this is where we stand gentleman, once again I can be reached on my cell - 832-656-4598
From: Khanna, Bharat
Sent: Sunday, October 14, 2001 7:10 PM
To: Ramesh, Ganapathy; Patel, Virendra; Mills, Bruce; McIntyre, Burton; Hayden, Frank; Abel, Chris; Basu, Nilay; Victorio, Tom; Keiser, Kam; Jia, Winston
Subject: Weekend Events.........
As most of you already know, it has been a crazy weekend for some of us. I'm going to address why all this happened, what we have done, and the ongoing problems we are having?
About 8 months ago Martin Cuilla took a position on a curve called GD-AECOUSD-DAIL. On Friday he wanted to move this position to another curve called GD-AECOCD/FRWKD. We had being seeing the position come in on the bench report but apparently VaR was not being captured on it. It was not in the Primary/Secondary curve mapping table in Risktrac and was not mapped to an intra-month curve either (OMICRON).
On Friday evening I asked Burton to map the new curve in the primary /secondary table and Virendra to map it to OMICRON_9. If this would have worked we expected Cuillas VaR to fall by approx. $800K given the risk structure of his positions. But the desired results were not achieved.
On Saturday with the help of Ramesh, Virendra and Burton we tried several different combinations of mappings with no positive results. The positions do not seem to be coming in as we are not picking up the book in Component Var. The book should be INTRA-ONTARIO-CAD2. It exists but, there are no positions in it.
The only reason this would happen if the curve had no Volatilities or Prices associated with it. We need to find out why this is so? Understand also that AECOCD/FRWKD is a derived curve so it has no Prices and Vols of it's own but takes it from AECO-CDN/IM. We have also mapped AECO-CDN/IM to CGPR-AECO/BASIS and to OMICRON_9 but still are not getting any changes.
Ramesh, Burton Virendra, and Bruce have all worked very hard this weekend, despite having other work related commitments, to try to resolve this, and have gone over and above what was expected of them. But despite all our efforts this is still a problem that needs addressing.
Cuilla is very concerned about this because Lavorato is watching him. I have to answer to him tomorrow and cannot use systems issues as an excuse. I want us as a team to look at this tomorrow and in the short-run fix Cuillas VaR. We also want curve mapping to become a more front-end solution (with very few users) because of their (the front office) knowledge of where the different curves need to be mapped. Kam Keiser will help us with this.
In the long-run I don't want any of us to have to do this again and want to identify a solution. Ramesh has rightly pointed out that this will take time and resources on his side. Given the speed of the work in this place none of us have either extra time or resources but, we have to find a way and, most of all be patient while we address these issues.
Please understand that tomorrow may be a little hectic for all concerned parties but to be patient while we try to reach the root of the problem. Nilay we will need your expertise about curve mappings, and OMICRONS, for this?
Thanks again for all your help this weekend.......