Sorry to hear about this one. 

I have some observations to make from my reading (and re-reading) of the 
account:

1 It looks like the old "undue influence" problem again. It seems some 
attempts were made by operations to follow up, but nothing happened. 

We tell our operations staff they have an obligation to dissent but we don't 
have any way to protect them if they do - there is still the fear that if 
they upset a trader they can get fired or held back in their career. And that 
is because twice a year, at the PRC, that can happen. This incident shows 
some of these attributes.

For the operations process to be effective there has to be (and be seen to 
be) little significant influence from Commercial Groups over the ranking of 
the operations staff. 

2 Whoever signed off the new deal lists in the DPR should be asked to confirm 
the basis on which they signed it.

3 Why wait for the traders to approve payments to brokers ? Brokerage is in 
the P&L and if operations are satisfied the trades are accurate and brokerage 
is at the agreed rate, they should simply pay the brokerage.

4 Two weeks to check telephone tapes is unacceptable - it should be done same 
day. I would not have waited for the traders to do it, but I would want to 
know why it took them so long.

5 Given that these (short) trades were absent from the portfolio during a 
highly volatile period of rising prices, I would wonder about motivation in 
addressing the position during this time. Essentially, if it had been correct 
we would have been reporting losses bigger than the ones we did.

I would pursue explanations to points 2, 4 and 5. 

DP





Fernley Dyson@ECT
06/12/2000 02:41 AM
To: James New/LON/ECT@ECT
cc: Richard Lewis/LON/ECT@ECT, Mike Jordan/LON/ECT@ECT, Robert 
Yeo/LON/ECT@ECT, Gail Hill/LON/ECT@ECT, Ian Sloman/LON/ECT@ECT, Steve W 
Young/LON/ECT@ECT, David Port/Market Risk/Corp/Enron@ENRON, Brent A 
Price/HOU/ECT@ECT 

Subject: Re: UK Gas late deal input and deal booking errors - o1.1m hit  

James,

Thanks for the note. As described, this is an intolerable situation. I agree 
your action points, but we need timelines and quantitative measures of 
compliance/improvement - please could you provide a weekly update on this 
area specifically and until we are satisfied that things are as they should 
be.

On a more general note - let's ensure that the 'Global Standards' address the 
lapses in timely action/follow-up identified here so that we can monitor 
performance across the business.

Fernley



James New
09/06/2000 10:03
To: Richard Lewis/LON/ECT@ECT, Fernley Dyson/LON/ECT@ECT, Mike 
Jordan/LON/ECT@ECT
cc: Robert Yeo/LON/ECT@ECT, Gail Hill/LON/ECT@ECT, Ian Sloman/LON/ECT@ECT 

Subject: UK Gas late deal input and deal booking errors - o1.1m hit


On the 7th of June Gas Risk Management became aware that two'late' trades had 
been entered into the gas trader's 'Blue book' which were both being booked 
significantly late and had a material negative P&L impact (negative o1.1m). 
The 'late' deals were done on the 3rd of February (Marcello Romano) and the 
30th of March (Haakon Olafson) and were both forward starting sells for 
periods in 2003-2004.

Background

1) Input errors

For some considerable time we have been concerned with the number of deal 
input errors that have been occurring in the Gas area with the most 
significant number coming form the UK Gas traders. Numerous errors have been 
found by documentation and risk management over the accuracy and timeliness 
of entering deals into the 'blue book' (the Gas traders are currently the 
only traders in Europe who do not enter their own trades into the valuation 
system - there is currently a big push to change this which has the support 
of John Sherriff with a deadline date of 31 July).

In the past documentation and gas risk management have repeatedly highlighted 
booking errors to the gas traders but the error rate has not fallen. A log of 
all deal input errors has been formally maintained since the beginning of May 
which highlights an error rate of 12%-15% of non EOL trades. This has been 
discussed with Richard Lewis and corrective action is expected.

2) Deal booking and reconciliation process (Non EOL trades)

Deals are initially written into the 'blue book' by the gas traders when the 
deal is done. This book is photocopied several times a day and the deals are 
the manually input into GasDesk by documentation on the same day.

Faxed broker statements are received (usually) by documentation the next day 
and these are reconciled to GasDesk with errors and amendments made by 
documentation liasing with the traders as necessary. At this time brokers 
were faxing one statement per trade (relevant later).

3) New Deal sign off

As part of the daily DPR process a separate new deal sheet is prepared and is 
signed off by the traders to formally agree the new deals done on the day and 
the implied origination P&L. For the 3rd of February Marcello signed off the 
new deals sheet and for the 30th of March Jonathan Whitehead signed off.

4) Documentation convention

In the gas market the convention is that the seller sends deal documentation. 
(relevant later). Documentation aim to get this out within one the day after 
trade booking.

5) Brokerage invoices

Brokerage invoices are usually received monthly by settlements who then 
reconcile the trades on the invoice to the valuation system to ensure the 
deal is valid before passing the reconciled invoice to the traders to 
authorise for payment. Any deals which cannot be found are raised with 
documentation who then investigate and follow up with the traders.


Specific details of these 'late' trades

1) Summary

For both these trades no evidence can be found of receiving, or can it be 
established that we did not receive, a broker statement which is a primary 
control in agreeing the existence and accuracy of new OTC deals. 

As Enron was the seller in both cases no documentation was received from 
either counterparty which would have alerted us to the trade.

The reconciliation of the brokerage invoices was carried out by settlements 
in April and was completed by the 20th with the invoices falling due for 
payment on the 27th. The work practice at that time was that brokerage 
invoices (including those with non reconciling trade queries) were batched 
together and passed to the traders. This would have taken place on the 27th 
of April. Details on exactly what was said and done in the 3 weeks between 
passing these invoices to the traders for approval (although there are claims 
that verbal follow up were made on a number of occasions) and when a mailed 
follow up was made on the 22nd of May are hazy. What is clear is that on the 
22nd of May Morgan Stanley contacted Enron to try to confirm their position 
which was when it was realised that an error could well have occurred.

Over the next two weeks tapes were then checked by the traders with the deals 
being booked on the 6th of June.

2) Deal specifics

EES Broker - Morgan Stanley trade on 3 Feb 2000
Trader Marcello
EES Invoice date 1 March; Invoice received before 15 March as settlements 
started checking it. (Keren Perrott - settlements started on 13 March)
Discrepancies sent to documentation on 23 May (Chris Ware to Emma Thomson)
Traders contacted  on 24 May after copy of broker confirm was sent on 22 May 
as requested by documentation to prove deal.
Deal entered 6 June after traders checked the tapes.


IEB Broker - NP trade on 30 Mar 2000
Trader Haakon
IEB invoice date 6 April; Invoice received before 27 April as settlements 
started checking it. (Keren Perrott)
Discrepancies sent to documentation on 23 May (Chris Ware to Robert Vaughn)
Traders contacted on 24 May after copy of broker confirm was requested.
Deal entered 6 June after traders checked the tapes.



 Who was at fault and where did the control break down ?

I do not feel it is constructive to try to attach blame for what happened but 
there are lessons to be learned and general observations and conclusions that 
can be made and drawn.

Firstly there are major concerns over the Gas trader's accuracy and 
prioritisation of booking trades. Despite numerous conversations on the 
subject very little improvement has been seen. It has now got to the stage 
where documentation are formally recording all errors including what the 
error was and who made it. This is now being communicated to Richard Lewis 
every two weeks with the aim being not only to reduce the amount of time 
trade support spend correcting errors and re running DPRs but also to try to 
avoid the incidence of errors such as this. The current error rate is 12% - 
15% of non EOL trades. Documentation currently estimate that in 9 times out 
of 10 the broker or counterparty are correct.

Secondly around a month ago we contacted brokers and asked for them to 
provide us with one statement per day rather than one per deal which will 
eliminate any lost statements or missed deals as it will be immediately 
obvious when a statement from a broker has not arrived. Most brokers have now 
changed but one or two still send one statement per trade. Reconciliation's 
of these daily statements to GasDesk are evidenced.

Thirdly it has been impressed on both documentation and settlements to follow 
up more vigorously (and by mail) errors and omissions on a more timely basis. 
There is however a backlog or broker invoices which have have been checked by 
settlements but have nto been checked or approved by the traders. This is 
causing concern at the brokers who are chasing payment.

Action Points

Trading

Traders to input their own deals directly into GasDesk

Traders to improve their accuracy of deal booking

Traders to approve invoices on a more timely basis

Trade Support

Settlements to clear backlog of broker invoices

Documentation to chase brokers to send daily deal statements not individual 
deal statements

All follow ups to be evidenced with mails used as appropriate

Documentation to check all trades since the beginning of the year versus the 
broker records

Documentation to perform a reconciliation with all counterparties to agree 
our forward position from July 2000 to September 2006 (the traded part of the 
curve)