Part 2

Question 1

What are the general risks

Answer 1

Risk a life drown cats

1. Reinsurance failure
2. Investment
3. Short-termism of senior management
4. Kompetition
5. Actions of distributors
6. Legal and regulatory.
   
   • treating customers fairly and sales processes. These could lead to fines and reputational damage.
   • New regulations might be introduced, eg maximum premium levels or restrictions on the company’s ability to use rating factors in pricing.
7. Inflation
8. Fraud
   
   • applicants may withhold significant information during the underwriting process
   • fabricated death claims could be made.
9. Expenses - There is a risk of expenses being higher than assumed in pricing.
10. The risk of future expense inflation being higher than assumed will contribute to this risk.
    
    • This is quite a large risk for term assurance business as expenses form a relatively large proportion of the premiums.
    • The effect of high expense inflation could be a particular risk for individual term assurances, where premiums are normally level and non-reviewable.
11. Data - There is a risk of having inadequate mortality data with which to price the business, resulting in inadequate premiums being payable.
12. Rates of mortality - The main risk to a life company of writing term assurance business is mortality risk, ie the risk of higher mortality rates than the company assumed in pricing.
13. Options and guarantees
14. Withdrawals (persistency) There is a risk on early withdrawal as the asset share will be negative in the early stages as a result of the high initial term assurance expenses (eg commission, underwriting expenses) and relatively low premiums. This may persist longer for decreasing term assurances, unless the premium-paying term has been shortened. This is because the claim costs are highest at early durations, preventing the asset share from becoming positive until later in the policy term.
    
    1. There is a mortality risk from selective withdrawal, as those in worse health are less likely to withdraw and so the average mortality of those remaining worsens.
15. New business - There is a risk of low new business volumes. This could result in overheads not being sufficiently covered by the per-policy loadings to cover them There is a risk of high new business volumes. This could cause capital strain that exceeds the capital available and could threaten solvency. The guarantees and high initial expenses of individual term assurances, in conjunction with typically small regular premiums, mean that new business strain could be significant.
16. There is also a risk in this scenario that admin capabilities will be over-stretched, leading to errors, poor customer service and reputational damage.
17. lf cross-subsidies exist, there is a risk from the mix of business sold, eg it is likely that large policies will subsidise small ones, so that there is a risk of the average sum assured being smaller than expected.
18. Controls
19. Aggregations
20. Tax There is a risk that the amount of tax payable will be different from expected, eg due to change in tax regime, causing losses. Changes in the ways that policyholder benefits and/or premiums are taxed could also lead to reduced demand for the product.
21. Selection

Question 2

What can the government do to increase the insurance market

Answer 2

They could lower the tax and provide tax breaks for customers on certain insurance products.

• Changing how insurance companies are taxed
• Changing how individuals are taxed (eg increasing tax relief on premiums and/or reducing the tax payable on benefits received).
• The effectiveness of changing insurance company tax in boosting sales depends on:
  
  • the extent to which the cost of life insurance products was the cause of low sales
  • the extent to which companies pass on the advantage to customers via lower premiums.
• lmproving the tax advantages of life insurance products is likely to be effective.
• Using available tax breaks is often a priority of customers who can afford to use them and the tax breaks may be used in marketing messages of companies and advisers. This will however depend on the taxation of life insurance compared to similar, competing products, eg unit trusts and other savings vehicles.

Compulsion

• The government could introduce elements of compulsion to purchase insurance products,
• Require employers to purchase a pension savings product for each employee
  
  • reqire an annuity to be purchased with the proceeds of pension savings
  • Require individuals to take out life cover for any mortgage borrowing.
• From a company’s point of view, large customer numbers may help with economies of scale (helping to keep costs down) provided the company can achieve sufficient market share.
• However, the premiums chargeable are also likely to be low in order to be affordable to the customers compelled to buy the product. This is likely to be effective for products and circumstances where it is justifiable, but there may only be a limited number of products / circumstances (as a reasonably high proportion of customers may not want or be able to afford the insurance).
• A less extreme approach than compulsion would be to rely on customer inertia, eg enabling employees to automatically be enrolled to various insurance products unless they took the deliberate step of opting out. c.

Encourage product innovation and new product designs

• The government could remove restrictions on the types of products that could be offered’
• This may enable product innovation, eg launching products that have been sold successfully in other countries.
• This may be effective in increasing sales if the new products meet customer needs that existing products do not meet.

Distribution channels and the selling process

• The government could increase insurance companies’ freedom to sell, eg by allowing the use of multiple distribution channels.
• This may be effective if it improves customer access to insurance products.
• The government could also change the rules about the information insurance companies provide to customers at the point of sale.
  
  • This may be effective if it was a lack of information (or too much information) that was preventing customers from buying insurance products.

lnvestment freedom

• The government could reduce the investment restrictions on insurance companies.
• This may enable them to invest in more risky assets with higher expected returns.
• This may be effective in improving sales if these higher returns are passed on to customers through lower premiums and if high premiums had been a cause of the decline in sales.

Premium or charge caps

• The government could introduce measures to make products cheaper,
  
  • eg by introducing charge caps.
  • This is likely to be effective if the problem is that insurance is too expensive for customers.
  • lt could be especially effective for certain customers (eg those with less disposable income) and certain products.
  • Its effectiveness could be limited if it led insurance companies to withdraw from the market (if charge caps made products unprofitable to sell) and so reduced the availability of certain insurance products.

Rating factors

• The government may relax the use of certain rating factors, eg gender, family medical history.
• This may be effective if it improved the ability of insurance companies to rate fairly and lead to lower prices on average (as companies had less need for prudence in the pricing given they no longer have to allow for the uncertainty about cross-subsidies resulting from the mix of business)
• Using more rating factors would result in insurance companies dividing customers into more groups for pricing purposes.
  
  • Some groups will gain and some will lose out from such a change, eg allowing gender as a rating factor would make annuities more attractive to males and less attractive to females. The move is therefore likely to be effective in promoting more purchases from the groups that benefit from lower premiums, but could change the mix of policyholders.

Underwriting

• The government could relax restrictions on the ability of insurance companies to underwrite products.
  
  • For example, insurers could be allowed to use genetic test results or allow them to decline policyholders with certain medical conditions.
• This tighter underwriting is likely to be effective in reducing premiums and so increasing sales.

Reserving and capital requirements

• The government could relax the regulations about the reserves and solvency capital that insurance companies are required to hold to demonstrate solvency.
• This may increase sales, particularly as products with marketable features, eg with more guarantees, may be especially capital-intensive.

Question 3

How to mitigate risks

Answer 3

1. Longevity risk

• ensuring that sufficient allowance is made for future mortality improvements when determining the value of assets to be transferred .
• Ensuring assumptions are tailored to the nature of the lives in the portfolio
• keeping up to date with the latest developments in projecting longevity improvements, eg with reinsurers’ assistance o after the transfer,
• Regularly monitoring the longevity experience .
• Using reinsurance, for example using quota share or excess of loss reinsurance or arranging a longevity swap.

Concentration risk

• Diversify by writing or buying blocks of business exposed to mortality risk, eg protection business, although there will be basis risk due to the different populations involved.
• The risk from the particular portfolio could be transferred,
  
  • eg using a reinsurance arrangement …
  • … or to the capital markets
  • a securitisation of the future profits on the in-force annuity portfolio.

Investment Risk

• The company may be able to hold more closely matching assets. lt could improve the portfolio matching after the transaction has taken place …
• … either by selling the assets it receives or by switching them with assets it already holds.
• It could hedge any currency risk exposure using derivatives.
• The company should monitor corporate bond performance and credit risk exposure, It could control credit risk by restricting the bonds it holds,
  
  • eg to those with sufficiently good credit ratings …
  • … and by using credit derivatives or credit insurance.

Expense

• The company should monitor its expense experience closely.
• It could look to improve operational efficiency, eg by investing in better systems or training or outsourcing.
• It could look to improve economies of scale eg by increasing the scale of its annuity book.
• It could seek to hedge its expense inflation by using real assets, eg index-linked bonds.

Data Risk

• ensuring the pension scheme provides the insurance company with an explanation of all the data items performing checks on the data before the transfer,
  
  • eg due diligence and/or doing a ‘trial run’data transfer .
  • agreeing that the pension scheme will bear any costs arising as a result of it providing poor quality data.

Other Risk

• The risk of regulatory or tax changes may be mitigated by keeping up to date with changes, responding to consultations on proposed changes and lobbying where appropriate.

Operational risks

• mitigated by thoroughly testing and documenting all systems and processes and by staff training.
• The risk that the assets transferred to cover the annuity liabilities are insufficient can be mitigated by ensuring a larger margin is allowed for in pricing the transaction … .
• .. or getting the proposed price reviewed by an independent party
• An appropriate pricing methodology should be used to determine the loading. Including appropriate assumptions.
• Add an extra margin for uncertainty.
• The increased mortality risk can be mitigated by means of reinsurance
• A minimal level of underwriting could be used. The margin for mortality could be an explicit margin for mortality…
• … or a reduction in the rate of improvement.

New business

• Ensure that the loadings are competitive relative to those charged by competitors for such options. Monitor competitors’ offerings.
• New business strain.
• Advertise/market the product.
• Thorough testing of the administration changes will be needed
• The sales literature for this product should be clear and not mis-leading
• Distributors should be trained on the new product.
• Administration staff should be trained on the new product.
• There should be clear process documentation and controls.
• External data/expertise could be used. e.g. from reinsurers.
• Expenses should be controlled carefully.
• Outsourcing could be used.
• Sufficient capital buffers should be put in place.
• The company could withdraw the option completely.
• Lapse and re-entry risk managed with appropriate margin, pricing or training

Question 4

Difference between intermediatary vs internal sales force

Answer 4

• Different social econmic selection
• Clients from intermediatary are more likely to be more sophisticated and wealthier
• Selective withdrawls could be higher through intermediataries,
  
  • Intermediary are more likely to compare rates around the market and reivew premiums More selective withdrawals could results in worse motality

Question 5

How premiums could be different in intermediary vs internal

Answer 5

Expense For policies in intermediatry they will need to consider the following

• Commissions
• Other costs and maintaining the business
• Cost of underwriting could be different
• Volumes could be different and so per policy expense could be different
• Persistency could be different
• Risk profile will be different
• Mix of business will be different

Question 6

Approach to modelling investment projections from the asset side

Answer 6

• Decide whether or not to adopt stocahstic or deterministic
• In determinisitic, assumptions are a point estimation of future investment returns
• The estimate could be the average return over all asset classes or it could be done for individual asset classes seperately, with averaging over the explicit mix of assets assumed in each future year period
• A central assumption could be supplemented by scenario or sensitivity testing
• A stochastic modelling, involves treating investment return as random variable
• A probility distribution must be determined for investment return
• This could be done for each class of assets seperately or as an aggregagte over the expliciit assmed mix of assets
• A model would be used by running a large number of simulations
• The average of all projections will be an estimate of the expected value of outcome
• The model will have to deal with
  
  • investment return from existing assets
  • reinvestment procceeds from existing assets
  • investment of future premiums on existing and new business.

Question 7

Features of a model

Answer 7

1. Needs to allow for all cashflows that may arise…. [1]
2. …which will depend on the nature of the contracts [½]
3. …in terms of premium and benefit structure [½]
4. …and any discretionary benefits (e.g. non-gteed surrender values)
5. Allow for cashflows from supervisory requirement to hold reserves… [1]
6. … and solvency capital
7. Cashflows needs to allow for interactions, [1]
8. particularly where assets and liabilities are modelled together [½]
9. Ability to use stochastic models and simulations need to be allowed for where appropriate… [1]
10. … to assess impact of financial options and guarantees.
11. The model being used should be adequately documented. [½]
12. The model should be easy to use, including communication of results. [½]
13. The outputs from the model should be capable of independent verification and signoff

Question 8

The difference between Pricing and reserving model points

Answer 8

Pricing

• A number of sample model points will be chosen [1]
• …to represent the expected new business under the product. [1]
• The profile of any similar existing product
• … [½] combined with advice from the company’s marketing department would be used. [½]

Reserving

• Real data on policies held will be used [1]
• This will either be ungrouped (i.e. policy by policy) [½]
• Or by using model points that are representative of the full data.

Question 9

What is the prime objective of building the model

Answer 9

1. Prime objective in building the model is to enable actuary advising company to give it appropriate advice
2. so that it can be run in sound financial way

Question 10

List 4 different types of life insurance company model

Answer 10

1. Profit test model
2. Business model
3. Existing business model
4. Full office model

Question 11

What are the basic requirements of a life insurance model

Answer 11

1. Model should project all cashflows that may arises
2. Cost of setting up supervisory reserves and required solvency capital should be allowed for in order to calculate profit flows
3. Proper allowance must be made for guarantees and options: stochastic model will probably be needed for this
4. Allow for interactions and correlationns between variables

Question 12

Advantages of stochastic model

Answer 12

1. Allow for probability distribution to be assigned
2. Positive liability can be calculated where deterministic approach might otherwise produce zero liability
3. Parameters maybe assumed to vary together as dynamic set, ie interactions can be explicityly modelled

Question 13

State 2 disadvantages of stocashtic models compared to determinisitic models

Answer 13

• TIme and computing constraints
• Possible spurious accuracy, ie results very sensitive to (determinisitic chosen) assumed values of parameter(s) involved

Question 14

Give two examples of circumstances in which determinisitic models might by appropriate

Answer 14

• The actuary is satisfied that similar results could be obtained as it a full stochastic projection was used eg. the possible outcomes form a symmetric distribution and information is only required on the expectation, or if a specific scenario is being tested within a simple cashflow
• A quick, independent tests is required to see that the results of a stochastic projection are reasonable

Question 15

Describe 2 approaches to calibrate stochastic models of economic variables

Answer 15

• Risk neutrals calibrations attempt to replicate market prices of financial instruments as closely as possible
• Real world calibration uses assumptions which accord to realistic “long term” expectations