Chapter 30: Data Flashcards

1
Q

What are the main uses of policy and claims data?

A
  • Administration
  • Accounting
  • Statutory returns
  • Setting investment strategy
  • Financial control
  • Management information
  • Risk Management
  • Setting provisions
  • Experience statistics and analyses
  • Pricing
  • Marketing

Ideally, all of these functions should be controlled through one integrated data system.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What are the main sources of data?

A
  • published company accounts
  • regulatory returns
  • internal experience investigations
  • industry data, including actuarial tables
  • national statistics
  • data from reinsurers
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Why might data be of poor quality or quantity?

A
  • Poor management control

- Poor design of data systems

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

How can good quality data from the proposal and claim forms be ensured?

A
  • well-designed and unambiguous questions that are in the same order that they will be processed on the administration system.
  • It is important that the result of any underwriting is captured;
  • and for general insurance, that the rating factors can all be separately identified.

The policy data should be used to check the validity of the claims data and vice versa.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

What are the data issues for employee benefit schemes?

A
  • The actuary may not have full control over the data, and so validation is particularly important
  • Additionally, the data might not be sufficiently detailed to provide all the information that is financially significant to the valuation
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

What are the sources of data that are useful in checking the current data from the scheme?

A
  • Data from the previous valuation, for reconciliation with the current data
  • Accounting data for checks on assets, benefit outgo and contribution income.
  • A full listing of the assets held to check whether they are permitted.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

What are the problems with using summarised data?

A

The reliability of the valuation will be reduced as full validation of the data is impossible.
Changes in the mix of the members may remain unidentified.
Summarised data is not useful for valuing individual options and guarantees.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Some countries have industry-wide data collection schemes.

What is this useful for?

A

Industry-wide data can be used

  • to increase the credibility of a country’s own data, or
  • for providing a comparison of a country’s own data against that of competitors.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Industry-wide data might not be directly comparable.

Why is this?

A

There will be differences in

  • target markets
  • geographical or socio-economic areas of operation
  • policy conditions
  • risk and rating factors used
  • underwriting and claims settlement
  • ways of storing data

In addition,

  • industry data may be less detailed or flexible than internal data.
  • It may be out of date,
  • not everyone contributes,
  • the quality may be dubious.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

4 Categories of checks to be carried out on data

A
  • Reconciliations
  • Cross checks
  • Reasonableness checks
  • Spot checks
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

3 Reconciliation checks on data

A

Include:

  • reconciling the total number of members or policies and changes in membership or policies using previous data and movement data
  • reconciling the total benefit amounts and premiums and changes in them using previous data and movement data.
  • reconciling shareholdings at the start and end of the period adjusted for sales and purchases and bonus issues.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

4 Cross checks on data

A
  • checking movement data against accounting data, e.g. benefit payments
  • checking membership data against accounting data, e.g. contributions
  • checking asset data against accounting data, e.g. investment return.
  • checking asset ownership records against the custodian’s records
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Reasonableness checks on data

A
  • Checking the average sum assured or premium looks sensible for the class of business, and against previous data.
  • Checking for unusual values or missing data.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

Spot checks on data

A
  • Checking individual members or policy records.
  • Checking that individual assets or liabilities exist on a given date.
  • Carrying out a full deed audit for certain assets, e.g. property title deeds
  • Checking that the correct value for an asset or liability is being recorded.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Main aim of risk classification

A

To obtain homogeneous data

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

Why is it important to have homogeneous data?

A

So that the experience in each group of risks is more stable, enabling the data to be used for projection purposes.

17
Q

What is the main practical problem of risk classification?

A

It may result in too little data in a group for a credible analysis.

18
Q

Spot check

A

Eg for lapse rates, check a policy to determine when it lapsed and check against when last premium was received

19
Q

Cross checks

A

• Check data against other sources
o Eg check accounting data vs data used to administer contracts
• Check for consistency between claims and exposed to risk data

20
Q

Reasonableness checks

A

• Check whether rates consistent with what was expected
• Check is rates reasonable compared to previous rates
• The drivers of rates should be analysed
o Consider how rates differ by SA, prem size, entry year, dist channel, calandar year etc
o Will help to explain any differences vs industry data
• Ensure no clustering of dates (claim/ lapse) – may suggest system error
• Consider rate measure – unweighted or weighted by eg prem size
• Check changes in business mix

21
Q

Reconciliations

A

• Reconcile latest results with previous results, eg for lapse data:
o Num pol in force at end of period= num at start + num new pol – num exits
• Could carried out for each product class and year
• Check results of an analysis of surplus