Chapter 32 Flashcards
(14 cards)
What is the purpose of policy data checks?
To ensure that the data are:
* Adequate
* Complete
* Accurate
* Maintained
* In the required format
What are the main checks that can be done on policyholder data?
- Reconciliation of current data with data from previous valuation using data movements
- Check data movements against accounting data
- Consistency checks
- Check for unusual values
- Spot checks for accuracy of individual records using other records and administration files
How does data reconciliation work?
- Where an investigation is being carried out on a regular basis, a reconciliation of the current data with those used for previous one can be attempted.
- Data are first grouped in a sensible way
- And then the following check is made for each group: (data at previous investigation) + (business come onto the books) - (business gone off the books) = (data at current investigation)
Which items of data should be checked for non-unitised business:
- Basic sum assured
- Office premium
- Amount of any attaching bonus
- Number of contracts
Which items of data should be checked for unitised business:
- Current benefits available
- Number of contracts
- Number of units actually allocated, subdivided by unitised fund
- Current premium payable
What is important to remember with regards to the movements data?
- The data must be verifiable and independent of the in-force data
- It should be checked against any appropriate accounting data, especially with regards to benefit payments
- The systems for producing the movements data must be checked periodically to ensure that they are working correctly and that the staff involved are following the procedures laid down
What are examples of consistency checks on non-unitised policies?
- The average sum assured, and premium should be sensible and consistent with the figure from the previous investigation
- The ratio of the basic sum assured to the premium (for each class of regular premium contract) should be sensible and consistent with the figure from the previous investigation
- Useful rules of thumb
– With-profits EA premiums are quite frequently > SA/term
– Without-profits EA premium will normally be < corresponding with-profit one
What are examples of consistency checks on unitised policies?
- The numbers of the units purchased by the premiums and encashed to pay benefits are consistent with the corresponding revenue accounts
- Check that the internal unit movements (e.g. charge encashments) are consistent with the surplus emerging during the year
– “Charge encashment” - refers to monetary expense and mortality charges that are deducted from the PH’s funds by cancellation of units
What are examples of unusual values that should be checked for?
- Very large or zero-unit values under the unitised contracts
- Impossible start dates
- Impossible DOB and retirement ages
- Missing data/fields
- Unusual special terms, e.g. massive increases to sum assured
What would we want to check for ages specifically?
Current and retirement ages (through DOB) - min, max, mean (currently and vs the previous valuation)
What is an alternative to looking at individual data values?
It may be possible to group items and look at how well distributed they are and compare this with the previous investigation, for example:
* Count policies by age
* Count policies by gender
How can we check computer held data?
Compare it with the information in the paper admin files. This is done on a spot check basis by randomly selecting a number of policies.
What checks can be done with an AOS and change in EV?
Identify any major discrepancy in the analysis compared to the previous analyses and this may indicate a problem with the data.
What is important for an AOS?
Make sure the results explained what happened throughout the year (ask managers what they experienced)
