Message-ID: <16003943.1075857239397.JavaMail.evans@thyme>
Date: Thu, 13 Apr 2000 07:15:00 -0700 (PDT)
From: benjamin.rogers@enron.com
To: russ.porter@enron.com
Subject: TECO-Answer to Chris Smith questions
Mime-Version: 1.0
Content-Type: text/plain; charset=ANSI_X3.4-1968
Content-Transfer-Encoding: quoted-printable
X-From: Benjamin Rogers
X-To: Russ Porter
X-cc: 
X-bcc: 
X-Folder: \Benjamin_Rogers_Dec2000_3\Notes Folders\Sent
X-Origin: Rogers-B
X-FileName: brogers.nsf

Here is the list of questions Yvan sent to Chris Smith and his responses.=
=20
Ben
---------------------- Forwarded by Benjamin Rogers/HOU/ECT on 04/13/2000=
=20
02:15 PM ---------------------------


YVAN CHAXEL@ENRON
03/21/2000 11:01 AM
To: Christopher Smith/HOU/ECT@ECT
cc: Benjamin Rogers/HOU/ECT@ECT, Berney C Aucoin/HOU/ECT@ECT, Carl=20
Livermore/HOU/ECT@ECT, Donald M- ECT Origination Black/HOU/ECT@ECT=20
Subject: TECO-Answer to Chris Smith questions

Chris,


Hopefully the following points should answer your questions. Feel free to=
=20
call me (34292) if you still have any question.

Yvan




1_  We will value Teco as a series of hourly Spread options.=20
Legitimate points of concern:=20
a) Are we valuing an index deal or a physical deal?
  b) If it is a physical deal can the plant realistically react on an hourl=
y=20
basis?

Answer: The deal Enron enters into is based on an spread between PJM index=
=20
and the appropriate OIL curve (New York Harbor + basis). However, the trigg=
er=20
for the default of the contract (which is really what we are concerned with=
)=20
is a function of the plant cash flows compared to the debt level. The plant=
=20
cash flows depend on the physical operations of the plant and its capacity =
to=20
dispatch during the most expensive hours of the year. Therefore the questio=
n=20
of whether the plant can or cannot ramp up quickly enough to dispatch=20
economically is very pertinent to our valuation.=20
I would argue that the plant will always (or almost always) be able to=20
dispatch economically for the following 2 reasons. In the first place, its=
=20
operational ramp up time is only 10 minutes. Secondly, the scalar curve we=
=20
use is fairly well behaved and doesn=01,t swing around a mean position (see=
=20
graph below), hence allowing sufficient preparation for the dispatchers. Th=
e=20
chances of dispatchers being caught by surprise when prices spike up or dow=
n=20
should therefore be relatively small, and the plant should be able to follo=
w=20
the real time prices pretty well.





2_ The power curve we use is PJM west. Even though the plant is located in=
=20
the south east end of PJM (i.e. the del Marva peninsula), we prefer not to=
=20
use the east curve which is too thinly traded to be meaningful.=20

4_ So far we have only used monthly vol for power. Should we start using=20
daily vol as well, then the blending formula would be:
'fvol =3D Sqr((dblMvol * dblMvol * (jDate - jToday) + dblDvol * dblDvol * 1=
5) /=20
(jDate - jToday + 15))

with  jDate the first of the considered month of dispatch
 dblMvol the monthly vol
 dblDvol the daily vol

6_ We are assuming 50% flat correlation for 20 years (a much more=20
conservative assumption than 15%). However we run sensitivities where the=
=20
correlation goes from 30% to 70%.

7_ Yes, we are deducting all operative expenses from the valuation.=20
Origination is currently trying to reconcile the differences in operations=
=20
expenses between their numbers and those used by credits.

8_ The valuation model (coded by Alex Huang, Research) is based on a=20
bidimensional tree, and therefore doesn=01,t use the SPRDOPT function. Howe=
ver,=20
we checked that the SPRDOPT function from the structuring model and the tre=
e=20
based model yielded the same values for the spread option.

9_ The strike for the spread option is the sum of the OIL price + the VOM.









***************************************************************************=
***
***************************************************************************=
***
*********************888








Please send me an excel spreadsheet which depicts the value that you see=20
associated with the heat rate spread option.  I want to make sure that I=20
understand exactly how you propose valuing the transaction and how this val=
ue=20
would be booked.  As such, please provide the following detail:

Will we be valuing/booking a series of monthly call options or a series of=
=20
daily call options?
What power curve are we valuing the deal against - PJM East or West Hub?
Please send me the current fuel curve that you are converting to $/MWh to=
=20
value the deal?
What volatilities are you using to value the options - monthly volatilities=
=20
or intra-day volatilities or a blend of the two (if blending please show me=
=20
what formula you using to blend the two vols)?
What expiration date are you using, i.e.: for monthly call options the 15th=
=20
of the relevant month, and for daily options?
What correlation are you assuming, 15% flat for twenty years?
Are you deducting all operating expenses from the value of the option?
Are you using the "SPRDOPT" Exotic Options function to value the option?
Are you using the $/MWh VOM dollar amount as your strike?
Has Don provided the fixed payment stream?  This stream should be covering=
=20
both P&I and not just principal.

Yvan, and Ben, please provide answers to these question via written=20
correspondence so that their is limited probability of misunderstanding. =
=20
Thank you both very much for your time and help thus far.  Furthermore, I=
=20
would like to reiterate that RAC's goal, prior to quoting any credit reserv=
e,=20
is to be 100% confident that: (i) the methodology that is being employed is=
=20
consistent among the internal groups; and (ii) the inherent value of the=20
price risk management contracts matches.  This enables RAC to manage the=20
associated risk during the life of the transaction both effectively and=20
appropriately.

Regards,

Christopher