Message-ID: <1055837.1075856188071.JavaMail.evans@thyme>
Date: Thu, 26 Apr 2001 01:45:00 -0700 (PDT)
From: joaocneves@email.msn.com
To: tanya.tamarchenko@enron.com
Subject: Interview - Numerical Methods & Finance
Cc: vince.j.kaminski@enron.com
Mime-Version: 1.0
Content-Type: text/plain; charset=ANSI_X3.4-1968
Content-Transfer-Encoding: 7bit
Bcc: vince.j.kaminski@enron.com
X-From: "JoaoCNeves" <JoaoCNeves@email.msn.com>
X-To: <tanya.tamarchenko@enron.com>
X-cc: <Vince.J.Kaminski@enron.com>
X-bcc: 
X-Folder: \Vincent_Kaminski_Jun2001_1\Notes Folders\All documents
X-Origin: Kaminski-V
X-FileName: vkamins.nsf

??? ??? Dear  Tanya:
?
??? 
??? ??? It was a  great pleasure to have met you. I very much enjoyed the 
interview and your  insightfull questions.
?
??? ??? I am keenly  aware that many of the methods that I discussed with you 
yesterday are unique,  new?and not reported elsewhere. This is true both 
about the work I did in  whole yield curve interest rate pricing?as well as 
GARCH. The innovations  stem from the extensive numerical analysis experience 
that I have both in  turbulence physics as well as finance. They entailed 
considering the problem  from its raw formulation, mathematical analysis, 
physical interpretation,  taylored numerical method development, software 
writting and develoment and data  management.
?
??? ??? As to why I  have not yet published anything the answer is that the 
driver in my work has  been adding value to the business not publishing. 
Publishing is however an  option that has always been open with my former 
supervisor who is aware of the  work that I did.
?
??? ??? I not however  that these results were possible only by exploring to 
the utmost extent the  mathematics, finance, software design and data 
managemnet aspects of the  problem. Absence of any of these aspects is likely 
to cripple performance and  execution.
?
??? ??? Please recall  that as good as they were the performance measures 
that I mentioned to you were  for a single processor machine. Vastly better 
can be achieved with both soft  parallelism (multithreading) as well as hard 
parallelism (heterogenous network).  This fo course allows us to step up the 
reach of the models used.
?
??? ??? In fact I  know for a fact that better can be done than what I 
mentioned in the interview.  From work that I have been doing on the 
integration of the swaption volatility  surface on the whole yield curve 
interest rate model ITM and OTM instruments can  be included in both the 
callibration, pricing and hedging. 
?
??? ??? I look  forward hearing back from you soon and particularly to the 
opportunity of us  cooperating.
?
??? ??? Best  regards
?
??? ???  Joao