MTF meeting minutes, 2008-04-25

["Brian Lynn" <brian.lynn@xxxxxxxxxxxxxxxxxxx>]

Attendees
=========
Irina , ISDA
Andrew, Handcoded
Marc G, ISDA
Lyteck, ISDA
Brian L, ISDA
Jason P, BGI
Harry McA, BNPP


Regrets
=======
Pierre L.
Christian N.
Robert S.


Review of outstanding actions
=============================
- IRD - Harry will send out the MTF issues to the IRD group.
- Andrew J. will finish a proposal on more strongly typing the IRS stream
(fixed vs. float streams) to be discussed following the other MTF issues.
- No status on Robert's action to produce message examples.  Robert, can you
please provide an update?
- Brian described a prototype he did of a validation script to validate that
all required messages are present.  Lyteck requested to see it.  Brian has
checked it into SVN, in the version 5.0 branch.  The script to run it is
run-msg.check.bat
- Lyteck developed a summary of FpML messages (discussed below).


FpML Message Summary
====================
We discussed the message summary developed by Lyteck in some detail.

Lyteck discussed the meaning of the various fields.
Adrew J stated that the existing messages haven't always been used as
intended.  For example, the RequestNovationConfirmation message was intended
to end the consent negotiation process, rather than to support confirmation
matching of novations.  This may explain the discrepancy between which
messages are currently defined and what would be required to support a full
post-trade event confirmation process.

Marc asked how to know whether there were correction and
cancellation/retraction messages for each request message.  We discussed the
possibility of adding columns to the spreadsheet to cover the correction and
cancellation message for each request/notification, but did not reach
consensus on this, in part due to the issue above.

Andrew volunteered to produce a simple UML model for a portion of the
messages (e.g. perhaps novation event) showing the various phases - consent
negotiation, execution, confirmation, notification, etc.

Brian volunteered to try to produce a high level summary table (a
"cross-tab") of the various business events vs. the support that FpML has
for them for various processing phases.


Identification Gap analysis
===========================
We spent a short time discussing the gap analysis, but we did not have time
to review it in detail.  No one had any immediate feedback to offer on it,
other than HM commented that the gap analysis seems fairly reasonable.

We agreed to discuss it in more detail next week, focusing on the first few
items, which is where the most serious issues are probably to be found.  For
each of at least the first few issues, Brian will try to propose several
alternative solutions to the issue, perhaps with some pros and cons.

Next meeting
============

May 2 - Discuss object identification gap analysis, messaging summary
classifications