Chapter 4: Life Annuities

Chapter4 Life Annuities

1. To understand the specifications of some simple life annuities

2. To calculate APVs for different life annuities

3. To draw relations between life annuities and life insurances

4. To apply recursion formulas for discrete life annuities

S. To relate continuous, discrete and mthly life annuities

The annuities you studied in Exam FM were annuities certain, in which all annuity payments
were guaranteed to be made. In this chapter, we will generalize annuities certain to life annuities.
Pensions, for example, are life annuities. A pensioner gets monthly payments, but the payments
will stop at the pensioner's death. Similar to annuities certain, life annuities could be continuous,
discrete or mthly. First, let us focus on continuous life annuities.

- 4. 1 Continuous Life Annuities

Whole Life Annuities

Let us begin with a review of continuous annuities certain, which you learnt in Exam FM. The
diagram below illustrates a continuous annuity certain with a payment rate of$ I and a term of n
years.

re: Actex 2013 I Johnny Li and Andrew Ng I SoA Exam MLC

 Chapter 4: Life Annuities

Payments at a rate of $1 per

annum are made continuously. (_

"""='"=""-====i----+• Time from now

0 I I+ dt n

Within a very small time interval t to

t + dt, a payment of $1 x dt is made

The present value (at time 0) of this annuity is given by

- r'= v dt = -1-v"
6
- .

The bar above a indicates that the payments are made continuously, while the subscript n I

indicates that the annuity has a duration of n years.

For a continuous whole life annuity on (x), the length of time over which the payments are made
is T,, the random future lifetime of (x). The payments from such an annuity (with an assumed
payment rate of$1 per annum) are illustrated in the following diagram.

Payments at a rate of $1 per

annum are made continuously

Time from now ( -

0 I+ dt
(Age x)
Within a very small time interval t to
t + dt, a payment of $1 x dt is made

As you see, we just replaced n with Tx. The present value of the payments is therefore
_ l-v 7'
Y=a-=--
r,1 t5 Tx <': 0.

The present value is random, because it depends on the future lifetime random variable T,. From
now on, we use Y to denote the present value random variable for a life annuity with a payment
rate of $1 per annum.

{:o Actex 2013 I Johnny Li and Andrew Ng I SoA Exam MLC

Chapter 4: Life Annuities

We are, of course, interested in values of E(Y) and Var(Y). First, let us focus on E(Y), the
actuarial present value. We can calculate E(Y) by integrating out the product of the random
variable Y and the probability density function of T,. That is,

This formula, however, is not the best way to calculate E(Y). By integration by parts, we can
show that E(Y) can be expressed alternatively:

E(Y) = r v',p,dt.

This alternative formula is much easier to memorize and to work with. It can be interpreted
easily as follows.
- The payment over a very small time interval t to I+ dt is $di.
- The appropriate discount factor for this payment is v'.
To get this payment, the individual has to survive to time I, with a probability 1Px·
- The APV of this payment is therefore v',p,dt.
- The APV of the entire annuity is the sum of the APVs of all payments. So we need to
integrate from 0 to oo.
Of course, if there is a limiting age, then you should replace oo with w-x.

We denote E(Y) for a whole life annuity on (x) with a payment rate of $1 per annum by a,. The
subscript x indicates the age when the annuity begins, while the bar above a indicates that the
payments are made continuously. It is important not to confuse symbols for life annuities with
symbols for annuities certain. As an example, a, 0 and a;Oi are two totally different things:

A continuous whole life A continuous annuity certain

annuity on a person age 30 with a term of 30 years

vs. a30I

The subscript 30 indicates The subscript 30 I indicates the

the age at issue is 30 length of the payment stream is 30

\:. Act ex 2013 Johnny Li and Andrew Ng I SoA Exam MLC

.......

Mathematically, a30 = E(ar

_l(I 1
) = f v',p dt, while a
30 301
= l-v". Remember their difference!
0

We then proceed to the calculation ofVar(Y). Sadly, we do not have E(Y 2) evaluated at O= E(Y)
evaluated at 20: That is,

Var(Y) *'a, -(a,)'.

To calculate Var(Y), we should consider the relationship between Y and Z, which we now
explain.

1'
Since Y = I-;' for T., 2. 0, we have the following relationship:

1-Z
Y =5- ,

1'
where Z = v ' for T, 2. 0 is the present value random variable for a continuous whole life
insurance of$! on (x). We call this the "Y-Zrelation".

The Y-Z relation is extremely useful. First, by taking expectation on both sides, we obtain

I-A
ax=--'
5
This formula allows you to calculate the APV of a whole life annuity when you know the APV r
of a whole life insurance, and vice versa. Second, by taking variance on both sides, we obtain

Var(Y) = Var(l-Z) = Var(Z) '/ix -/i'x

o o' o'
You should use this formula to calculate Var(Y).

:i) Actex 2013 Johnny L1 and Andrew Ng I SoA Exam MLC

Chapter 4: Life Annuities

Example 4.1

You are given:

(i) µx =µfor all x ;:>: 0.
(ii) The force of interest is o.
Express the following in terms ofµ and o.
(a) a,
(b) Var(Y), where Y is the present value random variable for a continuous whole life annuity on
(x) with a payment rate of$ I per annum.

- Solution
(a) We may approach this problem in two different ways. The first method is to calculate the
quantity by integration:

a,= r v',p,dt
= r e-b'te-J11 df = r e-(O+p)tdf

= = µ+o

The second method is to use the Y-Z relation and the fact that
- µ
A=--
x µ+o
when the force of mortality is constant. This gives

1-__1'._
- 1-A,
a=--·
µ+o I
x 0 o µ+o
(b) We make use of the Y-Zrelation, which gives

Var(Y)=
o'
=-I (
o' µ + 20
µ -(__1'
µ+o
.__)')·
[ END]

'i:; Actex 2013 I Johnny Li and Andrew Ng I SoA Exam MLC

 Chapter 4: Life Annuities

Example 4.2 [Course 3 Fall 2001 #3] L

For a continuous whole life annuity of 1 on (x), you are given:

(i) Tx is the future lifetime random variable for (x).
(ii) The force of interest and force of mortality are equal and constant.
(iii) ax = 12.50

Calculate the standard deviation of a:IJ.

(A) 1.67 (B) 2.50 (C) 2.89 (D) 6.25 (E) 7.22

- Solution
Note that Y = aT;i is the present value random variable for a whole life annuity of 1 on (x).

We are given thatµ= J. Hence,

l 1
ax =--=-=125
µ+S 2µ . '

which means p = o= 0.04. [Jn the above, we made use of the result from part (a) of Example
4.1.] Also,

A = ___l!._ = j!_ = ]__ and 2

A= µ = l:!:_ = -
x p+o 2µ 2 ' µ+20 3µ 3

As a result,

Var(Y) =
'Ax -A'' -l- (-
l --
l ) =52.083.
s' 0.04 2 3 2 2 (-

The standard deviation is ..)52.083 = 7.217.

[ END]

Temporary Life Annuities

An n-year temporary life annuity makes payments as long as the policyholder is alive, but the
period over which the payments are made is limited to n years. The diagram below illustrates
the payments from an n-year temporary life annuity of $1 on (x):

V Actex 2013 I Johnny Li and Andrew Ng I SoA Exam MLC

Chapter 4: Life Annuities

Case I: Death occurs before time n (i.e., T, s n)

Payments at a rate of $1
per annum are made
continuously until time T,,

Time from now

0 T, n
(Age x)

Case !I: Death occurs after time n (i.e., T, > n)

Payments at a rate of $1 per annum

are made continuously until time n

Time from now

0 n T,
(Age x)

,.
Hence, the present value random variable can be specified as follows:

1-v'
Y= ar:i =-o- Tx s n

1 a-=--
•I
I-v"
0
Tx > n.

We can calculate E(Y) by integrating out the product of the random variable Y and the
probability density function of T,. That is,

This lengthy formula is usually difficult to work with. Using integration by parts, we can show
that E( Y) can be expressed alternatively as

Summing the APVs of all Payment over the time

possible payments from interval t to dt is $di
time 0 to time n E(Y) = £' v',pxdt /
Discount factor
a period oft years
L Probability of survival to time t

\) Actex 2013 Johnny Li and Andrew Ng I SoA Exam MLC

 Chapter 4: Life Annuities

This alternative formula is much easier to work with! We denote E(Y) for an n-year temporary
(
life annuity of I on (x) by aml- . The subscript x, as before, indicates the age at issue, while the
subscript n I indicates that the payments will last for at most n years.

We also have a "Y-Z relation" for an n-year temporary life annuity of $1 on (x). Specifically, we

1-Z
have Y = - - , where Z = {v'; '11
T <n
'- is the present value random variable for an n-year
0 v , T_r: > n

endowment insurance of $1 on (x). The Y-Z relation implies that

1-A -
Zi-= x:nl
x.111
6

which allows us to calculate a,, 1

from Ax•I, and vice versa.

The Y-Z relation also implies that

Var(Y) = 02

which allows us to calculate Var(Y) for an n-year temporary life annuity straightforwardly.

Deferred Life Annuities

An n-year deferred life annuity 1s a life annuity in which payments start in n years if the F

policyholder is still alive at time n. In other words, if death occurs before time n (i.e., T, :> n),
then no payment will be made. If death occurs after time n (i.e., Tx > n), then payments will be
made as follows:
Payments at a rate of $1 per
annum are made continuously
from time n to time Tx

Time from now

0 n
(Age x)

Hence, the present value random variable can be specified as

.lo Actex 2013 [ Johnny Li and Andrew Ng [ SoA Exam MLC

Chapter 4: Life Annuities

It can be shown that E(Y), the actuarial present value, can be calculated with the following
formula:

Summing the APVs of all Payment over the time

possible payments interval t to dt is $dt
starting from time n E(Y) = [ v',p,dt /

Discount factor for

a period oft years
------J L Probability of survival to time t

We denote E(Y) for an n-year deferred life annuity of $1 on (x) by • ax. The subscript nl
1

indicates the period of deferral.

Unfortunately, we do not have a simple "Y-Z relation" for an n-year deferred life annuity. So,

we do not have a simple linear relation between , 1

ax and • .Ax. Also, to calculate Var(Y), we
1

need to use the first principles. That is, we first calculate

and then calculate Var(Y) by using Var(Y) = E(Y 2) - [E(Y)]2. Because the calculation ofVar(Y)
for an n-year deferred life annuity is so tedious, it is unlikely to appear in the exam.

We can similarly define an m-year deferred n-year temporary life annuity. The APV of such an
annuity is denoted by The subscript ml indicates that the payments are deferred by m

years, while the subscript n I indicates that payments will be made for at most n years.

i; Actex 2013 I Johnny Li and Andrew Ng I SoA Exam MLC

 Chapter 4: Life Annuities

Certain-and-Life Annuities
(

An n-year certain and life annuity (also called guaranteed annuity) makes guaranteed payments (

for n years, and if the policyholder survives more than n years, makes contingent payments until
his/her death. The payments from an n-year guaranteed annuity of $1 on (x) are illustrated
below:

Case I: Death occurs before time n (i.e., Tx s n)

Payments at a rate of$! per
annum are made continuously
until time n

"'l-======"'-"""'"""'""""""'""""'"""'"'1----------•• Time from now

0 Tx n
(Age x)

Case II: Death occurs after time n (i.e., Tx > n)

Payments at a rate of$ I per

annum are made continuously
until time Tx

0 n
(Age x)

Hence, the present value random variable can be expressed as

a- Tx sn
y = _"1
{ al;i T, > n
+:'ii--------,:
= : nr: : :

Present value for an

annuity certain with a
term of n-years
J : a-1: +:(ar1-a-1):
I
I
._
II'

__ ,
I
I n r •
I

Present value random

variable for an n-year
deferred life annuity

.[; Actex 2013 I Johnny Li and Andrew Ng I SoA Exam MLC

Chapter 4: Life Annuities

From the above, we see that an n-year guaranteed annuity is just a combination of an annuity
certain with a term of n years and an n-year deferred life annuity. As a result, we have the
following formula for calculating E(Y):

We denote E(Y) for an n-year guaranteed annuity of $1 on (x) by o,, 1

• The long bar over the

entire subscript is intended to indicate that the annuity continues to the maximum of n and the
remaining lifetime Tx.

Example 4.3

You are given:

(i) µ, = 0.05 for x ::". 30
(ii) b'= 0.10
Calculate
30:101

- Solution

First, we calculate 101

0, 0 :

101GJ0 = J: 1PJodt
V
1

= J: e--0.11e-o.05,dt

= J: e-0.1s1dt

e---0.151
= = 1.4875.
-0.15 10

We also need oliii, which is calculated as follows:

- l-e--0.1x10
aiiii= -6.3212.
0.1
Finally,

30:101
= 0-101 + ,01 0 10. = 6.3212 + 1.4875 = 7.8087.
[ END)

/:. Actex 2013 I Johnny L1 and Andrew Ng I SoA Exam MLC

 Chapter 4: Life Annuities

Increasing Annuities

For a continuously increasing annuity, the payment rate increases continuously and linearly with
time. That is, over a very small time interval from t to I+ dt, the annuity payment is $t x dt. The
probability of this payment is iPx and the applicable discount factor is v'. Hence, the APV of a
continuously increasing whole life annuity on (x) is given by

(la), = rtv', p,dt .

In the notation, I means continuously increasing. Similarly, the APV for a continuously
increasing n-year temporary annuity on (x) is given by

r tv' ,p,dt.

The following table summarizes the formulas for continuous life annuities.

%
Continuous Life Annuities %
%
%
%
% %
?,
% %
% I

% F-
%
%
%
% %
%
?,
%
%
%
%
%
%
%
?,
%
%
% %
%

1e Actex 2013 I Johnny Li and Andrew Ng I SoA Exam MLC

Chapter 4: Life Annuities

- 4. 2 Discrete Life Annuities (Due)

There are two kinds of discrete annuities: due and immediate. An annuity due makes payments
at the beginning of each time period, whereas an annuity immediate makes payments at the end
of each time period. In this section, we focus on life annuities due. The theory and formulas here
will parallel what has already been developed for the continuous case. If you have mastered the
continuous case, the discrete (due) case will be easy because it is similar.

Whole Life Annuities

Let us begin with a review of annuities due with a fixed term. The diagram below shows the
payments from an annuity due of$ I with a fixed term of n years.

n payments (made at the

beginning of each year)

$1 $1 $1 $1

Time from now

0 2 n-1 n

The present value of the payments is given by

.. , 1-v"
a;i =Liv =-d '
k=O

i
where d = -- is the effective rate or-discount. The two dots above a indicate that the annuity
I +i
is an annuity due (i.e., one with payments made at the beginning of each time period).

For a whole life annuity due, payments will be made at the beginning of each year as Jong as the
policyholder is alive. The diagram below illustrates the payments from a whole life annuity due
of$! on (x).

'( Actex 2013 I Johnny Li and Andrew Ng I SoA Exam MLC

 Chapter 4: Life Annuities

Death occurs
(no more payment beyond here)
Kx + I payments in total

$1 $1 $1 $1

Time from now

0 2 Kx Tx Kx+I
(Age x)

From the diagram above, you can see that the present value random variable can be specified as

.. I -v
Y= aK.+11 = d Kx = 0, I, 2, ...

The calculation of E(Y) can be accomplished by considering the APV of each possible payment:

$1 if(x) $1 if(x) $1 if(x)

$1
is alive is alive is alive
Time from now
0

(Age x)
(
f-
E(Y) =I+ vp, + v',px + v',px + ...

If there is a limiting age, we replace oo in the formula above with m- x - I. This is because if no

one can live to age m - x (where m is an integer), then T, must be strictly smaller than m - x,

which means that the largest possible value of Kx (the integral part of T,) would be m- x - 1.

We denote E(Y) for a whole life annuity due of$1 on (x) by iix. The two dots above a indicate

that the annuity is an annuity due, while the subscript x denotes the age of the policyholder

.i:' Actex 2013 I Johnny L1 and Andrew Ng I SoA Exam MLC

Chapter 4: Life Annuities

when the annuity begins. You should notice that the E(Y) formula here is highly parallel to that

for the continuous case, i.e., ax = r v',pxdt.

We can build a "Y-Z relation" for a discrete whole life annuity. Specifically, we have
1-Z
Y=-
d '

where Z = vK,+i is the present value random variable for a discrete whole life insurance of$ I on
(x). The Y-Z relation implies that

.. \-Ax
ax=-d-,

which allows us to calculate ii, when Ax is given, and vice versa. Also, we have

2
A -A'
Var(Y) = x x
d'

which allows us to compute Var( Y) easily.

Temporary Annuities

The payments from an n-year temporary annuity due of $1 on (x) are illustrated in the following
diagram:

Case I: Death occurs before time n (i.e., Kx < n)

Death occurs
(no more payment beyond here)

K, + I payments
in total

$1 $1 $I . . . $1

Time from now

0 2 Kx Kx+ I n
(Age x)

t. Actex 2013 I Johnny Li and Andrew Ng I SoA Exam MLC

 Chapter 4: Life Annuities

._, --
Case II: Death occurs after time n (i.e., Kx 2 n)
(-

n payments in total Death occurs

$1

0
$1 $1

2
$1

n-1 n
1 Time from now

(Age x)

Hence, the present value random variable can be specified as

iiK,+11 = 1-:K,.1 K_, < n

Y=
( .. 1-v"
a-=--, Kx 2n
111 d

To calculate E(Y), we can again consider the APV of each possible payment:

The last possible payment

$1 if(x)
$1
is alive
I
Time from now
i--
0 n-1 n

(Age x)

2 1
E(Y) = 1+ VPx +V 2 Px + ... +v"- n_ 1p,
n-1

= Iv'.Px·
k"'O

We denote E(Y) for an n-year temporary annuity of $1 on (x) by The subscript n I again

indicates that the annuity payments are made for at most n years.

l;' Actex 2013 I Johnny Li and Andrew Ng I SoA Exam MLC

 1-Z
It is easy to see that Y = - - , where Z =
{vK,+I ' K <
"
n is the present value random variable
d v", Kx2n
for a discrete n-year endowment insurance of$ I on (x). This Y-Z relation implies

I-Ax:11j- 'Ax;;r - (A,;;/

iix:nl- and Var(Y) =
d d'

Deferred Life Annuities

By considering the APV of each possible payment, it is easy to show that the APV of an n-year
deferred life annuity due is

The subscript nl is to indicate that the payments are deferred by a period of n years. If there is a
limiting age, we replace oo in the formula above with OJ-x- I. Note that there is no simple "Y-Z
relation" for deferred life annuities.

Certain-and-Life Annuities

Recall that an n-year certain-and-life annuity is just the combination of an n-year annuity certain

and an n-year deferred life annuity. Hence, we have iix•I = ii;;r +,1iix.

Increasing Annuities

We can construct an annually increasing whole life annuity due on (x) with payments $1, $2, $3
and so on. That is, the payment in year k +I is $(k +I), where k = 0, I, .... The discount factor
for the payment is l, and the probability for the payment is k{Jx- Hence, the APV of such an
annuity is given by

(Iii), = L (k + l)v\px .
kooO

l. Actex 2013 I Johnny Li and Andrew Ng I SoA Exam MLC

 Chapter 4: Life Annuities

In the notation, I (without a bar above) indicates that the annuity payments are increasing
{_
annually. Similarly, the APV of an annually increasing n-year temporary life annuity due on (x)
IS

11-\

= L (k + \)v'
k=O
1Px

The following table summarizes the formulas for various discrete life annuities (due).

u Discrete Life Annuities (Due)

?,
%:

%:

% %
?, ?,
%
% %
%
% %
?,
%
?,
%: %:
(---
%
%
?,

f' Actex 2013 I Johnny Li and Andrew Ng I SoA Exam MLC

Chapter 4: Life Annuities

Example 4.4 &#

You are given:
(i) The following life table:
x 90 91 92 93 94
!, 100 75 50 25 0

(ii) i = 0.06.
Calculate the following:
(a) ii90

(b) iioo,21

- Solution
(a) First of all, note that OJ= 94, since 193 >0and194 = 0. We have
94-90-1

ii90 = L vk kP9o
k=D

I 75 I 50 I 25
= l + - - x - + -2 - x - + - 3- x -
1.06 I 00 l.06 I 00 1.06 I 00
= 2.3625.
(b) We have
.. ,_, ' 1 75
·
= l:v
•=•
,p 90 =l+--x-=1.7075.
l.06 I 00
[ END)

- 4. 3 Discrete Life Annuities (Immediate)

An annuity immediate makes payments at the end of each time period. The good news here is
that we do not need to calculate annuities immediate from scratch. We can calculate the value of
a life annuity immediate simply by adjusting the value of the corresponding life annuity due.

C Actex 2013 I Johnny Li and Andrew Ng I SoA Exam MLC

 Chapter 4: Life Annuities

Whole Life Annuities

The diagram below compares the payments from a whole life annuity due and a whole life
annuity immediate of $1 on (x).

Death occurs
(no more payment beyond here)

$1 $1 $1

Time from now

0 2 Kx Kx +I

(Age x)

Immediate:
Death occurs
(no more payment beyond here)

--- '
'' $1 $1
' -- -- ,,'
Time from now
0 2 Kx Kx+ I

(Age x)

We observe that the only difference between these two annuities is the payment at time 0. Hence,
we have the following relation:

We use the notation ax to denote the APV of a life annuity immediate of $1 on (x). As with
annuity certain terminology, the symbol a without two dots denotes an annuity immediate.

t:· Actex 2013 I Johnny Li and Andrew Ng I SoA Exam MLC

Chapter 4: Life Annuities

Temporary Life Annuities

The diagram below compares the payments from an n-year temporary life annuity due and an n-
year temporary life annuity immediate of $1 on (x).

$1 if(x) $1 if(x) $1 if(x)

$1 is alive is alive is alive , ____ ,'
Time from now
0 2 n-1 n
'--y--J
(Age x) Yearn

Immediate:

- - - - $1 if(x) $1 if(x)
'' '' $1 if(x) $1 if(x)
, ____ ,' is alive is alive is alive is alive
'
Time from now
0 2 n -1 n
'--y--J
(Age x) Yearn

We observe that there are two differences: the first payment in the annuity due and the last
payment in the annuity immediate. Hence, we have the following relation:

In the formula, nEx is the APV of the last payment in the annuity immediate.

Deferred Life Annuities

By a similar reasoning, we can show that the only difference between an n-year deferred life
annuity due and an n-year deferred life annuity immediate of $1 on (x) is the payment that is
made at time n. Hence, we have the relation ,,1a, ,,1 ii, - ,,E,.

((! Actex 2013 I Johnny Li and Andrew Ng I SoA Exam MLC

 Chapter 4: Life Annuities

The following table summarizes all formulas regarding life annuities immediate.

::; Formulas for Calculating Life Annuities Immediate

%
%
?, ?,
%

?, ?,

Example 4.5 [Course 3 Fall 2001 #26]

You are given:

(i) Ax= 0.28
(ii) Ax+io = 0.40
(iii) 0.25

(iv) i = 0.05.
Calculate a"2oi.

(A) 11.0 (B) 11.2 (C) 11.7 (D) 12.0 (D) 12.3

- Solution
We always calculate the value of a life annuity immediate by adjusting the value of the
corresponding life annuity due. So, we need ii"2oi . Since we are given values of various

insurances, it is natural to consider the relation

1-Ax:201
-
ii.t:201
-
d

.C Actex 2013 I Johnny Li and Andrew Ng I SoA Exam MLC

Chapter 4: Life Annuities

1 5
We have d = - -·. = 0.0 , but the value of A -201 is not given in the question. To calculate A ·WI'
1+1 1.05 x. '

we consider
A.r:201
- =A'-+A
x:201 x:201
.

We are given that 0.25, so all then that remains is to calculate We calculate A.'.2o1

as follows:

This implies A'.2o1 = 0.28 - 0.25 x 0.4 = 0.18. Substituting, we obtain A,2oj = 0.18 + 0.25 = 0.43

d" l-0.4 3 1197

an a,2oj = 0.0511.05 = · ·

Finally,

a,2oj = ii,2oj -1+ 20 E,

LJ = a. \':2oj - 1+ A'
x:20i
= 11.97-1+0.25
= 1 l.22.
( END]

• 4. 4 mthly Life Annuities

Sometimes, a life annuity makes payments more frequently, say monthly or even weekly. For
this reason, we need to define mthly life annuities.

Before we get into mthly life annuities, let us have a quick review of mthly annuities certain,
which you learnt in Exam FM. As an example, we consider an mthly annuity certain due with a
fixed term of n years. This annuity makes payments at the beginning of each llm of a year. Each

":• Actex 2013 [ Johnny Li and Andrew Ng [ SoA Exam MLC

 Chapter 4: Life Annuities

payment is $1/m, and therefore the total payment over a period of one year is $1. The diagram
(
below illustrates the payments from such an annuity.

Total payment
in a year is $1 ) The last payment

$_!_ $_!_ $_!_ $_!_ $_!_

m m m m rn
Time from now
2 1 I
0 1-- n-- n
m m m m

We can generalize this to mthly life annuities due. For example, the APV of an mthly whole life

annuity due of $1 on (x) is denoted by a'."'>. By considering the APV of each possible payment,

we
. can calculate ax1"'> as follows:
1 1 _!_ 1 "-
ii(m)
x
=-+-vm p +-vm 2 p x + ···
- I x
m m -m m -m

We can define and ,,1a;"'> in a similar manner. Notice that the formula above contains

survival probabilities involving fractional ages, making it difficult to evaluate. It is unlikely that
you need to calculate the APV of an mthly life annuity with a summation-type formula. What is r
(
more important is the approximation of mthly life annuities, which will be discussed in Section
4.7.

Because each payment of an mthly life annuity is $ Ilm, we have the following adjustment
formulas for calculating mthly life annuities immediate:

(m) _ ··(111) 1 (m) _ "(m) 11 Ex

ax -ax - - , 111°x - 11iax ---
m m

er· Actex 2013 I Johnny Li and Andrew Ng i SoA Exam MLC

Chapter 4: Life Annuities

- 4. S Relating Different Policies

It is important to know how various life annuities are related to one another. You need to
remember two equations.

Equation 1

It is not difficult to see that a whole life annuity is just a combination of an n-year temporary life
annuity (which makes payments up to time n) and an n-year deferred life annuity (which will
make payments starting at time n). Hence, we have the following relation:

a=a-+
x x:nl
a. 111 .t

Equation 2

Suppose that you are now x years old and that you would like to start receiving life annuity
payments n years from now. You have two options:
The first option is to purchase at time 0 an n-year deferred life annuity for ,, ax amount of
1

money.
- The second option is to do nothing now, and then purchase a whole life annuity at time n (if
you survive to time n). The amount that you need to pay at time n will be a,.,,, because at

that time your age will be x + n. Therefore, at time 0, the expected present value of the cost

associated with this option is v" ,,pxax+,,.

Since the two options give you exactly the same payments, we have the following equation:

- II -
nlax =V ,,Px 0 x+n·

The logics behind the two equations are valid no matter if the life annuities are continuous,
discrete or mthly. Therefore, we have parallel equations for discrete and mthly life annuities.
The equations are summarized in the following table.

'"' Actex 2013 I Johnny Li and Andrew Ng I SoA Exam MLC

 Chapter 4: Life Annuities

?, Relations between Various Life Annuities

?,
?,
7, ?,
7,
?,
?,
?,
?, ?,
?, ?,
?,
%
?,
% %
?,
?, %
% %
%

Example 4.6

You are given:

(i) A30 = 0.6

(ii) = 0. J
r
(iii) = 0. 7

(iv) t5= 0.02

Calculate the following:

(a) a,o
(b) a,oiiij

(d) a,o

C' Actex 2013 I Johnny Lr and Andrew Ng I SoA Exam MLC

Chapter 4: Life Annuities

- Solution
(a) Using the Y-Z relation:
__ l-A30 _1-0.6_
G30 - - - 20 ·
0 0.02
(b) Using the Y-Z relation again:

_ 1-A,0,iQi
'°'"I 0
= l-(A;,iii1 +
0
1-(0.l+0.7)
=
0.02
= 10.

(c) Using Equation I developed in this section, we have

a'° =a,0 -a,0 iiii =20-10=10.

" 1

(d) Using Equation 2 developed in this section, we have

"a,,
1

I 0 - A30,;o1a 40
10 = o.1a, 0

a, 0 =14.2857.

[ END)

Example 4.7 [Course 3 Fall 2000 #3] ,

A person age 40 wins 10,000 in the actuarial lottery. Rather than receiving the money at once,
the winner is offered the actuarially equivalent option of receiving an annual payment of K (at
the beginning of each year) guaranteed for 10 years and continuing thereafter for life.
You are given:
(i) i = 0.04

(ii) A4o = 0.30

(iii) Aso = 0.35

t:' Actex 2013 I Johnny Li and Andrew Ng I SoA Exam MLC

 Chapter 4: Life Annuities

(iv) A;oiiir = 0.09

Calculate K.
(A) 538 (B) 541 (C) 545 (D) 548 (E) 551

- Solution

Our primary objective is to solve the following equation for K:

Kii-
40:101
= I 0000 .
We calculate i i - by decomposing it as follows:
40:101

ii-=
40:l0i
ii-+
101
ii
101 40 .

We have
1-VIO
iiiiij = -d- = 8.4353.

By using Equation 2 developed in this section, we have

We can calculate ii 50 as follows:

ii'°= 1-A,0 = 1-0.35 16 ·9 ·

d 0.0411 .04

All then that remains is the actuarial discount factor v 10 10p 40 , which can be calculated as follows:
r
A,o = A;oiiij + 101A<o (
10
=> A40 = A; 0 ,iiij + V 10P40Aso

=> 0.3 = 0.09 +0.35v'° 10 p 40

=> v'°10P<o = 0.6.

Substituting, we obtain i i - = 8.4353 + 0.6 x 16.9 = 18.5753. Hence,

40:101

K = 10000 538.35.
18.5753

[END]

\!':' Actex 2013 I Johnny Li and Andrew Ng I SoA Exam MLC

Chapter 4: Life Annuities

- 4. 6 Recursions

We also have recursions for life annuities. The recursion for a whole life annuity due with
annual payments is given by

The meaning behind this formula can be seen from the following diagram:

Time 0 Time I
(Age x) (Age x + 1)

Dead
qx No more payment

Alive
Continue to provide a
whole life annuity,
which has a value of
ii,+ 1 at time I
Pay $1 at time 0

The first payment of $1 at time 0 is certain.

- The policyholder could die during the first year (the interval from age x to age x + I) with
probability qx. In this case, no more payment will be made. The APV at time 0 in this case is
zero.
The policyholder could survive the first year with probability p,. At time I (i.e., age x + 1),
he/she would still have a whole life annuity with a value of iix+i. The APV at time 0 in this

case is vpx iix+I.

Using a similar reasoning, we can obtain the recursion for an n-year temporary annuity due:

ii-=l+vpii
x:nl
-.
x x+\:11-ll

t:. Actex 2013 I Johnny Li and Andrew Ng I SoA Exam MLC

 Chapter 4: Life Annuities

For mthly life annuities, each time step is l/m of a year and each payment is $1/m. We can
modify the above two recursions accordingly to obtain recursions for mthly life annuities. For
example, we have

I
··(m) 1 ; ·•(m)
ax =-+v iPx 0 I .
m -111 .r+-
m

?,
Recursions for Life Annuities ?,
% ?,
?, ?,
?,
?,
?,
?,
% ?,

• 4. 7 Relating Continuous, Discrete and mthly Life Annuities

l
You are required to know three methods for approximating continuous and mthly life annuities.
\

Method 1: Using the Y-Z relation and the UDD assumption

Using the Y-Z relation, we have a, = I-Ax . Analogously, we also have

0

"(m) I - Ax(m)
a.t
= d(m) '

where Jml = m[l - (I - d) 11'"].

Under UDO, we have Ax =!_A

6 x and A''") = -,i) A . Hence, we can approximate
x i"' x
ax and ii''") as
x

\C Actex 2013 Johnny Li and Andrew Ng I SoA Exam MLC

Chapter 4: Life Annuities

i !--'-· A
- {j x •(m) .r
( ) l
x {j
and ii"'=
x d(m)

respectively. This method is useful when we are given the value of Ax.

Method 2: Using the UDD assumption directly

Rearranging the previous equation (which is based on the UDO assumption), we obtain

ii;m) = a(m)iix - f3(m)'

id · -(m)
where a(m) = and f3(m) = . This formula is particularly useful when you know

the value of ii,. Note that you do not need to remember the formulas for a(m) and fl..m), as they

will be given in the exam. In particular, their values for i = 0.06 and various values of m will be
given in the exam. See the last page of the Illustrative Life Table.

- id i-0
When m oo, we have ax= a(oo)ax - fJ(oo), where a(oo) =Ji and a(oo) = y. The values

of a(ao) and fl..ao) for i = 0.06 will also be given in the exam.

For n-year deferred life annuities, we have the following formula:

Proof

= ,,E, (a(m)ii,., -f3(m))

= a(m),, ii, -f3(m),,E,.
1
D

By using the fact that ii;"') - ,A'."'), we obtain the following formula for n-year temporary
life annuities

a(m)iix,, 1
- f3(m)(I- ,,E,).

'\:; Actex 2013 I Johnny Li and Andrew Ng I SoA Exam MLC

 (c4-32)
------..1i...,'--"".,,..______________________________________________________________....i_
Chapter 4: Life Annuities

?r///,i7"//////////////,i7"//////////////,i7"/////////////h7"////////,//////,i7"///,//,///--%

% %
Approximating mthly Life Annuities Using the UDD Assumption

?,
?,
i ?,
?,
?,
?,
% %
% ?,
i

Example 4.8 [Course 3 Fall 2003 #9] ,,#>

For an annuity payable semiannually, you are given:
(i) Deaths are uniformly distributed over each year of age.
(ii) q69 = 0.03
(iii) i = 0.06

(iv) 1000 A.,0 = 530

Ca1cu 1ate a··(2)

69 . f-

(A) 8.35 (B) 8.47 (C) 8.59 (D) 8.72 (D) 8.85
- Solution - - - - - - - - - - - - - - - - - - - - - - - - -

Under UDO, we have

·-(>) -
a69 - a (2) a69
.. - /1(2) .

From the Tables for Exam MLC, we obtain a(2) = 1.00021 and /X_2) = 0.25739 when i = 0.06.
We need to find ii 69 from the information given.

Let us proceed from A.,0 • Under UDO,

\.' Actex 2013 I Johnny Li and Andrew Ng I SoA Exam MLC

Chapter 4: Life Annuities

which gives
= 0.535 = 0.53 ln(l .06)
A, 0 0.5147.
i 0.06
Using the Y-Z relation, we obtain

ii = 1-A,, = l-0.5147 =8.5735.

70
d 0.0611.06
What we need, however, is ii 69 instead. We can relate ii 69 and ii70 by using a recursion:

ii 69 = l + vp69 ii 10
1
= 1+-(l-0.03)(8.5736)=8.8455.
1.06
Substituting the value of ii
69 back into the first equation, we have

iili) = l .00021x8.8455-0.25739 = 8.5900.

[ END)

Method 3: Using Woolhouse's formula

You can approximate mthly life annuities by using Woolhouse's formula, which does not need
the assumption of UDD.

Woolhouse's formula (with three terms) for whole life annuities is given by

2
.. (m) _ .. m- l m - l (< )
ax - a' - 2m - l 2m' u + µx .

Note:
- Rather than the full formula, you may be asked in the exam to use Woolhouse's formula
with two terms only, that is,

This shorter formula, of course, will give a less accurate approximation.

- Ifµ, is not given, you should estimate it by using the following formula:

re Actex 2013 I Johnny Li and Andrew Ng I SoA Exam MLC

 Chapter 4: Life Annuities

Woolhouse's formula is also available for n-year temporary and n-year deferred life annuities.
The formulas are summarized in the following table.

Woolhouse's Formula
?,
?,
?, ?,
?,
?, %
?,
%
?,
?,

These formulas are not easy to memorize. We suggest that you only remember the formula for
whole life, because the other two can be derived easily by using the relations in Section 4.5.

Example 4.9 [Sample #284]

John approximates values of a:;l using Woolhouse's formula with three terms. His results are: I
I
= 8.29340 and ai:) = 8.16715.
Calculate using Woolhouse's formula with three terms and using the same mortality and

interest rate assumptions as John.

(A) 8.12525 (B) 8.10415 (C) 8.08345 (D) 8.06275 (E) 8.04135

- Solution

Woolhouse's formula with three terms is given by

.. Im) _ .. m -1 m' -1 (,. )

Ggo - 0 so - - - - - - 2 u +µso ·
2m 12m

f;c Actex 2013 I Johnny Li and Andrew Ng I SoA Exam MLC

Chapter 4: Life Annuities

Setting rn = 2, we have
··(2J -- 8 . 29340 --Ggo
.. 3
I _ _ (u+µso
< )
Gso __ .
4 48
Setting rn = 4, we have
3 15 c< )
0··(•J-816715-""
ao - · - 0 so _ _ _ _ u+µso ·
8 192
Subtracting a::i from we have

I 3 (-
8.29340-8.16715=--+-- 3 -15)
- (o+µ.)
4 8 48 192 ° ,
which gives i5 + µgo= 0.08.

Substituting rn = 2, = 8.29340 and o+µgo= 0.08 into the formula, we obtain

ii80 +2- (0.08) = 8.5484.
4 48
Finally, substituting rn = 12, o+ µgo= 0.08, and ii80 = 8.5484 into the formula, we obtain

ar8012 i =ii80 )
24 I 728 !lso
143
= 8.54840_.!_.!__ (0.08)
24 1728
= 8.08345.
Hence, the answer is (C).
[ END]

• 4. 8 Useful Shortcuts

Constant Force of Mortality

1
If the force of mortality is constant beyond age x, then a, = - -.
µ+o
We suggest that you remember the formula for a, only. Formulas for other APVs can be

derived straightforwardly without using integration. For example,

{:, Actex 2013 [ Johnny Li and Andrew Ng I SoA Exam MLC

 Chapter 4: Life Annuities

1 e-(6+µ)n
{j =V 11 p {j =e-One-µn _ _ = - - -
nl x n;rx+n µ +J µ +J

a- .-1 -
- -ax - nla.r
- -- _1_(1- e -(J+p)u) .
:r.11 µ+O

De Moivre's Law
(Da)- _
While it is not hard to show that ax = w-xl , it is not easy to compute (Da) w-xl and it is
1))-X

easier to calculate annuities using insurances.

I
The Illustrative Life Table (

(
The Illustrative Life Table gives values of iix. Below is a portion of the table, with that column

included. Note that the table assumes an interest rate of 6%.

x I, IOOOq, ax IOOOA, 1oooc2Axl

65 7,533,964 21.32 9.8969 439.80 236.03
66 7,373,338 23.29 9.6362 454.56 249.20
67 7,201,635 25.44 9.3726 469.47 262.83
68 7,018,432 27.79 9.1066 484.53 276.92

However, the APVs for temporary and deferred annuities are not given. We can compute them
quickly by using the relations introduced in Section 4.5. For example, (
.. 3 ..
3Ja6s =V 3PxG6s i--
I / ..
- - -3-68a68
- 1.06 / 65

=-1-x 7018432x 9 . 1066

1.063 7533964
= 7.1229.

Also,

t' Actex 2013 I Johnny Li and Andrew Ng I SoA Exam MLC