Chapter 32

Question 1

What are the four main types of policy data checks?

Answer 1

• Reconciliation checks – movements vs. in-force.
  
  • Consistency checks – averages and ratios.
  • Unusual values and spot checks – outliers/impossible values.
  • Analysis of surplus/EV – unexplained discrepancies.

Question 2

How can analysis of surplus reveal data issues?

Answer 2

• Large unexpected discrepancies compared to previous analyses suggest data problems.
  
  • The same applies to embedded value profit analysis.

Question 3

What unusual values or errors should be checked for in policy data?

Answer 3

• Zero or excessively large unit values.
  
  • Impossible dates of birth, start, or retirement.
  • Unusual clustering (e.g., all policies with same birth month).
  • Cross-check sample of electronic data vs. paper files.

Question 4

What policy data checks can be made on non-unitised business

Answer 4

Data reconciliation checks:
* # of contracts
* basic SA or equivalent benefit
* office premium
* attaching bonus amount

Consistency checks:
* average SA for each class must be sensible and consistent with previous investigation
* Ratio SA:Premium for each class must be sensible and consistent with previous investigation

Unusual values:
* like impossible dates of birth
* distribution of group items doesnt make sense

Question 5

What policy data checks can be made on unitised business

Answer 5

Data reconciliation checks:
* #of contracts
* # of units actually purchased by premiums, sub-divided by unit fund
* current premium payable
* current benefits available
* change in # of units allocated from switches between unit funds
* changes in premium payable and benefit

Consistency checks:
* # of units purchased by prems and uncashed to pay benefits are consistent with revenue account
* check internal unit movements are consistent with surplus emergence

Unusual values:
* very large or zero unit values