We are getting a lot of rounding errors; how do we fix them?

Currently there are two ways to deal with a calculation inconsistency known as a rounding error, which occurs due to rounded figures that do not add up correctly due to roundings - for instance writing 2 + 2 = 5 in place of 2.3 + 2.4 = 4.7.

  • The best way to deal with a rounding error is to simply correct your calculations so they do not produce erroneous outcomes such as 2+2=5.
  • The other and probably easier way to deal with rounding errors is to simply do nothing. Calculation inconsistencies caused by roundings are often non-blocking errors, which means you can file your ESEF report even with the errors. The downside to this approach is that anyone in possession of an iXBRL validation tool such as the ParsePort XBRL Inspector will get the same error message when they validate your report.

Why do we get rounding errors?

Rounding errors occur when your reporting includes an erroneous calculation which happens as a result of roundings. While roundings are an accepted accounting principle, they still produce errors when used in XBRL.

If you for instance want to report the following:

1.4 + 1.3 = 2.7

You would normally state in the beginning of your report that you are using roundings, and then turn the calculation into the following:

1 + 1 = 3

In your report the “1” could mean 1000$, in which case the .4 and .3 would be 400$ and 300$ respectively, which doesn’t make a big difference in an annual report. But if that same 1 instead meant 1,000,000,000$ then the omitted .4 and .3 would make quite the difference.

XBRL doesn’t differentiate between the two, and because of this, the system spits out a calculation inconsistency whenever a rounding doesn't sum up correctly.