Looks like Enpower is "on the side of the road" with respect to capturing and valuing retail deals.   As we discussed in yesterday's meeting, Plan B is a more robust spreadsheet that unbundles the components and values them separately and creates postions by risk component.

I know this is easier said than done, but we need a tactical solution that gives the group confidence that value and positions are correct, rather than a bundled black box batch value with hard coded assumptions.

 -----Original Message-----
From: 	Cross, Edith  
Sent:	Tuesday, June 12, 2001 5:47 PM
To:	Lavorato, John; Belden, Tim; Presto, Kevin M.
Cc:	Herndon, Rogers; Richter, Jeff
Subject:	FYI - Progress Report on testing an EES deal in Enpower


To all,

The following is documentation of a sample EES deal which was tested in the Enpower system.  This is for informational purposes only.  Please let me know if you have any questions, or need clarification on any points.  

Over the last few weeks, a few of us have been working to test an EES deal in Enpower (St. Luke's - Roosevelt, part of the Beth Israel deal).  The deal is a fixed price sale in New York City (Zone J) of Energy, Line Losses, Ancillary's, and ICAP.  The T&D is a pass through.  We only tested one site out of the ten.  Needless to say, this is probably one of the simplest structures available.  Obviously, we have worked on a few other things at the same time, so it has taken longer than we expected, but I hope this will help highlight how difficult it is to book EES deals in Enpower in it's current state.

Quick statistics for ONE site:

Each month of hourly data, takes 5-6 minutes to save.  We booked 100 months (approx 8.3 years), therefore each deal type (energy, ancillary, etc.) took 8-9 hours to book.
Since each deal type takes 8-9 hours to book, it takes a total of 25-26 hours to book energy, ancillary's (only 1), line losses, and ICAP (this is not booked hourly, therefore, it only takes 1 hour to book).
Although we booked only one ancillary position, most markets have at least four, therefore the total booking time should be closer to 50 hours.
Once we got a book to run, it took 25 minutes to calculate the value.
All of the above statistics assume there are no errors in entering the deals.

Recommendations:

Work with Anthony Dayao to set up a system that can handle these trades for position management, valuation, reporting, scheduling, settlements, accounting, etc.
In the interim:

Enhance RPS to make the valuation consistent with the batch valuation process.  This includes strengthening information for position and DPR reporting.  As well, create tools to help identify change in valuation due to curve shift, reg switch options, and new deals.
Enhance Intramonth valuation and reporting (East desk does not currently have an intramonth book).
Integrate the Tariff Valuation and Reporting.  Currently, this is done in a stand-alone spreadsheet application.
Support the Curve Database project which will allow the breakdown of curves into its components (as opposed to current bundled curves).
Move all the EPMI hedges to Enpower.  All hedges will become desk to desk instead of being entered in both RPS and Enpower.  Position and P&L reports will have be integrated between Enpower and RPS. 	


Here is a breakdown of what it took to 'rebook' this trade:

Week 1 (May 14-18)

Read Contract.
Get Batch model, as well as a breakdown of value between Energy, Ancillary's, etc.
Sit with Underwriting to answer questions about Load growth and contractual questions.
Work on Wholesale Full Requirements model (in Excel) to check valuation.

Week 2 (May 21-25)

Continue working on Full Requirements model.
Work with Risk Management (Stacey White's group) to set up delivery points and curves in Enpower.	
Begin to book deal in Enpower - Getting access to Enpower took some time.
Discovered that our assumption on hours for the 12x12 was incorrect.   We had gotten about 1/3 of the deal booked by the time we realized this.
At times, the database tables would run out of space, because we have never entered a deal of this size in the test database.  This would keep us from entering any new data.

Week 3 (May 29-June 1)

Finished Excel Full Requirements model.  The EES Batch and the Wholesale model tie out except for some minor differences on discounting and ICAP.  The current ICAP curves are            $/MW-day, and they are not calculated using the correct numbers of days (i.e. June and July have the same MW/day price, even though there are 30 and 31 days, respectively, in each month).  This can be fixed in the curve files.  In the future we should mark the curves as $/Kw-month.
Again, we ran into problems with the test database.
Completed booking in Enpower.

Week 4 (June 4-5)

Took 2 days to get a MTM valuation (portcalc) to work.
Discovered deal entry errors in booking.