Re: COUPDAYS Help needed

["Andreas J. Guelzow" <aguelzow taliesin ca>]

Neil Booth wrote:

> Andreas J. Guelzow wrote:-
>
>
> > Perhaps you can find a justification for the both of the 2  values:
> >
> > coupdays("25-Oct-2019","28-Feb-2023",1,1) == 365
> > coupdays("28-Feb-2019","29-Feb-2020",1,1)  == 366
> > this is in Act/Act, annually.
>
> Do these differ between Mac and Windows?  Or is this something you
> noticed?  (I don't have Excel available to play with at home).
>
>   coupdays("25-Oct-2019","28-Feb-2023",1,1) == 365
>
> I believe that should be 366 (OK, I'm sure it should be, since Excel
> defaults to RollEOM).  Do you agree?
>
>   coupdays("28-Feb-2019","29-Feb-2020",1,1)  == 366
>
> I reckon that's correct.
>
> The more I discover, the more useless I believe Excel's bond analysis
> functions to be.  I wonder if MS uses them in-house.



The numbers 365 and 366 are what Excel (both on Windows and on Mac) 
appears to evaluate them as. Since I don't have a windows version myself 
I ended up extracting the info from Jps files, and I can't be 100% sure 
I didn't mess up there. But on Mac we definitely get those two answers. 
To me they appear to be contradictory, which means it is pretty hard to 
match them both.

Andreas



--
Prof. Dr. Andreas J. Guelzow
http://www.math.concordia.ab.ca/aguelzow