[Discuss] Financial database / balance?

[richard.pieri at gmail.com (Richard Pieri)]

On Jan 15, 2012, at 4:13 PM, markw at mohawksoft.com wrote:
> 
> 0.999999999999999 for all intent and purposes is the same a one. And a 0
> with an infinite number of 9s after the decimal point equals one.

This is true for scientific math, but not for financial math.  I find it confusing that you make this statement immediately after making the point (which I agree with) that financial math is different from scientific math.

Financial math in the US usually is accurate to 3 decimal places, 1/1000th of a dollar, because that is the smallest unit of currency defined by the Coinage Act.  It's called the mill.  Banks and such don't recognize "a 0 with an infinite number of 9s after it" as valid.  This is why 0.9-bar dollars is not the same as 1.00 dollars.  There may be a mill missing, depending on how that figure was obtained.

A lone mill doesn't seem like much of an error, but remember: every calculation based on that figure will be off, and every calculation based on those calculations will be off, and so on.

Amortizing an IRA is definitely something that individuals do with personal financial software.  We ask "what if?"  What if I add X dollars every year for the next five years?  What if I add 3/4 X?  Will that be enough for my retirement?  A floating point rounding error will cascade through the amortization and throw off the long-term results by noticeable, possibly significant, amounts.

I reiterate my statement: wrong is not good enough, not even for personal finances.

--Rich P.