VaR numbers for the 26th

---------------------- Forwarded by Vince J Kaminski/HOU/ECT on 07/28/2000
05:09 PM ---------------------------

Enron Capital & Trade Resources Corp. - Europe

From: Kirstee Hewitt 07/28/2000 02:19 PM

To: Vince J Kaminski/HOU/ECT@ECT
cc:
Subject: VaR numbers for the 26th

Hi Vince,
I apologise that I did not send you the following mail ...
Kirsteee
---------------------- Forwarded by Kirstee Hewitt/LON/ECT on 28/07/2000
20:21 ---------------------------

Enron Europe

From: Kirstee Hewitt 27/07/2000 20:33

To: Andreas.Barschkis@mgusa.com
cc: Bjorn Hagelmann/HOU/ECT@ECT, Grant Masson/HOU/ECT@ECT

Subject: VaR numbers for the 26th

Hi Andreas,

I have run the VaR model for the 26th July and have attached a zip file of
the results:

The total VaR is $4,170,653 and the Cu position is $1,852,876.

I have had trouble getting hold of you so I thought I would summarised what I
wanted to talk about (also I thought
I would give your ears a rest!)
Basically it is wrt the second point in your mail yesterday (we briefly
discussed it earlier).
The fax you sent me to explain the risk calculation suggested that you use a
5 day period of adjustment to calculate
the risk (in this case it is called Capital at Risk). The VaR calculation for
our model is for a one day holding period which means that your risk factor
will be reduced by a factor equal the sqrt(5) or 2.24.
Since :

Old risk factor (%) = 1.65*(std of the price movement) = 3.99%

New daily risk factor = 3.99/2.24 = 1.78% which actually equates to std
of approx 1.1% a day.
I am happy with this as a estimate of the vol as the annualized spot vol we
are showing for Cu is approx 19% which equates to approx 1.2% daily.

Using this new risk factor for the daily VaR the Cu positions would give a
VaR (for the 19th) of approx $2m which is less that the
figure we estimated ($3,100,568).
The other thing is that by taking net numbers we are disregarding the term
structure of the price curve/vol curve and position
curve and are hence collapsing everything into a one factor model which means
that it is difficult to compare the numbers.
I hope that this helps to explain our number.

Hopefully we can talk tomorrow,
Cheers
Kirstee