19 - Data

Question 1

What is the main concern of using / transferring data across international borders?

Answer 1

The legislation around data handling may be more stringent in one of the two countries and organisations need to take extra care to not breach local standards.

Question 2

List the eight conditions of the POPI Act in South Africa.

Answer 2

1. Accountability
2. Processing limitation
3. Purpose specification
4. Further processing
5. Information quality
6. Openness
7. Security safeguards
8. Data subject participation

Question 3

Describe the POPI Act condition of accountability.

Answer 3

The party responsible for processing the data is also responsible for compliance with POPI.

Question 4

Describe the POPI Act condition of processing limitation.

Answer 4

Information must be processed in a fair, lawful and relevant manner, after consent is given by the data subject.

Question 5

Describe the POPI Act condition of purpose specification.

Answer 5

Personal information must be collected for a specific purpose. Record keeping to be destroyed when personal data is no longer relevant or authorised to be held.

Question 6

Describe the POPI Act condition of further processing limitation

Answer 6

Further processing must be compatible with the initial collection prupose.

Question 7

Describe the POPI Act condition of information quality

Answer 7

Data completeness, accuracy and updates to be ensured by holder of the data.

Question 8

Describe the POPI Act condition of openness

Answer 8

Documentation to be maintained on all processing operations and maintaining transparency on data use.

Question 9

Describe the POPI Act condition of security safegaurds

Answer 9

Integrity and confidentiality of personal data must be secured and all processing done only by authorised operators. Notification to be done on security compromises.

Question 10

Describe the POPI Act condition of data subject participation.

Answer 10

The data subject may request confirmation of personal data held and request correction or deletion of any inaccurate, misleading or outdated information held.

Question 11

Aside from criminal action and fines, what is another damaging effect of data breaches occurring within a company’s data bases?

Answer 11

Damage to reputation and the ability to retain and attract clients.

Question 12

Give the aspects that a data governance policy should aim to cover.

Answer 12

1. The specific roles and responsibilities of individuals in the organisation with regards to data.
2. How an organisation will capture, analyse and process data.
3. Issues with respect to data security and privacy
4. The controls that will be put in place to ensure that the required data standards are applied
5. How the adequacy of the controls will be monitored on an ongoing basis with respect to data usability, accessibility, integrity and security.

Question 13

Give the data governance risks.

Answer 13

1. Legal and regulatory non-compliance
2. Inability to rely on data for decision making
3. Reputational issues
4. Incurring additional costs

Question 14

Give a data concern around mergers and acquisitions.

Answer 14

1. Should data be combined into one system
2. Which company’s system to use
3. Data aggregation issues.

Question 15

Give the main risks associated with risks.

Answer 15

1. The data are inaccurate or incomplete
2. The data are not credible due to being insufficient volume, particularly for the estimation of extreme outcomes.
3. The data are not sufficiently relevant to the intended purpose
4. Past data may not reflect what will happen in the future.
5. Chosen data groups may not be optimal
6. The data are not available in an appropriate form for the intended purpose.

Question 16

Why may past data not be an accurate reflection of future experience.

Answer 16

1. Past abnormal events
2. Significant random fluctuations
3. Future trends not being reflected sufficiently in past data
4. Changes in the way in which past data was recorded
5. Changes in the balance of any homogeneous groups underlying the data
6. Heterogeneity with the group to which the assumptions are to relate
7. The past data may not be sufficiently up to date
8. Other changes

Question 17

State the data protection principle which it can be difficult to meet when using big data

Answer 17

Personal data should be adequate, relevant and not excessive for the purposes concerned.

Question 18

How can companies avoid big data being excessive for the given purpose?

Answer 18

Anonymisation can be used to ensure that the data is not considered to be personal data.

Question 19

List the main uses that actuaries make of data.

Answer 19

1. Premium rating, product pricing and determining contributions
2. Setting provisions
3. Experience analysis
4. Risk management - underwriting and reinsurance
5. Investing
6. Accounting
7. Management information
8. Marketing
9. Administration

Question 20

List the key data required for active members when valuing a pension scheme

Answer 20

1. Membership ID / number
2. Date of birth
3. Date of joining employer
4. Date of joining the scheme
5. Date / age of retirement
6. Current salary
7. Salary scale / growth assumptions
8. Category of membership
9. Dependents - marital status
10. Age of dependants
11. Data from previous valuations for reconciliations

Question 21

Outline the design features of a good proposal form.

Answer 21

1. Collects data at an appropriate level - including data that is not currently used but may be used in the future
2. Be clear and unambiguous - to capture the correct information
3. Have inputs that are quantitative as far as possible

Question 22

Give a design feature of the claims form in order to store good quality data.

Answer 22

Should be clear and unambiguous and link to the proposal form - to cross check information

Question 23

Give features of data inputting processes that can ensure that good quality data is stored by a company.

Answer 23

1. Inputs should be in the same order as the proposal form
2. Staff that are inputting data should be trained
3. Financial incentives for accurate inputting
4. Data systems should have data validation checks - blank entries and sensible entry values
5. Send policyholders copies of the key information in order to check all values are captured correctly

Question 24

Give the data system features that can help ensure that good quality data is stored by an insurance company.

Answer 24

1. The system should be capable of storing information so that historical data is available for future pricing exercises
2. System should be robust yet flexible
3. System should be secure - restricting access of people who can manipulate data
4. Regular checks of data movements and changes
5. Single integrated systems can make data handling easier

Question 25

List reasons why claims data might not be directly comparable between different general insurance companies.

Answer 25

1. Organisations operating in different geographical or socio-economic sections of the market 2. Different policies being sold - policy conditions or perils 3. Products being sold by different sales methods or to different target markets 4. Differing underwriting standards at the initial claims stage 5. Different companies assessing risk differently - different rating factors 6. Data being stored or recorded differently or relating to different time periods

Question 26

Describe the issues with using industry data.

Answer 26

1. Supplied data may be inaccurate or incomplete 2. May be out of date 3. May not be relevant to the intended purpose 4. Data may not be available in intended format 5. Chosen data groups within the industry-wide data may not be optimal 6. The coding used for the factors by which the data is split may vary between pension schemes 7. May have different definitions of benefits 8. May not be detailed enough data 9. Data may be less flexible than is required 10. Data may not be credible at extreme conditions 11. Data may not reflect what will happen in the future

Question 27

Describe how industry data may not be relevant for a benefit scheme.

Answer 27

1. Not originally collected for the intended purpose 2. May not be accurately comparable 3. Employer operates in a different geographical region 4. Has a different socio-economic mix of members than the industry average 5. The nature of the benefits is very different to industry averages 6. The extent of voluntary vs compulsory membership differs from the industry average

Question 28

List the reasons that past data may not reflect future experience.

Answer 28

1. Past abnormal events 2. Significant random fluctuations 3. Future trends not reflected in historical data 4. Changes in the way in which past data was recorded 5. Changes in the balance of nay homogeneous groups underlying the data 6. Other changes such as medical, social or economic.

Question 29

Give the advantages of algorithmic decision making.

Answer 29

1. Faster 2. More efficient 3. Possibly lower dealing costs on trades 4. It is possible to implement complex trading strategies

Question 30

Give the disadvantages of algorithmic decision making.

Answer 30

1. Could be errors in the algorithm or the data or parameters - especially when a large number of trades could be implemented in a short period 2. Algorithm may not operate in adverse conditions 3. Trading may be suspended before algorithmic trading can be completed in turbulent markets 4. Could have adverse impacts on the financial systems

Question 31

Give the main sources of data for an insurance company.

Answer 31

1. Publicly available data 2. Internal data 3. Industry-wide data collection 4. Reinsurance data

Question 32

Give the main issue with using internal data.

Answer 32

There may not be sufficient data for launching or pricing a new product or target markets.

Question 33

Give some advantages of using industry-wide data

Answer 33

1. Can compare experience with that of the industry as a whole 2. Can provide an indication of the ways in which competitors attract policyholders through differences between the company and the industry-wide data.

Question 34

Describe the main data checks that actuaries can use.

Answer 34

1. Reconciliations of member / policy numbers 2. Reconciliations of benefits and premiums 3. Movement data against accounts 4. Validity of dates 5. Consistency of contributions and benefit levels 6. Consistencies between average sum assured and premium for each class when compared to previous investigations 7. Consistency of asset income data and accounts 8. Full deed audit for certain assets 9. Consistency between start and end periods 10. Records picked at random for spot checks

Question 35

Give the main circumstances under which there may be a lack of ideal data.

Answer 35

1. Insufficient volume to provide credible results | 2. Data captured at insufficient details

Question 36

Describe the main issue with using summarised data.

Answer 36

The records cannot be validated which makes error detection impossible

Question 37

Outline the issues with provisions that are too large

Answer 37

1. The funding level will appear to be lower than it actually is 2. Capital may not be used efficiently

Question 38

Outline the issues with provisions that are too small

Answer 38

1. Over time it will become apparent that additional money is required 2. Worst case is insolvency 3, Profits will be recognised earlier and the payment of tax will be accelerated 4. Inappropriate business decisions can be made.