On May 22, 2001 HoustonStreet representatives (Frank Getman, John Wyman and David Cuadros) came to Enron to give us a demonstration of the gas interfaces.  HoustonStreet was allowed to load the required .cab files on the demonstration computers prior to the demonstration.  HoustonStreet provided us with a script to follow relative to the demonstration as well as a list of test and production mappings.  During the demonstration we experienced the following:

1 - 8 second delays between a price update on EnronOnline and a price update on HoustonStreet.  With just five products active we experienced a delay of up to 5 seconds.  Due to the delay in price updates, transaction attempts on the HoustonStreet platform resulted in a high percentage of failed transactions with Enron.
When we activated Product ID 16612 (US Gas Basis PEPL Jun-Oct01) the product showed up on EnronOnline but did not show up on HoustonStreet.  HoustonStreet provided us with a list of products which had been mapped in the test environment and this product was included on the list.  HoustonStreet indicated that they had verified the mapping of all the products on this list.  We reviewed the mapping of this product by Enron and confirmed that it was active and mapped.  Incorrect mapping of products has the potential of allowing transactions to occur with mismatched product types.
When we attempted to transact on Product ID 16600 (US Gas Basis NGI Chicago Jun-Oct01) we were unable to complete a transaction even when the bid and offer prices were not changing.  These transaction attempts should have resulted in successful transactions.  We also experience instances where the bid and/or offer price would disappear after a failed transaction attempt and not reappear even thought the price was still present on EnronOnline.  In many cases the price would disappear for up to 30 seconds or more.  Although we did not witness any incorrect transactions, general functionality failure may or may not result in incorrect transactions.  At a minimum this type of behavior would result in frustration on the part of the counterparty when a transaction attempt results in a failure with no subsequent price move or a disappearing price.
A review of the transactions we recorded versus the transactions indicated by HoustonStreet reveals some discrepancies.  In an email from John Wyman dated 5/23 (forwarded seperately), John Wyman indicated that they had cleaned their database just prior to 3:00 PM and gave us just the trade count beginning at 14:58 on 5/22.  Although there has not been a deal by deal reconciliation of the trades, we compared our total trade count beginning at 14:58.  If the deal counts do not match this would result in substantial financial exposure.  Below I have highlighted the HoustonStreet transactions by counterparty versus the EnronOnline transaction count:

			HSE	EOL

	Arizona		 12	 14
	Avista		 03	 03
	Adams		 17	 19
	BP		 57	 58
	
	Total		 89	 94
	
Based upon my review of the demonstration, I continue to believe there significant potential exposure associated with the HoustonStreet interfaces.



Bob Shults
EnronOnline LLC
Houston, Texas 77002-7361
713 853-0397
713 825-6372 cell
713 646-2126
bob.shults@enron.com