8.5 Issues with Assumptions
Three assumptions that are likely reasonable
Must select a basic limit \(B\)
Needs an ultimate claim size model
Or use ILFs
Just have to accurate around the basic limit and policy limit since that’s where we calculate expectations
Need to create a triangle of losses at basic limit and one cost level
- We do this in the step converting the base triangle
Two assumptions that are more tenuous
Requires robust claim size model at each age
- Ultimately we need the ratio of limited losses at different layers, less import to get the absolute values correct
Trend
Trend should be applied to the incremental losses
Not clear how to apply trend to reported losses (by day of reserve or payment?)