Ch 32: Policy data checks

Question 1

Why are policy data checks important? (2)

What would be the qualities of ‘good data’? (5)

List 5 types of policy data checks (5)

Answer 1

Why are policy data checks important?

1. actuary will typically make extensive use of data in their work
2. need to ensure that data used is ‘good data’ so that correct results/conclusions may emerge/be drawn from the data

Through a number of tests, we can ensure that data used is good ie

1. adequate
2. accurate
3. complete
4. all data items required are being maintained by company
5. can be accessed in required format

5 types of policy data checks

1. Data reconcilliation checks
2. Consistency checks
3. Unusual values
4. Spot checks
5. Analysis of surplus and/or analysis of embedded value profit

Question 2

For what purposes might data reconcilliation checks be useful? (2)

What are the steps involved in the data reconcilliation? (5)

Answer 2

Data reconcilliation checks might be quite useful where

1. an investigation is being carried out on a regular basis
2. a reconcilliation of current data with those used for previous one can be attempted

Data reconcilliation may be carried out as follows

• first group data sensibly eg by year of entry within each broad contract type
• using data similarly grouped relating to business that has
  
  • come onto insurer’s books and gone off
  • between the dates of the two investigations
• the following check is done for each group
  
  • data at previous investigation + busines come into books - business gone of books = data at current investigation

Question 3

What kind of data reconciliation checks can we do for non-unitised business? (4)

What kind of data reconciliation checks can we do for unitised business? (6)

What other 3 factors are important when doing data reconciliations? (3)

Answer 3

Data reconcilliation checks for non-unitised business

1. number of contracts
2. basic sum assured - or the equivalent benefit depending on contract’s nature
3. office premium
4. for with-profits contracts, the amount of any attaching bonuses

Data reconcilliation checks for unitised business

1. number of contracts
2. number of units allocated, sub-divided by unitised fund
3. current premium payable
4. current benefits available eg amount of death cover
5. if relevant, recon must also allow for items such as
   
   1. changes in number units allocated arising from switches betwen unitised funds
   2. changes in premium payable and benefits under existing contracts

When doing data reconciliations, it’s important that

1. systems for producing movements data are checked periodically to ensure
   
   • they’re working correctly
   • staff involved are following procedures laid down
2. movements checked against any appropriate accounting data, especially regarding benefit payments
3. movement data independant
   
   • ​​ie seperately generated from the in force data
   • use the actual movement numbers, don’t use data start, end and new business to derive the movement data ie. do not use the movement data as a balancing item.

Question 4

Give examples of checks of consistency for non-unitised contracts (3)

Give examples of checks of consistency for unitised contracts (2)

Answer 4

Consistency checks for non-unitised contracts

• Average sum assured or average premium per class of business should be
  
  • sensible and consistent with the figure for the previous investigation
• Ratio of basic sum assured (or equivalent benefit) to premium payable per class of regular premium contracts should be
  
  • sensible and consistent with the figure for the previous investigation
• Cashflows vs revenue account
  

Consistency checks for unitised contracts

• Number of units purchased by premiums and encashed to pay benefits are
  
  • consistent with corresponding revenue account items
• Internal unit movements eg for charge encashments are
  
  • consistent with surplus emerging during year

Note: charge encashments generally refer to monetary expense/mortality charges that are deducted from PH’s funds by the cancellation of units

Question 5

What kind of checks can be made for unusual values? (5)

What kind of checks for unusual values can be performed in addition to looking at just individual data? (2)

What kind of checks can be done for computer held data? (2)

Answer 5

Checks for unusual values

• very large or zero unit values under unitised contracts
• impossible:
  dates of birth,
  retirement ages, and
  start dates

In addition to looking at only individual data it may also be possible to

• group items and see how well distributed they are
• eg an unusually high clustering of birth month may represent a data input error needing further investigation

May be useful to compare an extract of computer held data with information in paper administration files -
* can be done on a spot check basis, by randomly selecting a number of policies

Question 6

What is an analysis of surplus? (1)

What in particulare are we looking for when it comes to the AOS and checks on data? (3)

Answer 6

An analysis of surplus (AoS) is

• an investigation that seeks to explain reasons for the change in valuation result (ie Va - Vl) between one valuation date to another

When conducting an AoS, we should be close attention to

• A major discrepancy in the analysis compared to previous analyses
• that may indicate a problem with the data
• the same will also be true of large discrepancies in any analysis of the change in EV