18. Modelling

Question 1

Obtaining models

Answer 1

• Commercial purchase
• Modify existing model
• New model

Question 2

Constructing models

Answer 2

1. Must fit purpose
2. Operational issues to be considered
   - Well documented
   - Easily communicable, with clearly displayed results
   - Have sensible behaviour of variables
   - Capable of independent verification
   - Not be overly complex/time consuming to run
   - Be capable of development and refinement
   - Be capable of being implemented in range of ways
   - Have appropriate time period between projected cashflows, balancing reliability of output with speed of running model
3. Stochastic vs Deterministic / combination
4. Must be dynamic

Question 3

Dynamic model

Answer 3

Allows for interaction between parameters and variables affecting cashflows

Question 4

Steps in developing model

Answer 4

1. Purpose
2. Obtain and adjust data
3. Set parameters/assumptions incl dynamic links
4. Construct model cashflows
5. Check accuracy and fit; amend if necessary
6. Run many times
7. Output and summarise results

Question 5

Model point

Answer 5

Representative policy in a relatively homogeneous group of policies

Question 6

Discount rate

Answer 6

Used to discount future cashflows

Allows for:

• Required return
• Level of statistical risk (assessed by sensitivity analysis, stoch model or comparing market data)
• Can use stochastic discount rate
• Can use different rates for each cashflow to better reflect risks, but not done due to simplicity of using one

Question 7

Considerations after modelling premiums/ charges

Answer 7

1. Product design
2. Distribution channel(s)
3. Profit requirement
4. Size of market
5. Whether to launch product
6. Appropriateness given business strategy and capital requirements

Question 8

Uses for models

Answer 8

• Setting premiums/charges
• Setting future financing strategies by benefit schemes
• Risk management to determine capital requirements
• Valuations
• Pricing options and guarantees

Question 9

Sensitivity analysis

Answer 9

Changing one parameter at a time to determine sensitivity of results to that parameter

• llustrates potential variability of results and identify impact of mis-estimation of parameter values

Question 10

Scenario testing

Answer 10

Changine many parameters together to see effect on results

Question 11

Reducing model error

Answer 11

Goodness of fit tests

Question 12

Alternative ways of allowing for statistical risk associated with parameter values

Answer 12

• Allow for in discount rate and/or

- margins in parameter values