1'11.

1' Ihtrnduation,
llli.2 The. MBtHematias of Carn~~undIht&x!st'

1 1.5.41 Insuranoe. P@able& : t h hrlbmmtrafDeath

1 1 . Analysis and>Valuation-ofhuities,Eoans3 Bonds and4OtHel-r-Securities

1g1.8 1ntrodtlation:to 9tbdiasti~Interest b t a Mbdkls andiktuarial~Appllcatihns.

1 1 9 Let Us-S-umTiTp~

1'1112 h s w or: Hints-t b ) C h d , H o u l . B f a ~ s

Ih.l i k i n s m m anntmnt,. the. h a f i t [of i n d i msistb Q$ a+sing&~qynent:and;

I tlia sum,a s 4 1 'liHe:tirnwand;ttla m r m n t ~ a f l ~ m t ~ a r a . f i d i m ~ a f ' ~ e ~ m d b m ~
variattlk T (or,. F[cc$)~
wlie;le. T is, tha. futw life t i m e of) pmanS of?age xi l?hw
 TePhniques-ri present vdue of the payment is denoted by Z. The expected value of the payment
E(Z) is the net single premium of the contract. The premium does not reflect the
risk to be caqied-by the insurer. In order to assess this, one requires the different
characteristics o f 2. I

We are introducing the models of stoch~sticinterest rate and some simple uses in
actuarial applications. I

11.2 THE MATHEMATICS OF COMPOUND

11.2.1 EffectivecInterestRates
. An interest rate is always related with a basic time unit like an annual interest rate
I
of 5%. With the interest rate, the time interval at the end of which interest rate is
credited or "Compounded" has to be stated, which is known as conversion period.
An interest rate is called effective if the basic time unit; and the conversion period
are identical. 1n this case interest.is credited at the end df basic time unit.
Let i be the effective interest rate and let Fobe the ilitial capital. Let rk be the
additional amount invested at the end of year k , k = 1,2,. . .,n.
We want to find out the balance at the end of n years.
Suppose Fkis the balance at the end of year k includin rk, interest credited to the
previous year's balance is i Fk-,.
So,.
,
Fk = Fk-,+iFk-, +rk,k= 1,2,..., n (1 22.1)
which implies ;

~ u l t i ~ l ~ i n g - ( l 1 . 2by ) + i r - k and summing over all values of k we obtain,

. 2 (1
n n-k

Fn = (l+i)^ F,+ (l+i) rk

The powers of (1 + i) are called accumulation factors. The accumulated value of

an initial capital C after n years is (l+i)" C. Equation (1 1.2.3) gives the capital at
the end of interval which is the surn of value of initial capital plus the surn of
accumulated~valuesof the intermediate deposits.
1
The discount factor is v = -
1+ i
n
. Equation (1 1.2.3) can be written as vn Fn= Fo+ z v k rk
k = l

Hence the present value of a capital C due at time h is vhc . '

From equation(l1.2.1), we have, Fk- Fk-,= i 4-I+ rk

 C : i h Flow Projection
Now summing over k we get Fn - F, = 2 ( i F' - ,+ r k )
i.e., the increment of the find is the sum of the total interest credited and thqtotal
deposits made.

11.3 NOMINAL INTEREST RATES

If the conversion period does not coincide with the basic time unit, then the
interest rate is called nominal.
An interest rate of 6% with conversion period 3 months means that interest of
(y) = I - 5% is credited at the end of each quarter. Thus an initial capital of 1

increases to (1 - 015)4 = 1 - 06136 at the end of one year. Therefore an annual

nominal interest rate of 6% is equivalent to an annual effective interest rate of
6..136% .
11.3.1 Relation between Effective Interest Rate and Nominal
Interest Rate
Let i denotes a given effective interest rate. We define ilm)as the nominal interest
rateconvertible m times per year which is equivalent to i.
Equality of accumulation factors for 1 year leads to the following relationship

The limiting case correspond to continuous compounding.

Let

It is called the force of interest equivalent to i.

So fiom equation (1 1.3.2) we have

i.e., 6 is a derivative of the function (1 + i)" at x = 0 .

so,
f'ctaarin' Tec'lniqaes-k' ~ r n mthis we conclude that the accumulntion factor for a period of h years is
(1 + ir = e" and discount. fmtor for the same period*is v h = e-& where h, the
length of the period is any real number.
Exarnple:lA d~posit;afllK)0;000'is made into a newly estabijshed fund. The fund
pays nominal interestpof; 12% canvmtibie quartorip. At tha and of each six
months. a1 withdrawal is- mada h m tHe hndi The. first withdiawl is- X, the
second is 2X, tha.thirxli~3Xandlso on. TNe lnst is sipcah withdrawal whidi exmly
exhausts tHe fund: C a l d h t e X.
Solution: Taking the notation of the taxt to hale yeam we get F, = 100,000 and
f( = ~ , ( 1 . 0 3 y- X and so on. This can be taken using (11.1.2). The 6th
withdrawall exactly exhausts the h d . So F, = 0 . Solving this we get X = 6128.

One of the t~chniquesis give* by the equation (1 1.2.4) where the interest is
credited at the endi of ea& oanwsrsian period; in arrears. The ather technique. is
given below, where the interest1 is credited at the beginning of the conversion
period. Interest credited in this way is also referr~dto as discount and. the
correspondingrate is called discount rate or rate of intarest in advance. Let d be
an annual effective discount rate. A person investing an amount of C will' be
credited interest equal to dG 4mmediately and the invested aqital C will be
returned1at the end of the peiiod, Ihvesting the interest dC at the same conditions,
the investor will receive additional intemst of d ( d ~=) d 2 C , and. the additional
invested amount wills be retumedlat the end opthe year; reinvesting, the intemst
yields additional interest~ofd ( d 2 c ) =d 3 C and sa on.
Repeating this prooms infinitely we. find tha investor will receive the total sum of

at the end of'the year in wiu.m,fbrinvesting the initial aapital C. The equiwdant
effective intbms4 rate i is given by. tHe equation.

which im~lies.

This result has the follb-wingint~mtatim.If 11unit of capital: is invested, d (the

interst payable1 at!the b g n n i n g o f t i i y~ e ) is. the~disco~nted~value
of the interst;;i
to be paid at the end:ofyearear

Thus the iritmeit: pystttlh- a t t the1 end of the year is the accumdhted valhe of
interest payable; atithe h@nnihg:ofi the yean
Actuaria'Techniques-ll which is the net single premium calculated at twice the origins1 force of interest.
Thus calcullating the variance 'is no more difficult than calculating the net single
premium.
An insurance which provides for payment only if death occurs within n years is
known as term insurance of duration n. For example: 1 unit is payable only if
death occurs during the first n years, the actual time of payment still being the end
of the year of death. So
vK+'for K = 0 , 1 , 2 , ..., n-1

0 for K = n , n + 1, n + 2 , ...
The net single premium is denoted by

and
E ( Z ~ ) = E [ 2L6 ( K + I ) ]
11.5.2 Pure Endowments
A pure endowment of duration n provides for payment of the sum insured.only if
the insured is alive at the end of n years,
0 for K = 0 , 1 , ...,n-1
( 1 1.5.9)
vn for K =n; n + l , n + 2 , ...

The net premium is denoted by A , and is given by

x,
;;i

A I =vnn~x (1 1.5.10)
+,
;;i

The formula for the varianck of a Bernoulli random variable gives . >

11.5.3 Endowments
Assume that the sum insured is payable at the end of the year of death, if this
,,ccurs within the first n years, otherwise at the end of the nth year
VK +I
for K = O,l,2,...,n-1

vn for K = n , n + 1, n + 2 , .,.

The net single premium is denoted by A I J . Denoting the present value of ( 1 1 S.6)
by ZI and (1 1.5.9) by Z2, we have
 . ,

Cash Flow Projection

and'
v(z)=v(z,)+v(z2)+2cov(z,,z,) ,

The product Z I . Z2 is always zero, hence

C o v ( Z , , 2,) = E ( z , , z , ) - ~ ( z , ) , E ( Z 2 )

As a consequence of the last identity the risk in selling an endowment policy,

measured by the variance is less than that in selling a term insurance to one
person and a pure endowment to another.
Let us finally consider an m year deferred whole life insurance. Its present value
is
' o fo> K = o , ~...,
, m-I

vK+' for K = m,m+l,m +2,...

The net single premium is denoted bym,Ax.Alternative formula for its net single
premium are

rn1A.r = n, ~ x ~ ~ ' A x + r n
and

m,AI= A x -

The second moment E(z,)again equals the net single premium at twice the
original force of interest.
11.5.4 Insurance Payable at the Moment of Death
In previous sections we have discussed that the sum insured was payable at the
en'd of the year of death; This assumption does not reflect insurance practice in a
realistic way, but has the advantage that the formulae may be evaluated directly
from a life table.
Let us now assume that the sum insured is payable at the instant of death. i.e. at
time T. The present value of a payment of payable at the instant of death is at time
T. .'The present value of a payment of 1.payable immediately on death is
The net single premium is denoted by A,. Using the :formula from force of
mortality

The force of mortality at age x is defined as the ratio of instantaneous rate of

1 d Id
decrease in 1, to the value of I, . is . px = - - - 1, = - - (log 1,). I,, is the
I , dx dx
number of person's living at age x .
One practical approximation of T is given by

where K is the number of complete future years lived by (x) and S is the fraction
of a year during which ( x ) is alive in the year of death.
Making use of assumed independence of K and S, as well as the uniform
distribution of S, so that

we find

Thus calculation of Ax is a simple extension ofA,.

A similar formula may be derived for ten& insurance. For endowments the factors
i
- is only used in the term insurance part.
6

Let us finally assume that the sum insured is payable at the end of the mth part of
the year in which death occurs i.e, time K + $m). The present value of a whole life
insurance of 1 unit then becomes

and K + s(")= (K + 1) - (I - sI"))

so,

and we obtain
 Cash Flow Pro.jection

Example: 2,is the present value random variable for an n-year continuous
endowment insurance of 1 issued to (x).2, is the present value random variable
for an n year continuous term insurance of I issued to (x).Given
Var (2,)= 0.01, v n = 0.3, p, = 0.8,E (2,)= 0.04 ,calculate Var (2,).
Solution: Let Z3be the present value random variable for the pure endowment
Z,= Z, + Z,
we want to find out ~ar(2,).
Hut var(2,)= V U ~ ( Z+~ 2,)
= Var (2,)+ 2 c o v ( ~ , 2,)
, + ~ar(&).

v n . 1 with probability 1

0 otherwise
Now
E(&) = v n ,,P,=(0.3)(0-8)=(0-24)

var(z3) = v2" 2,(1-,~,)

= (0.3)~(0-8) (0 2)

Check Your Progr-ess I

1) What is the difference between effective interest rate and nominal interest
rate? Find the relation between them.

...........................................................................................
-)
?
For m = 1,2.3,4,6,12,and oo calculate i(") for i = 6%.
Actuarial Techniques-]]
3) Z1 is the present value random variable for an n-year continuous
endowment insurance of 1 issued to (x). Z2 is the present value random
variable for an n-year continuous term insurance of 1 issued to (xj. Given
values Var(Z2) , px,E(Z2) obtain a formula to find out Var(Z1).

11.6 ANALYSIS AND VALUATION OF ANNUITIES,

BONDS, LOANS AND OTHER SECURITIES
To have a clear idea on annuity and its analysis we introduce certain types of
perpetual streams known as perpetuities. The resulting formulae are very simple
and will later be useful for calculating the present value of annuities with a finite
term.
11.6.1 Perpetuities
First we consider perpetuities consisting of annual bayments of 1 unit. If the first
payment occurs at time zero, the perpetuity is called perpetuity-due and its present
value is denoted by a4 . Thus

If the first payment is made at the end of yea? 1, we call the perpetuity an
immediate perpetuity. Its present value is denoted by a4 and is given by
el
1

i
1
Let us now consider perpetuity where paymedts of - are made rn times each
rn
1
year. If the payments are made in advance (first payment of - at time O), the
rn
present value is denoted by ii$) and is

1
If the payments ire made in arrears (first payment of - at time O), the present
m 1

value is denoted by a$') and given by

I
 Cash Flow Projection
!

J
j L
\
The results in (1 1.6.3) and (1 1.6.4) lead to an interpretation of the identity
(1 1.4.7). Since a perpetuity-due and an immediate perpetuity differ only by a
1 1
payment of - at time 0, their present value differ by - .
m m
Let us now consider a continuous perpetuity with constant rate of payment r = 1
and starting at time 0. Its present value is denoted by a4 and given by

Letting m + oo in (1 1.6.4) or (1 1.6.3) we will get the same result.

The systematic pattern in formulae (1 1.6.1) to (1 1.6.5) is evident.
A certain type of perpetuities with increasing payments is defined by two
parameters m and q. i.e. the number of payments per year and the number of
increase per year. We assume that q is a factor of m. if m = 12, and q = 4, it means
that payments are made monthly and increase quarterly. In general, the payments
of such as increasing perpetuity due are defined as follow.

Time Payment

and so on

m k
In particular, the last - payments of year k are - each.
4 m

z ) ~. We can
(m)
Let us denote the present value of such a perpetuity by ( I ( ° )
calculate it by representing the' sequence of increasing payments as sum of
1
perpetuities ~ t constants
h payments of - payable m times per year and
(mq)
1 2
beginning at times 0,-, -, .... Thus we obtain the follawing formula:
4 4
 ( 1
( I y ) = ,6;) '
, L
l + v Y+ v 4 + . . .

The corresponding immediate annuity differs only in that each payment is made
one mth year'later, thus giving

A superscript 1 is always omitted. For instance, ihe present value of a perpetuity

due with annual payments of 1,2,.. . is

( I )( I )a =
(,)(I)
-1
d2
. (11.6.8)

For continuous payment streams

without actually calculating the integrals.

But when the payments are arbitrary with annual payments ro , r,, r, ,.. . (at times
0,1,2,.. .), its present value is denoted by a .
a = r O+vr, + v 2 r2 (1 1.6.1 1)
Such a variable perpetuity may be represented as a sum of constant perpetuities in
the following way.

Annual Payments Starts at time

ro 0

5 - ro 1

r2 - 'i 2

and so an

The present value'of this perpetuity may therefore be expressed as

 Cash Flow Projection
We can use (1 1.6.1 I) to calculate the present value of exponentially growing
payments with r, = erk for k = 0,1,2, ... and obtain (1 1.6.13)

provided r < S .
Annuities
In practice annuities are more frequently used than perpetuities. An annuity is
defined as a sequence of payments of a limited duration, which we denote by n.
Here we consider some standard type of annuities or annuities-certain.
The present value of an annuity due with n-annual payments of 1 starting at time
0, is denoted by a d . It is given by

Representing, the annuity as the difference of two perpetuities (one starting at

time 0, the other at time n), we find that

'This result can be verified by evaluating the geometric sum ( 1 1.6.15). Similarly
from ( 1 1.6.2), (1 1.6.3) and (1 1.6.4) the following formulae can be obtained

Note that only the denominator varies, depending on the payment made
(immediate 1 due) and frequency. Note that n must be an integer in (1 1.6.16) and
1
(1 1.6.17). and a multiple of - in (1 1.6.18) and (1 1.6.19).
m

The final or accumulated value of annuities is also of interest. This is defined as

the acculnulated value of the payment stream at time n, and the usual symbol used
is S. The final value is obtained by multiplying the initial value with the
accumulator factor
Actuarial Techniques41

Another simple relation between the initial value and final value of a constant
annuity may be verified.

Let us now consider an increasing annuity due with parameter q and m.

Time Payment

Such an increasing annuity can be represented as an increasing perpetuity starting

at time 0, minus an identical increasing perpetuity starting of time n, minus a
constant annuity starting at n.

Substituting (1 1.6.6) and (1 1.6.3) and using (1 1.6.19) we obtain the equation

Similarly the present value of the corresponding immediate annuity is calculated

as I
In these equations n must be a multiple of l/q. Special cases are the combinations Cash Flow Projection
.
o f m = 1 a n d q = 1 , m = 1 2 a n d q = 12,rn=ooandq= l , a n d m = o o a n d q = m .
The annuities just considered are known as standard increasing annuities Q.
Standard decreasing annuities (D) are similar, but the payments are made in the
reversed order. The sum of these two annuities is a constant annuity. This relation
carries over to the present values and we obtain

Using (1 1.6.26) and (1 1.6.27)

we get

The standard decreasing annuity due may be interpreted as a constant perpetuity

with mth ly payments of d m , minus a series of deferred perpetuities-due, each
1 2
with constant mrhly payments of and starting at times -, -, ..., n.
4 4
Loans, Bonds and Other Securities
Let S be the value at 0 of a loan that is to be repaid by payments r,, rz,.. .,r,,,
made at the end of years k = 1,2, ...,n. Then S must be the present value of the
payments.

Let S k be the principal outstanding i.e., the remaining debt immediately after r k
has been paid. It consists of the previous years debt, accumulated for one year
minus Q.

It consists of two components, one is interest on the running debt and reduction of
principal. Substituting -Skfor Fk(1 1.6.32) is equivalent to (1 1.2.2). From (1 1.2.3)

From this one can get S,, = 0 and Sk= v rk+,+ v2rk+,+ ...+ ~ ~ - ~ (1 1.6.35)
r ~
Formula (1 1.6.34) is the retrospective formula and (1 1.6.35) is the prospective
formula for the outstanding principal.
The payments rl, r2, . .., r k may be chosen arbitrarily, subject to the constraint
(1 1.6.3 1). By proper choice of payment steam the following formula of annuities
can be obtained. For example, S = 1 can be repaid by the payments,
I Actuarial Techniques-I1
The payments rl, r*, ..., rk may be chosen arbitrarily, subject to the constraint
(1 1.6.31). By proper choice of payment steam the following formula of annuities
can be obtained. For example, S = 1 can be repaid by the payments,

In this case only interest is paid for the first (n - 1) years and the entire 'debt
together with the last year's interest, is paid of the end of the nth year. From
(1 1.6.31) l=iaJ +vn (1 1.6.37)
The debt S = 1 may also be repaid by constant payments of
.

As an alternative to repaying the creditor at times 1,2,.. ..n-1, one could pay only
the interest on S as in (1 1.6.36). In order to cover the final repayment one can
make equal deposits to a fund that is to accumulate to 1 of the end of n years.
1
from this it is obvious that the annual deposit must be -. Since the total annual
sn
outgo must be the same in both the cases and we arrive at

Suppose now that we repay a debt of S = n, so that the principal outstanding

decreases linearly to D, S k = n - k for k = 0, 1,2,. . .,n. From (1 1.6.33) it is evident
thatr, = i ( n - k + > 1) + l.Using(11.6.31)

I'his result is a special case (m =q = I ) of (1 1.6.30)

The loan itself may consist of a series of payments. Assume that equal payments
of 1 are received by the debtor at times 0,1,2,. ..,n-1. At the end of each year
interest on the received amounts is paid and the total amount received is repaid at
time n .
r, = ik for k = 1,2,...,n-1, r, = i n +n (1 1.6.4 1)
From the equality of present values

For security, suppose an investor pays a price p which entitles him to n future
payments. The payments are denoted by rl, r2, . .,, r,, and payment rk is due at time
Tk, for k = 1,2,.. .,n. What is the resulting rate of return?
The present value of the payment stream to be received by the investor is a
function of the force of interest 6 . Define
Let I be the solution of the equation Cash Flow Proiection

P (1 1.6.44)
The internal rate of return or investment yield is defined as i = e' - 1 .
Equation (1 1.6.44) can be solved by standard numerical methods such as interval
bisection or the Newton-Raphson method. We shall present a more efficient and
simpler method than other methods.
Consider

here r = r, + ... + r, denotes the undiscounted sum of payments. It is not difficult

to verify that

The last inequality may be verified by interpreting f W ( 6 ) as a variance.

Interpreting fo
Ci
as the slope of the secant and noting that f is a convex function

by (11.6.46) we see that -

f ( 6 ) is an increasing function of 6 . Hence for
6

giving -.f ( [ ) ,y <, <-.f ( t ) U

f (4 f (4
Thus we have proved that

ln (" ) s <l<-
ln (P;) U
(1 1.6.49)
a[+r ' ) I(. " ( r u )~

The process may be iterated yielding the following algorithm. Starting with do,
calculate

recursively.

21
Actuarial Techniques-11
For 6, < t , the resulting sequence will be monotonically increasing and will
converge to t. For 6, > t , the sequence will decrease monotonically to t. Thus any
arbitrary positive value may be chosen for 6,.
Example: 'Security = Rs.50001- (Face amount) yearly coupons = Rs.3001-.
Assume that the security has been bought for Rs.5250 for a remaking running
time of 9 years. Thus we have

r = 7700.
Assuming we know that the current yield of similar securities lies between 5%
and 5.5% we have s = In 1.05, u = In 1- 055. The bounds for the investment
yield is

and

The algorithm defined by (1 1.6.50) may be used to obtain greater precision.

11.7 YIELD CURVES

There are ,many classes of bonds in financial markets. For example, we can
restrict ohrselves to various risk classes, such as Treasury bonds, municipal
bonds, corporate bonds or junk bonds. Given the class of Treasury bonds, we see
Treasury bills that mature within a year, Treasury notes with maturities between
one and five years and Treasury bonds with maturities extending to 30years.
If risk-fee instantaneous interest rates were constant at r, the time t price of a pure
discount bond with maturity u would be given by
i

~ ( ut ), = 100e-~("-') (1 1.7.1)

where BC) denotes the present value of 100 discounted at a rate r. A pure
discount bond does not make any coupon payment, does not contain any implicit
option and has a par value of 100. The function e-r(u-I) at time t plays the role of a
discount factor. At time

the bond matures and the exponential function equals 1. At all t < u, it is less
than 1.
The bond price formula depends on the whole spectrum of future short rates.
Hence the yield curve at time t contains all the available information concerning
future short rates and the bond prices depend on the whole yield curve or on the
term structure of interest rates, which we define below.
 Cash Flow Projection
Definition: Assume that at time t there exists zero coupon bonds with a spectrum of
. their price be ~ ( u , t )and their yield be given by P;'. Then
maturitiesu E [ t , ~ ] Let
the spectrum of yields (R,",u E [t, T]} is called the term structure of interest rates.

Here the yield R," is defined as the number that satisfies the equality
B ( ~ ,=~1)OOe-4"(~t-t)
,t <U (1 1.7.3) ,

log B .(u, t) - log 100 (1 1.7.4)

or R~ =- -

t-u
Examples of yield curves
In practice, there are two ways of proceedings. Often a market participant
assumes a functional form for the yield curve and from there obtains the implied
forward rates. A second method is to go the opposite direction. One can assume a
dynamic behavior for the forward rates and from there obtain a yield curve. In
most common case one assumes that the yields R(-) depends on one variable rt
which represents the short rate observed now, and they are given by the functional
form.

Functions ~ ( u , t and
) c(u,t) can be constructed in various ways, so that the yield
curve can be upword, downward sloping or hump shaped.

11.8 INTRODUCTION TO STOCHASTIC INTEREST

RATE MODELS AND ACTUARIAL
APPLICATIONS
The interest rate that will apply in future years is of course not known. Thus it
seems reasonable to ask why future interest rates have not been modelled as
stochastic process. Two reasons are there: (1) Life insuranc'e is particularly
concerned with the long term development of interest rates and no commonly
employed stochastic models exist for making long term predictions. (2) A
reasonable assumption is that the remaining life times of the insured lives are
essentially independent random variables. The probability distribution of
aggregate loss can be obtained by convolution. The variance of the aggregate loss
is the sum of the individual variances, which facilitates the use of normal
approximation. Stochastic independence between policies would be lost with the
introduction of stochastic interest rate, since all policies are affected by the same
interest development.
The mathematics of derivative assets assume that time passes continuously. As a
result, new information is revealed continuously and decision makers may face
instantaneous changes in randomness. Hence technical tools for pricing derivative .
products require ways of handling random variables over infinitesimal time
intervals. To have a clear understanding on stochastic interest rate, we must have
clear concept on continuous payments.
Continuous Payments
Let us consider that payments are made continuously with an annual
instantaneous rate of payment of r(t). Thus the amount deposited to the fund
during the infinitesimal time interval from t + dt is r(t) dt. Let F(t) denote the
balance of the fund at time t. We assume that interest is credited continuously,
according to a time-dependent force of interest ~ ( t )Interest . credited in the
infinitesimal time interval from t to t + dt is F(t) 6 ( t ) dt.
The total increase in the capital during this interval is thus

To solve the corresponding differential equation

we write

Integration with respect to t from 0 to h gives

Thus the value at time 0 of a payment to be made at time t (i.e., its present value)
-I
I

6(s) d.s
is obtained by multiplication with the factor e "
From (1 1.8.4) we obtain

F ( h ) = eo
I
Thus the value at time h of a payment made at time t < h (its accumulated value),
is obtained by multiplication with the factor
Is
h
(s) dr
eO

In the case of a constant force of interest i.e., ~ ( t= )6 , the factors e " and
h
IS(.^) dr

eO are reduced to the discount factors and accumulation factors.

Now we consider equation (1 1.7.1) where B(.) is the present value of 100
discounted at a rate 'r'. When interest rates become stochastic, the formula
(11.7.1) would change. Rt would represent the risk free rate earned
instantaneously, and to obtain the price of a discount bond with par value equal to
100, we would have to use
 First of all, we need to use the conditional expectations operator E[.J.], since r, Cash Flow Proiection
represents the instantaneous rate at some future time s > t and consequently is not
known with certainty at time t, it can only be predicted.
!
- Secondly, there is the issue of which probability measure is used to calculate these
expectations. One can argue that Treasury bonds are risk - free assets and that
there is no need to use the equivalent martingale measure. This would not be
correct. As interest rates become stochastic, prices of Treasury bonds with larger
maturities are subject to more shocks and, everything else being the same, longer
I
maturities are riskier. In order to eliminate the corresponding risk premia, we need
i to use equivalent martingale measure in evaluating the expressions shown in
a (1 1.8.6).
Third, note a heuristic explanation for where the integral in (1 1.8.6) comes from.
Consider the case of a 3 year bond. In discrete time frame-work the price of a 3-
period discount bond will be given by

where rl is the current short rate, r2 is the unknown short rate during year 2 and r3
is the (unknown) short rate during year 3. According to equation (1 1.8.7) the
bonds price is equal to the expectation under risk neutral probabilities of the
discounted value-of the par value. The discount factor is obtained by multiplying
the one period discounts during the present and the two future years. Since future
discounts are unknown, an expectation operator needs to be used.
This example may be more useful in interpreting the cbntinuous-time bond price
given in (1 1.8.6). As time becomes continuous, discrete-time discount factors
need to be replaced by the exponential function. But because interest rates are
continuously changing an integral has to be used in the exponent.
,
Check Your Progress 2
1) Define annuity. Represent annuity as the difference of two perpetuities
one at time 0 and the other at time n. Show that

Prove that in repayment of debt, each payment consists of two

components, one is interest on running debt and the other is reduction in
principal.
Actuarial Techniques-11 Discuss the method of continuous payments. Give an expression for yield.
3)

11.9 LET US SUM UP

This unit discusses the Cash-Flows like interest, annuities, bonds, loans and other
payments which arise due to insurance contracts. An important theme covered is
the projection of interest in deterministic and stochastic framework. While dealing
with the deterministic component, we have included the derivation of effective
and nominal interest rates. These rates are extended to include 'the techniques of
discounting to evaluate their present values. The general Cash-Flows in insurance
is discussed taking various life insurance forms such as whole life and pure
endowments.

11.10 KEY WORDS -

.

Annuity: A sequence of payments of a limited duration.

Complete Expectation of Life or Complete Future Life Time: The average
number of years a person of given age can be expected to live under the
prevailing mortality conditions. It gives the number of years of life entirely
completed and includes the fraction of the year survived in the year in which
death occurs which on the average is taken as '/z year. So complete future life time
= curtate future life time + !h .

Conversion Period: The time interval at the end of which interest is credited or
compounded. .
Curtate Future Life Time: The average number of completed future years lived
by the group attaining age x.
Discount Rate: When interest is credited of the beginning of conversion period. '

Effective Interest Rate: When the conversion period and basic time unit are
~dcnricalinterest rate is credited at the end of basic time unit.
. I".

I[.ife Table: A conventional method of expressing the most essential facts about
the age distribution of mortality in a tabular form and it gives the life history of a
hypothetical population which is gradually diminished by deaths. It is a powerful
tool for measuring the probability of life and death at various age sectors.
Net Single Premium: The expected value of present value of the payment.
Nominal Interear Qgte: When the conversion period does not coincide with basic
time unit.
Perpetuity: A sequence of payments of an unlimited duration.

11.1 1 SOME USEFUL BOOKS

-

Batten,R.W.(1978) Mortality Table Construction. Prentice-Hall Englewood diffs,

New Jersey
I
Cash Flow Projection
Butcher, M.V., Nesbitt C.J. (1971) Mathematics of Compound Interest. Ulrich's
I
1
Books, Ann Arbor, Michigan
t Jordan, C.W. (1967) Life Contingencies, Second Edition, Society of Actuaries
,I Chicago, I11
1
!
Salih, N. Nefici (1996), An Introduction to the Mathematics of Financial
Derivatives, Academic Press. San Diego-California, USA

'11.12 ANSWER OR HINTS TO CHECK YOUR

PROGRESS
Check Your Progress 1
1) Read 11.2 and 11.3 and answer.
2) Use equation 1 1.3.4 and tabulate the result.
3) Read 11.5 with the example and answer
Check Your Progress 2
1) Read 1 1.6 and use equation and 11.6.15 and 11.6.16 and answer.
2) Read 11.6 and use equation 11.6.32 and 11.6.33 and answer.
3) Read 11.8 and answer.

11.13 EXERCISES

2) A deposit of 100,000 is made into a newly established fund. The fund pays
nominal interst of 12% convertible quarterly. At the end of each six
months a withdrawal is made from the fund. The first withdrawal is X, the
second is 2X, the 3rd is 3X and so on. The last is sixth withdrawal, which
exactly exhausts the fund, Calculate X.
3) Given: (i) A a = 4 , (ii) A : : ~= y ; (iii) A,,. = z . Determine the value of
Ax in terms of u, y and x.
4) Z I is the present value random variable for an n year continuous
endowment insurance of 1 issued to x. Zz is the present value random
variable for an n year continuous term insurance of 1 issued to x. Given
(i) ~ar(~,)=0.01, (ii) v n = 0 - 30 .

(iii) ,,p, = 0.8, (iv) E(z,) = 0.04.

Calculate var(Z,).
5) A loan of 1000 at a nominal rate of 12% convertible monthly is to be
repaid by six monthly payments with the first payment due at the end of
one month. The first three payments are x each, and the final three
payments are 3x each, calculate x.