Fixed-point also has guaranteed imprecision for many operations, because you only have a finite number of digits after the decimal point.

e.g, with two decimal digits: (2.83 * 0.10) = 0.283, which is stored as 0.28.