Chp 29-32, 48

Question 1

1) Requirements of a good model (Acronym – VARIABLE )

Answer 1

Valid
Adequately documented
Rigorous
Inputs to parameter values appropriate
Arbitrage free (economic interpretation)
Behaviour reasonable
Length/expense of run not too long/high
Easy to understand

Question 2

1) Requirements of a good model (Acronym – CRISPS )

Answer 2

Communicable workings and outputs
Reflects risk profile of contracts modelled
Independent verification of outputs
Sensible joint behaviour of variables
Parameters allow for all significant features
Simple (parsimonious) but retain key features

Question 3

1) Requirements of a good model (Acronym – CARD)

Answer 3

Clear results
A range of implementation methods
Refineable
Developable

Question 4

2) Cashflows to include in a model

Answer 4

a) Cashflow arising from any supervisory or commercial requirement to hold reserves and to maintain adequate solvency capital
b) Cashflow need to allow for any interactions, particularly where assets and liabilities are being modelled together
c) Potential cashflows from options and take-up rate need to be allowed for
d) More frequently the cashflows are calculated, the more reliable output from the model, less frequently the cashflows are calculated the faster the model can be run and results obtained

Question 5

3) Approaches to modelling and factors determining their merits - approaches

Answer 5

a) Approaches to modelling
i) Purchase commercial modelling product
ii) Reuse existing model with modification
iii) Develop a new model in house

Question 6

b) Merit depends on

Answer 6

Cost of each option
Level of accuracy required
In-house expertise available
Number of times use
Desired flexibility of the model

Question 7

a) Deterministic model +/-

Answer 7

i) Pros
(1) More readily explicable to a non-technical audience
(2) Clearer what economic scenarios have been tested
(3) Easier to design and quicker to run
ii) Cons
(1) Requires thought as to the range of economic scenarios that should be tested

Question 8

b) Stochastic model +/-

Answer 8

i) Pros
(1) Tests a wider range of economic scenarios
(2) Quality of results may be higher than deterministic modelling
ii) Cons
(1) Programming is more complex and run time is longer
(2) Danger of spurious accuracy if number of runs too small

Question 9

a) Steps in running a deterministic model

Answer 9

Specify the purpose of the investigation
Collect, group, and modify data
Choose the form of the model, identifying its parameters and variables

Ascribe values to the parameters using past experience and appropriate estimation techniques
Construct a model based on the expected cashflows
Check that the goodness of fit is acceptable – run past year and compare model with actual results

Attempt to fit different model if first choice does not fit well
Run model using selected values of variables
Run model using estimates of values of variables in future
Run model several times to assess sensitivity of results to different parameter values

Question 10

b) Steps in running a stochastic model

Answer 10

Specify the purpose of the investigation
Collect, group, and modify data
Choose a suitable density function for each of the variables to be modelled stochastically

Specify correlation between variables
Construct a model based on the expected cashflows
Check the goodness of fit is acceptable – run past year and compare model with actual results

Attempt to fit a different model if the first model does not fit well
Run the model many times, each time using a random sample from the chosen density function(s)
Produce a summary of the results that shows the distribution of the modelled results after many simulations have been run

Question 11

6) Ways of assessing the variability of experience (scenario testing and stochastic modelling)

Answer 11

a) Scenario testing – various scenarios can be tested by varying the parameter values in the model and assessing their results to see how the experience varies with varying each parameter
b) Stochastic modelling – This will give a probabilistic distribution of experience, whereby the expected experience and deviation from the expected experience can be analysed

Question 12

8) Sources of data

Answer 12

Tables – e.g. actuarial mortality tables
Regulatory reports and company accounts
Abroad – data from overseas contracts
Industry data
National statistics
Experience investigations on the existing contract
Reinsurers
Similar contracts

Question 13

9) Problems relating to data quality and how to overcome them

Answer 13

a) Problems relating to data quality
i) Result of poor management control
ii) Poor verification processes
iii) Poor design of data systems

b) How to overcome them
i) Information on proposal form
(1) Questions need to be well designed and unambiguous
(2) Cross check against claims information

Question 14

10) Data checks

Answer 14

Investment income – consistency between asset data and accounting data

Random spot checks on data
Average sum assured or premium for each class of business – should be sensible and consistent with previous investigation
Number of members/policies and changes – using previous data and movement data

Shareholdings – at start and end of period adjusted for sales, purchases and bonus issues
Audit of certain assets – e.g. checking title deeds to large real property assets
Assets held by third party –reconcile between beneficial owner’s and custodian’s records

Movement data – check against appropriate accounting data

Benefit amounts and premiums – reconcile total and changes using previous data and movement data
Unusual values – e.g. impossible dates of birth, retirement ages, start dates
Salary-related contributions and in-payment benefit levels – compare membership data and figures in accounts

Question 15

11) Advantages and disadvantages of industry-wide collection schemes (Acronym - DR DONEQ)

Answer 15

Detail insufficient
Reporting formats differ

Differences in target markets, underwriting, geographical area, sales processes, contract wordings, claim settlement, rating factors
Out of date
Not everyone contributes
Errors
Quality only as good as that of contributors

Question 16

12) Description of risk classification

Answer 16

a) Main aim is to obtain homogeneous data
b) Group data according to risk factors which have most influence on experience under interest
c) Experience in each group is then more stable and appropriate for projection purposes
d) If data is scarce, groups may be too small – in this case, allow for a bit of heterogeneity so that groups are large enough to be credible
e) Balance must be struck between homogeneity and credibility

Question 17

13) Things to consider when setting assumptions

Answer 17

a) Consider the use to which the assumptions will be put
b) Focus on assumptions with most financial significance
c) Achieve consistency between various assumptions
d) Consider legislative or regulatory constraints
e) Consider the needs of the client

Question 18

14) Examples of historical and current data that can be used to help setting assumptions

Answer 18

a) Investment returns – past dividend yields, total returns on relevant asset classes may be useful
b) Salary growth – past data on salary levels
c) Demographic assumptions – past mortality/morbidity data
d) Inflation – relationship between yields for fixed and index-linked bonds
e) Economic factors – policy statements by governments or controlling banks

Question 19

15) Why this data may need adjusting to be suitable for setting future assumptions

Answer 19

a) Abnormal fluctuations
b) Changes in the experience with time
c) Random fluctuations
d) Changes in the way in which the data was recorded
e) Potential errors in the data
f) Changes in the balance of any homogeneous groups underlying the data
g) Heterogeneity with the group to which the assumptions are to relate

Question 20

16) Definitions of fixed, variable, direct and indirect expenses

Answer 20

a) Fixed – expenses fixed in real terms e.g. building maintenance, staff expenses in short term
b) Variable expenses – expenses which vary directly according to level of business, e.g. staff expenses in long-term varying to meet changing levels of new and existing business
c) Direct –arise from department dealing purely with one class of business, if occur with more than one class then time sheets can be kept to help split costs
d) Indirect – departments concerned not related directly to any particular class of business but form a support function, charging out basis can be used for computer time , premises costs can be allocated by floor space

Question 21

a) Fixed expenses

Answer 21

expenses fixed in real terms e.g. building maintenance, staff expenses in short term

Question 22

b) Variable expenses

Answer 22

expenses which vary directly according to level of business, e.g. staff expenses in long-term varying to meet changing levels of new and existing business

Question 23

c) Direct expenses

Answer 23

arise from department dealing purely with one class of business, if occur with more than one class then time sheets can be kept to help split costs

Question 24

d) Indirect expenses

Answer 24

departments concerned not related directly to any particular class of business but form a support function, charging out basis can be used for computer time , premises costs can be allocated by floor space

Question 25

17) Categories in to which expenses need to be allocated for an expense analysis

Answer 25

a) Securing new business
b) Maintaining existing business (renewal and investment)
c) Terminating business (including claims)

Question 26

18) How to allocate typical expenses such as commission, salaries, property costs, computer costs, investment expenses, and one-off capital costs

Answer 26

a) Commission –initial and renewal, take into account clawback period, differ for each distribution channel
b) Salaries – Timesheets can be used
c) Property costs – by floor space, charging notional rent to each department
d) Computer costs – first amortised over its useful lifetime, then split by hours of usage
e) Investment expenses – These should form part of the products, for custodianship split expenses proportionally for each asset held
f) One-off capital costs – amortised over its useful lifetime, then allocated to overheads as one-off cost

Question 27

19) The ways in which expenses can be loaded

Answer 27

a) Fixed amount per contract
b) Percentage of premium/contribution charged
c) Percentage of sum assured
d) Combination of above

Question 28

20) Main expense items for a provider

Answer 28

a) Salary related expenses
b) Property costs (rent, property taxes, heating, lighting, cleaning)
c) Computer costs
d) Investment costs (investment department, stamp duty, commission, custodian)

Question 29

21) Reasons for monitoring experience

Answer 29

Update assumptions as to future experience
Monitor adverse trends in experience so as to take corrective action
Provide management information

Question 30

22) How to monitor demographic and economic assumptions, including data issues, the time period involved, the use of the results

Answer 30

a) Data needs to be divided into sufficiently homogeneous risk groups according to relevant risk factors
b) This ideal must be balanced against danger of creating data cells with too little data
c) Volume of data will indicate whether analysis will produce meaningful results
d) Must have data on exposed to risk as well as contingency
e) Questions whether there is any reason why any past trends may continue into the future

Question 31

24) Reasons for monitoring investment experience

Answer 31

a) Liability structure may have changed significantly – e.g. writing of a new class of business, takeover, benefit improvements, legislation
b) Funding or free asset position may have changed significantly
c) Manager’s performance may be significantly out of line with that of other funds

Question 32

25) How to compare investment performance – issues surrounding benchmarks and constraints on the fund managers

Answer 32

a) Return will be judged relative to that achieved by other managers for similar funds
b) The more severe the restrictions placed on the managers as to the assets or asset classes that can be held, the less appropriate it is to set performance targets that relate directly to the generality of funds