32. Policy data checks

Question 1

What are the types of checks that can be done?

Answer 1

• Data movements reconcile between start and end period
• Consistency (numbers sensible?)
• Unusual values
• Analysis of surplus (unex[lained items?)

Question 2

For recon checks what must the data at current investigation be equal to?

Answer 2

Data at previous investigation + business come onto the books - business gone off the books

Question 3

Which items of data can be reconciled for movement checks?

Answer 3

Non unitised
* Basic sum assured- or equivalent benefit
* Attaching bonuses for w/p

Unitised:
* # of allocated units subdivided by unitised fund
* Benefits available e.g. death benefit
* Changes in number of units allocated due to switches between funds
* Changes in premiums and benefits

Both:
* # of contracts
* Premium payable
* Check systems periodically to ensure working correctly and staff following procedures
* Check against accounting data esp benefit payments

Question 4

Give examples of consistency checks for non-unitised contracts

Answer 4

• Ave sum assured or premium must be sensible and consistent with previous investigation
• Ratio of basic sum addured to premium for each class of reg premium contracts must be sensible and consistent with previous investigation

Question 5

Give examples of consistency checks for unitised contracts

Answer 5

• Numbers of units bought by premiums and encashed to pay benefits must be consistent with revenue account
• Internal unit movements (e.g. charge encashments) must be consistent with surplus emerging during year

Question 6

Give examples of spot checks

Answer 6

• Very large or zero unit values for unitised contracts
• Impossible dates of birth or retirement ages or start dates
• Group items to see distribution e.g. high clustering of birth month may suggest data entry error
• Randomly compare and spot check computer data extract with paper admin files