Joe and Jerry, thanks both very much for your input. Looks like I've got a
bit of work to do to prove categorically how the interest is calculated,
particularly as regards rounding. I never realised there could be so many
different approaches to when and how rounding occurs. Leave it with me,
thanks.

"joeu2004" wrote:

On Mar 3, 4:41 am, Jerry W. Lewis wrote:
"joeu2004" wrote:
Note: The bank might use "banker's rounding" rules, which always
rounds a half-cent to even cents.
[....]
I would be very interested in any evidence that this rounding rule has ever
been used in banking.

I would be very interested in any evidence to the contrary; or in
evidence that banks use the "normal" rounding rules. I would also be
interested in evidence that God exists, or that she does not exist.

As Harlan pointed out, some banks might still be using software
written in computer languages that might implement "banker's rounding"
together with decimal arithmetic -- for example, COBOL and RPG, as I
recall vaguely. That is sufficient "evidence", I think.

But the fact is: the difference between "banker's rounding" and
"normal" rounding arises only when the "fraction to the right" (for
want of a better description) is __exactly__ 5, and then only when the
digit to the left is odd. I suspect the situation arises almost never
in real-life daily and monthly interest calculations, unless the
bank's software truncates the intermediate amount to the "decicent"
before rounding, which I think is unlikely.

So notwithstanding the correctness of my original comment, it was
probably frivilous to mention because it is unlikely to make any
difference. Sometimes I get "obsessed" with being unduly complete.