joeu2004 wrote :
"Bruno Campanini" wrote:
I suppose the interest rates 6.25, 7.63, 8.9, 10 and 8
are effective annual rates to be converted in effective
infra-annual rates as follows:
i2 = (1+0.0625)^(1/2)-1 FV = 5000(1+i2)^24
i4 = (1+0.0763)^(1/4)-1 FV = 5000(1+i4)^24
i12 = (1+0.089)^(1/12)-1 FV = 5000(1+i12)^120
i365 = (1+0.1)^(1/365)-1 FV = 5000(1+i365)^1095
[....]
These calculations may differ from the Excel's ones.
In fact Excel uses the standard US banking calculations
which take into account bank profits instead of
mathematical principles.

That assertion is patently incorrect. In fact, FV(i2,24,0,-5000) returns
exactly the same result as 5000*(1+i2)^24. The same is true for the other
three examples.

Yes of course! But you must separately calculate i2 from i
with the formula i2 = (1+0.0625)^(1/2)-1

The standard Excel financial functions (PV, FV, NPER, PMT and RATE) do not
take any "standard US banking" practices into account, much less calculations
based on "bank profits instead of mathematical principles".

They do!
In my country (Italy) you can negotiate a financial operation with a
bank at 5% annually compounded.
Then if you change your mind and ask for semi-annually compounded
operation you see that the rate is <2.5% as per i2 = (1+0.05)^(1/2)-1
I've been told that normally that's not the case in US.

Users of those functions might, however. As I noted: "there are two schools
of thought on how to convert an annual rate to a periodic rate. You need to
do it according to the method used by the particular investment. Sometimes
that is dictated by regional law".

I don't think there are regional laws in any civilized country stating
that an annual rate in converted in quarterly rate with a simple
division by four!
May be there are conventions that the single small operators cannot
change against the bank system.

I cannot say the same with impunity about Excel bond functions (e.g. YIELD
and PRICE). They might or might not rely on some US-specific bond
conventions. Honestly, I don't know since I am not familiar with bond
conventions the world over.

In any case, I am quite sure that any such US-specific bond conventions are
not motivated by "bank profits instead of mathematical principles". More
likely, they were developed in pre-computer times in order to make
computations easier to do by hand or with a mechanical calculator using
predetermined tables.

You are too much optimistic!
In my country, many years ago, all banks calculated the interest to pay
to clients using the convention of 1 year = 360 days.
Well, such method allowed easier computations, of course.
But when they calculated interest to be payed by clients they (the
banks) were happy to effort more complicated calculations taking into
account 1 year = 365 (or 366) days...
Such a behaviour has a name: robbery!

Bruno