The attempt is unsuccessful, since rounding as well as further casting to DECIMAL type (which is redundant in the given context, though) is carried out for a value which is already integer. In consequence, we manage to get not more than two zeroes after decimal point.