Vince,
Sorry for getting back a little late. I have been in touch with Kirstee to 
figure out how to treat positions data which is the primary cause for 
fluctuations. Apparently, as of 7/19, MG still has July position. This may be 
spot or balance of the month, or some settlement issue. This indeed is what 
we are trying to find out/verify (it may be an error too). Kirstee is going 
to get back to me on this. If it is not an error, then Anjam's latest results 
($5.5 million total VAR) are reasonable given all the assumptions and the 
missing pieces. I will keep you posted.
Cantekin




Vince J Kaminski
07/26/2000 10:12 AM
To: Cantekin Dincerler/HOU/ECT@ECT
cc: Vince J Kaminski/HOU/ECT@ECT, Ted Murphy/HOU/ECT@ECT 
Subject: Re:  

Cantekin,

Can you figure out the reason for cocoa beans VaR fluctuations?
The same is true of aluminum. I assume this is the position change.


Vince



Cantekin Dincerler
07/26/2000 09:28 AM
To: Anjam Ahmad/LON/ECT@ECT, Kirstee Hewitt/LON/ECT@ECT
cc: Vince J Kaminski/HOU/ECT@ECT 
Subject: 

Anjam and Kirstee,


As I have suspected the position as of 6/30 vs 7/19 makes the difference. The 
first table first column is my VAR number with 6/30 position and gold&silver 
prices, the second column is your VAR with 7/19 position and dummy 
gold&silver prices. The second table first column is my VAR with 7/19 
position and 6/30 gold&silver prices, the second column is as before.

I would ask you to plug the gold and silver prices and see what kind of 
numbers you get in order to verify we are on the same page. Please refer to 
MODELVAR2000.xls that I have sent you for gold & silver prices and 
volatilities. 

Thank you,
Cantekin


  TABLE 1



  TABLE 2