Work Update

Vasant,

A quick update on a number of on-going projects:

1. JCC Forward Curve/Hedging Strategy (Enron Japan)

h An update of the historical P&L model was completed and included both commodity and foreign exchange risks. The standard deviation of historical simulations for single month swaps was significantly reduced with the foreign exchange hedging strategy (two long yen/$ positions and 1 short yen/$ position).

h Enron Japan, in cooperation with Enron London, developed an alternative hedging strategy based on Brent $ value. No P&L historical P&L simulations were conducted for the alternative strategy.

h The original Brent hedge ratios were based on an average of 3 monthly preliminary JCC values offered by METI. Presumably, the swap will be settled off of the finalized METI JCC price which is released two months after the fact. Rousseau forwarded those numbers from 1998 to present this morning. Unfortunately, our model uses historical JCC values going back to August 1991. Rousseau will hopefully obtain the required historical JCC time series early next week.

h I haven't yet heard about Jeff Shankman's response to the memo that was forwarded on Wednesday.

2. Ethanol Forward Curve (Enron Global Markets)

h A meeting was held with Marc de la Roche, Darren Maloney, and Sanjay Patel to discuss the development of a forward curve for ethanol. A first-cut curve based on NYMEX unleaded gasoline, NYMEX WTI crude, and EIA fuel ethanol inventories was constructed and backtested. Rolling monthly backtesting (using data from 2000 to 2001) suggested that, on average, the modeled and actual ethanol prices were off only by a few cents for time periods up to 6 months (e.g., 6 month swap). Unfortunately, the 1 SD value for the historical projections was quite large (+/- $0.30 on a $1.40/gal commodity).

h The changing dynamics of the ethanol market offer the greatest challenge to forward curve construction. That is, ethanol will likely replace MTBE as the fuel oxygenate of choice over time, which will probably signal a stronger relationship with unleaded gas price. Unfortunately, the current model uses historical data from 1998 to present which do not necessarily reflect such a strong relationship. These concerns have been noted to the ethanol guys.

h The ENE ethanol pricing index has been updated with August, September, and October prices and continues to generally track other industry publication values (e.g., Oxy-fuel news).

h Marc and Darren are in talks with Petrous as a potential counterparty for an ethanol swap.

h An ethanol status meeting will be held Wednesday or Thursday of next week with Randy Maffett.

3. Plastics Forward Curves (Enron Global Markets)

h Mauricio Mora and I have focused on the development of forward curves for a suite of plastics including propylene, ethylene, polypropylene, HDPE, and LLDPE (see earlier email for details).

h Initial models for each of these plastics have been completed and will be compared to the current model forward curves and Alan Engberg's expected values. Alan did forward the most recent curves and personal outlooks to us today.

h Note that the methologies behind the creation of the old forward curves are not known. Evidently, one can use various rankings (say 1 to 5 depending on market conditions) to manipulate the model forward curve--I don't know whether regression was involved in these forward curve constructions.

4. Butadiene-Natural Gasoline Hedge (Enron Energy Services)

h A request was made to EGM Fundamentals by Trey Comiskey (Industrial Services origination) to develop hedge ratios in natural gasoline to offset a position in butadiene. The request was forwarded to Research.

h Stationarity tests and cointegration and regression analyses were conducted on the two series. The two series could not be cointegrated and the R2 on the differences model was roughly 37%. Note that there appears to be little month-to-month movement in the historical butadiene series back to 1996. Analysis of a more recent time series might offer better results--I am also not sure why natural gasoline was selected as the hedge commodity.

h Trey was contacted to discuss the results of the analysis.

5. Emissions Technology Investment/Clean Air Solutions (Enron Global Markets)

h An informal meeting was held with Vince Middleton, Scott Affelt, and Reese Soares (Coal Desk/Clean Air Solutions Group) to discuss a pilot-scale emissions reduction technology evaluation with the Tennessee Valley Authority (TVA). The group is nearing completion of the testing with potential full-scale implementation of the technology at a number of sites.

h One of the effluents exiting the system (fly ash) is being discharged into a holding pond. The holding pond water is exhibiting high levels of nitrogen and the TVA has proposed that treatment of the ponds (9 total) will cost $9 million, half of which will be paid by Enron.

h The group has asked me to contact the TVA engineers directly to discuss possible treatment options and to see if costs may be further minimized. No information has yet been forwarded to me.

6. Textiles Cross-Selling (Enron Energy Services)

h A meeting was held with Jay Sparling (Industrial Services) and Rodney Derbigny to discuss cross-selling opportunities to the textiles sector. The initial purpose of the meeting was to discuss the intricacies of Enron's agricultural effort given that cotton is a key feedstock in this industry.

h The focus of the discussion shifted to cross-selling opportunities and the suite of products currently being offered by Enron. We discussed most of the EGM commodities (e.g., diesel fuel, weather derivatives, plastics, truck freight capacity) with an eye toward perhaps bringing some of the parties together at some point in the near future. In the near term, the most likely cross-selling candidate would be diesel fuel swaps for the trucking fleet.

7. Traffic Volume Analysis (Enron Freight Markets)

h Pravas Sud continues to work part-time on analyzing the DataPave database traffic studies with a goal of developing some understanding of freight volume throughout a given year.

h The construction of the dataset using the information sent by SAIC has been extremely difficult. In conjunction with IT, we have had to reconstruct the data series multiple times.

h Data for the I-10 corridor from Houston to Los Angeles and the I-5 corridor from Sacramento to Seattle constitute the initial routes to be investigated. Several testing points in each state are located along the routes and 5-axle and above traffic volume will be evaluated for monthly and daily trends (1998 data only).

8. Propane Demand Model (Enron Global Markets)

h Sarah Wooddy has completed the US propane monthly demand model and forwarded the results to EGM Fundamentals for inclusion into a larger model for Lee Jackson. She has conducted backtesting analysis on the monthly data and also forwarded this information to Fundamentals to give them a sense of model error.

h Propane demand modeling work will continue to meet the requests of the Weather Research Group.

9. Heating Oil Demand Model (Enron Global Markets)

h Shalesh Ganjoo is continuing to work on monthly heating oil demand for the PADD 1 states. The monthly models for Connecticut and Maine have been completed (note that time series were checked for stationarity prior to modeling).

h The short-term goal for this work is the completion of monthly demand models for all of PADD 1 by the end of next week. We concluded that it would be best to finish up the monthly models for the entire PADD before embarking on the daily model analysis.

10. Pulp and Paper Forward Curves (Enron Industrial Markets)

h Week-to-week discussions have been held with Maria Theresa Aguilera-Peon (EIM Fundamentals) on pulp and paper forward curves. She is currently beginning work on a pulp forward curve.

Nelson