0% found this document useful (0 votes)
243 views46 pages

Chapter 2 Profit Testing Tak Kuen (Ken) Siu PDF

This document provides an outline and overview of key concepts in profit testing. Profit testing is used to determine the appropriate premium for life insurance contracts by examining the profit profile or signature. It allows sensitivity analysis of assumptions. The outline includes sections on emerging cash flows, allowance for reserves, profit measurement, and actuarial bases. An example calculates profits incorporating reserves held according to the valuation basis. Profit testing provides more informative profit criteria than traditional approaches, such as expected profit at the risk discount rate.

Uploaded by

ulfa dianita
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
243 views46 pages

Chapter 2 Profit Testing Tak Kuen (Ken) Siu PDF

This document provides an outline and overview of key concepts in profit testing. Profit testing is used to determine the appropriate premium for life insurance contracts by examining the profit profile or signature. It allows sensitivity analysis of assumptions. The outline includes sections on emerging cash flows, allowance for reserves, profit measurement, and actuarial bases. An example calculates profits incorporating reserves held according to the valuation basis. Profit testing provides more informative profit criteria than traditional approaches, such as expected profit at the risk discount rate.

Uploaded by

ulfa dianita
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 46

Chapter 2: Prot Testing

Tak Kuen (Ken) Siu


Department of Actuarial Mathematics and
Statistics
School of Mathematical and Computer
Sciences
Heriot-Watt University

Term III, 2006/07


Prot testing can be described as the process
of determining the appropriate premium for a
life contract by examining its prot prole or
prot signature. It is a very useful tool to
price life contracts, such as unit-linked con-
tracts, which are too complicated to for actu-
arial formulae. It also provides a very conve-
nient method to perform sensitivity analysis for
the prot signature to various scenarios cor-
responding to dierent actuarial assumptions.
In practice, actuaries perform prot testing for
real contracts with the aid of computer. In this
chapter, I will give an introduction to some es-
sential concepts in prot testing and illustrate
the procedure to prot test life contracts. You
will learn various important concepts and tech-
niques in prot testing, such as the emergence
of prot, allowance for reserves, prot mea-
surement and actuarial bases, etc. The outline
of this chapter is listed as follows:
Outline of Chapter 2

• Section 2.1: A review of the traditional


approach

• Section 2.2: Emerging cashows

• Section 2.3: Allowance for reserves

• Section 2.4: Prot measurement

• Section 2.5: Actuarial bases

• References: Course Note, Volume 4, Unit


4, Pages 7-50
Section 2.1: A review of the
traditional approach

• Example: Consider a non-prot endowment


assurance with the following contractual
features

 Term n = 5 years

 Age at issue x = 60

 Sum assured SA = £10, 000 paid at the


end of year of survival or earlier death

 Mortality: Assured lives mortality table


AM92 @ 4% (i.e. Interest Rate = 4%)

 Initial expenses: £100

 Renewal expenses: 5% of premium per


year
 Problem: Evaluate the annual premium
P paid at the beginning of each year
such that the expected prot is £50 us-
ing the traditional approach

• Solution:

 Recall that for the traditional approach

1. Equation of value: P äx,n = eäx,n +


Ax,n

2. Require the valuation of dierent ele-


ments in the equation of value sepa-
rately

 Use actuarial functions:

0.95P ä60,5 = 50 + 100 + 10000A60,5

 ä60,5 = 4.550 and A60,5 = 0.82499 (AM92


@ 4%, Yellow Tables, Page 101)

 P = £1943.30
Section 2.2: Emerging cashows
• Convention: Estimate the expected cash-
ows in ANY future year per policy in force
at the start of the year

• Cash ow diagram:

• Cash ow in year t

 CFt: Cashow expected t − 1 → t given


that the policy is in force at t − 1

 Pt: Premium at t − 1

 Et: Expenses at t − 1

 it: Investment return t − 1 → t

 qx+t−1: Probability of death age x + t −


1→x+t
 px+t−1 = 1 − qx+t−1

 Dt: Benet on death t − 1 → t

 St: Benet on survival to t

 CFt = Pt − Et + it(Pt − Et) − qx+t−1Dt −


px+t−1St

• Consider the last example again:

YR Prem E i D Total
1 P 100 0.038P 80.22 0.988P
+0.05P −4 −184.22
2 P 0.05P 0.038P 90.09 0.988P
−90.09
3 P 0.05P 0.038P 101.12 0.988P
−101.12
4 P 0.05P 0.038P 113.44 0.988P
−113.44
5 P 0.05P 0.038P 127.16
 In year 5: Survival Benet
= £10, 000p64 =
£10, 000(1−0.012716) = £9872.84 and
Total = £0.988P − £10, 000

 i1 = (P −100−0.05P )×0.04 = 0.038P −


4

 D1 = 10, 000q60 = 10, 000×0.008022 =


80.02

 D2 = 10, 000q61 = 90.09

 CFt = Cashow at t given in force at


t−1

 Golden rule: EPV = Amount × Prob ×


Discount

 Fill in the missing items


YR CFt t−1 px vt EPV
1 0.988P 1 −177.13
0.96154
−184.22 +0.95P
2 0.988P 0.991978 0.92456 −82.63
−90.09 +0.9061P
3

 t−1 px can be calculated based on AM92


table (Page 78) with x = 60

 v t @ 4%

 Suppose the required expected prot is


£50. Then, P =

 Note that in theory, P is the same as


the result from the traditional method;
Slightly dierent because of the round-
o error
• Given that P = £1943.30

YR PREM EXP INT D S CFt


1 1943.30 197.17 69.85 80.22 0 1735.76

 In year one, EXP = 0.05 P + 100

 INT = 0.04 × (PREM - EXP)

 Dt = 10, 000q60+t−1 from AM92 table

 For t = 1, 2, 3, 4, St = 0; S5 = 9872.84

 CFt = PREM - EXP + INT - D - S


Section 2.3: Allowance for
reserves

• Prot testing

 Denition: The process of estimating


future expected cashows and the cost
of setting up reserves in terms of a rate
of return on capital

 Application: Determine the appropriate


premium for a life contract by examining
its prot prole

• Prot vector PRO

 PROt: The prot made in the year t−1


to t per policy in force at the start of
the year, after allowance for setting up
reserves according to our basis
 PRO = (PRO1 , PRO2 , . . . , PROn), for
some year n

• PROt: Prot vector in year t − 1 → t

 tV : Reserve held per policy in-force at


time t

 Then,

PROt = Pt − Et + it(Pt − Et) − qx+t−1


Dt − px+t−1St + t−1V (1 + it)
−px+t−1 × tV

 Hence,

PROt = CFt + it × t−1V − (px+t−1 × tV


−t−1V )

 it × t−1V = Interest on the reserve held


at the start of the year t
 px+t−1 × tV − t−1V = Increase in reserves

• Prot signature σt

 Denition: The prot made in the year


t−1 → t per policy sold at time zero,
after allowance for setting up reserves
according to our basis

 σt = t−1px × PROt

 Expected prot = t t−1 px ×PROt ×v t =


P
t
t σt × v
P

• Calculating reserves

 Use a strong basis => High reserves

 Prudential or Cautious: A larger amount


of money is available when an unex-
pected cost occurs
 Example: Suppose x = 60; n = 5; Mor-
tality: AM92 @ i = 4%; SA= £10, 000
; The interest rate on reserves is 4%.
Calculate the net premium reserve

 Solution:
!
A60,5
1. Net Premium NP = £10, 000× ä =
60,5
£10, 000 × 0.82499
4.550 = £1, 813.17 @ i=
4%

2. Calculate reserves recursively:

5V = 0

(tV + N P )(1.04) = 10, 000q60+t +


t+1 V × p60+t + St+1 × p60+t

3. Hence, 4 V = £7802.22

4. Fill in the missing items in the follow-


ing tables:
Table 1.

YR CFt t−1 V Interest on Increase in


Reserves Reserves
1 1735.76 0 0 1805.21

2 1829.89

3 1818.86

4 1806.54

5 −8080.02 7802.22

Table 2.

YR PROt σt
1 −69.45 −69.45

 Interest on reserves = 0.04t−1V (i.e. Not


the interest rate in the valuation basis)
 PROt = CFt + Interest on reserves -
Increase in reserves

 σt = t−1px × PROt

 Expected prot =
P5 t
t=1 σtv = 50.26 ≈
50 @ 4%, which is the same as the case
without reserves

 Typical pattern of prots: σ1 < 0 and


σt > 0, for all t > 1

• New business strain:

 Equal to −σ1

 The rst premium is insucient to pay


for both the initial expenses and setting
up the reserve at the end of the year

 The insurer needs a source of funds to


pay for the initial reserve or loss
• The capital of the insurer

 Additional assets above those needed


for reserves

 Total Assets = Reserves + Capital

 Reduced by new business strain

 Increased by later prots

 Provided by the owners or shareholders


of the insurance company

 The owners will expect a good rate of


return on the capital used
Section 2.4: Prot Measurement

• Traditional approach

 Evaluate the expected prot discounted


at the premium basis interest rate ip us-
ing actuarial functions

 ONLY method available using actuarial


functions

 The premium basis interest rate ip is the


ONLY appropriate discount rate using
actuarial functions

 In practice, ip is Unlikely to be an ap-


propriate discount rate

• Four more informative prot criteria can


be introduced with prot testing
 Expected prot at RDR

 Prot margin

 Discounted payback period

 Internal rate of return

• Expected prot at RDR

 Risk Discount Rate (RDR): An investor's


Required rate of return allowing for the
risk of the investment

 Then, expected prot =


P t
t σtv @ RDR

 The expected prot at RDR is the same


as that in the traditional approach if
RDR = ip

 Usually, RDR > ip


 ip = Expected return on our assets (i.e.
An estimate of the actual average re-
turn)

 Normally invest in safe assets (e.g. Gilts)


which provides a low rate of return

 The insurer's capital is invested in a riskier


asset - Insurance policies

 What are the major risks?

1. Interest

2. Mortality

3. Expenses

 Capital Asset Pricing Model


Use of the
(CAPM) or Arbitrage Pricing The-
ory (APT) to evaluate the required dis-
count rate given the risk
 Typically, 10% < RDR < 15%

 Consider the last example:

1. Expected Prot (EP) @ 10% = 33.56

2. RDR > ip = 4% => EP @ RDR <


EP @ ip = 50.26

 The rewards for selling the policy

1. Split between the salesperson and the


insurer

2. Saleperson receives commission

3. Insurer's prot should be related to


commission

 Example: Suppose we (i.e. the insurer)


only require an expected prot equal to
half of the initial commission. If the ini-
tial commission = 0.5×Initial Expense =
50, then we must pass the test since
EPV @ 10% = 33.56
• Prot Margin

 Denition: The proportion of each pre-


mium taken as prot

 Prot Margin EP @ RDR


= EPV of Prem @ RDR

 Can be used to set our prot criteria


(e.g. Prot criteria may be that the
prot margin exceeds 1 %)

 Example: Consider EP @ 10% = 33.56.


EPV of premiums = 4 t p =
P
t=0 P (1/1.1) t x
7972.25. Then,
33.56
Prot Margin = = 0.421%
7972.25
 Prot is a very small percentage of pre-
miums
• Discounted payback period (DPP)

 The time it takes to repay the new busi-


ness strain with interest at the RDR

 The rst time t such that EPV of prots


in the rst t years

(1 + RDR)−sσs ≥ 0
X
=
s≤t

 The faster the strain is repaid (i.e. the


smaller t), the sooner the insurer can
invest in a new project
 Example: Suppose the prot signatures
σt are given as before; RDR = 10%.
Fill in the missing items in the following
table
YR σt P
s≤t (1 + RDR) σs
−s

1 −69.45 −63.14

2 34.01

3 33.70

4 33.36

5 32.99

1. DPP = 4 years

2. Less than half of the new business


strain has been repaid by year 2
• Internal rate of return (IRR)

 The interest rate at which the present


value of the expected prots is zero

 IRR = i such that


P
t σt(1 + i)
−t = 0

 Cannot always be used since there may


be no solutions or more than one solu-
tions to the above equation

 Prot criteria: IRR ≥ RDR (i.e. Ac-


cept the contract with yield at least our
required rate of return)

 Example: Suppose RDR is 10% and


IRR = i. Given σt in the following table:
YR σt
1 −104.52

2 16.78

3 35.00

4 53.33

5 71.73
Then, i can be determined by the fol-
lowing equation of value:
−104.52 16.78 71.73
0= + + ··· +
(1 + i) (1 + i)2 (1 + i)5

 Hence, IRR = i = 19.7% > RDR

• Note that DPP and IRR have widely been


used for evaluating prots from investment
projects in corporate nance

• Prot measurement in practice

 Evaluate the four prot measures when


designing a new product or setting pre-
mium rates

 Adjust premium/benets until the prot


criteria are passed
 Marketing considerations: Very impor-
tant to ensure that we are able to sell
the contract (e.g. competitors' premi-
ums)
Section 2.5: Actuarial bases
• Reference: Course Notes, Volume 4, Pages
30-40

• Recall that
P ROt = Pt − Et + it(Pt − Et) − qx+t−1Dt −
px+t−1St + t−1V (1 + it) − px+t−1
×t V
σt = t−1px × PROt

• Calculate reserves, premium, PROt and σt

• Specify the unknown future levels of in-


terest rate, mortality and expenses

• Actuarial Basis: A set of assumptions used


to carry out one or more actuarial calcula-
tions
• Estimate reserves, premium, PROt and σt
using an Actuarial Basis whose elements
include assumptions on:

 Interest rate

 Mortality

 Expenses

• Carry out the calculations based on these


assumptions, but not the expectation of
their actual future levels
• Experience Basis

 A set of assumptions about the actual


outcome

 Suppose the premium and reserves are


given. We can calculate PROt and σt
under ANYset of assumptions of future
experience

• Scenario Testing: Given the same premium


and reserves, we perform prot test using
several dierent sets of experience

• Sensitivity Analysis: Investigate the conse-


quences of changing only one item of expe-
rience with other items of experience being
xed
• Example: Assume i = 5% (i.e. the ex-
perience basis interest rate); The interest
rate on reserves is 5%; The RDR is 10%.
Fix expenses and mortality; Premium and
reserves are given as before.

YR PREM EXP INT D S CFt


1 1943.30 197.16 87.31 80.22 0 1753.22

2 1943.30 97.16 92.31 90.09 0 1848.35

3 1943.30 97.16 92.31 101.12 0 1837.32

4 1943.30 97.16 92.31 113.44 0 1825.00

5 1943.30 97.16 92.31 127.16 9872.84 −8061.56

YR t−1 V Int on Res Increase PROt σt


Given 0.05t−1 V in Res
1 0 0 1805.21 −51.99 −51.99

2 1819.81 90.09 1868.39 70.95 70.38

3 3721.73 186.09 1933.44 89.96 88.44

4 5712.94 285.65 2000.77 109.88 106.92

5 7802.22 390.11 −7802.22 130.77 125.81

 EP @ 10% = 228.49

• Question: What will be the eect of a


change in the Experience basis on the ex-
pected prot?

 High interest => High prot

 High expense => Low prot

 High mortality: Depends on what kind


of products

1. Term assurance: Low prot

2. Annuity: High prot

• Premium Basis: Calculate premiums using


either

1. The traditional approach

2. A prot test

• Example: The premium basis can be 4%


AM92
• Once the premium is calculated accord-
ing to a given premium basis, we can test
the prot under dierent experience bases
with the premium being kept xed

• Question: What will be the eect of a


change in the premium basis on the ex-
pected prot (with the experience basis be-
ing held constant)?

 High interest => Low premiums =>


Low prot

 High expenses => High premiums =>


High prot
 High mortality

1. Term assurance => High premium =>


High prot

2. Annuity => Low premium => Low


prot

 High premiums => Higher prots =>


Strong premium basis

 Marketability => Premiums cannot be


too high in order to sell the contracts

• Valuation basis

 Need to know the size of the reserves


to calculate the prot vector PROt

 The valuation basis only species the


size of the reserve tV
 All other cashows are given by the ex-
perience or premium basis

 The valuation basis is specied by leg-


islation and has to be prudent or strong
=> Highly safe reserves

• Question: How are the reserves aected by


a change in the valuation basis?

 High interest => Low reserves

 High expenses => High reserves

 High mortality

1. Term assurance => High reserves

2. Annuity => Low reserves

• Question: How does a change in reserves


aect the prot signature?
• Example:

 Experience basis: 4% AM92

 Valuation bases

1. 4% AM92 => Net premium (NP) =


A60,5
10, 000 ä = 1, 813.17
60,5

2. 6% AM92 => N P = 10, 000×( 0.75152


4.39 ) =
1, 711.89

 Fill in the missing items in the following


table

Valuation Basis 4% AM92

YR CFt t−1 V Int on Res Increase PROt σt


Given 0.04t−1 V in Res
1 1735.76 0 0 1805.21 −69.45

2 1829.89

3 1818.86

4 1806.54

5 −8080.02
Valuation Basis 6% AM92

YR CFt t−1 V Int or Res Increase PROt σt


Given 0.05t−1 V in Res
1 1735.76 0 0 1734.73 1.02

2 1829.89

3 1818.86

4 1806.53

5 −8080.02

• Remarks:

 High valuation basis interest rate =>


Lower reserves

 Setting up larger reserves leads to smaller


prots (larger losses) in the early years
since the prot vector includes the in-
crease in reserves as an item of loss

 Setting up larger reserves leads to larger


prots in the later years since larger re-
serves early on
1. More interest on reserves

2. Smaller increase in (or larger release


of ) reserves in the nal year

 The 6% valuation basis is too weak


=> Reserve is too small to pay the ben-
ets in the nal year and σ5 = −47.2 < 0
(a loss)

• Question: How does a change in reserves


(i.e. choice of valuation basis) aect the
EPV of prots?

• Theorem: Suppose the RDR is the ex-


perience basis interest rate (i.e. RDR =
i). Then, the EP is NOT aected by the
choice of valuation basis
• Example:

Consider the last example. Suppose RDR


= 4%. Then, the expected prot is 50.26.
For valuation bases 4% and 6%, calculate
the expected prots and check them by
yourself

• Proof of the theorem:

 EPV of Prots
n
(1 + RDR)−t
X
= × t−1px × PROt
t=1

 But, RDR = i, the experience basis in-


terest rate

 EPV
n
(1 + i)−t × t−1px × [(Pt − Et)
X
=
t=1
×(1 + i) − qx+t−1 × Dt − px+t−1 ×
St + t−1V (1 + i) − px+t−1 × tV ]
 Note that the premiums and reserves are
given and that all other items are esti-
mated on the experience basis

 EPV
n
(1 + i)−t ×t−1 px × CFt +
X
=
t=1
n
(1 + i)−t+1 × t−1px × t−1V
X

t=1
n
(1 + i)−t × tpx × tV
X

t=1

 Let s=t−1

 EPV
n
(1 + i)−t × t−1px × CFt +
X
=
t=1
n−1
(1 + i)−s × spx × sV
X

s=0
n
(1 + i)−t × tpx × tV
X

t=1
 However, nV = 0V = 0

 Therefore, EPV
Pn
=
t=1 (1+i) −t×
t−1 px ×
CFt, which does not depend on the re-
serves

• Require a high rate of return (RDR) on


capital for setting up large reserves

• Why?

 The insurer needs to borrow a large amount


of capital to nance large reserves

 There is a risk that actual experience is


worse than the premium basis, and so,
the capital may never be returned

 Require a high rate of return (RDR) on


capital as compensation for the risk
• Example: Expected prots @ RDR = 10%
for dierent valuation bases are evaluated
as follows:

VALN BASIS 4% 6%
EPV 33.56 49.45

• Hence, the EP is reduced by setting up


large reserves

• Question: What will the eect of a change


in the valuation basis be on the expected
prot?

 High interest => Low reserves => High


prot

 High mortality

1. Term assurance => High reserves =>


Low prot
2. Annuity => Low reserves => High
prot

• In practice, we need to strike a balance of


the tradeo between

 Low reserves => More prot

 High reserves => More prudent

• Actual Experience

 The valuation, premium and experience


Assumptions
bases are all of the Un-
known Future experience
 Once the contract is nished, the actual
experience is the Known past interest,
mortality and expenses
 Ideal situation: Our best estimates or
assumptions are as close as the actual
experience as possible

 In practice, Estimation Error does ex-


ist

• Emerging Cashows v.s. Actuarial Func-


tions

 Advantage of actuarial functions

1. Quick and easy to calculate by hand

 Disadvantages of actuarial functions

1. Ignore the cost of setting up reserves

2. Provide no information on the timing


of cashows

3. Dicult to use to price complex ben-


ets (e.g Unit-linked policies)
 Advantages of emerging cashows

1. Easy to implement using a computer,


in particular, spreadsheet

2. Allow for the cost of reserves

3. Provide the expected cashows for each


year => Convenient for nancial plan-
ning, especially choice of assets

4. Can be used for valuing even the most


complex contracts or benets
 Disadvantages of emerging cashows

1. Very slow to calculate by hand

2. Even quicker to calculate commuta-


tion functions by computers

End of Chapter 2

You might also like