[Quantlib-dev] [ quantlib-Bugs-1177912 ] More Calendar problems on all Libors

[SourceForge.net <no...@so...>]

Bugs item #1177912, was opened at 2005-04-06 18:31
Message generated for change (Settings changed) made by lballabio
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=112740&aid=1177912&group_id=12740

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: None
Group: None
>Status: Closed
>Resolution: Fixed
Priority: 5
Submitted By: Daniele De Francesco (qq7te)
Assigned to: Luigi Ballabio (lballabio)
Summary: More Calendar problems on all Libors

Initial Comment:
After looking for a proper definition of how the Libor
fixing works, 
I have finally found an authoritative source:
http://www.bba.org.uk/bba/jsp/polopoly.jsp?d=225&a=1412

It seems that any calendar we pick will be wrong. 
Their method for determining the value date from the
fixing date is quite different from anything I've seen.
It's a strange combination of two calendars that is
explained fully in the above mentioned page. 

Picking the London calendar for ALL the libors will be
a very good first-order approximation that will make
all the libor classes much more accurate. 

Eventually a fix is needed to have an
"OddlyJoinedCalendar" or something like that :-)

I would fix it myself, but I will have to do it outside
of work because of my emplyment agreement, so it will
be a bit delayed. If anyone with access to CVS wants to
at least do a quick fix to move to London Calendar that
would be a great first step

-Daniele

----------------------------------------------------------------------

>Comment By: Luigi Ballabio (lballabio)
Date: 2005-06-07 17:57

Message:
Logged In: YES 
user_id=75450

The version in CVS should follow the right conventions. It
will appear in version 0.3.11.


----------------------------------------------------------------------

Comment By: Daniele De Francesco (qq7te)
Date: 2005-04-13 22:58

Message:
Logged In: YES 
user_id=288761

Careful using the JointCalendar. 
the actual day calculation would not be accurate using
either the London calendar alone or a joint one. We need to
come up with something that meets the specification
published at the link that I posted. 
I will contribute something as soon as I get a chance.. :-) 

----------------------------------------------------------------------

Comment By: Luigi Ballabio (lballabio)
Date: 2005-04-08 09:17

Message:
Logged In: YES 
user_id=75450

Thanks for the link. The CVS version already uses the London
calendar for all LIBORs. We'll have to check whether we can
use JointCalendar, but that will have to wait for 0.3.10 as
0.3.9 is being released.


----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=112740&aid=1177912&group_id=12740