Questions in frequency-severity

[Adithyan]

I have the following questions pertaining to F-S approach:

We don't take in to account the IBNR claim amounts in the severities (These are basically ultimates of open claims and paid claims trended to period for which rate is needed - Chap 20 mentions this way) to which we fit a distribution. We are taking ultimate claim counts (which is a combination of Reported and IBNR counts) in the frequency distribution. We are here not accounting for IBNR claim amounts in severity. Doesn't it become inconsistent? (Because IBNR claim count accounted in frequency distribution but no IBNR component in severity). What is the reason behind not including IBNR claim amounts in Severity distribution?

I have a small question regarding trending severities

Are we just aggregating losses by cohort year and then looking for a trend in those numbers?

Does the order in which trending and development of losses matter?

Thanks in advance for your time and efforts to help!

 

[Hemant Rupani]

IBNER is included, pure IBNR is not.
severity is a loss per claim, as pure IBNR is not yet reported we can't take into account of distribution. even if you take you would assume pure IBNR severity as same as the severity of reported claims.