C.2 Univariate Classification

Question 1

Main distortion in the Pure Premium Approach to
class ratemaking

Answer 1

It assumes that there is no correlation between exposures for different rating variables (i.e., distributional bias). If the exposures of the levels of the rating variable you are analyzing are correlated with the exposures of the levels of another rating variable, then this approach will “double count” the experience of those levels.

Question 2

How the Loss Ratio Approach corrects for

distributional bias

Answer 2

Since it uses premiums instead of exposures, the premiums reflect the higher (or lower) levels of premium obtained within a class as a result of correlation with other higher (or lower) rated levels of other rating variables.

Question 3

Advantages and a disadvantage of the Adjusted Pure

Premium Approach

Answer 3

It doesn’t require on-leveling the premiums (as required in the Loss Ratio Approach). Also, it attempts to correct for distributional bias by adjusting exposures based on the levels of other rating variables.

A disadvantage is that calculating the weighted average
relativities can be cumbersome in a rating plan with many
variables.