Zo PDF
Zo PDF
Graham Luffrum
Timetable – Day 1
Professionalism
Life insurance products
z Living benefits
z Whole life assurances
z Unit-linked contracts
z Term assurances
z Index-linked contracts
z Convertible or z Methods of distributing
renewable term profit
Endowment assurances
consumer needs
z Expenses
z Withdrawals (surrenders)
z Capital requirement
design of contract
frequency of payment of premium
relationship between pricing and reserving
bases
level of initial expenses
Fixed term contracts
z Consumer needs
certainty of payment at a known date
z Risks
investment, expense and withdrawal as for
endowment
mortality – how long will premiums be paid
capital requirement – as for endowment
Whole life assurances
z Consumer needs
benefit on death – for funeral or to pay tax
z Risks
investment and mortality – depend on age
at entry and duration in force
expenses – inflation risk at long durations
withdrawals and capital requirement – as
for endowment
Term assurances
consumer needs
z Mortality
main risk
large anti-selection risk (except for group)
selective withdrawals
z Expenses – as for endowment
z Withdrawals – asset share can often be
negative
z Capital requirement
basic requirement small
BUT increased by required solvency margin
Convertible/renewable term
consumer needs
z Cancer
z Heart attack
z Stroke – cerebrovascular incident
z Coronary artery bypass graft
z Multiple sclerosis
z Kidney failure
z Major organ transplant
z Total and permanent disability
Critical illness contracts
risks
z Consumer needs
premiums used to buy units which change in value
according to the value of the “fund” of assets in
which they are invested
specific deductions made to cover expenses and
the cost of death, critical illness cover etc
higher expected benefit than under comparable
non-linked contracts
greater flexibility in type and size of benefits and
ability to vary premium
Description
Unit-linked contracts 2
z Risks
investment, mortality and expense risks
usually carried by insured, if no guarantees
high marketing risk due to higher variance
of benefit as compared with other
comparable non-linked contracts
Capital requirement depends critically on
the design of the contract
Index-linked contracts
z Consumer needs
consumer obtains a benefit that is guaranteed to
move in line with the performance of a specified
investment index
z Risks
main risk relates to investment – company may not
be able to invest to match movements in the index
this contrasts with unit-linked where this risk is
usually borne by the insured
Methods of distributing profit
z European methods
(i′ − i )
Savings profit t +1V
(1 + i )
1 + i′
Mortality profit ( S − t +1V )[qt − qt′ ]
1+ i
European methods 2
z Distribution channels
z Regulatory and fiscal regimes
z Professional guidance
Framework of points to consider by
actuaries in order to maintain professional
standards
Interpretation of Government Regulations
Distribution channels 1
z Own salesforce
employees of a life insurance company
paid by commission or salary or both
actively seek new clients
z Direct marketing
telephone selling
press advertising
internet
Distribution channels – effect 1
z Contract pricing
level of underwriting and hence mortality etc
assumptions depend on sales channel
class selection also affects these assumptions
withdrawal assumptions affected by class selection
and who initiated the sale
competition
– brokers need competitive contracts
– a bank must consider its good name
– an own salesforce is not in a competitive situation
Regulatory regime 1
z Further restrictions
amount of business a company can write through
requirements for reserves and solvency margins
types of asset in which company can invest
amount of any particular type of asset
z Different regulatory regimes for institutions
offering similar savings contracts, e.g. banks
and insurance companies
z May be effects on contract design
Fiscal regime 1
z Sources of risk
z Maintained by company
⇒ actuary has no direct control over them.
z Risk that data will not be adequate,
accurate and complete
⇒ result of valuing liabilities incorrect.
z Sometimes need model of policy data
risk that model does not represent the
data.
Other data
z Three risks
“model risk” → model may not be
appropriate
“parameterrisk” → parameters for the
model may not be appropriate
“random fluctuations risk” → actual
experience different from expected even
though model and parameters correct.
Mortality etc assumptions
z Fraud
z Mis-selling
z IT issues
z Systems and control failures
z Management failures
Counterparties
End
Unitised contracts
z Policy charge
z Unit allocation rate
z Bid-offer spread
z Fund management charge
z Benefit charges
z Surrender penalties
z Actuarial funding/unit cancellation rates
Policy charge
z Requirements of a model
A model must be valid, rigorous and adequately
documented.
Any model points used must be chosen to reflect
adequately the distribution of the business being
modelled.
Parameters used in a model must allow for all the
features of the business being modelled that could
significantly affect the advice being given.
Models in life insurance 4
z Sensitivity analysis
The results obtained from the models discussed
depend on
– the choice of model points
– the values assigned to the parameters of the model
We may need to test for model point error –
although we should choose an adequate set of
model points to start with
Mis-estimation of parameters can be investigated
using sensitivity analysis.
Valuing the liabilities
NP reserve =
(Expected present value of future benefits) –
(Expected present value of future net
premiums)
= St Ax +t :n−t − NP ax +t :n−t
S 0 Ax:n
NP =
a x:n
Net premium method 2
z Liabilities consist of
expected future benefits Bs
expected future expenses Es
expected future premiums (negative liability) Ps
z If Ls is the “expected liability outgo” at time s,
Ls = Bs + Es – Ps
z Gross premium reserve =
∑ s
L
s
(1 + i ) −s
= ∑ s
B
s
(1 + i ) −s
+ ∑ s
E
s
(1 + i ) −s
− ∑ s
P
s
(1 + i ) −s
Gross premium method 2
z For an endowment contract we have reserve
at time t is
St Ax +t :n −t + E c Ax' +t :n −t + E r a x' +t :n −t − P (1 − E p )a x +t :n −t
z Retrospective accumulation of
gross premiums paid to date
less actual expenses
less cost of benefits provided
using actual investment return achieved.
z It can be used
to help determine surrender values
in some profit distribution systems to determine the
profit to give to policyholders.
EAS – Investment return
z Special characteristics
contracts usually costed yearly
may be premium rate guarantee or with profit
z In theory need
unexpired premium reserve
incurred but not reported reserve
deficiency reserve (if have rate guarantee)
reserve for experience refund (if with profit)
z In practice may reserve 1 year’s premium
Critical illness benefits
z Book value
z Market value
z Historic cost
z Readily ascertainable
z Stable value
• Realistic value
• Objective
Market value 2
⇒ subjective
∞
At
∑
t =1 (1+ i)
t
z We need assumptions
⇒ subjective
z Gives a more realistic long-term value?
z Discount rate ignores risk
⇒ leads to inconsistencies
⇒ can lead to inappropriate investment
decisions
z Complicated
Value of assets for supervisory
purposes
z Risk-based capital
Calculate capital requirements which reflect the
size and overall risk exposures of an insurer.
Similar to formula method in that sub-results are
established by applying factors to exposure
proxies, e.g. asset risk, underwriting risk etc.
But uses more risk proxies and factors than the
formula method and these are combined with a
more sophisticated mathematical formula.
Calculation of solvency margin
requirements 3
z Risk-based capital
Better than formula method.
Not simple to apply, but does not need complex
systems and model to calculate results.
Calculation based on factual and historical data ⇒
no subjectivity.
Capital requirements still to some extent arbitrary.
Does not cover all risks or interactions between
them.
Not dynamic.
Calculation of solvency margin
requirements 4
z Scenario-based approaches
Attempt to analyse the impact of specific risk
variables to company specific exposure.
Capital requirements based on the worst-case
outcome from a set of scenarios applied to the
insurance company’s financial model.
Financial model is typically dynamic (but could be
static) and produces deterministic cash-flow and
balance sheet projections.
Scenarios will usually consist of various sets of
future inflation and interest rates, returns on assets
etc.
Calculation of solvency margin
requirements 5
z Scenario-based approaches
Allows for a straightforward and intuitive
interpretation of results.
Capital requirements more clearly defined.
Scenarios provide flexibility in the scope of risks
considered.
Framework for considering risk-interaction.
Can be dynamic and forward looking.
Result dependent on the specific set of scenarios
used.
Models can be complex and more sophisticated
versions impose considerable data requirements.
Calculation of solvency margin
requirements 6
z Probabilistic approaches
Attempt to cover the full range of risk variables
which are sampled from statistical distributions
using simulation.
Therefore consider a wider range of outcomes, the
likelihood of adverse development and the
interaction of risk variables.
Give the full probability distribution of possible
outcomes.
Capital requirements calculated from features of
the distribution using ruin-probability or expected
policyholder approaches (or other risk measures).
Calculation of solvency margin
requirements 7
z Probabilistic approaches
Greater flexibility.
Specifically covers interaction of risks.
Attempts to combine and refine distinct risk
categories.
Results much more difficult to understand.
Most complex approach with significant data
requirements.
Where data not available can be quite subjective.
Difficult to codify or standardise.
Solvency II
Minimum capital requirements set for Assessment of the strength and Disclosures recommendations and
firms generally using a risk based effectiveness of risk management requirements create
approach assessed by reference to systems and internal controls including transparency by allowing market
underwriting information, and assets review of: participants to assess key information
and liabilities in the financial - exposures (including the on scope of application, capital, risk
statements. reinsurance programme); exposures, risk assessment and
- internal risk models; management processes, and capital
Options for firms to graduate to - stress testing of technical adequacy of the insurance
scenario approaches and provisions and assets; undertakings.
internal (probabilistic) models. - fitness and propriety of senior
management; Disclosures on risks:
Group solvency requirements taking - asset/liability mismatch. - risks;
account of additional risks at group - key sensitivities and scenario
level. analysis on assets and technical
provisions.
Other prudential rules (assets and
liabilities).
Risks at company level
z Market changes
Risks faced by the economy
z Discount these
∞
expected profits at a EV ( NBt )
GV = ∑
suitable rate of t =1 (1 + j ) t
EPV – expenses
and commission = αB + βPa&
&x:n + γBa&
&x:n
n −1 n
EPV – benefits = B ∑ t q x v + n p x v
t +½
t =0
= B Ax:n
Equation of value - European
P =
(
B Ax:n + α + γ a&
&x:n )
(1 − β ) a&
&x:n
Equation of value - Britain
B Ax:n + Ei + Er (a&
&′x:n − 1) + Ec Ax′:n
P =
(1 − cr )a&
&x:n − (ci − cr )
Immediate annuity - Britain
12 An a (12 )
x
′
+ Ei + E p a x(12 )
P =
1 − ci
Equation of value method
z Simple
z Works well for with profit contracts.
z Not good otherwise
Does not quantify profit.
Ignores reserving requirement.
Difficult to allow for cancellations.
Very difficult to use with unit-linked
contracts.
Emerging costs method 1
Example
ECM – risk discount rate 1
Example
ECM - Advantages
z Complicated
z Can price any contract
z Allows for reserves and cancellations
z Quantifies the expected profit from the
contract
z Can investigate the sensitivity of the
expected profit
Sensitivity – Investment return
z Competition
Assumptions - mortality
z Expenses as before.
z Investment return not important as one
year contracts.
z Mortality assumptions based on
occupations involved
geographical location
free cover limit
Assumptions - Group life 2
z Product differentiation
z Remove expensive guarantees or options
z Sell to a different target market
z Change sales channel
z Use “marginal expense costing”
z Persuade shareholders to accept lower profit
z Abandon the project
Competition - what not to do
Do not change
actuarial
assumptions
Managing risk - reinsurance
z Why reinsure?
z Main types of reinsurance contract
z Specifying the amount to reinsure
z Facultative and treaty
z How much to reinsure – retention limits
Why reinsure?
z New business
reduce financial risk from new business by
– increasing available capital
– reducing financing requirement
original terms reinsurance or some form of
financing reinsurance can be used
z Technical assistance
underwriting, product design and pricing, and
systems design
particularly useful for new company or new line of
business
Main types of reinsurance
contract
z “Facultative” = may
z “Obligatory” = has to
z Hence we have (insurer/reinsurer)
facultative/facultative
facultative/obligatory
obligatory/obligatory
z Last two types are usually formalised in
a “Treaty” – the first may be.
Retention limits 1
z Sources of evidence
questions on the proposal form
reports from applicant’s usual doctor
medical examination
specialist medical tests
z Mortality options
z Investment guarantees
Mortality options 1
z Examples
option to increase death benefit at standard rates
under a term assurance (increase option)
option to take out a new term assurance at
standard rates when an existing contract expires
(renewal option)
option to convert a term assurance into an
endowment or whole life contract at standard rates
(conversion option).
Mortality options 2
z surrender values
use simulation as for maturity value of a
unitised contract
calculatepremium required in respect of
“worst case” guaranteed surrender value
Investment guarantees - Pricing 3
z annuity option
assume company invests to meet maturity value
derive a stochastic interest rate model
use model to simulate interest rate at maturity
for each simulation calculate cost at maturity of
taking annuity
cost at maturity of guarantee calculated such that
probability of making loss is less than a certain
figure.
Discontinuance terms
or
z Ease of application
Itshould not be too complicated to
calculate the values.
Conversion to a paid-up form
z Underwriting needs
Limiting conditions and
constraints 1
z Absolute matching
income from assets exactly matches outgo
in respect of liabilities
company is not at risk from future changes
in investment conditions
in practice requires that net liability outgo is
always positive, which is not usually true.
z Immunisation Skip mathematics
Immunisation – Mathematics 1
Define
At = asset proceeds at time t
Lt = liability outgo at time t
δ = ruling force of interest at time t = 0
∞
∫ Ae
−δ t
VA = t dt
0
∞
∫Le
−δ t
VL = t dt
0
Immunisation – Mathematics 2
∫
−δ t
z Expanding the ( At − Lt ) t e dt = 0
above
0
differentials
we obtain the
two conditions ∞
∫ (A − L )t e
2 −δ t
to the right
t t dt ≥ 0
0
Immunisation
ln Y (t ) = µ y + α (ln Y (t − 1) − µ y ) + ηt
where Y(t) = interest rate at time t
µy = assumed average “force” of interest rate
α = auto-regressive parameter
ηt = normally distributed r. v. with zero mean
Profit distribution and risk
z Policyholder expectations
z Provision of capital
Margins for future adverse
experience
z Additions to benefits
method has greatest degree of flexibility
the more profit is given in terminal bonus
form the greater is the deferral
super-compound RB defers more than
compound which defers more than simple
as used in the UK the method can lead to a
significant deferral of profit .
Provision of capital 5
z European methods
itis usual to distribute all the profit as it
arises
z Contribution method
sometimes with this method companies do
not distribute all the profit as it arises, but
give a terminal dividend in addition to the
regular ones
z Analyse by
type of contract
age
sex
duration from entry
degree of underwriting
source of business
smoker/non-smoker status
cause of claim (critical illness)
A of E – Withdrawals 1
z Big problem
z No correct way of dealing with them
if load too much - not competitive
if too little - may become insolvent
z Earnings affected by
spreadbetween sectors
company’s investment policy
z Consider in analysis
current earnings by asset type
expected reinvestment returns
likely future spread of investments by type
Analysis of surplus
Example
Analysis of change in EV
Example
Using the results
Professionalism