## VAR

ankuragrawal.nit
Finance Junkie
Posts: 64
Joined: Sun Apr 01, 2012 10:52 am

### VAR

The Var things is confusing me in two aspects:

1. when we are calculating % var we shd take std deviation in % right?
Then when we are converting it to daily var this % var must be multiplied by base
reference to q1, page 8 (link provided below)
The % Var calc is 0.016 Acc to me this shd be 1.6 %
and the formula for daily VAR which says VAR(in %)
hence , 1.6 must be multiplied by 2.33 instead of 0.016 being multiplied by 2.33
This is confusing...

2. In the example in pristine's notes on VAR
i.e. notes : "VAR-I.pdf"
Example 3
I have Wa= 0.4
Wb = 0.6
(do not know why pristine has taken 40 mn and 60 mn as Wa and Wb, shdnt it be .4 and 0.6?)
Secondly the value that one gets aftr applying % VAR formula is in % so portfolio VAR shd be in %
Even If we require the absolute VAR i.e. to say Daily VAR of portfolio then cant we multiply this % VAR by 100 mn ( the total value of portfolio)
Below is the link to the file:

http://www.edupristine.com/infrastructure/file.php/134/FRM-July-2011/VaR-I.pdf
page 10
pristine has taken 40 mn and 60 mn as weights and then they have calc the daily var of protfolio as 7.22 mn
If we carefully observe for Daily Var of A and B we see that daily VAR of A = (1.96*.055*40 mn)
{assuming tht we convert 5.5%= 0.055 which I hvnt understood why?}
daily VAR of B similarily comes out to be 4.99 mn and tht of A to be 4.32mn
whereas Var for portfolio is cmng out to be 7.22 mn
How is it possible?

Are we trying to say tht a correlation of 20% has resulted in more risk?
How is tht possible?
Diversification always leads to lesser risk

I think I am definitely missing some potential point in here if Pristine's calculations are correct :(

Tags: