[rory]

It clearly does fit the rule because the year doesn't start in February. There's no need to change reality so you were right.

It's just a rule of thumb, not some profound maxim of economics.

[OJFord]

The point is that if all the bad happened in half the quarter, that can make quite a difference, and the quarter start/end date is entirely arbitrary.

That said, you can apply the same to GP's suggestion of months, if the timescales are even shorter.

[rory]

Yeah, the cutoff points for any rule of thumb are going to be fairly arbitrary. I don't understand why people are treating this like a profound observation.