Subject: re : your comments on metals var model
dear andreas ,
thanks for the very useful response and information on positions . i have
handed over primary responsibility for metals var to kirstee hewitt ( at least
as far as london is concerned ) and she will follow up on the points you
kindly reported , although i am of course available to assist where
necessary . tanya will remain the point of contact for var modelling in
houston , and of course kirstee and tanya will work together on this to
resolve these and further issues . i would be happy to help set up meetings
for you in the london office when you visit next week , with kirstee , myself
and anyone else that you would like to meet with and we very much look
forward to your visit - please let me know if you need any help with the
arrangements .
regards ,
anjam
0207 783 5383
- - - - - - - - - - - - - - - - - - - - - - forwarded by anjam ahmad / lon / ect on 27 / 07 / 2000 08 : 40
- - - - - - - - - - - - - - - - - - - - - - - - - - -
enron capital & trade resources corp .
from : andreas . barschkis @ mgusa . com 27 / 07 / 2000 00 : 36
to : anjam . ahmad @ enron . com , kirstee . hewitt @ enron . com
cc : bjorn _ hagelmann @ mgusa . com
subject : re : varmodel _ live . xls
anjam ,
thanks for your var model .
i would like to point out the following :
1 ) position data : i noted that the outright ( longs + shorts ) for copper
include positions which we should not use , i . e . r . wolff quantities . this
positions are imputed into the system at their integrity ( some legs / hedges
are missing )
this affects copper and lead position for now . the issues should be have
been sorted out by the end of this month , ie next week . so in copper you
have 25 , 487 mt to much and with lead you have 13 , 579 mt to much . ( as of june
19 )
2 ) looking at copper position you calculate a var of - 3 , 100 , 568 with a
total outright of 63 . 765 mt .
this seams too low . if i calculate 63765 x 1807 x 3 . 99 % = 4 . 218 mil . us $ . var
( outright qty x price levelx risk factor , riskfactr as per mercur ) .
in my view we have been understating var in mercur because we do not
consider the spread position correctly ( i . e . in detail ) . on a position like
that i would expect a figure of around 6 mil us $ .
i guess the issue is in the volatility and in the holding period .
3 ) your correlation matrix for the commodities is not final i assume as
many fields are blank .
lets talk tomorrow on the phone .
attached please find the position summary for the last week # 30 ( since june
19 ) as requested by kirstee .
( see attached file : mgposw 30 . xls )
andreas barschkis
mg metal & commodity corp .
520 madison avenue , 28 th floor
new york , ny 10022
tel : + 1 . 212 . 715 . 5628
cel : + 1 . 917 . 679 . 8287
fax : + 1 . 212 . 715 . 5608
e - mail : andreas . barschkis @ mgusa . com
anjam . ahmad @ e
nron . com to : andreas . barschkis @ mgusa . com
cc : kirstee . hewitt @ enron . com
07 / 26 / 2000 subject : zipped varmodel _ live . xls
11 : 37 am
hi andreas ,
this is the semi - final spreadsheet - have only to include price curves for
gold and cocoa . kirstee and i would welcome your comments .
regards ,
anjam
- mgposw 30 . xls