Chapter 20 - page 25

Hi,

I am not very clear in terms of the approach to price proportional covers. Can I clarify if the below is right based on my understanding of the core reading for page 25-26?
1. Collect the triangulated data (both premium and claims (paid and incurred))
2. group them into homogeneous groups
3. project both the triangulated premium and claims data to ultimate position for each historical year
4. calculate resulting loss ratio for each year (by dividing the ultimate claims over premium)
5. apply trend to the loss ratios (why did we not trend the claims and premium data prior to developing to ultimate like frequency-severity approach? also, when do we trend the premium data to on-level premium as the core reading did not mention when this is done)
6. average loss ratios to get estimate of expected loss ratio for period of cover (or consider other trends of loss ratios e.g. upward/downward over time)

 

Thanks.
If I had the data to trend losses and premiums separately, should I do it before step 3 i.e. before developing them to ultimate? Or it does not matter if I do it before or after I develop to ultimate?