Multi Dimensional Credibility to Estimate Class Frequency - Couret & Venter

Question 1

3 methods for injury type ratios

Answer 1

1. The hazard group injury type ratios (i.e., V_h)
2. The raw sample data injury type ratios (i.e., V_i)
3. The injury type ratios resulting from the credibility procedure (i.e., V_{i est})

Question 2

Holdout Sample

Answer 2

Split a data set into 2 parts; 1 part used as the sample for analysis, and 1 part used as a holdout sample for testing the results of the analysis.

Question 3

Reasons Couret & Venter give why initial SSE test didn’t show much improvement for their procedure

Answer 3

1. Estimators derived from the even year data are designed to fit that data.
2. Class data by year is volatile.

Question 4

Quintiles Test steps

Answer 4

1. Sort the injury type relativities produced by the credibility procedure for all classes in increasing order (i.e., the V_{i est}). We will use these to classify classes into quintiles.
2. Group the classes into 5 groups (quintiles) based on the sorted relativities. The size of the quintiles should be set so that each quintile has about the same number of TT claims. Note that the same class can be in different quintiles for different injury types.
3. Calculate the average injury type relativity for the classes in each quintile, and divide this by the injury type relativity for the hazard group. Do this step for each of the 3 methods and for the holdout sample using their respective relativities. Note that for the hazard group method, this will equal ’1’ for all quintiles, since it assumes the same injury type ratios for all classes within the hazard group.
4. Calculate the SSE for each of the 3 methods.
5. The method that produces the lowest SSE is deemed best, and you would use this method’s injury type ratio for that particular injury for that particular class.