RE: [office-comment] ODFF: XIRR

["Dennis E. Hamilton" <dennis.hamilton@acm.org>]

[moving to opendocument-user list for continuity with remarks there]

Here:
http://lists.oasis-open.org/archives/opendocument-users/200809/msg00002.html
I did manage to find some resources to check out, and tracing references
reveals some authoritative sources.  I agree that these vary from
jurisdiction to jurisdiction and particular regulatory structures on
accountancy and financial activities.  I'm also learning that financial
mathematics is not static, and experts have to deal with using formulae that
applied in particular time periods.

I remain hopeful that there is enough to at least establish what the
foundation is for a given approach, whether one specified in OpenFormula or
in OOXML or any other specification that should be disclosing the *specific*
financial functions its algorithms implement.

One should not be surprised (even though I often am) that context is
everthing in these matters, and that offering context-free solutions is not
helpful.  I am aware of this with regard to treatment of dates and times in
computer software, and I look forward to the results David Wheeler's
efforts.

 - Dennis

PS: I'm with you about basic principles.  It would be useful if we knew what
the basic principles, if any, are that a specification intends to have be
applicable, and then that an implementation claims to satisfy.

-----Original Message-----
From: David King [mailto:openoffice@drking.plus.com] 
http://lists.oasis-open.org/archives/office-comment/200809/msg00019.html
Sent: Tuesday, September 09, 2008 03:22
To: office-comment@lists.oasis-open.org
Subject: RE: [office-comment] ODFF: XIRR

Hi Dennis

> Can you recommend authoritative sources

In short, I'm afraid not - there are many books about, but because I'm not a
specialist I think it would be unwise for me to recommend. I found mistakes
in one book... And many sources skirt around the detail.

The good news is that we now have a pretty good understanding of what the
financial functions actually do - and I think ODFF will nail it down so that
future implementors can succeed (in a way that OOXML has not). The main
outstanding issue is probably the calendar systems, which David Wheeler is
working on I think.

The bad news is that it's hard to know how useful (and used) the financial
functions are. Two of the Calc functions always return an error, and no-one
complained (I've raised an issue now). I suspect that despite the huge
effort of implementation, not many people find them useful, except for
students 
doing assignments. Personally, as things are at the moment, I'd often be
more inclined to go back to basic principles, so I could be sure of the
result...

> I'm concerned that we have authoritative subject-matter expertise for 
citation and also comparison with what various programs do. 

Agreed, a good point, but it's hard to find the right *international*
expertise. Or at least I've found it so. I think the financial functions are
quite biased towards the US.

We also have to remember that a spreadsheet is a general tool, and that
specialist software will always do more for that area. Like R and
statistics.

Not sure it that helps at all...

David

[ ... ]