along with an increase in the age of policyholders. In the case with the age categorical, is also noted the
concave shape of the fitted frequency of claims, obtaining high values for the category of
young drivers, an obvious decrease over the years, but a slightly increase in elderly drivers
category. Taking into account that among certain age groups the estimated frequencies of
claims do not differ significantly, this variable could be grouped into fewer categories
considering the breakpoints that can be easily observed on the right side of the graph.
Therefore, based on this graphic representation, the policyholder’s age could be grouped on
year intervals as follows