other reserving

Question 1

general order of diagnostic triangles

Answer 1

paid counts / disposal rates -> settlement rate

rptd counts -> reporting pattern and/or freq

avg paid on closed counts -> severity/inflation

avg reported on reported counts -> strengthening

avg C/O -> strengthening

Question 2

why are tail factor selections important?

Answer 2

selection affects all accident years reserve needs, thus has a disproportionate leverage on the total reserve need

Question 3

difficulties in selecting a tail factor

Answer 3

generally little relevant data that can be used in the tail

Plus, due to the age, claims trends, benefits levels, and policy limits may have changed, so this must be considered in the selection.

Question 4

estimating ultimate losses or IBNR for non latest AY using CC

Answer 4

calculate ECR @ loss levels for the AY you are estimating for

ie. your calc ECR is for 2012 but estimating 2010 losses

need to use ECR @ 2012/(tort factor and loss trend used to adjust 2010 to 2012 levels)

Question 5

disposal rate technique

Answer 5

disposal rate = cumulative closed/ultimate claim counts

proj incremental closed counts = (ult-diag)/(1-DR) * (DR2-DR1)

need trended incremental avg paid on closed claims

=incremental paid on closed claims/incremental closed claim and then use trend

Question 6

considerations when selecting the ultimate claims:

understanding of the underlying data

Answer 6

This includes trends and changes in the data, such as mix of business, underlying limits, reinsurance contracts, and external changes, such as laws and regulations. Other important changes to consider include operational changes such as changes in case adequacy, claims closure rates, and significant changes in the definition of counts.

Question 7

considerations when selecting the ultimate claims:

retrospectively testing

Answer 7

Whenever there is a sufficient loss history, we should retrospectively test each method to determine its past accuracy and lack of bias in projecting future losses. To do this, we can use past years data to project the most recent year’s data. Methods can be adjusted to remove any detectable bias.

Question 8

considerations when selecting the ultimate claims:

diagnostic review of projected ultimate values

Answer 8

This includes such a review of claim ratios, frequency, severity, pure premium, and average unpaid claims. These can be used to check the reasonability of the ultimate selections by looking at the various year to year progressions. Any anomalies should be explained. Differences in projections may be due to factors such as large losses, highly leveraged development factors, changes in the underlying assumptions