entire history is normal. But the fact that there are no interdependent periods is not very normal) At least the crises should have given the same behavior.

Impulses synchronize the assets, but we are not looking for a perfect 0 correlation
It's a 0.1-0.2 range.
 
CHINGIZ MUSTAFAEV:
Pulses synchronise assets, but we're not looking for perfect 0 correlation
Deto 0.1-0.2 range
how many people have died on these formulas....
 
Renat Akhtyamov:
how many people have died on these formulas....
Reply on the subject or flub in other threads.
123456789101112
New comment
Reason: