MBF 839

NATIONAL OPEN UNIVERSITY OF NIGERIA
SCHOOL OF MANAGEMENT SCIENCES
COURSE CODE: MBF 839
COURSE TITLE: QUANTITATIVE TECHNIQUES FOR BANKING

& FINANCE
Course Code MBF 839
Course Title Quantitative Techniques for Banking
and Finance
Course Developer/Writer Mr. E. U. Abianga

National Open University of Nigeria
Lagos.
Course Editor Mr. S.O. Israel-Cookey

Lagos.
Course Co-Ordinator Mr. E. U. Abianga

Lagos.
Programme Leader Dr. 0.J Onwe

Lagos.
NATIONAL OPEN UNIVERSITY OF NIGERIA

MB!' 839 QUANTITATIVE TECHNIQUES FOR BANKING AND FINANCE

Headquarters
14/16 Ahmadu Bello Way
Victoria Island
Lagos
Abuja Office
No. 5 Dar es Salaam Street
Off Aminu Kano Crescent
Wuse II, Abuja
Nigeria
e-mail: centralinfori4nou.edu.ng
URL: www.nou.edu.ng
Published by:
National Open University of Nigeria 2008
First Printed 2008
ISBN: 978-058-365-3
All Rights Reserved
Printed in Nigeria by:

Dail y Graphi cs Ni geri a Ltd.,
Tel: 0803-3108965, 02-7522693, 0804-4104608,
0803-3449198, 0805-2281869,
E-mail: dailygrafix@yahoo.co.uk
For:
National Open University 0 r Nigeria.
II
MBF 839 QUANTITATIVE TECHNIQUES FOR BANKING AND FINANCE
CONTENTS PAGE
MODULE! ......................................................................... 1
Unit 1 Introduction to Basic Mathematics 1 — 18

Unit 2 Functions 19 — 29
Unit 3 Sequences and Series .......................... 30 — 38
Unit 4 Investment Appraisals I .......................39 — 46
Unit 5 Investment Appraisals II .....................47 — 58
MODULE 2 ......................................................................... 59
Unit 1 Valuation of Bonds and Shares .............. 59 — 71

Unit 2 Linear Programming I ........................... 72 — 91
Unit 3 Linear Programming II .......................... 92 — 109
Unit 4 Inventory Control I ............................... 110— 119
Unit 5 Inventory Control II ............................. 120— 129
MODULE 3 ........................................................................ 1 3 0
Unit 1 Inventory Control III ........................... 130— 137

Unit 2 Time Series Analysis ............................ 138— 155
Unit 3 Correlation ................................... 156 — 174
Unit 4 Regression Analysis ............................. 175— 185
Unit 5 Project Planning 186— 198
ill
MODULE 1
Unit 1 Introduction to Basic Mathematics
Unit 2 Functions
Unit 3 Sequences and Series
Unit 4 Investment Appraisals I
Unit 5 Investment Appraisals II
UNIT 1 INTRODUCTION TO BASIC MATHEMATICS
CONTENTS
1.0 Introduction
2.0 Objectives
3.0 Main Cont ent
3.1 Numbers and basic arithmetic operations
3.1.1 Number
3.1.2 Basic Arithmetic Operations
3.2 Exponents and Roots
3.3 Algebraic Expressions / Operations
3.3.1 Addition and Subtraction of Expressions
3.3.2 Multiplication Expressions
3.3.3 Division Expressions
3.4 Ratio and Proportion
3.5 Variation
3.6 Advantages of Quantitative Skills for Management
4.0 Conclusio n
5.0 Summary
6.0 Tutor - Marked Assignment
7.0 References /Further Readings
1.0 INTRODUCTION
This unit is to assist you recapitulate your knowledge of numbers and
basic arithmetic operations of addition, subtraction, multiplication and
division of numbers. You are expected as Banking and Finance student
to be reasonably grounded in numerical studies. As you go through the
examples and exercises, you need a calculator to assist you in
facilitating your calculation. This unit will give you the needed
foundation and to appreciate mathematics of finance as it is applicable to
business in later units.
As mathematics is full of symbols e.g. x, y, z and expressions involving

powers, roots and the basic arithmetic operations, you will go through
methods of manipulating such expressions in order to present them in
1
their simplest forms. There will be combination of symbols to obtain

algebraic expressions (basic arithmetic operations).
In this units too, ratio, proportion as well as variation will be revised.

These are basic concepts which appear frequently in business and
financial calculations. For example, liquidity ratios, profitability ratios
and activity ratios are normally used for determining how well a
company's fund is managed etc. etc.
2.0 OBJECTIVES
On completion of this unit, you should be able to:
 perform basic arithmetic operations on numbers like conversion

between fractions, decimals and percentages;
 simplify numerical expressions involving exponents and roots;
 identify like terms in algebraic expressions and perform basic
algebraic operation including long division;
 distinguish between variations (direct and indirect).
3.0 MAIN C ONTENT
3.1 Numbers and Basic Arithmetic Operations
3.1.1 Numbers
Numbering in word and expressing them correctly in figures is

important in daily business. For example, if you write out a cheque to
draw N10,000.00, the amount in words is ten thousand naira. In the
figure, each digit has a place value which enables the correct magnitude
to be attached e.g. 16,790,406 — "sixteen million, seven hundred and
ninety thousand, four hundred and six".
There are ten digits (base 10 or decimal system) namely 0, 1, 2, 3, 4, 5,

6, 7, 8 & 9, all integers can be formed using these ten digits.
There are positive as well as negative integers. Numbers such as 10, 20,
35, 78 and 300 are positive integers while numbers such as — 50, —38,
and —137 are negative. A number without a sign is always assumed a
positive number. Numbers can be arranged in word form or figures as
below:
 Billions—. Hundred Millions—. Ten Millions—> Millions—>

Hundred Thousands Ten Thousands—> Thousands—>
Hundreds—. Tens—> Ones.
2
3.1.2 Basic Arithmetic Operations
The basic arithmetic operations are addition (+), subtraction (—),

multiplication (x) and division (+). When these are present in a single
mathematical expression, they are performed in a specific order:
(i) perform all operations within brackets (parentheses);

(ii) perform all operations with exponents (e.g. 2 3);
(iii) multiply, divide, add or subtract in the order that numbers are
presented (left to right).
Example 1
Simplify 15 + 3+23 x 1 —4 x 2 (3 + I)
Step I: open the brackets:

15+3 +2 3x 1-4 x2 x 4
StepII: the term with exponents:
15+3 + 8 x 1-4 x 2 x 4
Step III: multiply and divide in the order they appear:

(15 + 3)+(8 x 1)—(4 x 2 x4)
5 + 8 — 32 (Add and subtract as they appear)
13—32 = —19
Fractions and Percentages
Proper and Improper Fractions
Fractions are common phenomenon when working with numbers.

Fraction helps you in terms of parts of the whole.
For example; One-third, meaning one out of three, written 'A or 1/3,
while one-quarter, meaning; one-fourth, written as 'A or 1/4. In
general, a fraction is an expression of the form a or a/b where a and
b are integers. This definition obviously excludes the case b = 0
(why?) Imagine dividing a number by 0 (if is allowed), what will be the
values of, say 5/0 or 18/0? Suppose 5/0 = k; then one must have 5 = Ox
k i.e. 5 = 0, which is absurd.
In m/n fraction, m is the numerator while the number n is denominator.

The fraction min is a proper fraction if m < n (m is less than n) and
known as an improper fraction if m n (i.e. if m is greater than or equal to
3
MBE 839 QUANTITATIVE TECHNIQUES FOR BANKING AND FINANCE
n). E.g. %, 4/7 and % are proper fractions while 7/3, 8/8 and 4/3 are
improper fractions.
7/3 may be written as 2% (mixed number).
Note: Every mixed number is the sum of an integer and a proper

fraction. Conversely, every mixed number is an improper fraction.
Finally, note that the reciprocal of a fraction is obtained by simply

interchanging the numerator and the denominator.
Every answer in mathematics in the form of fraction is presumed to be

expressed in the lowest term (form) or simplest or reduced form, unless
otherwise directed.
Example 2
70 = 14 = 2A (why?)
105 21
is the lowest form.
SELF ASSESSMENT EXERCISE 1
( i) Convert 9/4 to a mixed number.
( ii) Convert 4 3 to an improper fraction.

5
(iii) A store carries 22 different styles of girl's shoes, 16 different

styles of boy's shoes, 12 different styles of men's shoes and 24
different styles of women's shoes. What fraction of all shoe
styles is for men?
Arithmetical Operations with Fractions
Addition and Subtraction
When two fractions are with a common denominator, they can be added
by simply adding their numerators;
a+b = a + b e.g. 4+3 = 4+3 =
cc c 55 5
4
A similar rule applies for subtraction
a—b = a—b e.g. 5— 3 = 5-3 = 2 = 1
cc c 12 12 12 12 6
However, when you must add or subtract two fractions with unequal
denominators, then the fractions must first be re-written with a common
denominator, (usually the least common multiple (LCM) of the unequal
denominator or simply the product of the unequal denominator).
Example 3
5+3=
64
Approach I:
10 + 9 (since LCM of 6 and 4 is 12)

12 12
10+9 = 19 = 17
12 12 12
Approach II (Alternatively):
5+3 20 + 18 (24 is the product of 6 and 4)

64 24 24
20 + 18
24
= 38 = 19 = 11
24 24 12
For mixed fractions, deal with the whole numbers first and then treat the
fractional part as in example below:
Example 4
Simplify 4% + 21/4
Solution:
Step I: Deal first with whole numbers:

4+2=6
5
Step II: 2A + % = 8+3= 8+3= 11

12 12 12 12
4% + 2% 611
12
Multiplication and Division
axe = ac
bd bd
Irrespective of whether the fractions are proper or improper, it is always

the p r od uct of the nu mer ators d ivid ed b y the p r od uct of the
denominators.
Example 5
lx 5 = 1x5 = 5
39 3x9 27
To divide one fraction by another, multiply the first fraction by the

reciprocal of the second.
a±c = axd = ad
bd bc bc
3 x 3 =
Example 6
3x5 = 15 5
75 7x3 21 7
Decimal Fractions
Another common expression of fraction is in decimal form like:
3 = 0.3; 27 = 0.27
10 100
Example 7
3= 3÷ 8= 0.375
8
6
Percentage
Percent means of one hundred and is denoted by the symbol %. For

example, 5% is read as five percent, meaning 5 of one hundred, or five
hundredths.
Percentage also indicates proportion; it is a fraction with a denominator

of 100 i.e. 75% is equal to 75 = 3..
100 4
Example 8
In a firm of 1,250 employees, 575 are male. Find the percentage of

female employees.
Solution:
Number of Female employee 1,250 less 575

675
Percentage of female = 675 x 100

1250 1
54%
(i) Convert the following percentages to decimal numbers: 6.5%,

123/4%, 147%.
(ii) Convert 80% to a fraction reduced to its lowest terms.
3.2 Exponents and Roots
Exponents
The product x. x. x... = x3 because there are three (3) of the x multiplied
together. The 3 in x 3 is the exponent (power), while x is known as the
base. In general:
n times
E.g. 25 = 2x2x2x2x2 = 32
7
i.e. '2 raised to power 5 is equal to 32'.
State 3 in descending order as follows:
33= 27 (31x 31x 31) = 31 ++1
3 2 1+1
3
9 (3'x35
3 (31)
3 0
1 (3 + 3) = 1 = 3 I-I
3-1 1 (1 + 3) = 1 = 3 1 -1
31 31
3 -2 (1 + 3) =1 = lx1 1
=
31 32 31 31 32
3-3 = 1 (1+3) = 1 = lx lx 1 = 1
33 32 33 31 31 31 33
And so on.
This pattern leads one to a definition of x" when n is zero or a negative

integer:
—1 1 for x 0
x
n
x. x. x. ..........
Example 9
(a) 1.0753= 1.075 x 1.075 x 1.075 = 1.2423
Use calculator with the key (y") to calculate easily.
(c) 1 45 = 4 x4 x4 x4x 4 = 1,024

4-5
(d) 1 — 1.02-3 1— 1/1.023 = 1 —O.942 =

2.88
0.02 0.02 0.02
8
Roots
Roots are the opposite of exponents. If r" = x, the r is the nth root of x to
be written as r = x.
E.g; 52 =25;
So 5 is a second root (normally called the square root of 25). (-5)+ =
25 also, therefore —5 is also a square root of 25. (Generally, the square
of any real number is expected to be positive).
Conventionally, 24x is often written as ix without the 2, e.g. 24 36 = 4

36 = 6. If x is a positive number, then the expression qixP means the
same as xpig where p and q are integer and q is positive (i.e. q> 0).
Example 10
( a ) X 3 1
( b ) 9-1/4 249-1 = 41/9= V3
(c) —
72 = —3

From what you have learnt, so far, simplify the following expressions:
(ii) (2x3)3
3.3 Algebraic Expressions/Operations
In algebraic expression, numbers are expressed in symbols like x, y and

z. They are combined by operations of addition, subtraction,
multiplication, division on extraction of roots. For example, 2x 2_ 3x +
7 is an algebraic expression in the variable x while 3x 2y — 2x + 1 is in
the variables of x and y.
In the example 2x 2 - 3X ± 7, there are three terms: 2x 2, - 3X and 7. In

the term 2x 2, 2 is called the coefficient of the literal part x 2. Similarly,
in the term — 3x, —3 is the coefficient of the literal part x. The third term
7 does not have a literal part. Try and be familiar with the following
9
MBE 839 QUANTITATIVE TECHNIQUES FOR RANKING AND FINANCE
properties of addition and multiplication, cumulative property,

associative property and distributive property before studying the
combinations of algebraic expression.
Cumulative Property:
x+y y + X andx+(—y)— —y+x,
E.g. 4+5 = 5 + 4 = 9; 4 +(-5) = —5+4 = —1
3x6 = 6x3 = 18; 3 x(-6) = (-6)x 3 = —18

Associative Property:
x-f(y+z) = (x+y)+z and xx(yxz) =

(xxy)xz
E.g. 2 + (3 + 5) = (2+3)+5 = 10; 2 x(3 x 5) = (2 x 3)x
5 =30
Distributive Property:
X + + -= xxy+X+z;
X x (y z) = xxy—xxz
Multiplication is distributive over addition, i.e. multiplication can be

distributed over addition. More so, multiplication is also distributive
over subtraction. In view of the cumulative law, x x (y + z) is the same
as
(y + z) x x. Thus, ( y +z ) X =y x X + z x X = X + y + X + z a s well.
E.g. (2 + 3) x 4 = 2 x 4 + 3 x 4 = 8 + 12 = 20.
—2(3 —5) = (-2) (3) — (-2) (5) =-6 +l6=4

3.3.1 Addition and Subtraction of Exprosion
When adding or subtracting algebraic expressions, the terms must be
arranged such that all like terms are grouped together. That simplifies
computation.
Example 1.11
Simplify 5a + 7b — 3 + 313 — 2a + 8
10
Solution: Re-arrange the terms:

5a+ 7b-3 +3b-2a
(5-2)a±(7+3)b+(-3 +8)
3a+ 10b+ 5
3.3.2 Multiplication Expressions
This is straight forward in algebraic expressions. Just use the

distributive property repeatedly and put like terms together as shown
below:
Example 12
(a) —3 (x — 2y + 7z3) = (-3) x + (-3) (-2y) + (-3) (7z3)

= —3x +6y — 21z3
(b) x2 (x 2 + 3x _ 5 y3) = x 2 yx2+ x 2y 3x + x2 y (-5y 3)

= x 4y 3x3y 5x2y4
(c) (3x — 4) (6x 2 - 5x + 2)
 (3x — 4) (6x2 - 5x + 2)— 4 (6x2 - 5x + 2) (3x) (6x2) +

(3x) (-5x)+(3x) —3 (z)+ (-4) (6x2) + (-4) (-5) + (-4) (2)
 18x3— 15x2 + 6x — 24x2 + 20x — 8

Like terms together:
 18)(3— 15x2+ 6x — 24x2 + 20x — 8

18x3— 392+ 26x — 8
3.3.3 Division Expressions

Recall that a + b a+b =
cc
This will be used when dividing one algebraic expression by another.
11
MBF 839 QUANTITA I IVE TECHNIQI P S FOR BANKING AND FINANCE
Example 13
(a) 2x2+ 6x = 2x2 + 6x = x+3

2x 2x 2x
3 2
(b) 25(1 — 12q + 15q - 6
3q
25q3— 12q2+ 15q— 6 = 252— 4q — 2

3q 3q 3q 3q 3
In the division of one algebraic expression by another, the expression

that is being divided is the dividend and the expression by which we are
dividing the dividend is called the divisor. If the power or degree) of the
divisor is less than or equal to the power of the dividend, then the long
division method can be used to carryout the division as will be shown
below in
Example 14
Note: The result of an algebraic division normally consists of two parts

— quotient and remainder — written thus:
Dividend = Quotient + Reminder

Divisor Divisor
Divide 2)(3— 14x —5 by x —3
Solution:
Here, the dividend 2x 3 —• 14x — 5 and the divisor is x — 3 using long

division method. Therefore, to avoid errors, we arrange both dividend
and divisor in decreasing powers of x and fill in missing powers with
zero coefficients.
X3 - 14x —5 (quotient)
divisor x 3 2)( 3 -0— 14x —5 (dividend)
2x3-6x
6x2 — 14x
— 6x2 — 14x
4x —5
— 4x — 12
—7 (remainder)
12
MBE 839 QUANTITATIVE TECIINIQUES FOR BANKING AND FINANCE
So 2x3 — 14x — 5 = 2x2 + 6x + 4 + 7

x —3 x—3
In the algebraic expressions below, carry out the operations indicated

and simplify your answers:
L 7x — 4y + 3 -4- 3x — 2y — 1
ii. (x 2 — 3) (x 3 — 4x 2 — 4x + 1)
iii. t3 + 7t2 — 5t + 4
i. t 2
3.4 Ratio and Proportion
Ratio
Business information often compares related quantities as ratio. For

example, if someone invests a total of N22,000 in two popular stocks in
the stock market — stock A and stock B, in the stock market, investing
N6,000 in stock A and N16,000 in stock B, then
Investment in stock A N6,000 = 6 = 3
Investment in stock B N16,000 16 8
is known as the ratio of investments in the two stocks.
There are three ways to express ratios:
(1) three to eight or 3 to 8 (word/figure description)

(2) 3:8 (colon)
(3) 3/8 (fraction)
However, if you are comparing more than two quantities at the same
time, then use colon e.g. compare 6kg, 2kg and 7kg would be written as
6 : 2 : 7.
Note: such comparisons must be made only if the quantities being

compared are expressed in the same unit. For example, if the ratio A : B
where A is 1 hour and B is 50 minutes, then first convert B to hours;
then
A:B = 1 : 50/60
13
MBE 839 Q U A N T I TA T I VE TE C I IN I Q U E S F O R B A N K I N G A N D F I NA NC E
Ratios can also be obtained when comparing quantities relating to

different things but expressed in the same units. For example, suppose a
product is in terms of cost components, made up of N200 of material,
N100 of direct labour and N25 of overhead. The ratio of the cost
elements is then 200:100:25.
Now, just as fractions are reduced to their lowest terms, ratios can also
be reduced to their lowest terms; simply divide out by their common
factor.
Example 15
200 : 100 : 25 = 8:4:1
27 : 45 = 3:5
Ratio is also tool for allocation.
Allocation problems require the division of a whole into a number of

parts according to a specified ratio. The number of parts into which the
whole is divided is the sum of the terms of the ratio.
Example 16
A business suffers a fire loss of N210,000. It is covered by three

insurance policies and the claims these insurance companies have to pay
are in the ratio 1/3 : 1/8 : 5/12. Find the amount paid by each company.
Solution:
Since ratios are normally expressing using integers, it is necessary to

convert 1/3 : 1/8 : 5/12 to a ratio involving integers only. This is done
by multiplying the fractions by the LCM of their denominators, namely
24. Thus,
1/3 : 1/8 : 5/12 1/3 (24) : 1/8 (24) : 5/12 (24)
The ratio now 8 : 3 : 10
The sum of the terms of the ratio is 8 + 3 + 10 = 21; then divide

N210,000 into 21 parts in the ratio i.e. N80,000, N30,000 and N100,000
respectively:
(8/21 of 210,000) N80,000,
(3/21 of 210,000) N30,000
14
(10/21 of 210,000) N100,000
Proportion
A proportion is formed when two ratios are equal.

For example: 2 :3 = 8 : 12
2/3 = 8/12 = 4/6
1 :3 : 9 6 : 18 : 54
Example 17
Solve for x if 5/16: 14/3 = x : 21/10

Solution:
Expressing both ratios as fraction, we have

5
16 = x (multiply both sides by 21/10)
14 21
3 10
 5
16 x 21
14 10
3
 9
64 Example
18
A grocery store charges N100 for 5 packs of oranges. Assuming there is

a no quantity discount, how much will be charged for 18 packs of
oranges?
Solution:
Let x represent what is charged for 18 packs of oranges.
That is: =100

18 5
100 x 18
5 1
15
Thus, the grocery shop will charge N360 for 18 packs of oranges.
3.5 Variation
Variation is related to ratio and proportion because it is also concerned
with the comparison of two (or more) variables. Two kinds of variation
are most common: direct or indirect variations. In direct variations, the
ratio y : x is constant i.e. y/x = k, or y = kx where k is a constant.
In indirect (or inverse) variation, the product yx is constant i.e. yx = k

for a constant k. Note that in direct variation, y = kx implies that as x
increases in value, y also increases and as x decreases in value, y
decreases as well.
However, in indirect or inverse variation, yx = k implies that as x

increases in value, y decreases.
Conversely, as x decreases in value, y increases. In example 1.18, as the

number of packs of oranges increases, the amount charged increases;
this is direct variation. Actually, the price for one pack of oranges
remains constant; in this case at N100/5 = N20 per pack. This is why
we obtain:
360 = 180x x 360= -

18 180 18 1
We can express this as y = kx, where y is the amount charged, x is the

amount charged, x is the number of packs of oranges and k = 20 is the
price per pack.
Example 19
If it takes 12 days for eight workers to complete a job, how long does it
take 15 workers to complete the same job? Assume that the ability of
each worker is the same.
Solution:
With more workers — the job will be done in less time.

So this is the case of inverse variation. Let y be the number of days
needed and let x be the number of workers. The inverse variation
equation is yx = k for some constant k. Now, when x = 8, y = 12, since
16
MBE 839 QUANT ITA TIVE TECH NIQUES FOR BANKI NG A ND FI NAN CE
it took 8 workers 12 days to complete the job. Therefore, from yx = k, it

follows that k = 12, 8x = 96, and it is in order to state that yx = 96.
If there are now 15 workers, i.e. x = 15, we obtain the corresponding

value of y, y x 15 = 96.
96/15 = 32/5 = 62
5
Thus, it will take 15 workers 62/5 days to complete the job.
( i) S o lve f or x if 4 : 1 4 = x : 9 .
(ii) The scale of a map is 1.5 cm for 300 km. How many kilometers
are represented by 5 cm.
3.6 Advantages of Quantitative Skills for Managers
Executives at all levels in business and industry make decisions at every

course of daily activities. Quantitative methods provide scientific basis
for decision-making to the executive and enhance his ability to make
long range plans and to solve everyday problems of running a business
and industry with greater efficiency and confidence.
The advantages of the study of quantitative methods include:

(1) Definiteness: general statements are presented in a precise and
definite form.
(2) Condensation: it simplifies large data and presents meaningful
information from them.
(3) Comparison: it enables comparisons between past and present
results with a view to ascertaining the reasons for change.
(4) Formulation of policies: provides the basic material for framing
policies in business, etc.
(5) Formulating and testing hypothesis: useful in formulating and
testing hypothesis or assumption or statement and to develop new
theories.
(6) Prediction: for framing suitable policies or plans and then for
implementation it is necessary to have the knowledge of future
trends.
17
MI3F 839 QUANTITATIVE TECHNIQUES FOR BANKING AND FINANCE
4.0 CONCLUSION
The understanding of the basics of the nature of numbers and the basic
operations of addition, subtraction, multiplication and division
determine the efficiency in quantitative skills. It is then necessary to
become familiar with algebraic expressions and operations in order to
cope with problems involving unknown quantities. This marks the
conclusion of our review of the basic concepts of mathematics involving
arithmetic and algebra, which you will encounter in business, banking
and finance calculations.
5.0 SUMMARY
This unit is a review of numbers and various ways numbers are

combined with the basic mathematical (arithmetic) operations of
addition, subtraction, multiplication and division. It covers fraction and
percentages, algebraic operations, the use of symbols in place of figures,
and establishes relationship between ratios, fractions, proportion and
variations. These are all the tools for business calculations.
6.0 TUTOR - MARKED ASSIGNMENTS

( 1 ) 3/7 3/5
( 2 ) 9-1/4
( 3 ) Allocate N600 in the ratio of 3 : 2 —
( 4 ) (x2 + 2) (x2— 2) =
7.0 REFERENCES AND FURTHER READINGS
Indira Gandhi National Open University (2000). School of Management

Studies. MS — 8 Quantitative Analysis for Managerial
Applications, Basic Mathematics for Management I.
National Open University of Nigeria (2004). MBA 717: Basic

Mathematics and Statistics.
18
UNIT 2 FUNCTIONS
CONTENTS
1.0 Introduction
2.0 Objectives
3.0 Main Content
3.1 Definitions of Terminologies
3.2 Types of Function
4.0 Conclusion
5.0 Summary
6.0 Tutor-Marked Assignment
7.0 References / Further Readings
1.0 INTRODUCTION
For decision problems which use mathematical tools, one would first
identify or formally define all significant interactions or relationships
among variables relevant to the problem. These relationships usually
are stated in equation form (or sets of equations) or in-equations. Such
type of simplified mathematical relationships help the manager — the
decision-maker to understand (any) complex management problems.
For example, the decision-maker knows that demand of an item is not
only related to price of that item but also to the price of the substitutes.
Thus, if specific mathematical relationship (model) exist and can be
defined, then the demand of the item in the near future can be
forecasted. In this unit, you will study mathematical relationships
(functions) in the context of managerial problems.
2.0 OBJECTIVES
At the end of this unit, you should be able to:
 Explain functional relationships in business calculations;

 Define the terminologies relative to function;
3.0 MAIN CONTENT
3.1 Definitions of Terminologies

From unit 1, you were exposed to basic mathematic expressions that will
enhance your business and financial calculations. You correlate your
previous knowledge vis-à-vis this exposition to more dimension.
19
MBE' 839 QUANTITATIVE TECHNIQUES FOR BANKING AND FINANCE
Variable
A variable is something whose magnitude can vary or which can assume

various values. The variables in applied business mathematics include:
sales, price, profit, cost, etc. Since magnitude of variables can vary,
therefore, are represented by symbols (such as x, y, z etc.) instead of
specific number.
P = price;
9 = quantity;
c = cost;
s = saving or sales
d = demand etc. as the case may apply.
Variables can be classified in a number of ways. For example, a

variable can be discrete (bias to counting e.g. 2 houses, 3 machines, etc.)
or continuous (bias to measurement, e.g. temperature, height, etc.)
Constant and Parameter
A quantity that remains fixed in the context of a given problem or

situation is called a constant.
An absolute (or numerical) constant such as V 2, 7t, etc. retains the same
value in all problems whereas an arbitrary (or parametric) constant or
parameter retains the same value throughout any particular problem but
may assume different values in different problems, such as wages rates
of different category of labourers in an industrial unit.
Function
There are situations in which two or more variables are related to each
other. For example, demand (D) of a commodity is related to its price
(P) mathematically expressed as:
f (P) (1)
i.e. "Demand is function of price" or simply "f of p". It is not D equals f

times P. The function has two variable D and P. These are known as
variables because they can take on different numerical values.
Let us consider a mathematical relationship that contains three variables.

Assume that the demand (D) of a commodity is related to the price (P)
per unit of the commodity, and the level of advertising expenditure (A).
Then the general relationship among these variables can be expressed
as:
20
D = f (P, A) (2)
The functional notations of (1) and (2) above are meant to give a general
idea that certain variables are related. However, for making managerial
decisions, we need a specific and explicit, not a general and implicit
relationship among selected variables. For example, for the purpose of
finding the value of demand (D), we make the general relationship (2)
more specific as shown in (3) below:
4 + 3p — 2pA + 2A2 ........................(3)
Note: For any given values of p and A, the value of D can be calculated
using the relationship (3). This means that the value of D depends on
the values of p and A. D is the dependent variables and p and A are
called independent variables. That is, a rule of correspondence
established between the dependent variable and independent variable(s).
As soon as values are assigned to the independent variable(s) the
corresponding unique value for the dependent variable is determined by
the given specific relationship. This is why a function is sometimes
defined as a rule of correspondence between variables.
Other examples of functional relationships are as follows:
(I) Sales volume (V) of the commodity is a function of price (P) i.e.
V = f (P);
(H) Total inventory cost (T) is a function of order quantity (Q), i.e. T
=f (Q);
(iii) The net present value (Y) of an investment is a function of net
cash flows (Cr) in different time periods; project's initial cash
outlay (B), firm's costs of capital (P) and the life of the project
(N), i.e. y = f (Ct, B, P, N).
To understand the nature of model, (mathematical relationship) between

independent variable(s) and dependent variable you must be familiar
with such terms as parameter, constants and variables.
Example 1
Suppose an industrial worker gets N25 per day. If he works for 26 days
in a particular month, then his total wage for this month is N25 x 26 =
N650. During some other month, he may have worked a total of only 25
days, and then he would have earned N625. Thus, the total wages of the
worker, assuming no overtime; is calculated as follows:
21
MB': 839 QUANTI TATIVE TECHNI QUES FOR BAN KING AND F INAN CE
Total wages , 25 x number of days worked
T = Total wage
D , number of days worked
then T = 25D
This represents the relationship between total wages and number of days
worked. In general, the above relationship can also be written as:
T = KB
where k is a constant for a particular class of worker(s), to
be assigned or determined in a specific situation. Since
the value of k can vary for a specific situation, problem or
context (parameter).
3.2 Types of Function
They include: linear polynomial, absolute value, inverse step, algebraic

and transcendental (exponential and logarithm).
(1) Linear Functions: A linear function is one in which the power

of independent variables is 1, the general expression of linear
function having only one independent variable is:
y= f(x) = a + bx
where; a and b are given real numbers and x is an independent

variable taking all numerical values in an interval. A function
with only one independent variable is also called single variable
function. A single-variable function can be linear and non-linear.
For example:
y = 3 + 2x, (linear single-variable function) and
y = 2 + 3x + 3x — 5x2 ± x2 (non-linear single-variable function)
A linear function with one variable can always be grouped in a

two dimensional plane (or space). This graph can always be
p lotted b y g iving d iffer ent valu es to x and calcu latin g
corresponding values of y. The graph of such functions is always
a straight line.
22
Example 2
Plot the graph of the function, y = 3 + 2x
For plotting the graph of the given function, assign various values to x
and then calculate the corresponding values of y as shown in the table
2.1 below:
x 0 1 2 3 4 5
Y 3 5 7 9 11 13
Table2.1
The graph of the given function is shown in figure 1:
A
13 - (5, 13)
II - (4, 11)
9- (3,9) y 3 + 2x
7 -
(2,7)
5- (1,5)
3 (0.3)
1 2 3 4 5 6
Figure .1
A function with more than one independent variable is defined, in

general form, as:
y = f (xi, x2, ............ xn ) = a n, a l x, a2 x2 ± ........ ax„
Where an, al, a2, ...., an are given real numbers and xl, x2, x„ are
independent variables taking all numerical value in the given intervals.
Such functions also called multivariable functions. (It can be linear and
non-linear). For example:
y = 2 + 3x1 + 5x2 (linear multi-variable function) and
y = 3 + 4x1 + 15x,x2 + 10x22 (non-linear multivariable function)
Multivariable function may not be graphed easily because these require

three-dimensional plane or more dimensional plane for plotting the
23
graph. In general, a function with n variables will require (n + 1)

dimensional plane for plotting its graph.
2. Polynomial Functions: A function of the form:
y = f (x) = alx" + a2x"-I + + ax' +
where a,'§ (i = 1, 2, ..., n + 1) are real numbers, ..................... (4)
al # 0 and n is a positive integer is called a polynomial of degree

n.
(a) If n = 1, then the polynomial function is of degree 1 and is

called a linear
function. This is, for n = 1, function (4) can be written as:
y = alx] + a2x0 (al # 1)
This is usually written as:
y = a + bx (.' = 1)
where 'a' and 'b' symbolize a 2 and al respectively.
(b) If n = 2, then the polynomial function is of degree 2 and is

called a quadratic function. That is, for n = 2, function (4)
can be written as:
y = a1x2 + a2x1 + a3 (a, # 0)
This is usually written as:
y -= ax2 + bx + c
where al = a, a2 = b and a3 = c
(3) Absolute Value Functions: By absolute value, we mean that

whether x is positive or negative, its absolute value remains
positive.
For example /7/ = 7 and 1-6/ = 6.
(4) Inverse Function: Take the function y = f (x). Then the value
of y, can be uniquely determined for given values of x as per the
functional relationship. Sometimes, it is required to consider x as
a function of y, so that for given values of y, the value of x can be
24
uniquely determined as per the functional relationship. This is

called the inverse function and also denoted by x = f1 (y). For
example, consider the linear function:
Expressing this in terms of x
x= y — b = y — b= c y+ d
a aa
where c = 1/a, and d = —b/a
This is also a linear function and is denoted by x = f -1(y)
(5) Step Function: For different values of an independent

variable x in an interval, the dependent variable y = f (x) takes a
constant value, but takes different values in different intervals. In
such cases, the given function y = f (x) is called a step function.
For example:
y = f (x) y k if 0 x<50
Y2, if 50 x < 100
y3, if 100 < x < 150
This can be shown graphically in figure 2.2 below for y3 <Y2 <
Yi:
Y1
y2
Y3
50 100 150
Figure 2
(6) Algebraic and Transcendental Function:
In unit 1, you were treated to funct ions wit h respect to

mathematical operations (addition, subtraction, multiplication,
division, power (exponents) and roots) associated with functional
relationship between dependent variable and independent
25
variable(s). When only finite numbers of items are involved in a

functional relationship and variables are affected only by the
mathematical operations, then the function is called an algebraic
function, otherwise transcend function.
The sub-classes of transcendental functions are as follows:

Exponential and logarithm i.e. functions. The exponential has
been introduced in unit 1, you will now be introduced to
logarithm function — expressed as:
y = logs x
where a 1 and > 0 is the base. It is read solution of functions.
The value(s) of x at which the given function f (x) becomes equal
to zero are called the roots (Or zeros) of the function f (x) for the
linear function:
y ax + b
the roots are given by:

ax + b = 0 or x = —b
a
Thus, if x = —b/a is substituted in a given linear function y = ax +

b, then it becomes equal to zero.
In the case of quadratic function
y = ax2 + bx + c,
to solve the equation ax 2 bx + c = 0; a 0 to find the roots of

y. The general value of x for which the given quadratic function
will become zero is given by:
—b± b2 —4ac
2a
Thus, in general, there are two values of x for which y becomes
zero. One value is:
—b + b2 — 4ac
2a
26
and other value is:
—b — b2 — 4ac
2a
 It is very important to note that the numbers of roots of the given

function are always equal to the highest power of the
independent variable.
In supply and demand functions, cost, profit, revenue, production
and utility functions etc. are applied in mathematical operation,
examples will be constructed to show these functions an their
solutions.
Example 3
A company sells x units of an item each day at the rate of N50 per unit.
The cost of manufacturing and selling these units is N35 per unit plus a
fixed daily overhead cost of N1000. Determine the profit function.
Solution:
The revenue received by the company per day is given by:
Total Revenue (R) = price per unit (P) x number of items sold (S)
 50 x x (50.x)
The total cost of manufactured items per day is given by:
Total Cost (C) = variable cost per unit (V) x number of items
manufactured (N) + fixed daily overhead
cost (FOC)
 35.x+ 1000
Thus, total profit (p)= Total revenue (R) — Total cost (C)
 50.x — (35.x + 1000) — 15.x— 1000
Example 4
Let the market supply function of an item be q = 160 + 8p, where q
denotes the quantity supplied and p denotes the market price. The unit 27
cost of production is N4. It is felt that the total profit should be N500.
Solution:
Total profit function can be constructed as follows:
Total profit (P) = Total revenue — Total cost
Price per unit x Quantity supplied

p.q — c.q
= (p — c).q
Given that c = Nq and q = 160 + 8p, then total profit function becomes:
P = (p —4) (160 + 8p)
= 8p2 + 128p — 640
If p = 500, then = 500 = 8p2 + 128p — 640
Thus becomes = 8p2 + 128p — 1140 = 0
Writing as ax2 + bx + c = 0 (the quadratic equation formula)
—b ± b2 — 4ac
2a
—128 ± (128) 2— 4 x 8 x (a-1140)

2x8
m — 128 ± 229.92
16
6.36 or —22.37
Since negative price has no economic meaning, therefore, the required
price per unit should be N6.36.
SELF ASSESSMENT EXERCISE
(i) Solve the equation 2x 2 — I lx + 22 = 10 using the quadratic

equation formula.
(ii) Explain the role of variables in describing a business transaction

and give four (4) examples of each terminology.
28
4.0 CONCLUSION
In this unit, effort has been put in place to explain the unique role of
functional relationship among decision variable. It will facilitate your
understanding of basic mathematics in business calculation.
5.0 SUMMARY
We started with the mathematical concept of function and defined terms
such as constant, parameter, independent and dependent variable.
Various examples of functional relationships are mentioned. Types of
functions normally used in managerial decision-making are sited along
with suitable examples, their graphs and solution procedure, with some
examples.
6.0 TUTOR MARKED ASSIGNMENTS

-
(1) Name various types of functions used in calculus (mathematical

calculations) and explain two you know in detail.
(2) If the profit function of business is p2 — 4p + 10 = 6, what is the

required price per unit?
7.0 REFERENCES AND FURTHER READING


Applications, Basic Mathematics for Management.
Collins, Eliza G.0 & Devanna, Mary Anne (1993). The Portbale MBA.
Spectrum Books Limited: Ibadan.
29
MBE 839 QUANTI TATIVE TECHNI QUES FOR BAN KING AND F INAN CE
UNIT 3 SEQUENCES AND SERIES
CONTENTS
1.0 Introduction
2.0 Objectives
3.0 Main Content
3.1 Sequence
3.2 Series
3.3 Arithmetic Progression (AP)
3.4 Geometric Progression (GP)
4.0 Conclusion
5.0 Summary
7.0 References/ Further Readings
1.0 INTRODUCTION
Sequence of numbers — when a set of numbers is arranged in a

recognizable pattern so that there is a first, second and third number and
so on, then the set of numbers is called a sequence of numbers, e.g. 2, 4,
6, 8, ..... Such numbers are arranged according to a definite rule and
each number in the sequence is called a term of the sequence. The terms
are usually expressed as xl, x2, x3 ... x„ ... where x, is the first term, x 2 is
the second term, etc. The nth term x„, is the general term. The two
kinds of sequences useful for business applications, especially in
financial and banking (interest) calculations are: arithmetic progression
(arithmetic sequence) and geometric progression (geometric sequence).
2.0 OBJECTIVES
After studying this unit, you should be able to:
 define the arithmetic and geometric progression;

 calculate sequences of numbers in arithmetic or geometric
progression in business;
3.0 MAIN CONTENT
3.1 Sequences
If for every positive integer (number) n, there corresponds a number an

is related to n by some rule, then the terms a l, a2..., an ... are said to
form a sequence. A sequence is denoted by bracketing its nth term i.e.
(an).
30
Example of some sequences:
(i) If an = n 2 , then sequence (a n ) is 1, 4, 9, 16, .... n 2 , ....

(ii) If an = 1/n, then sequence (a„) is I, 1/4, 1/4, 1/4, ...., 1/n ...
(iii) If an = n2/n+1, then sequence (a n) is 1/4, 4/3, 9/4, ...., n2/n+1,
Note: The concept of sequence is very useful in finance and banking —

simple and compound interest, annuities, present value and
mortgage payments.
3.2 Ser ies
A series is formed by connecting the terms of a sequences connoting

plus or minus sign. Thus, if a n is the nth term of a sequence, then: a, +
a 2 ... a n is a series of n terms.
3.3 Ar it hmet ic Progressio n ( AP)
A progression is a sequence whose successive terms indicate the growth

or progress of some characteristics. An arithmetic progression is a
sequence whose term increases or decreases by a constant number —
common difference of an A.P, denoted by d. That is, each term of the
arithmetic progression after the first is obtained by adding a constant d
to the preceding term. The standard form of an A.P is written as:
a, a + d, a + 2d, + a + 3d.
where 'a' is called the first term. Thus, the corresponding

standard form of an arithmetic series becomes:
a + (a + d) + (a + 2d) + (a + 3d) +
Example I
Suppose we invest N100 at a simple interest of 15% per annum for 5

years. The amount at the end of each year is given by 115, 130, 145,
160, 175.
This forms an arithmetic progression.
The nth term:
The nth term of an A.P is also called the general term of the standard
A.P. It is given by:
T„ = a + (n — 1) d; n = 1, 2, 3... sum of the first n terms of an A.P.
31
MBF 839 QUANTITATIVE ma INIQUES FOR BANKING AND FINANCE
Consider the first n terms of a, a + d, a + 2d, a + 3d ..., a + (n — 1) d.

The sum, S,, of these terms is given by:
= a + (a + d) + (a + 2d) + (a + 3d + + a + (n — 1) d
= (a -I a ) ... 1- a) F d {1 + 2 -I- 3 I ....± (n — 1){
= n.a + d n (n — 1) } (using formula for the sum of first (n— 1)

2 natural numbers)
n { 2a (n — 1) cl}
2
Example 2
Suppose Mr. Sola repays a loan of N3,250 by paying N20 in the first
month and then increases the payment by N15 every month. How long
will be taken to clear his loan?
Solution:
Since Mr. Sola increases the monthly payment by a constant amount,

N15 every month, therefore d = 15 and first month installment is, a =
N20. This forms an A.P. If the entire amount be paid in a monthly
installment, then we have:
n{ 2a+(n— 1)d}
2
or 3,250 = n { 2 x 20 + (n — 1) 15}
2
3,250 x 2 = n { 2 x 20 + (n — 1) 15}
6,500 = n{25 + 15n }
15n2 -I- 25n —6500 = 0 a quadratic equation in n.

Find n applying formula as previously discussed in unit 2.
—b ± b2 — 4ac
2a
—25 ± 4(25)2— 4 x 15 x ( — 6500)
2x 15
—25 ± 625 = 20 or —21.66

30
32
The value n = —21.66 (negative, hence meaningless). Mr Sola will pay

the entire amount in 20 months.
(i) Find the 15th term of an A.P. whose first term is 12 and common
difference is 2.
(ii) A retailer's revenue and costs exactly balanced on December 31,
2001. During the following year, his costs rose by N150 each
month and his revenue rose by N200 each month. What were; (a)
the cost incurred and (b) the revenue received during the month
of December 2002?
3.4 Geometric Progression (G.P)

A geometric progression (G.P) is a sequence when each terms increases
or decreases by a constant ratio called common ratio of G.P and is
denoted by r. In other words, each term of G.P is from the first by
multiplying the preceding term by a constant r. The standard form of a
G.P is expressed as a, ar, ar2..., where `a' is called the first term.
Thus, the corresponding geometric series in standard form is

a + ar + ar2 +
Example 3
Suppose we invest 14100 at a compound interest of 12% per annum for
three years. The amount at the end of each is calculated as follows:
(i) Interest at the end of first year:
 100 x 12 =- 1412
100
Amount at the end of first year:
 Principal + Interest
 100 + 100 (12/100)
 100 0 + 12/100)
33
This shows that the principal of N100 becomes N100 (1 +
12/100) at the end of first year.
MBE 839 QUANTITATIVE TECI INIQUFS FOR BANKING AND FINANCE
(Principal at the beginning of second year) ={ 1 + 12 }

100
 100 (1 + 12/100) {1 + 12/100)
100 {1 + 12/100}2
Amount at the end of third year:
 100 {1 + 12/100) {1 + 12/100) {1 + 12/100}

 100 {1 + 12/100}3
Thus, the progression giving the amount at the end of each year
is:
 100 {1 ±12/100); 100 (1 ±12/100)2; 100 {1 +12/1O0};
This is a G.P with common ratio r = {1 -I- 12/100)
In general, if P is the principal and the compound interest rate per

annum, then the amount at the end of first year becomes P (1 + 1/100).
Also, the amount at the end successive years forms a G.P.
= P(1 + 1/100); P (1 + 1/100) 2; ........... with r = P (1 -I- 1/100)
The nth term of:
The nth term of G.P. is called the general term of the standard G.P. It is
given by
Tn arn — In = 1, 2, 3....
,
Note here that the power of r is one less than the index of T n, which
denotes the rank of this term in the progression.
Sum of the First n Terms in G.P
Consider the first n terms of the standard form of G.P.
a, ar, ar2, arn- '
34
MB]: 839 QUANTITATIVE TECIINIQUES FOR BANKING AND FINANCE
The sum, S„ of these terms is given by:
s„ a + ar + ar2 arn 2 + a?' I ..... (1)
Multiplying both sides by r, we get:
rS„ = ar -F ar2 ar3 + + arn-2 -I- a? (2)

Subtracting (2) from (1), we have:
S„, —rS„ a—
S„ (1 —r) a (1 —?)
Or S = a (1 — ; r #1 and < 1
(1 — r)
Changing the signs of the numerator and denominator, we have:
S„= a (rn — 1) ; r #1 and > 1

(r — 1)
(a) If r = 1, G.P becomes a, a, a, ... so that S„ in this case is S r, = n.a.
(b) If number of terms in a G.P are infinite, then:
S„= a;r<1
(1 — r)
For r I, the sum tends to infinity.
Example 4
A car is purchased for N80,000. Depreciation is calculated at 5% per

annum for the first 3 years and 10% per annum for the next 3 years.
Find the money value of the car after a period of 6 years.
Solution:
(i) Depreciation for the first year:
80,000 x 5
100
The depreciated value of the car at the end of first year is:
80,000 — 80,000 x 5
100
80,000 (1 — 5 )
100
35
MBE 839 QUANTITATIVE TED INIQUES FOR BANKING AND FINANCE
( ii) D epr eciation f or the s econd year :
(Depreciated value at the end of first year) x (rate of

depreciation for second year)
80,000(1 — 5 ) 5
100 100
Thus, the depreciated value at the end of the second year is:
(depreciated value after first year) — (depreciation for

second year)
— —80,000 (1 — j) /51
80,000 (1 100 100
100
80,000 0 — 5) 0 — 5)
100 100
80,000 (1
— 5)2
100
Calculating in the same way, the depreciated value at the end of

three years is:
(depreciated value after second year) — (depreciation for

third year)
80,000(1 — 5) — 80,000 (1 —5) (5) (5)

100 100 100 100
80,000 (1 — 5)0 —5) (1 -5)

100 100 100
80,000(1 _)3
100
( iii) D ep r ec iat io n f or f ou r t h ye ar :
80,000(1 _)3 (10)

100 100
Thus, the depreciated value at the end fourth year is:

(depreciation value after three years) x (depreciation for
fourth year)
36
80,000(1 _)3 — 80,000 (1 _)3 (L))

100 100 100
Calculating in the same way, the depreciated value at the end of

six years becomes:
= 80,000(1 —5f — 80,000 (1 _)3 (1 0)3

100 100 10
N49,980.24
(i) Find the 6th term of the geometric progression 72, —24, 8, ....
(ii) You decide to increase your saving by N200 per annum. If you
save N500 in the first year, how long will it take to save N5,300?
4.0 CONCLUSION
In this unit, you can define sequences of numbers in arithmetic or
geometric progression and examples of their application in some
business calculations.
5.0 SUMMARY
This unit has directed you to define, distinguish and establish managerial
applicability of arithmetical and geometrical progressions in business
operations.

-
(1) Find the general term and the sum of first 20 terms, of the
arithmetic progression 2, 6, 10 .....
(2) Suppose that people spent 0.9 and save 0.1 of any money
received. If N1,000 is received, then N900 is spent and N100
saved. The people receiving the N900 will spend N810 and save
N90. This process will continue. Find the maximum sum of a
geometric progression.
37

Applications, Basic Mathematics for Management.
Owen, F. and Jones, R. (1975). Modern Analytical Techniques, Polytech

Publishers Limited, Stockport.

38
UNIT 4 INVESTMENT APPRAISAL I
CONTENTS
1.0 Introduction
2.0 Objectives
3.0 Main content
3.1 Compound Interest
3.2 Increasing Investment
4.0 Conclusio n
5.0 Summary
1.0 INTRODUCTION
In unit 3, you were led through the progressions (arithmetic and
geometric) as they are need in business and financial calculations. This
unit will introduce geometric progression relationship with investment
appraisal through compound interest problems solving. The compound
interest rate used for discounting cash flows is also called the discount
rate.
2.0 OBJECTIVES
 Identify how close geometric progression is to

solving compound interest problems;
 Solve problems in invest ment
appraisals.
3.0 MAIN CONTENT
3.1 Compound Interest
Compound interest is the interest that is received on the original amount

(principal) but not withdrawn during earlier periods. Compounding is
the process of finding the future values of each cash flow by applying
the concept of compound interest.
As stated above, geometric progressions are closely linked to compound
interest calculation as could be stated below.
Suppose N100 is invested at 5% per annum, then the interest earned

after the first year is:
39
N100 x 5 = N5
100
and the value of the investment at the end of the first year would be
N100 + N5 =14105
If the rate of interest is expressed as a proportion rather than a

percentage, the value of the investment can be calculated thus:
14100 -i 100 x 0.05
N100 (1 + 0.05) = 14105

Expressing the rate of interest as a proportion is more convenient than
expressing it as a percentage. Instead of a rate of interest of r% per
annum, we shall let the rate of interest be 100r% per annum. This may
appear a little difficult.
If the rate of interest is 7% per annum, then

100r = 7
7 = 0.07
100
Suppose NP is invested at 100e/0 per annum compound interest. At the

end of the first year, the value of investment is
P + Pr = P (1 Fr)
At the end of the second year, the investment is
P0 10 ± rP (1 r)
 (1+r) (P + rP)
 (l+r) (1 +r) P
 P (1+02
At the end of the third year, the value of the investment is

40 P (1 Er)2 -F rP (1+r)2
= (1+02 (P + rP)
(1+02 P (1+0
P (l+r)3
It should be obvious to you from the pattern emerging that the value of
the investment at the end of the nth year is
P (1 + r)" (1)
This formula involves calculating P r)" and from the basic
knowledge of basic mathematics in unit I and 2, where exponential
function mare discussed, you would be able to calculate with ease and
with the use of scientific calculator. In most books that involves
mathematics of finance, discount table are normally provided to
assist. It gives values of (I It 0 -IT which is a more convenient form
than (I I-On. The discount table is in table I and gives sufficient vahaes
for all calculation related to investment appraisal (compounding),.
Rates of interest are measured horizontally and time is measured
vertically. (Most books of logarithms include a reciprocal table. The
reciprocal of a number is the number divided into unity:
The reciprocal of x =1 = x-1
The reciprocal of x -1
=I=
x-t
Refer to unit 1, hence reciprocal of (1 + r) = (1 +
Example 1
Suppose N1,000 is invested at rA per annum compound interest. The

value of the investment at the end of the tenth year is
1000 (1 + 0.07)10
= 1000 (1.07)10
1000 x(1.967)
N1,967.00
OR
1000 (1 + 0.07)10 (From discount tables (1 -I- 0.07)-1° is

0.5083)
41
1 + 0.07)1° = 1
0.5083
1.967
Hence, the value of the investment is 1000 x 1.967 = N1,967.00.
Example 2
Suppose that N1,000 are placed in the savings account of a bank at 5

percent interest rate. How much shall it grow at thc end of the three
years? It will grow as follows:
= 1000 (1 + 0.05)3
1000 (1.05)3 In the scientific calculator, you enter 1.10,
press y" key, press 5 and then equal to key
1000 (1.1576) (---) etc.
N1,157.60
OR Breakdown
Year 1 = 1000 + 1000 x 5%
 1000 + 50
N1,050.00
Year 2 = 1050± 1050 x 5%
1000 + 52.50
 N1,102.50
Year 3 - 1102.50± 1102.50 x 5%

1000 + 55.10
N1,157.60
42
3.2 Increasing Investment

Suppose that you decided to invest N1,000 on the first of January in a
certain year, and to invest a further NI00 at the end of each year. If
interest is compounded at 10% per annum, we can deduce the following:
The amount invested at the end of the first year is
1000 (1 + 0.10) + 100
The amount invested at the end of the second year is
1000 (1 + 0.10)2 + 100 (1 + 0.10) + 100
The amount invested at the end of the nth year is
1000 (1 0.10)" + 100 (1 F ± 100 (1 + 0.10r 2 + 100
At this point, it is useful to have a general function. If an amount P is

invested at the beginning of a year, and a further amount a is invested at
the end of each year, and ifs is the sum invested after n years then:
P (1 + r)" + a (1 + + a (1 + rf2 + .......... + a
If the first term of the right hand side is ignored, then the remainder
forms a geometric progression with a first term a (1 + r) -I and a
common ratio 1/1+r. Substituting in the G.P formula, the series can be
summed for n years:
P 0 ± on + a 0 + + {1 _
1 r
1—1
1+r
Now to tidy up this expression, writing the denominator as a single
fraction gives:
1+r
1—1
1+r
OR
P (1 + + a (1 + rri + (1 (1)" 1 (1 -Fr)

—
Using the rules of indices, in mathematical expressions (calculation):
43
MEW839 QUANTITATIVE TECHNIQUES FOR BANKING AND FINANCE
 P (1 +r)"+a(1 +0' 1 + (1 —(1)"} (1+4
 PO +0"+a(1 +r)" —a
Removing the common factor (1 + rf
P (1 + r)" + (1 + r)" — a (2)

Example 3
You can now use the formula to solve the original problem. Suppose
you wish to know the sum invested after 4 years:
Then S = (moo + oc) + (1 + 0.04 — oo
0.1 0.1
 2000 (1.1)4— 1000
 2000 (1.464)
N1,928.00
Often, you wish to know how much must be invested now to give a
specified income for a specified period, given the rate of interest.
Clearly, the sum invested at the end of the period will be zero as
follows:
0 = (P — a) (1 + r)" + a
(NOTE: that as withdrawal of fund is made, 'a' will be negative).
0 = P(1 + r)r— a (1 -I On -I- a
P(1 F r)" a (1 + —a
P(1 I r)" a [ (1 + r)" — 1 I
—(1 + r) (3)
44
Example 4
Suppose an investment is required to yield N1,500 at the end of each

year for five years, and money can be invested at 5% per annum
compound interest, then P, the sum that must be invested at the
beginning of the first year is:
1500 11 —(1 +0.05F5
1.
-5
0.05
1500 x 0.2165
0.05
N6,495.00
SELF ASSESSMENT EXERCISE I
(i) Your uncle died and leaves you N1,500 invested in unit trusts.
Your financial adviser stated that the value of your investment
can be expected to grow by 7% per annum (income and capital
growth). Estimate the value of the investment after 10 years.
(ii) A Chief Executive Officer is due to retire at the end of the year,
and the board vote that an income of N2,000 per annum be paid
to him or his family for 10 years. The accountant is instructed to
set aside a sum of money from which the income will be paid. If
the fund can be invested at 8% per annum, how much should the
account set aside?
4.0 CONCLUSION
In this unit, you have leant the link between the progression (especially
geometric) to compound interest calculation and how they are used in
investment analysis, appraisals as well as calculations.
5.0 SUMMARY
Attention has been focused on investment, savings, pension fund etc.
and how they are calculated. The role of compounding is emphasized;
noting how it relies on the concept of compound interest and how they
are related to investment appraisal, and calculation.
45
MI3F 839 QUANTITATIVE TECHNIQUES FOR BANKING AND FINANCE
6.0 TUTOR-MARKED ASSIGNMENT

Suppose N20,000 is invested at the beginning of a year at 5% per annum
compound interest, and N2,000 is withdrawn at the end of each year for
four years. Note that a is a withdrawal and will be negative.

Pandcy, I.M. (2005). Financial Management, UBS Publisher's

-
Distributors pv r LTD.
46
MI31. 839 QUANTITATIVE TECHNIQUES FOR BANKING AND FINANCE
UNIT 5 INVESTMENT APPRAISAL II
CONTENTS
1.0 Introduction
2.0 Objectives
3.0 Main Content
3.1 Present Value Concept
3.2 Investment Decisions
3.3 Investment Projects
3.4 Annuities
3.5 Sinking Funds
3.6 Discount Rate
4.0 Conclusion
5.0 Summary
1.0 INTRODUCTION
As a follow up from unit 4, you will be shown how compounding can be
used to increase an investor's analytical power to compare cash flows
that are separated by more than one period, given interest rate per
period. With the compounding technique, the amount of present cash
can be converted into an amount of cash of equivalent value in future.
Present value of a future cash flow (inflow or outflow) is the amount of

current cash that is of equivalent value to the decision maker.
Discounting is the concept of determining present values of a series of

future cash flows.
2.0 OBJECTIVES

 explain present value technique
(discounting) in financial decisions;
 calculate present and future values of investment appraisals;
47
M131, 839 QUANTITATIVE TECHNIQUES FOR BANKING AND FINANCE
3.0 MAIN CONTENT
3.1 Present Value Concept
Were someone to offer you the choice of receiving N1,000 cash now, or
alternatively, N1,000 cash in twelve months' time, it is highly likely that
you would prefer to take the money now, as most people would. People
have a strong preference for holding cash now as against the prospect of
receiving cash in future. Economist calls this preference, 'liquidity
preference'. To the mathematician, who in many aspects of his work is
try to quantify human behaviour, the concept implies that people place a
higher value on money received now than they do on the same amount
of money received in the future (time value of money).
The recognition of the time value of money and risk is extremely vital
in financial decision-making. If the timing and the risk of cash flows is
not considered, the person (manager etc. may make decisions that may
allow it to miss its objective of maximizing the owners' welfare. The
welfare of owners would be maximized when wealth or net present
value is created from making a financial decision.
What is Net Present Value? How is it computed?
We may say that the N1,000 we receive today has a greater value to use
than the N1,000 we are going to receive next year. It is still the same
value, but the mere fact that you can express a preference is sufficient
evidence that these one Naira notes have a different value according to
when they are received. Preference alone, however, is of little use in
mathematics. If this aspect of human behaviour is to be analysed, and if
the analysis is to be of assistance, one must find some means of
qualifying the preference. That is of saying by how much we prefer
N1,000 now to N1,000 next year or the year after.
Fortunately, the use of the interest formula developed earlier in unit 4

gives us a simple method of assessing the difference in value between
money now and money in the future. You know that if we were to
invest N1,000 today at say 10% interest, at the end of twelve months it
will have grown to N1,100 and at the end of two years to N1,210. It is
thus possible to say that N1,000 is the present value both of N1,100
received one year hence, and also of N1,210 received two years hence,
when the rate of interest is 10%. Generalizing present value may be
defined as follows:
The present value of Na receivable n years hence is that sum of money

which, invested at the current rate of interest roA, will amount to Ilia after
the expiry of n years.
48
MB!: 839 QUANTITATIVE TECHNIQUES FOR BANKING AND FINANCE
This concept is more manageable if you consider the formula already

developed. You know already that if NP is invested for n years at r%
expressed as:
a = P(l+r)n (1 )
From this, we can deduce that
a = a (1 + r) (2)
(1 + r)"
These two equations {(1) and (2)) are the basis of every investment
appraisals. You will have realised that they are in reality merely two
different forms of the same equation, but with this important difference
tells:
(1) The terminal value of a given sum of money invested at interest
over time;
(2) What to invest to achieve at a given terminal sum over a given
period over a given time at a given interest rate
It shows in other words, what is the present value of a given sum of

money received some time in the future.
Example 1
Suppose we wish to find the present value of N50 receivable in one

year's time, given the rate of interest as 10%. The present value, P =
N50 (1 + 0.1)-1 i.e.
50 = N45.45
1.1
Recall: the meaning of present value, the logic of this result should be
obvious.
N45.45 invested at 10% for 1 year would yield interest of N4.545.

Thus, the value of the investment after 1 year would be N45.45 +
N4.545 = N50.
Example 2
Suppose we wished to calculate the present value of N1,000 receivable
in seven year's time, assuming a 5% rate of interest. Substituting in the
formula for present value, you have
49
= 1000 (1 + 0.05) -7
You will remember the discount tables that give values of (1 + r)
(1 + 0.05) -7
= 0.7107
hence P = 1000 x 0.7107
N7,107.00
(i ) Calculate from the first principles the present value of N10,000

receivable one year hence given a rate of interest of :
(a) 3%
(b) 20%
(ii) If you were buying a television set, would you prefer to pay:
( a) N 20 0 now in comp lete s ettlement;

OR
(b) N100 now and a second installment of N110 in twelve
months' time.
(in) What is the present value of Ni receivable one year hence, given
a rate of interest of:
(a) 4%
(b) 6%
3.2 Investment Decisions
The investment decision rules may be referred to as capital budgeting

techniques, or investment criteria. A sound appraisal technique should
be developed to measure the economic worth of an investment project.
The essential property of a sound technique is that it should maximize
the shareholder's wealth. One of the characteristics of a sound
investment evaluation criterion is that "it should consider all cash flows
to determine the true profitability of the project".
In the above examples, you have calculated the present value of a future
flow of income and of costs met at different points of time. These two
calculations form the basis of most investment decisions (cash flow to
determine profitability).
50
So far, as income is concerned, we would prefer to undertake that

project which yields revenues having the greatest present value.
Assumption
You will assume that the objective of the firm considering investment
projects is the maximization of profit. This assumption may not be
true at all times for all firms, but it is a good working approximation.
You can say then that the aim of the firm is to maximize the
difference between the present value of the future flow of income and
the present value of the flow of costs.
Example 3
An equipment costing N1,000 has an expected life of 5 years. It is

estimated that the cash flow resulting from the use of the machine will
be N400 a year. The rate of return expected from capital of this type is
15%. Is the investment worthwhile?
The appraisal would be as follows:
Year Income (1 + 0.15) —n Present Value
1 400 0.8696 347.84

2 400 0.7561 302.44
3 400 0.6575 263.00
4 400 0.5718 228.72
5 400 0.4972 198.88
1,340.88
Thus, for an investment of N1,000 now, you will receive a return over
the next five years which has a present value of N1,340.88. What does
this signify? It can mean one the one hand that to obtain a return of
N400 a year for five years, you would have to invest N1,340.88 at 15%
and hence the machine is a good investment. Alternatively, you can
look at it this way. Had the rate of return to capital been 15% in this
case, the present value of the cash flow discounted at 15% would have
been N1,000.
The fact that we have a surplus return of N340.88 means that the rate of
return to this investment is greater than 15%. The machine is yielding a
rate of return greater than capital of this type generally.
51
This appraisal depends on three variables:
( i) t he co st o f t he equip me nt ;
( ii) t he expect ed inco me ;
( iii) t he expect ed rat e of ret urns.
For present purposes, however, it is enough to note that if you are

comparing two projects with different costs of purchase, both of which
yield a surplus return, you can compare them by expressing the surplus
return as a percentage of cost price. Thus, in the above example, the
surplus return is N340.88 which is 34.088% of cost.
Note: Ultimate costs depend on each factor as development costs. In

real life, these costs could be over or understated and this will influence
the rate of return which is normally obtained from the type of capital
applied although expectations of industrialists can vary over time.
Consideration is made of appropriate discount rate too in certain
circumstance.
The unknown factor is the expected flow of income. At times, industry

is often too optimistic about cash flow resulting from an investment.
3.3 Investment Projects
In practice, the decision to compare investment projects is not simply to

invest or not to invest. Usually, there is a range of investment projects
which can be undertaken, and the problem is to find the best
Example 5.4
A firm is faced with two alternative investment plans. Plan I will cost
N750 and Plan 2 N950. Both plans involve the purchase of equipment
the life of which is four years, and the current rate of return on capital is
expected to be 20%. The estimated cash flows resulting from the
projects are:
Year
1 2 2 4
Plan 1 N300 N400 N300 N200
Plan2 N500 N400 N300 N300
The present value of these expected returns assuming a rate of return on

capital of 20% is:
52
Year Return (1+ 0.2) ' Present value

1 2 1 2
1 300 500 0.8333 249.99 416.65
2 400 400 0.6944 277.76 277.76
3 300 300 0.5787 173.61 173.61
4 200 300 0.4823 96.46 144.69
797.82 1,012.71
Both projects will yield a surplus.
With plan 1;
A return with a present value of N797.82 for a current cost of N750, a

surplus of N47.82.
Plan 2 will yield N1,012.71 for a current cost of N950 — a surplus of

N62.71.
Both are yielding a return greater than 20%.
The only way in which the two projects can be compared easily is by
expressing the surplus as a percentage of the cost of the projects.
Project 1 has a surplus percentage return of almost 6.4%, while project 2

has a surplus percentage return of 6.6%. Thus, although there is little in
it, project 2 would be the preferred project to undertake.
(i) A further examination of the two projects in Example 5.3 shows

that the equipment bought for project 1 will have a scrap value of
N150 at the end of its life, while the equipment bought for project
2 will have a scrap value of 14100. Assuming the equipment is
sold for cash, will the new information affect the investment
decision?
(ii) The Government now decides to give an investment grant of 40%

of the original cost of the equipment bought, payable one year
after purchase. How will this affect the decision?
(a) ignoring scrap value; (b) allowing for scrap values as

above?
53
MI31 839 QUANTI TATIVE TECHNI QUES FOR BAN KING AND F INAN CE
,
3.4 Annuities
Annuity is a fixed payment (or receipt) each year for a specified number
of years. If you rent a flat and promise to make a series of payments
over an agreed period, you have created an annuity. The equal
installments loan from the house financing companies or employers are
common examples of annuities.
What is the present value of future sums of money? Or what sum of

money investment now at r% for n years would accumulate to a given
total? Or how much would I have to invest now at a given rate of
interest r% in order to receive an income of Na per year for n years?
Refer to example 5.4. It meant that if you invested N797.82 at 20%

compound interest, drawing N300 at the end of the first year, N400 at
the end of the second year, and so on, by the time you drew N200 at the
end of the fourth year you would just have exhausted our original
deposit and the interest earned.
Now this is exactly what an annuity is — an investment which while

earning interest, enables us to draw a given sum of money for n years,
by which time your capital plus interest is just exhausted.
Annuity Formula
Suppose we consider an annuity yielding Na per year when the rate of

interest is r%, that is:
Na
Na receivable one year, hence has a present value of 1 +r
Na
Na receivable two year, hence has a present value of (1 + r)2
Na receivable n years, hence has a present value of Na

(1 4 On
Thus, the present value of an annuity yielding Na receivable two year,

hence has a present value of Na per year for n years is:
P.V = a + a + a + .... a
(1 + r) (1 + r) 2 (1+ r) 3 (1 + r)"
As you can see, this is a geometric progression of which the first term is
a/I 1 r and the common ratio 1/1 + r. The sum to n terms of a geometric
progression is as you know,
54
Sn = a (1 —r") where a is the first term and r the common

ratio.
1 —r
Thus, in summing the progression, you have a = a and r = 1
1+r 1+r
The present value = the sum of the geometric progression
a [1 — (1)1
1-Er 1-Er
1— 1
1+r
a — ___ x 1 + r
1+rL1+r
P.V = al1 — ( 1+r) l
Notice the semblance of formula derived in a different fashion in

Example 4.4 applied to equation (3) in unit r.
Example 5
An annuity yields N100 per annum for 7 years. How much does it cost
if the current market rate of interest is 6%?
The present value of the annuity:
100 [ 1 — (1.06) -7 ]
100 (1 — (0.6651) = 100 x 0.3349
0.06 0.06
(i) Effa, an employee of We Limited retired from his employment at

the age of 65. He received from his employers a terminal gift of
NL500 or the alternative of a pension amounting to N400 per
year for life. Assume a rate of interest of 5% and that a man aged
65 has a life expectancy of 7 years, which alternative should Effa
choose?
55
MEW 839 QUANTITATIVE TECHNIQUES Mk RANKING AND FINANCE
(ii) Had Effa adopted another course of action (using his lump sum to
buy an annuity), how much income would he have gained or lost
per annum?
3.5 Sinking Fund
A related type of problem to annuities is found in the concept of a

sinking fund. Sinking fund is a fund, which is created out of fixed
payments each period to accumulate to a future sum after a specified
period. For example, companies generally create sinking funds to retire
bonds (debentures) on maturity.
Obviously, this is a sum of money put aside at regular intervals usually

monthly or annually in order to achieve a given sum at the end of a
predetermined period. This concept may be found in the provision for
replacement of capital, or in, say, unit trust investment of a few pounds
per month by the individual saver. You will recall in unit 4 deriving the
formula
P I. (1 + On- 11
giving us the terminal value of an annual investment of EP at r% for n

years.
It can easily be seen that if you are given the required terminal sum and
wish to know the annual savings, we can modify the formula into:
Sr
(1 + r)" 1
Example 6
A firm wishes to make provision for the replacement of certain items of
capital equipment which will wear out in 8 years' time. The estimated
cost of replacement is N5,000. If the rate of interest is 8%, what annual
provision must it make to ensure funds being available?
The annual provision NP = 5000 x 0.8

(1+ —1
(1 + r)n = 1 = 1 =
1.851
(1 + r) —n 0.5403
N400
1 851 — 1
N470.00
56
( )
Suppose an investor wants to find out the present value of
N50,000 to be received after 15 years. Her interest rate is 9
percent. What is that present value?
(ii) A man invests N100 in a unit trusts and a further N100 at the end
of each year thereafter for the next ten years. What will be the
value of his holding at that time assuming a rate of growth in the
value of the units of 3%?
3.6 Discount Rate

In this unit, you have been shown elements of Mathematics of Finance —
the two most common methods of adjusting cash flows for time value of
money:
(i) Compounding — the process of calculating future values of cash

flows, and
(ii) Discounting — the process of calculating present values of cash
flows.
Choosing an Appropriate Discount Rate
It should be quite obvious by now that in any of the techniques

discussed the result obtained will depend largely on the rate of discount
applied to future receipts. If this is not appropriate the results can not be
a guide to policy. The precise rate you choose must of course depend on
the problem in hand.
If we are evaluating capital investment, the most generally useful rate is

the rate of return generally expected from capital of this nature. That is,
capital investment is just worthwhile if when you discount cash flows,
the present value of future cash is just equal to the cost of the
investment. If there is a surplus above this, it implies that the rate of
return is greater than the discount rate used. That is, the capital is
earning a greater percentage return than you would expect. If there is a
deficit, it is earning less than comparable capital elsewhere.
On the other hand, if you are dealing with the lending or borrowing of
cash funds, you can usually find some well established market rate of
interest for a loan of that type and that degree of risk which will give a
more than useful guide you with a reasonable knowledge of finance, will
soon be able to assess comparability of loans and interest rates.
57
In all these techniques, however, there is nothing like sound common

sense. The methods outlined have been proved over time, but no
technique, however good, is better than the user.
4.0 CONCLUSION
In this unit, you have learnt how investments can be analysed to

compare cash flows at different time value — present and future.
Illustrations of the two most common method of adjusting cash flows
for time value of money. Compounding and discounting were used as
appropriate.
5.0 SUMMARY
It is hoped that this unit has made you to be able to apply the present
value technique in financial and investment decision. The calculations
of the present and future values methods are well presented and
explained to enable you apply them practically in business problems.
6.0 TUTOR-MARKED ASSIGNMENTS
(1) KC Bank pays 12 percent and compounds interest quarterly. If

N1,000 is deposited initially, how much shall it grow at the end
of 5 years?
(2) A firm purchases a machinery for N800,000 by making a down

payment of N150,000 and remainder in equal installments of
NI 50,000 for six years. What is the rate of interest of the firm?
7.0 REFERENCES / FURTHER READINGS
Pandey, I.M. (2005). Financial Management, 9th Edition, VICAS

Publishing House Pvt Ltd.

58
MEW 839 QUANTI TATIVE TECIIN IQUES F OR BAN KIN G AN D F INAN CE
MODULE 2
Unit 1 Valuation of Bonds and Shares

Unit 2 Linear Programming I
Unit 3 Linear Programming II
Unit 4 Inventory Control I
Unit 5 Inventory Control II
UNIT! VALUATION OF BONDS AND SHARES
CONTENTS
1.0 Introduction
2.0 Objectives
3.0 Main Content
3.1 Concepts of Value
3.2 Features of a Bond
3.3 Bonds Values and Yields
3.4 Valuation of Preference Shares
3.5 Valuation of Ordinary Shares
3.6 Equity Capitalisation Rate
3.7 Linkages between Share Price, Earnings and Dividends
4.0 Conclusion
5.0 Summary
1.0 INTRODUCTION
In this unit, you shall be led through the explanation of fundamental
characteristics of ordinary shares, preference shares and bonds (or
debentures).
Assets can be real of financial; securities like shares and bonds are
called financial assets, while physical assets like plant and machinery
are called real assets. The concepts of return and r isk, as the
determinants of value, are as fundamental and valid to the valuation of
securities as to that of physical assets. The unpredictable nature of the
security prices is, in fact, a logical and necessary consequence of
efficient capital markets. This will help the financial manager to know
the variables which influence the security prices. The bottom-line is that
it should be appreciated that ordinary shares are riskier than bonds (or
debentures) and also that some shares are more risky than others.
59
2.0 OBJECTIVES
 explain the fundamental characteristics of ordinary shares,

preference shares and bonds (or debentures); earnings and
dividends;
 apply the present value concepts in the valuation of shares and
bonds;
 use price-earnings ratio
3.0 MAIN CONTENT
3.1 Concepts of Value
Earnings per share (EPS) and price-earnings (PIE) ratios are the most
frequently used concepts by the financial commun ity. From our
previous unit on investment, it is emphasized that the present value is
the most valid and true concept of value. There are many other concepts
of value that are used for different purposes. They are named as
follows:
 Book Value, Replacement Value, Liquidity Value, Going

Concern Value, Market Value
3.2 Features of a Bond

A bond is a long-term debt instrument or security. Government bonds
do not have any risk default as it is always honoured. The private sector
companies also issue bonds (or debentures) — secured or unsecured.
Their interest is generally fixed and known to investors. The principal
of a redeemable bond or bond with a maturity is payable after a specific
period called maturity period.
The main features of a bond or debenture are: Face Value, Interest Rate,
Maturity, Redemption Value and Market Value.
3.3 Bonds Values and Yields

It is relatively easy to determine the present value of a bond since its
cash flows and the discount rate can be determined. If there is no risk of
default, then there will be no difficulty in estimating the cash flows
associated with a bond. The expected cash flows consist of annual
interest payments plus repayment of principle. The appropriate
capitalization or discount rate would depend upon the risk of the bond.
60
Bonds are classified into three, namely:
 Bo nd w it h mat ur it y;
 Pure Discount Bonds, and
 P e r p et u a l B o nd s .
(a) Bonds with Maturity:
The government and companies issue bonds that specify the interest rate
(coupon) and the maturit y period. The present value of bond
(debenture) is the discount value of its cash flows, that is, the annual
interest payments plus bond's terminal, or maturity value. The discount
rate is the interest rate that investors could earn on bonds with similar
characteristics. By comparing the present value of a bond with its
current market value, it can be determined whether the bond is over-
valued or under-valued.
Example 1 — Value of Bond with Maturity
Suppose an investor is considering the purchase of a five-year NI,000

per value bond, bearing a nominal rate of interest at 7% per annum. The
investor's required rate of return is 8%. What will he be willing to pay
now to purchase the bond if it matures at par?
The investor will receive cash of N70 as interest each year for 5 years
and N1,000 on maturity (i.e. at the end of the fifth year).
The present value can be determined as follows:
Bo = 70, +70, + 70, + 70,+ 70, + 1,000

(1.08)1 (1.08)2 (1.08)3 (1.08)4 (1.08)5 (1.08)5
Observation:
N70 is an annuity for 5 years and N1,000 is received as a lump sum at
the end of the fifth year.
Using the present value tables in appendix, given at the end of this book,
the present value of bond is:
Bo = 70 x 3.993 + 1,000 x 0.681
 279.51 + 681
61
This implies that N1,000 bond is worth N960.51 today if the required
rate of return is 8 percent. The investor would not be willing to pay
more than N960.51 for bond today. Note that N960.51 is a composite of
the present value of interest payments, N279.51 and the present value of
the maturity value N681.00.
Since most bonds will involve payment of an annu9ity (equal interest

payments each year) and principal at maturity, we can use the following
formula to determine the value of a bond.
Bond value Present value of interest + Present value of

maturity value

Bo INT, + I NT ? + INT, + 13,
L(1)-Kd )1 (1+Kd)2 (11-1c)_ (1+1Q)"
INT, + B. (1)
=I (1 f Kd)' ( +1(d)n
- - 1
Note: Bo is the present value of a bond (debenture),

INT, is the amount of interest in period t (from year 1 to n),
Kd is the market interest rate or the bond's required rate of return,
Bn is bond's terminal or maturity value in period in n and
n is the number of years to maturity.
In equation (1), the right-hand side consists of an annuity of interest

payments that are constant (i.e. INT, = INT 2 = INT) over the bond's
life and a final payment on maturity. Thus, an annuity formula can be
used to value interest payments as shown below:
Bo TNT X INT, + _________ + 130 ...... (2)

(1+Kd) I (Kd) Kd(l+Kd):, (1 +1(d)n
Yield to Maturity (YTM) — the measure of a bond's rate of return that

consists both the interest income and any capital gain or loss. YTM is
bond's internal rate of return. The yield-to-maturity of 5 year bond,
paying 6 percent interest on the face value of N1,000 and currently
selling for N883.40 is 10 percent as shown below:
883A = 60, + 60, + 60, + 60, + 60+ 1,000

(1+YTM) I (1+YTM) 2 (1+YTM) 3 (1+YTM) 4 (1+YTM) 5
It is, however, simpler to calculate a perpetual bond's yield-to-maturity.

It is equal to interest income divided by the bond's price.
62
For Example:
If the rate of interest on N1,000 per value perpetual bond is 8 percent,

and its price is N800, its YTM will be:
n=co
130 INT = INT
-1 (14-ICV (1Cd)
Kd INT 80 0.10 = 10% (3)

800
Bonds Value and Semi - Annual
In practice, some companies pay interest on bonds (or debentures) semi-

annually. The formula for bond valuation can be modified in terms of
half-yearly interest payments and compounding periods as given below
— ref. equation (1) giving it semi-annual approach.
2xn
BO 1/2(INTtl = lin
=1 (1+L)t + &ix o
2 2
20
Bo 60 1 000
,
=I (1.06)t (1.06)2°
60 x Annuity factor (6%, 20) + 1000 x PV factor (6%, 20)
60x 11.4699+ 1,000 x 0.3118
688.20 + 311.80
N1,000.00
(b) Pure Discount Bonds:
Pure Discount Bonds do not carry an explicit rate of interest. It provides

for the payment of a lump sum amount at a future date in exchange for
the current price of the bond. The difference between the face value of
the bond and its purchase price gives the return or YTM to the investor.
For Example:
A company may issue a pure discount bond of N1,000 face value for
N520 today for a period of five years. Thus, the debenture has:
63
MBE 839 QUANTITATIVE Turf INIQUES FOR BANKING AND FINANCE
(a) purchase price of N520;

(b) maturity value (equal to the face value) of N1,000 of five
years
The rate of interest can be calculated as follows:
520 = 1,000
(1+YTM)5
(1 + YTM)5 = 1,000 = 1.9231

520
1.923l"-1 = 0.14 or 14%
( c ) P e rp e t u a l B o n d s :
Perpetual bonds also known as consols, has an indefinite life and

therefore, it has no maturity value. Perpetual bonds or debentures are
rarely found in practice. In the case of the perpetual bonds, as there is
no maturity, or terminal value, the value of the bonds would simply be
the discounted value of the infinite stream of interest flows.
Example:
Suppose that a 10 percent NI,000 bond will pay N100 annual interest
into perpetuity? What will be the value of the bond if the market yield
or interest rate were 15 percent?
The value of bond is determined as follows:
Bo = INT = 100 = N667

Kd 0.15
From the above example, calculate the value of the bond if the yield or
interest rate were 5%, 10%, 20%, 25% and 30%.
3.4 Valuation of Preference Shares
A company may issue two types of shares: ordinary shares and

preference shares. Owners of shares are called shareholders, and capital
contributed by them is called share capital.
64
Preference shares have preference over ordinary shares in terms of

payment of dividend and repayment of capital if the company is wound
up. They may be issued with or without a maturity period.
 Redeemable preference shares are shares with maturity.

 Irredeemable preference shares are shares without any maturity.
The holders of preference shares get dividends at a fixed rate. With

regard to dividends, preference shares may be issued with or without
cumulative features. In the case of cumulative preference shares, unpaid
dividends accumulate and are payable in the future. Dividends in
arrears do not accumulate in the case of non-cumulative preference
shares.
Features of preference and ordinary shares include claims, dividends,

redemption and conversion.
Example:
Suppose an investor is considering the purchase of 12-year, 10 percent

N100 par value preference share. The redemption value of the
preference share on maturity is N120. The investor's required rate of
return is 10.5 percent. What should she be willing to pay for the share
now?
The investor would expect to receive NIO as preference dividend each

year for 12 years and N110 on maturity (i.e. at the end of 12 years).
The present value annuity factor can be used to value the constant
stream of preference dividends and the present value factor to value the
redemption payment.
r --,
Po , 10 x 1— 1 + 120
0.105 0.105 x (1.05)12 (1.105)12
‘... ..."
10 x 6.506 + 120 x 0.302
-= 65.06 + 36.24 = N101.30
Note that the present value of N101.30 is a composite of the present

value of dividend, N65.06 and the present value of the redemption value
N36.24. The N100 preference share is worth N101.30 today at 10.5
percent required rate of return. The investor would be better off by
purchasing the share for N100 today.
65
MBI, 839 QUANTITATIVE TECIENIQUES FOR BANKING AND FINANCE
A formula similar to the valuation of bond can be used to value

preference shares with a maturity period:
Value of preference similar to the valuation of bond can be used to value

preference shares with a maturity period.
Po = PDIV I + PD1V 9 + + PD1Vt, + P

(1+Kp)T (l +K7 ( 1 4-1(p
)11
(1+Kpf
n
i.e. Po = / PDIVI + P_ ........................... (4)
t =1 (1 + Kp)t
t
( 1 + K p )
PD1Vt is the preference dividend per share in period t, K p the required

rate of return of preference share and P n the value of the preference
share on maturity. Since PD1V is an annuity, the equation (4) can also
be written as follows:
r 
Po = PD1V x 1 — 1 +p
(Kr) n
ICp(1+1Cp) , 1
(1+Kp) "
Valuing Irredeemable Preference Share:
Consider that a company has issued N100 irredeemable preference share

on which it pays a dividend of N9. Assume that this type of preference
share is currently yielding a dividend of 11 percent. What is the value of
the preference share?
The preference dividend of N9 is perpetuity. Therefore, the present

value of the preference share is:
Po = PD1V = 9 = N81.82
Kp 0.11
Yields on Preference Share:
If the price of the preference share is N81.82, what return do investors

require? In that case, we will have to solve the following equation:
81.82 9
K P
Kp= 9 = 0.11 or 11 percent.

81.82
66
3.5 Valuation of Ordinary Shares

The valuation of ordinary or equity shares is relatively more difficult
because of two factors:
( 1) T he r at e o f divide nd o n equit y s har es is not kno wn

(discretionary);
( 2) The earnings and dividends on equity shares are generally
expected to grow.
Dividend:
The general principle of valuation applies to the share valuation. The
value of a share today depends on cash inflows expected by investors
and risk associated with those cash inflows (dividend).
Single Period Valuation
Example:
Assuming: an investor intends to buy a share and will hold it for one
year. Suppose he expects the share to pay a dividend of N2 next year
and will sell the share at an expected price_of N21 at the end of the year.
If the investor's opportunity cost of capital or the required rate of return
(Ks) is 15 percent, how much should he pay for the share today?
The present value of the expected dividend per share at the end of the
first year, D1V1, plus the present value of the expected price of the share
after a year, PI,
Po = DIATI + PI (6)
1 + Ks
Po = 2+21 N20
1.15
Note: In practice, there could be a difference between the present value

and the market price less than the share's present value. On the other
hand, an over-valued share has a market price higher than the share's
present value.
It may be seen in the example that theshare value after a year represents
an expected growth or capital gain of 5 percent:
21—20 = 0.05 or 5 percent
20
Pi — Po
Po
67
An investor can, thus, represent his expectation with regard to the future
share price in terms of expected growth. If the share price is expected to
grow at g percent, then we can write PI as follows:
P1 Po (1 + g)
Can be re-written equation (5) as:
Po = DIV, + P0(1 + g) (7 )
I + K,
Simplify Equation (7), we obtain a simple formula for the share

valuation as follows:
Po (8)
Ke — g
In words, the present value of a share is determined by its expected

dividend discounted (divided) by the difference of the shareholders
capitalization or required rate of return (K,) and growth rate (g). In the
example, if the investor would have expected share price to grow at 5
percent, the value of the share today using equation (8) will be:
Po = 2 2= N20
0.15 —0.10 0.10
3.6 Equity Rate
One must know the expected dividend and the required rate of return
(the opportunity cost of capital or capitalization rate) — will depend upon
the risk of the share.
In a well-functioning capital market, the market price is fair price of a

share. Therefore, the shareholders expect the share to earn a minimum
return that keeps the current share price intact. For firms for which
dividends are expected to grow at a constant rate indefinitely, and the
current market price is given, Equation (9) is modified to estimate the
capitalization or the required rate of return of the share:
Po D1V,
K0 —g
DIV, + g
Po
68
Example:
A company's share is currently selling for N50 per share. It is expected
that a dividend of N3 per share after one year will grow at 8 percent
indefinitely. What is the equity capitalization rate?
The equity capitalization rate is given as follows:
Ke DIV, +g 3 + 0.08
Po 50
0.14 = 14 percent
3.7 Linkages between Share price, Earnings and Dividends

Why do investors buy shares? Do they buy them for dividends or for
capital gain?
Investors may choose between growth shares or income shares. Growth
shares are those which offer greater opportunities for capital gains.
Dividend yield (i.e. dividend per shares as a percentage of the market
price of the shares) on such shares would generally be low since
companies would follow a high retention policy in order to have a high
growth rate.
Income shares, on the other hand, are those that pay higher dividends,
and offer low prospects for capital gains. Because of the high payout
policy followed by companies, their share prices tend to grow at a lower
rate. Dividend yield on income shares would generally be high.
Is there a linkage between the share price and earnings and dividends?
The following is an example:
Suppose a company estimates its earnings per share after a year (EPS) at
N6.67, it follows a policy of paying 100 percent dividend (i.e. its
retention ratio, b is zero). Thus, the company's earnings per share
(EPS), and its earnings and dividends would not grow since it does not
reinvest any earnings.
What would be the price of the company's share if the opportunity cost
of capital were 12 percent?
The following formula suffices:
Po DIV, =EPS, (1 — b)--- 6.67 (1 — 0) = 6.67 = N55.58
Kerb0.12 0.12
69
Note that since retention ratio, b, equals to zero, then D1V 1 = EPS, and
g = rb = 0 and P o is given by the earnings per share divided by the
opportunity cost of capital i.e.
Po EPS,
K,
Suppose that the company would pay a divided of Rs 4 share in the first
year and reinvest the retained earnings (RE) at a rate of return (r = ROE)
of 20 percent. What is the company's payout ratio, retention ratio and
growth rate?
Payout ratio = D1V1 = 4 = 0.6 = 60%
EPS, 6.67
Retention ratio =1 — payout 1 — 0.6 = 0.4 = 40%
Growth rate = Retention ratio x ROE

6xr = 0.4 x 0.2 = 0.08 = 8%
Assuming that the company will follow a constant policy of retaining 40

percent earnings (i.e. payout of 60 percent) at 20 percent rate of return,
then its earnings and dividends will grow perpetually at 8 percent (g = rb
= 0.2 x 0.4). What would be the price of the company's share? It is
calculated as follows:
Po = D1V1 = EPS, (1 —1:) = DIV,

IC,—rb K, — g
6.67 (1 —0.4) = 4.00

0.12— 0.2 x 0.4 0.12 — 0.08
N100.00
You may note that without retention of earnings (b = 0), the company
has no growth (g = 0) and the price of its share is N55.58. But when 8
percent growth is expected (from reinvestment of retained earnings), the
price of the company's share is N100. Thus, the difference — N100 —
N55.58 = N44.42 is the value of growth opportunities.
70
SELF ASSESSMENT EXERCISE I
(i) Calculate the price of a share if EPS = Rs 2.5, b = 0.4, K e = 0.10

and ROE = r = 0.20. What shall be the price if r = =
0.10?
(ii) The Federal Government is proposing to sell a 5-year bond of

N1,000 at 8 percent rate of interest per annum. The bond
amount will be amortised (repaid) equally over its life. If
an investor has a minimum required rate of return of 7 percent,
what is the bond's present value for him?
4 . 0 C O N C LU S I O N
In this unit, you have been taken through the application of the concept
of present value to explain the valuation of bonds shares in quantifiable
manner. The specific positions as they affect the value of bonds, shares
are variably discussed and applied.
5 . 0 S U M M AR Y
In this unit, concept of present value have been applied, to explain the
value of bonds — maturity, yield, payments, pure discount and perpetual.
Valuation of preference shares, valuation of ordinary shares, equity
capitalization rate and linkages between share price, earnings and
dividends.
6 . 0 TU TO R - M AR K ED AS S I G N M EN T
The Chief Executive of a company decides that his company will not
pay any dividends till he survives. His current life expectancy is 20
years. After that time, it is expected that the compan y could pay
dividends of N30 per share indefinitely. At present, the firm could
afford to pay N5 per share forever. The required rate of this company's
shareholders is 10 percent. What is the current value of the share?
What is the cost to each shareholder of the managing director's policy?
7.0 R EFER ENCES /FURTH ER R E AD IN GS
Owen, F. and Jones (1975). Modern Analytical Techniques. Polytech

Pandey, I.M. (2005). Financial Management, 9th Edition Vikas

Publishing House PVT LTD. New Delhi, India
71
UNIT 2 LINEAR PROGRAMMING I

CONTENTS
1.0 Introduction
2.0 Objectives
3.0 Main Content
3.1 Linear Programming (LP) — Allocation, Definition
3.1.1 Expressing LP problems - Limitations
3.2 Graphical LP Solution
3.2.1 Production and Material Limitations
3.2.2 The Valuation of Scarce Resources
4.0 Conclusion
5.0 Summary
7.0 Reference / Further Reading
1.0 INTRODUCTION
Linear programming typically deals with the problem of allocating the
limited resources available to an organisation among competing
activities in the best possible or optimal way, that is, in a way minimizes
the returns from performing them. The problem of resource allocation
arises whenever one must select the level of certain activities that
compete for scarce resources to perform them.
2.0 OBJECTIVES
 Explain the techniques of linear programming (LP);

 Describe the objective functions and the constraint of LP;
 Solve the minimizing problems graphically.
3.0 MAIN CONTENT
3.1 Linear Programming (LP) Allocation, Definition

A variety of business situation exist to which description such as the
following applies; the allocation of production facilities to different
products, product mix, portfolio selection for the allocation of
investment funds, quality control, inspection problems, among others.
72
Linear programming uses a mathematical model to describe the problem

of concern. The term "linear" suggests that all the mathematical
functions in the model are required to be linear functions. The word
"programming" is not synonymous with computer programming.
Rather, it means planning. Thus, linear programming involves the
planning of activities to obtain an optimal result, that is, a result that,
among feasible alternatives, reaches the specified goal best according to
the mathematic model. LP is a mathematical technique concerned with
the allocation of scarce resources.
3.1.1 Expressing LP Problems Limitations—
Before considering the detailed methods of solving LP problems, it is

necessary to be able to express a problem in a standardized manner.
This not only helps the calculations required for a solution but also
ensures that no important element of the problem is over-looked. The
two major factors are:
 T he o bje ct ive s , a nd
 The limit at ions or const raint s
Objectives: The first step in LP is to decide what result is required, i.e.

the objective. This may be to maximize profit or contribution, or
minimize cost or time or some other appropriate measures. Having
decided upon the objective, it is now necessary to state mathematically
the elements involved in achieving this. This is called the objective
function.
Example: 1
A factory can produce two products, A and B. The contributions that can
be obtained from these products are:
A contributes N20 per unit, B contributes N30 per unit and it is required to
maximize contribution.
The objective function for this factory can be expressed as:
Maximise 20x1+ 30x2
where xi = number of units of A produced

and x2 = number of units of B produced
Note: This problem has 2 unknowns, xi and x2 . These are sometimes

known as the decision variables. Only a single objective at the time (in
73
the case, to maximize contribution) can be dealt with in a basic LP

problem.
The objective function is:
Example: 2
A farmer mixes three products to feed his pigs. Feedstuff M
costs 20p per kilo, feedstuff Y costs 40p per kilo and feedstuff
Z costs 55p per kilo. Each feedstuff contributes some
essential part of the pigs' diet and the farmer wishes to feed
Minimise 20x1 + 40x2 55x3
where x, = number of kilos of M

X2 = number of kilos of Y
X3 = number of kilos of Z
Alternatively, if the costs were required in N's, the objective function

could be expressed as follows:
Minimise 0.2x1 + O.4x2 + 0.55x3
Note: This example has three decision variables. The number of

decis io n var iab les can var y fr om tw o to man y hu nd reds. F or
examination purposes, four or five is the maximum that is likely to be
encountered. Linearity has been assumed in both examples above and is
assumed in al those that follow.
Limitations or Constraints
Circumstances always exist which govern the achievement of the

objective. These factors are known as limitations or constraints. The
limitations in any given problem must be clearly identified, quantified,
and expressed mathematically. In a problem concerned with the
allocation of scarce resources restrictions take the form:
• Resources used< Resources available
The resources used must be expressed in linear form and the resources
available form part of the given data.
74
MBF 839 QUANTITA FIVE TECHNIQUES FOR BANKING AND FINANCE
Limitation Examples:
A factory can produce four products and wishes to maximise contribution. It
has an objective function as follows:
Maximise 5.3x1 + 2.7x2 + 6.0x3 +4.1x4
Where xl = number of units of A produced

x2 = number of units of B produced
X3 = number of units of C produced
3(4= number of units of D produced
and the coefficients of the objective function (i.e. 5.5, 2.7, 6.0 and 4.1) are the
contributions per unit of the products.
Therefore, Products
A B C D
Skilled hours 5 3 1 8
Unskilled hours 5 7 4 II
The limitations as regards to labour can be stated as follows:
Skilled: 5x1 • 3x2 • x3 8x4 • 8000
Unskilled: 5x1 • 73(2 • 4x3 + 11 x 4 • 6000
In addition a general limitation applicable to all maximising problems is that it is

not possible to make negative quantities of a product, i.e.
X > 0, x2 > 0, x3 > 0, x4 > 0

Notes:
The resource limitations in this maximising problem follow a typical
pattern being of the less than or equal to type (1;
The formal statement of the non-negativity constraints on the unknowns

(x 1, x2, etc.) has to be made for computer solutions but is normally
inferred when solving by manual means;
This above restriction applies to labour hours. Machine hour restrictions

would be dealt with in a similar fashion.
3.2 Graphical LP Solution

Graphical methods of solving LP problems can only be used for
problems with two unknowns or decision variables. Problems with
75
MI31: 839 QUANTI TATIVE TECHNI QUES FOR BAN KING AND F INAN CE
three or more unknowns must be solved by techniques such as the

simplex method. Graphical methods are the simplest to use and should
be used wherever possible.
(a) Limitations. Graphical methods can deal with any number of

limitations but as each limitation is shown as a line on a graph, a
large number of lines may make the graph difficult to read. This
is rarely a problem in examination questions.
(b) Types of problems and limitations. Both maximization and

minimization problems can be dealt with graphically and the
method can also deal with limitations of the 'greater than or equal
to' (>) type and the 'less than or equal to' (<) type.
(c) Graphical example. The method of solving LP problems

graphically will be described step by step using the following
maximising example as a basis.
Example: 3
A manufacturer produces two products, Klunk and Klick. Klunk has a

contribution of £3 per unit and Klick £4 per unit. The manufacturer wishes to
establish the weekly production plan which maximizes contribution.
Production data are as follows:
Per unit
Machining Labour Material
(Hours) (Hours) (kgs)
Klunk 4 4 1
Klick 2 6 1
Total available per week 100 180 40
Because of a trade agreement, sales of Klunk are limited to a weekly maximum of

20 units and to hojnour an agreement with an old established customer at least 10
units of RHO( must he sold ner week.
Step 1: Formulate the LP model in the standardized manner

described earlier above.
Maximise 3x1+ 4x2
Subject to constraintA 4x1 + 2x2 < 100 (Machining hours constraint)

B 4x1 + 6x2 < 180 (Labour hours constraint)
C x, + x 2 < 40 (Materials constraint)
x, < 20 (Klunk sales constraint)
x2> 10 (Klick sales constraint)
x2? 0
76
where xi = number of units of Klunk

X2= number of units of Klick
Note:
As it is impossible to make negative quantities of the products, it is

necessary formally to state the non-negativity constraint (i.e. xi > 0).
The resource and sales constraints include both types of restrictions (i.e.
> and <).
This is a problem with only two unknowns (i.e. xi and x2 ), it can be

solved graphically. The number of limitations does not exclude a
graphical solution.
Step 2: Draw the axes of the graph which represent the unknowns,
xi and x2 thus:
xi
0 x2
Figure 1
Note: The scales of the axes are best determined when the lines for the
limitations are drawn. Each axis must start at zero and the scale must be
constant (i.e. linear) along the axis but it is not necessary for the scales
on both axes to be the same.
Step 3: Draw each limitation as a separate line on the graph.
Sales Limitations: These normally only affect one of the products at a

time and in Example 1, the sales restrictions were:
xi < 20 and x2> 10

These limitations are drawn on the graph as follows:
The Klunk Sales constraint (i.e. xi <20).
77
MB!' 839 QUANTITATIVE no INIQUES FOR BANKING AND FINANCE
Constraint D
xl
60
50
40
30
20
10
0 10 20 30 40 50 60 70
Figure 2
Note: The horizontal line represents xl = 20 and the batched area below
the line represents the area containing all the values less than 20.
The Klick sales constraint (constraint E) (i.e. x 2 ?_ 10) is now entered

thus:
x,
6 0 -
5 0 -
-
3 0 -
2 0 -
10 -
X2
10 20 30 40 50 60 70
Figure 3
Note: The vertical line represents x 2 = 10 and the hatched area to the
right of the line represents the area containing all values greater than 10
(i.e. x, 10). These two sales limitations can be shown on the same
graph thus:
78
MBE 719 QIIAN TI TATIVE TECHNI QUES FOR BAN KING AND FI NA NCE
xl
61) -
10 20 30 40 50 60
Figure 4
Note: The hatched area represents the area of possible production (i.e.
which does not violate the constraints drawn) and is called the feasible
region. The areas on the graph marked (El) violate one or both of the
constraints.
3.2.1 Production and Material Limitations
In a similar fashion to above, the other restrictions should be drawn on

the graph. Because these restrictions involve BOTH unknowns they
will be sloping lines on the graph and not horizontal or vertical lines like
thc sales restrictions.
The three remaining restrictions are all of the same type and area dealt
with as follows:
The matching constraint, constraint A, 4x 1 i 2x2 :5. 100 is drawn on the

graph as 4x1 + 2x2 = 100.
Therefore, when xi = 0, x2= 50 (i.e. 100 )

2
and when x2 = 0, xi =25 (i.e. 100 )

4
and so a line can be drawn from 25 on the x i axis to 50 on the x 2 axis

thus:
79
MBE 839 QUANTITATIVE TEC1INIQUES FOR BANKING AND FINANCE
xl
60
50
40
30
20
10
. I 1 0 J.:1111:1
C I 2 0 I C"
30 I 1 40 I I 50 60 70 x2
Figure 5
Note: As previously, the batched area represents the area containing the
'less than' values. The other constraints are dealt with in the same
manner.
i.e. The labour constraint B 4x 1 + 6x2 180 is drawn on the graph as 4x 1

+ 6x2 = 180.
Therefore, when xi = 0, x2 = 30 (i.e. 180)

6
and when x 2 = 0, x i = 45 (i.e. 180)

4
and so a line can be drawn from 45 on the x i axis to 30 on the x2 axis.
The materials constraint (C), x i + x2 40

when x, = 0, x 2 = 0, xi = 40
and a line can be drawn from 40 on the xi axis to 40 on the x2 axis.
All of the constraints (sales, production and material), can now be drawn
on a single graph and the resultingfeasiNe region defined.
80
Feasible Region
60
50
40 B
30
20
10 A
0 10 30 40 50 60 70
Figure 6
Note:
(a) The feasible region is the area which does not contravene any of
the restrictions and is therefore the area containing all possible
production plans.
(b) The non-negativity restrictions (i.e. x l 0, x2 0) are

automatically included in the graph because the graph quadrant
used in the earlier figure only shows positive values. It should be
noted that as more restrictions are plotted, the feasible region
usually becomes smaller.
(c) It will noted that the material constraint C (line 40, 40) does not
touch the feasible region. This is an example of a redundant
constraint, i.e. it is non-binding.
Step 4: Now that the feasible region has been defined, it is necessary to
find the point in or on the edge of the feasible region that gives the
maximum contribution which, it will be recalled, is the specific
objective.
This is done by plotting lines representing the objective function and

thereby identifying the point in the feasible region which lies on the
maximum value objective function line that can be drawn. These
objectives function or contribution lines are straight lines representing
different combinations of Klunk and Klick which yield the same
contribution. For example:
 20 units of Klunk and zero units of Klick yield £60 contribution

 12 units of Klunk and 6 units of Klick yield £60 contribution
 8 units of Klunk and 9 units of Klick yield £60 contribution
 zero unit of Klunk and 15 units of Klick yield £60 contribution,
etc.
81
MBE 839 QUANTITATIVE TECIINIQUES FOR RANKING AND FINANCE
Very many other contribution lines could be drawn and if a number of

these lines were drawn on the graph, it would be noticed that:
(a) They are parallel to each other with the same shape, which is
determined by the relative contribution of the products.
(b) The further to the right they are drawn, the higher value of
contribution they represent.
It therefore follows that the contribution line furthest to the right but
still touching the feasible region shows the optimum production plan
to provide the maximum possible contribution, thus:
x2
30
Point X
10 20 30 40 50 XI
Figure 7
Optimum solution at point X: i.e. 15 units of xl and 20 units of x2

yielding a contribution of £125 (i.e. £3 x 15 + 4 x 20).
Notes:
(a) The lines marked I to III are three of the many contribution lines
that could be drawn and represent the following contributions
3x1 + 4x2 = 60
3x1 + 4x2 = 90
3x1 + 4x2 = 125
The contribution line has a slope of 4/3 which is the ratio of

the coefficients of x 1 and x2. The intercept on, say, the x 1
axis is fou nd b y d ivid ing the contr ibution b y the x l
coefficient, and vice versa. The intercepts for line II in figure
..., for example, are as follows:
Intercept on xl axis = contribution = 90 = 30

xl coefficient 3
Intercept on x2 axis = 90 = 22.5
4
82
MBE 839 QU AN TI TA TIVE TE C HN IQU E S F OR BAN KI NG AN D l • I N . \ N I T
(b) Only parts of lines I, II and III are feasible. As we require a

maximise contribution, we are only interested in Point X where
line III touches the feasible region. It will be noted that the
optimum is at a vertex or corner of the feasible region. This is
ALWAYS the case.
(c) Various contribution lines have been drawn on earlier page for
instructional purposes. For examinations, it is sufficient to draw
only the contribution line representing the optimum position, i.e.
in the example above, line III.
(d) The contribution lines are sometimes termed iso-profit lines.
(e) A simple way to check your answer is actually possible is to

insert the values of the unknowns in the constraints and check
whether the constraints are satisfied, e.g. the optimum solution of
example found from figure .... is
xi = 15 units
x2 = 20 units
These values can be inserted into the constraints thus:
Constraint A: (4 x 15) -I (2 x 20) = 100 constraint satisfied, no spare

Constraint B: (4 x 15) + (6 x 20) = 180 constraint satisfied, no spare
Constraint C: (1 x 15) + (1 x 20) = 35 constraint satisfied, 5 below
maximum
Constraint D: Sales of x, = 15 Constraint satisfied, 5 below
maximum
Constraint E: Sales of x2 = 20 Constraint satisfied, 10 above
maximum
It will be noted that the two constraints which intersect at the optimum
vertex (see figure ...) are constraints A and B. These are the only
constraints fully satisfied with no spare values. This is a general rule.
They are known as binding constraints.
Minimization Example
Provided that they only have two unknowns, minimization problems can
also be dealt with by graphical means. The general approach of drawing
the axes with appropriate scales and inserting lines representing the
limitations is the same as for maximizing problems but the following
differences between maximizing and minimizing problems will be
found.
83
(a) Normally in a minimizing problem the limitations are of the

greater than or equal to type so that the feasible region will be
above all or most of the limitations.
(b) The normal objective is to minimize cost so that the objective

function line(s) represent cost and because the objective is to
minimize cost the optimum point will be found from the cost line
further to the left which still touches the feasible region, i.e.
the converse of the method used for maximizing problems.
Example: 4
A manufacturer is to market a new fertilizer which is to be a mixture of

two ingredients A and B. The properties of the two ingredients are:
Ingredients analysis
Bone meal Nitrogen Lime Phosphates Cost/kg
Ingredient 20% 30% 40% 10% 1.2p
A 40% 10% 45% 5% 0.8.
Ingredient
B
It has been decided that:

(a) the fertilizer will be sold in bags containing 100 kgs
(b) it must c onta in at leas t 15% nitr ogen
(c) it must c onta in at leas t 8% phos phate
(d) it must c onta in at leas t 25% b one meal
Solution:
Because of the basic similarity between graphing minimizing and

maximizing problems (except for the two differences mentioned above),
the detailed intermediate steps given in example ... above will be
repeated.
The problem in standardized format is as follows:

Objective Function:
Minimise 12x1 + 8x2 (cost expressed in tenths of a penny) subject to:
Constraints A x, + x 2 = 100 (weight constraint)

0.3x1 + 0.1x2 15 (nitrogen constraint)
C 0.1x 1 + 0.08x 2 8 (phosphates constraint)
0.2x1 + 0.4x2 25 (bone meal constraint)
84
MBI: 839 QUANTITATIVE TECHNIQUES FOR BANKING AND FINANCE
xl ?_ 0 and x2 ?. 0
Where xl kgs of ingredient A
x2 = kgs of ingredient B
Feasible
Cost line
Figure 8 x2 kgs
20 40 60 80 100 120 140 160
Notes on Figure
7.8 above:
Optimum solution at point X i.e. 60 kgs of x l and 40 kgs of x2 giving

a cost of:
12 x 60 + 0.8 x 40 = £1.04 per 100 kilo bag.

a) Only one cost line (12x1 + 8x2 = 1,040 in tenths of a penny) has
been drawn so as to show the optimum position.
b) Each restriction on the graph is labeled A, B, C, D and can be

cross referenced to those stated in the standardized format above.
c) The optimum position in the feasible region is the furthest point

to the left touched by the cost line. This is because it is a
minimizing problem. Note again that the optimum is at a corner
or vertex of the feasible region.
85
Using Simultaneous Equations
An alternative method of finding the solution values of the constraints at

any intersection of the graph, including the optimal vertex, is to solve
the simultaneous equations of the relevant binding constraints. To
illustrate this, the answers to the maximizing problem and the
minimizing problem in Example... will be re -calculated using
simultaneous equations below;
Reworking Example
The two constraints which intersected at optimum were constraints A

and B (see figures ..... and ....) thus:
 C o n s tr a i n t A 4x1 + 2x2 = 100

 C o n s tra i n t B 4x1 + 6x2 = 180
Solving by deducting A from B gives:

4x2 = 80
.*. x2 = 20
and by substitution, x l = 15, which confirms the answer found by

reading off the intercepts of the graph.
Reworking example....
The intersecting constraints were A and C (see figure ....) thus:
Constraint A x1 + x2=100
Constraint C 0.1x1 + 0.05x2 = 8
A — 10C = 0.5x2 = 20
x 2 = 40
and substituting x I = 60, again confirming the results from the graph.
If it was known what constraints appeared at the optimum intersection

before the graph was drawn, then the problem could be solved by
simultaneous equations without having to draw a graph. However, this
is unlikely to occur but it is sometimes useful to calculate the exact
solution values using the equations rather than use the approximate
values obtainable from a graph.
3.2.2 The Valuation of Scarce Resources

The graphing of an LP problem not only provides the optimal answer
but also identifies the binding constraints, alternatively called the
86
limiting factors. For example, in example ... machining hours and

labour hours are the binding constraints, the shortage of which limits
further production and profit. Not all factors in a given problem are
limiting factors. For instance, in example ... constraints C, D and E are
not at their maxima and therefore not scarce or limiting.
It is important management information to value the scarce resources.

These valuations are known as the dual prices or shadow prices and are
derived from the amount of the increase (or decrease) in contribution
that would arise if one more (or one less) unit of scarce resource was
available. Only a scarce resource can have a positive dual price and the
calculated price assumes that there is only a marginal increase or
decrease in the availability of the scarce resource and that all other
factors are held constant.
Finding the Shadow or Dual Prices
There are two methods of calculating these prices — the arithmetic

method and the dual formulation — and both are illustrated using the data
and results from Example ..., reproduced below:
Original Problem:
Maximize 3x1 + 4x2
Subject to:
Constraint A 4x1 + 2x2 100 machining hours

Constraint B 4x 1 + 6x2 180 labour hours
Constraint C + x2 40 materials
Constraint D x l 20 sales
Constraint E x2 10 sales
Where xl = units of klunk

x2 = units of klick
The solution was Produce 15x1 and 20x2 giving a contribution of £125.
Constraints A and B are binding
The problem now is to find the shadow prices of the two binding
constraints, machine hours and labour hours, i.e. what is the valuation of
one more (or less) machine hour and one more (or less) labour hour?
87
MBF 839 QUANTITATIVE TED INIQUES FOR BANKING AND FINANCE
Arithmetic Method of Finding Shadow Prices
Dealing first with machine hours, we assume that 1 more machine hour
is available (but labour hours are constant at 180) and calculate the
resulting difference in contribution, thus:
The binding constraints become:
Machine hours 4x1 ± 2)(2 = 101 (i.e. original 100 + 1)
Labour hours 4x1 + 6x2 = 180 (unchanged)
Solving these simultaneous equations, new values for x l and x, are

obtained: xl = 15.375 and x, = 19.75 and substituting into the objective
function gives a new contribution.
3(15.375) + 4(19.75) = £125.125

original contribution = £125
Difference = £0.125
Thus 1 extra machine hour has resulted in an increase in contribution of

£0.125 which is the shadow price per machining hour.
A similar process for labour hours is shown below:
New constraints with extra labour hour: (but machine hours constant at
100).
Machine hours 4x1 + 2x2 = 100

Labour hours 4x1 + 6x2 = 181
and solving gives xl = 14.875 and x2 = 20.25
New contribution 3(14.875) + 4(20.25) = £125.625

Original contribution = £125
Difference = £0.625
shadow price per labour hour = £0.625
Notes:
(a) Similar results would be obtained in each case if 1 less hour had
been used in the calculations. Verify this yourself
(b) The shadow prices calculated above only apply whilst the
constraint is binding. If, for example more and more machining
ours became available, there would eventually be so many
88
machining hours that they would no longer be scarce and some

other constraint would become binding. This point is developed
further below.
Dual Formulation Method for Shadow Prices
Every LP problem has an inverse or dual formulation. If the original

problem known as the primal problem, is a maximizing one then the
dual formulation is a minimizing one and vice versa. Thus, as the
Example ...primal formulation is a maximizing problem, its dual is the
minimizing problem. Solving the dual problem gives the shadow prices
of the binding constraints, hence the alternative term, dual prices.
The stages in finding and solving the dual for the solution to Example...
are shown below.
The relevant parts of the original fill formulation which appear in the
solution are:
Maximize 3x1 + 4x2 (objective function)
Subject to 4x 1 + 2x 5. 100 (machine hours)

4x1 + 6x2 5 180 (labour hours)
Constraints C, D, E (non-binding so do not appear in the solution).
The minimizing dual problem is formed by inverting the above

formulation, i.e. making the columns into rows and the rows into
columns thus:
Dual formulation:
Minimize 100M + 180L (i.e. originally the constraint column)
Subject to 4M + 4L 3 (originally the x, column)

2M + 6L 4 (originally the x2 column)
What are now the constraints can be solved by simultaneous equations?
4M + 4L = 3
2M + 6L =4
Solving gives M = 0.125 and L = 0.625, which will be recognized as the

valuations already calculated earlier above.
If these dual prices are inserted into the objective function of the dual,
exactly the same value is obtained as in the primal problem.
89
i.e. 100 (0.125) + 180 (0.625) = £,125
This result is identical to the Primal problem and this will always be so.
Notes:
(a) The dual formulation above is a contraction of the full dual

formulation. This is possible here because the problem had
already been solved and it was then known that three of the
constraints (C, D and E) were non-binding.
(b) The dual formulation is developed further in subsequent chapters.
Interpretation of Shadow Prices
The shadow price of a binding constraint provides valuable guidance

because it indicates to management the extra contribution they would
gain from increasing by one unit the amount of the scarce resource. As
an example, the shadow price of labour hours calculated above is £0.625
per hour. This means that management would be prepared to pay up to
£0.625 per hour extra in order to gain more labour hours. If the current
labour cost was £8 per hour then management would be prepared to pay
up to £8.625 per hour for extra labour hours, perhaps from overtime
working.
It is important to realise that only binding constraints have shadow or
dual prices. Those that are not binding have zero shadow prices. This
accord with common sense for there would be little point in paying more
to increase the supply of a resource of which you already have a surplus.
In example ... constraints C, D and E have zero shadow prices.
What is the 'objective function' of linear programming?
4.0 CONCLUSION
In this unit, you have been introduced to linear programming as a

method of solving problems where an objective has to be optimized
subject to constraints and how LP is a resource allocation technique.
90
5.0 SUMMARY
Linear programming is a solution method to problems where an
objective has to be optimized subject to constraints. All factors
concerned have to be numeric and there must be linear relationships.
LP is a resource allocation technique where some objective, for instance,
to maximize contribution, is required to be optimized subject to resource
constraints.
LP problems can be maximizing or minimizing problems.

6.0 TUTOR MARKED ASSIGNMENT
-
A firm produces two products X and Y with a contribution of N8 and

N I 0 per unit respectively. Production data are (per unit):
Labour hours Material A Material B

X 3 4 6
Y 5 2 8
Total available 500 350 800
Formulate the LP model in the standardized way. Solve the model in

the above question using the graphical method.
Lucey, T. (2002). Quantitative Techniques 6th Edition, (LLST)

Educational Low-Priced Sponsored Texts.

Pandey, I.M. (2005). Financial Management, UBS Publisher's

Distributors PVT LTD.
91
UNIT 3 LINEAR PROGRAMMING II

CONTENTS
1.0 Introduction
2.0 Objectives
3.0 Main Content
3.1 Simplex Method
3.2 Mixed Limitations
3.3 Comparing Simplex and Graphical Solutions
4.0 Conclusion
5.0 Summary
7.0 Reference / Further Reading
1.0 INTRODUCTION
When considering linear programming methods, one has so far confined

one's attention to production problems — maximizing profit product —
mix or minimum cost combination for inputs and blending problems.
The time is now to widen the horizons of linear programming techniques
and introduce various techniques which will simplify the method under
certain conditions.
2.0 OBJECTIVES
 use the simplex method for solving maximizing LP problems;

 in ter p r et t he s imp lex tab leau .
3.0 MAIN CONTENT
3.1 Simplex Method
A step by step arithmetic method of solving LP problems whereby one

moves progressively from a position of, say, zero production, until no
further contribution can be made. Each step produces a feasible solution
and each step produces an answer better than the one before, i.e. either
greater contribution in maximizing problems, or less cost in minimising
problems. The mathematics behind the Simplex method are complex
and this unit does not try to explain why the method works but it does
describe how to use the technique.
92
Formulating the Simplex Model
To use the Simplex method, it is first necessary to state the problem in

the standardised manner. It will be recalled that this results in an
objective function and a number of constraints which are inequalities
either of the or < type. Having stated the problem in the standardised
format, the inequalities must be converted to equations. For example,
if a manufacturer made two products A and B, which took 3 and 5 hours
respectively to machine on a drilling machine which was available for
up to 320 hours per period the constraint would be written in the
standardised format as follows:
3x1 + 5x2 < 320
where xl units of A
X2 units of B
This is converted into an equation by adding an extra variable called a

slack variable thus
3x1 + 5x2 + x3 = 320
The slack variable, x3 in this case, represents any unused capacity in the
constraint and can thus take any value from 320 hours (i.e. the position
of zero production and therefore maximum unused capacity), to 0 hours
(i.e. the position of the machine being fully utilised and therefore, zero
unused capacity).
Notes:
(a) Each constraint will have its own slack variable;
(b) Once the slack variable has been incorporated into the constraint
the Simplex method automatically assigns it an appropriate value
at each iteration.
A Simplex Maximizing Example
The following maximizing example will be used to provide a step by

step approach to the Simplex method.
93
Example: 1
A company can produce three products, A, B and C. The

products yield a contribution of £8, £5 and £10 respectively. The
products use a machine which has 400 hours capacity in the next
period.
Each unit of the products uses 2, 3 and 1 hour respectively of the

machine's capacity. There are only 150 units available in the
period of a special component which is used singly in products A and
C.
200 kgs only of a special alloy is available in the period. Product A

uses 2 kgs per unit and Product C uses 4 kgs per units. There is an
agreement with a trade association to produce no more than 50 units of
Product B in the period.
Step 1: Express the problem in the standardised format thus:
Objective function:
maximise 8x1 + 5x2 + 10x3

subject to 2x1 + 3x2 + X3 < 400
x, + X3 < 150
2x1 + 4x3 < 200
X2 < 50
xj 0, x2 0, x3 0
= where x1 = no. of units of Product A

X2 = no. of units of Product B
X3 = no. of units of Product C
Step 2: Make the inequalities in the constraints into equalities by

adding a 'slack variable' in each constraint, thus:
maximize 8x 1 + 5x 2 + 10x 3
subject to 2x1 + 3x2 + x3 + x4 400
x1 x3 + X5 = 150
2x1 4x3 + X6 = 200
X2 ± X7 = 50
Note: x4 , x5 , x6 , x7 are the slack variables and represent the spare

capacity in the limitations.
94
Step 3: Set up the init ial Simplex Tableau by arranging the

objective function and equalized constraints from Step 2 in
the following form.
leau Table 8.1

Solution Products Slack variables Solution
variable xt X2 X3 X4 X5 X6 quantity
X7
X4
2 3 1 1 0 0 0 400
x5 1 0 I 0 1 0 0 150
X6 2 0 4 0 0 1 0 200
x7 0 1 0 0 0 0 1 50
Z 8 5 10 0 0 Q 0 0
Notes:
(a) It will be seen that the values in the body of the table are the
values from the objective function and constraints in Step 2.
(b) The variable 'Z' has been used for the objective function and
represents total contribution.
(c) The tableau shows that x.3 = 400, x6 = 150, x6 = 200, x7 = 50 and
Z = 0.
(d) The tableau shows a feasible solution, that of nil production, nil
contribution, and maximum unused capacity as represented by
the values of the slack variables x4, x5, x5 and x7.
Although feasible, this plan can obviously be improved and this

is done as follows:
Step 4: Improve the previous feasible solution by making as many

as possible of the product with the most contribution, i.e.
the highest figure in the Z row. The number that can be
made will be limited by one or more of the constraints
becoming operative thus:
Select highest contribution in Z row — i.e. 10 under x3.

Divide the positive numbers in the x3 column into the solution quantity
column.
i.e. 400 + 1 = 400
150 + 1 = 150
200 4 = 50
50 + 0 (ignore)
95
MBE 839 Q UANT ITAT IVE T ECHNIQUES FOR BANKING AND FINANCE
Select the row that gives the lowest answer (in this case the row
identified x 6). Ring the element which appears in both the identified
column (x3) and the identified row (x 6), this element is known as the
pivot element thus:
Initial Simplex tableau reproduced, Table 2

Variable X1
X2
X4
X5 X6 quantity
x3 x7
x4 2 31 0 0 400
xs 1 0 150
I 0 0 I 0 200
x7 1 40 50
2 0 0 0 1
0 1 0
0 0 0 0
1
Z 8 50 0 0 0
10 0
Step 5: Divide all the elements in the identified row (x 6) by the

value of the pivot element (4) and change the solution
variable to the heading of the identified column (x 3) thus:
Old row
X6 2 0 4 0 0 1 0 200
Divide by 4 and re-titling the new row becomes
New row
I
X3 A 0 1 0 0
/4
I
0 50
Enter this row in a new tableau
Second Simplex tableau Table 3

Row Solution Products Slack variables Solution
Nos. variable xi X2 X3 X4 X5 X6 X7 quantity
2 3 1 1 0 0 0 400
- N M
x4
x5 1 0 1 0 1 0 0 150
x6 1/2 0 1 0 0 IA 0 50
x7 0 1 0 0 0 0 1 50
5 Z 8 5 10 0 0 0 0 0
96
Notes:
(a) The row numbers have been included to aid the explanatory
material which follows and form no part of the Simplex Tableau.
(b) It will be seen that this second tableau is identical to the first
except for row 3 which was calculated above.
(c) Row 3 means that 50 units of x 3 are to be produced.

Step 6: As a consequence of producing 50 units of x3, it is
necessary to adjust the other row so as to take up the
appropriate number of hours, components etc. used and to
show the contribution produced for the 50 units of x3.
This is done by repetitive row by row operations using
Row 3 which makes all the other elements in the pivot
elements column into zeros. To maintain each row as
equality, it is, of course, necessary to alter each element
along the row on both sides of the equality sign. Using
Row 1 in the Second Tableau as an example, the process is
as follows:
Row 1 x4 2 3 1 1 0 0 0
400
Minus Row 3 x3 1/2 0 1 0 0 1/4 0 50
Produces a
new row xi: 11/4 01 0 0 —1/4 0 350
Notes:
(a) This new row will be inserted into a third tableau along with all
the other altered rows and Row 3 from the second tableau.
(b) The aim of this row operation was to produce the zero (marked
*). In this case, a simple subtraction was all that was necessary
but to make a zero in other cases may require further operations
using Row 3 as a basis.
The other rows in the second tableau are operated on a similar fashion.
Row 2 xs 1 0 1 0 10 0 150
1 0 0
Minus Row 3 x3 1/2 0 50
Produces a 0
0*
new row x5 V2 0 0 1 0 100
97
Row 4 needs no operation because the element in column x 3 is already

zero.
Row 5 Z 8 5 10 0 0 0 0 =0
Minus Rcnn/ 3x1 5 0 10 0 0 21/2 0 = 500
Produces a new
row Z 11/2 0 1 0 0 —21/2 0 =-500
Notes:
(a) To produce the required zero (0 *), it was necessary to multiply

the Row 3 by 10 and then subtract from Row 5.
(b) The `-500' at the end of the new Z row is the contribution earned
by 50 units of x3 at £10 i.e. £500. The negative sign is merely a
result of the Simplex method and the fact that the contribution is
shown as a negative figure can be disregarded.
Step 7: When all the row operations have been done, a third
tableau can be produced thus:
Third Simplex tableau, Table 4

Nos. variable xl x2 x3 x4 x5 x6 x7 quantity
6 (Row 1 — Row 3) x4 11/2 3 0 1 0 —1/4 0 350
7 ( Row 2 — Row 3) xs 1/2 0 0 0 1 —1/4 0 100
8 (i.e. Row 3) x3 1/2 0 1 0 0 1/4 0 50
9 (i.e. Row 4) x7 0 I 0 0 0 0 1 50
10 (Row 5 — 10x Row 3) Z 3 5 0 0 0 —21/2 0 —500
Notes:
(a) All the new rows produced by the row operations in Step 6 have
been inserted into the third tableau.
(b) The rows have been consecutively numbered again and a

summary of the operations carried out in Step 6 to produce the
new lines has been given against the new row numbers. e.g.
Row 10 was produced by multiplying Row 3 by 10 and
subtracting it from Row 5.
Step 8: To pr odu ce subs equent tab leau x and eventu ally an

optimum solution, steps 4 to 7 are repeated until no
positive numbers can be found in the Z row. From Row
10 it will be seen that the maximum contribution is 5.
x2 column is chosen
98
The positive numbers in the x2 column are divided into the solution
quantities and the lowest result selected.
i.e. Row 6 350 ÷ 3 = 116%
Row 7 100 ± 0 = ignore
Row 8 50 + 0 = ignore
Row 9 50 ÷ 1 = 50
Row 9 is selected and the pivot element identified and the

solution variable altered to x2 thus
Row 9 x2 0 1 0 0 0 1 = 50
As the pivot element is already 1 no further action is necessary on it but
the other elements in the pivot element column (x2) must be made into
zeros by using row operations based on Row 9 thus
Row 6 x4 1% 3 0 1 0 — 1/4 0 = 350
Minus 3 x
Row 9 x, 0 3 0 0 0 0 3 = 150
Produces a
new row x4 1¼ 0* 0 1 0 _IA —3 = 200
Note:
It was necessary to multiply Row 9 by 3 to produce the zero in column

x2 (0*) Rows 7 and 8 need no operation because the elements in column
x2 are already zero.
Row 6 Z 3 5 0 0 0 — 21/2 0 = 500
—
Minus 5 x
Row 9 x2 0 5 0 0 0 a 5 = 250
Produces a
0*
new row x4 3 0 1 0 — 21/2 — 5 = 750
—
The new row produced can now be entered into a fourth tableau as
follows:
Fourth Simplex tableau, Table 8. 5

Nos. variable xl x2 x, x4 x5 x6 x2 quantity
11 (Row 6 —3x Row 9) x4 11/2 3 0 1 0 —1/4 —3 200
12 (asRow 7) x5 1/2 0 0 0 1 —1/4 0 100
13 (i.e. Row 8) x, /2
1
0 1 0 0 1/4 0 50
14 (Pivot Row as Row 9) x2 0 1 0 0 0 0 I 50
15 (Row 10 — 5x Row 9) Z 3 0 0 0 0 —21/2 —5 —750
99
Notes:
(a) The above tableau shows that 50 units of Products B and C could
be made (x2 and x3= 50).
(b) As a result of this amount of production £750 contribution would

be gained (Z = —750.
Step 9: Because there is still a positive number in the Z row (3

under column x i ) the iterative process is repeated in
precisely the same manner. Column x l is chosen and the
positive numbers in the x l column are divided into the
solution quantities and the lowest number selected.
i.e. Row 11 200 ÷ 11/4 = 1331/4

Row 12 100 + 1/4 = 200
Row 8 50 + 'A = 100
Row 9 50 ± 0 = ignore
Row 13 is selected and the pivot element identified and the
solution variable altered to x l thus
Row 13 x l 0 0 /2
1
0 0 0 = 50
The pivot element must be made into a 1 so the whole row is multiplied
by 2 thus
Row 13 x l 0 2 1 0 1/2 100
The rest of the elements in column xl must now be made into zeros by
the usual row operations.
Row 11 x4
Minus 1 'A x 11/2 0 0 1 0 0 = 200
11/2 0 3 0 0 3 = 150
Row 13 xi %
Produces a 0 1 0
new row x4 0 * —3 —1 —3 =
50
0 0 1
Row 12 x5 1/2 0 100
100
M131: 839 QUANTITATIVE TECHNIQUES FOR BANKING AND FINANCE
Minus 54 x
Row 13 xi % 0 1 0 00 = 50
Produces a
n ew r ow x6 0* 0 —1 0 10 = 50
Row 14 needs no operation because the element in column x l is already

zero.
Row 15 Z 3 0 0 0 0 —21/4 —5 = —750
Minus 54 x
Row 13 xi 3 0 6 0 0 —11/2 0 = 300
Produces a
new row x5 0* 0 —6 0 0 —4 —5 = —1050
The new rows produced can now be entered into a fifth tableau thus
tableau Table 6
Row Nos. variable X, x2 xs x4 xs x6 x, quantity
16 (Row 11 — 11/2x Row 18) x4 0 0 —3 1 0 —1 —3 50
17 (Row 12 — 1/2 x Row 18) xs 0 0 —1 0 1 —1/2 0 50
18 (Pivot Row 2 x Row 13) X3 1 0 2 0 0 1/2 0 100
19 (as Row 14) x2 0 1 0 0 0 0 1 50
20 (Row 15 — 3x Row 13) Z 0 0 —6 0 0 —4 —5 —1050
As there are no positive values in the Z row the optimum solution has
been reached.
Step 10: All that remains is to obtain the maximum information

from the fifth tableau.
Dealing first with the solution variables: xl, x2, x4 and x5,
Optimum product mix
XI 100 i.e. produce 100 units of Product A

X2 50 i.e. produce 50 units of Product B
Value of Slack Variables
50 i.e. there are 50 machine hours unused at Optimum
X5 50 i.e. there are 50 components unused at Optimum
101
Note:
It will be seen that the other two slack variables, x 6 and x7, do not have
values in the Solution Quantity column. Their values are both zero,
which means x 6 and x7 , representing the alloy and sales constraints
respectively, have no unused capacity at optimum and the constraints
they represent are fully utilised.
Contribution and Resource Valuations
It will be seen from Row 20 in Table 6 that Z has the Solution Quantity
of —1,050. This means, at optimum, the maximum contribution is
£1,050 (this can be confirmed by calculating the contributions of the
optimum product mix, i.e. (100A and 50B = (100 x 8) + (50 x 5) =
1,050).
The value of —6 under Product x 3 , the product that was not in the
optimum plan, means that if any unit of x 3 was produced then overall
contribution would fall by £6.
The values of Row 20 for the slack variables are of great importance.
These are the valuations of resources and are known as shadow prices.
These have the following meanings:
X4 = 0 i.e. there are no value to be gained from increasing
machine hours
x5 = 0 i.e. there are no value to be gained from increasing

components
x6 = —4 i.e. for every extra kilo of alloy available £4 extra

overall contribution would be gained
X7 = —5 i.e. for every extra unit of B that was allowed to be

produced contribution would increase by £5
It will be seen that only the constraints x6 and x7 that are binding, i.e. are
fully utilised, have non-zero shadow prices. This is a general rule, for
there would be no value in increasing the availability of a resource
already in surplus. Thus, in this example, machine hours and the supply
of components have shadow prices of zero.
The shadow prices of the binding constraints can be used to confirm

again the overall contribution thus:
102
Alloy availability 200kg x £4 = £800

Sales constraint 50 units x £5 = £250 •
£1050
When solving LP problems by graphical means, the shadow prices have

to be calculated separately. When using the Simplex process they are an
automatic by product.
Note:
(a) Each variable in the final solution variable column has a specific
meaning which is detailed above.
(b) Using step 10 as a guide interpret the meanings of the solution

variables an the valuations in the Z row of the intermediate
tableaux.
(c) Alternative names for shadow prices are: shadow costs, dual
prices or simplex multipliers.
(d) Take heart. To work through a normal Simplex problem with 3

unknowns is a very quick process once the foregoing steps are
mastered.
3 . 2 M ix ed Limitatio ns
The maximizing example given above had constraints all of which were
of the 'less than or equal to' type (<). This is a common situation but on
occasions the constraints contain a mixture of < and varieties. The
usual cause of one or more 'greater than or equal to' (?) constraints is
the requirement to produce at least a given number of certain products.
In such circumstances, the simplest approach is to reduce the capacity of

the other limitations by the amounts required to make the required
number of the product(s) specified. Then maximize in the normal way
and add back the quantities which were required to be produced, to the
optimum solution found by the normal Simplex method. The method is
shown in the following example.
Example Contained Mixed Constraints
Assume that an LP problem had been set up in the usual standardised

format as follows:
103
MI3F 839 QUANTITATIVE TECIINIQUES FOR BANKING AND FINANCE
Example: 2
Objective function
Maximize 5x1 + 3x2 + 4x3

Subject t o a. 3x, + 12x2 + 6x3 < 660 (machine hours constraints)
b. 6x1+ 6x2 + 3x3 < 1,230 (labour hours constraints)
c. 6x1 + 9x2 + 9x3 < 990 (component constraints)
d. X3 > 10 (sales constraints)
where xl, x2 and x3 represent units of Products A, B and C
Only one of the constraints d. is of the variety so we decide to make the minimum quantity possible
to satisfy constraint d. i.e. 10 units of x 3. The resource requirements for 10 units of x3 must be
subtracted from the total available in constraints a, b and c thus:
Constraints a. 3x1 + 12x2+ 6x3 < 660 —(6 x 10)

b. 6x1+ 6x, + 3x3 < 1,230 — (3 x 10)
c. 6x1+ 9x, + 9x3 < 990 — (9 x 10)
Notes:
(a) In each case the expression in the bracket represents the

coefficient of x 3 in each constraint multiplied by the 10 units
being made.
(b) An alternative way of dealing with the constraint is to multiply

both sides by —1 and change the inequality sign.
(x3 10) x —1
i.e. —x3 —10
This constraint can then be used in the normal manner.
Having eliminated constraint D and made the appropriate reductions in

the other constraints, the initial Simplex Tableau can be set up.
Initial Simplex tableau Table 8.7

Modified Solution Products Slack variables Solution
Constraint, variable xl x2 x3 x4 x5 x6 quantity
A x4 3 12 6 1 0 0 600
B xs 6 6 3 0 1 0 1,200
C X6 6 9 9 0 0 1 900
Z 5 3 4 0 0 0 0
104
Note:
This is now a normal maximizing Simplex model with a slack variable

for each constraint and is solved by exactly the same process covered in
the earlier part of the chapter.
After carrying out the usual Simplex procedure, the final tableau
becomes:
Final Sim lex tableau Table 8

variable x, x3 X4 Xs X6 quantity
xt 0 7% 1% 1 0 —% 150
x5 0 —3 —6 0 1 —1 300
X6 1 11/4 N 0 0 1/6 150
Z 0 0 —6 0 0 —5/6 —750
Solution from final tableau —150 units of xl producing £750 contribution

plus production to satisfy constraint d. 20 units of x 3 producing £40
contribution.
Therefore total solution is 150 units of xl and 10 units of x3 giving £790

contribution.
As previously described in Step 10, the tableau can be interpreted as

follows:
Spare capacity x4 = 150 means that there are 150 spare machining
hours
xs = 300 means that there are 300 spare labour hours
Products Not Being Made
From the final Tableau table it will be seen that x2 and x3 do not appear
in the Simplex solution and they have valuations in the Z row of —41/2
and —31/2 respectively. We already know that 10 units of x3 will be made
and the tableau informs us that the 10 units of x3 which we have had to
make have cost £35 in reduced contribution. Any units of x2 that were
made would similarly reduce contribution by £4.50 per unit.
Shadow Prices
Only one constraint (constraint C, components) is fully utilised and thus

has a non-zero shadow price. The tableau shows that for every extra
component that could be obtained, contribution would be increased by
£216.
105
As previously, the shadow price of the binding constraint can be used to

confirm the overall contribution as follows:
Modified component constraint = 900 x £216 = £750.
An alternative method of dealing with mixed limitations uses what are

termed artificial variables as well as slack variables. This method is
outside the scope of this course material.
3.3 Comparing Simplex and Graphical Solutions
The Simplex method can be used for problems with any number of
unknowns — even those with only two unknowns that can also be solved
graphically. To illustrate both solution methods, example is shown
below using the Simplex method (graphically).
Example reproduced from Unit 2
Maximise 3x1 + 4x2
4x1 + 2x2 100 A

4x 1 + 6x 2 180
x 1 + x 2 40
xi
<20
x2 10
This problem is inserted into the initial tableau with a slack variable for
each of the five constraints and with constraint E multiplied by —1 to
reverse the inequality sign.
Initial Simplex tableau Table 9

variable xi X2 X2 X4 X5 X6 X2 quantity
'7 '7
X1 1 0 0 0 0 100
1/4.0
X4 0 I 0 0 0 180
x5 0 0 1 0 0 40
--0
x6 0 0 0 1 0 20
x7 0 0 0 0 I —10
Z 3 40 0 0
This problem is then solved by the usual Simplex iterations. Each

improves on the one before and the process continues until optimum is
reached. The following tables show the position after each one is cross
referenced to figure ... which is a reproduction of the graphical solution.
106
First iteration equivalent to point 1 on Table 10

variable xi X2 X3 X4 X5 X6 X7 quantity
0 'Cr `71" •-•

X4 0 0 0 0 —1 10
0
x4 1 0 0 0 2 80
0
x5 0 1 0 0 6 120
0
x5 0 0 1 0 1 30
0
x7 0 0 0 1 0 20
Z 3 0 0 0 0 0 4 —40
This shows 10x2 being produced and £40 contribution. The first four
constraints have surpluses of 80, 120, 30 and 20 respectively. Not
optimum, as there are still positive values in Z row.
Second iteration equivalent to point 2 on Table 11

variable xl X2 X3 X4 X5 X6 X7 quantity
x4 0.667 1 0 0.167 0 0 0 30
X4 2.667 0 1 —0.333 0 0 0 40
x5 —0.333 0 0 —0.167 0 0 0 10
X6 1 0 0 0 1 1 0 20
x7 0.333 0 0 0.167 0 0 1 20
Z 0.333 0 0 —0.667 0 0 0 —120
This shows 30x2 being produced and £120 contribution. All constraints
have surpluses except labour hours. Not optimum as there is a positive
value in Z row.
Third iteration equivalent to point 3 on Table 12

Solution Products Slack variables - Solution
variable x1 X2 X3 X4 xs x6 x7 quantity
X4 0.375 0.125 0 0 0 15
0 0 0 0
-
X4 —0.25 0.25 0 0 0 20
0
X 5 1 0 0 5
0
—0.125---0.125
—0.375 0125
0
X6 0 1 0 5
0
X7 —0.25 0.25 0 0 I 10
Z 0 0 —0.125-0.625 0 0 0 —125
This is the optimum and shows all the information obtained previously.
107
Point 2 (Table I
Point 3 optimum (Table 12)
10 20
30 Point 1 XI
(Table 10
Point 0
(Cabin9)
Figure 1 Graphical/Simplex comparison
Solution:
Optimum production:
 xl = 15 i.e. produce 13 units of Klunk

 x 2 = 20 i.e. produce 20 units of Klick giving an overall
contribution of £ 125
Shadow prices of binding constraints:
x3 = £0.125 i.e. each additional machine hour would produce

£0.125 extra contribution labour.
X4 £0.625 i.e. each additional labour hour would produce

£0.625 extra contribution labour.
The non-binding constraints represented by x 5 , x6 and x2 have 5, 5 and

10 spare respectively.
Note:
By seeing how the Simplex interactions move from one vertex to an

improved vertex on the feasible region, it is possible to gain a better
understanding of the process.
Summary of Simplex method
Flow chart 1 (on page ....) provides a concise summary of the Simplex
method and is cross referenced to key points in the Unit.
108
(i) What is the dual of a minimizing LP problem?

(ii) How is the final Simplex Tableau interpreted?
4.0 CONCLUSION
This unit introduces special techniques for linear programming e.g.
transformation in a simple approach.
5.0 SUMMARY
The simplex flow chart method shows the steps of maximizing. The
simple transformation given may be applied to al minimization problems
with all constraints of the greater than or equal type. The simplex
method can be applied to all combinations but only a restricted range of
problem is covered.

-
Form the dual of the following problem:
Minimise: 30A + 608 + 20C
Subject to: 5A + 10B + 15C > 2,000

2A + 38 + IC > 300
8A + 6B + 4C > 650
Lucey. T. (2002). Quantitative Techniques 6th Edition, (LLST)

Owen. F. and Jones. R. (1975). Modern Analytical Techniques, Polytech


109
UNIT 4 INVENTORY CONTROL I
CONTENTS
1.0 Introduction
2.0 Objectives
3.0 Main Content
3.1 Types of Inventory
3.2 Reasons for holding Stocks
3.3 Inventory Management Techniques
3.4 Stock Costs
3.4.1 Cost of holding stock
3.4.2 Cost of obtaining stock
3.4.3 Stock out costs
3.4.4 Cost of stock
3.5 Types of Control System
3.5.1 Re-order level system
3.5.2 Periodic review system
4.0 Conclusion
5.0 Summary
1.0 INTRODUCTION
This unit will draw from the business world that irrespective of the
nature of a firm, a businessman will need to hold stock. Retailers need
stocks of components, raw materials and spare parts of machinery if
production is to be continuous. A manufacturer may also need stocks of
finished goods to satisfy his customers. Even an Accountant must hold
stocks of stationery!
You will be introduced to different types of cost involved in stock

(inventory) management like ordering costs, set-up cost, holding cost,
stock-out cost, etc. Thus, you will see the need that a policy to minimize
inventory costs is a rational objective.
2.0 OBJECTIVES
 define the need for and nature of inventor y;

 differentiate the four costs associated with inventory;
 identify inventory control systems and periodic review system
 distinguish the principles of re-order or two bin system;
110
3.0 MAIN C ONTENT
3.1 Types / Nature of Inventory
Inventories are stock of the product of a company is manufacturing for

sale and components that make up the product. The various forms of
inventories that exist in a manufacturing company are conveniently
classified into: raw materials, work-in-progress and finished goods.
 Raw materials — the materials, components, fuels, etc, used in the

manufacture of products.
 Work-in-progress (WIP) — partly finished goods and materials,
sub-assemblies, etc. held between manufacturing stages.
 Finished goods — completed products ready for sale or
distribution.
The levels of three kinds of inventories for a firm depend on the nature
of business. What would be classified as a finished product for one
company might be classified as raw material for another. For example,
steel bars would be classified as a finished product for a steel mil and
raw material for a nut and bolt manufacturer.
3.2 Reasons For Holding Stocks
The main reasons for holding stocks can be summarized as follows:
(a) to ensure that sufficient goods are available to meet anticipated

demand;
(b) to absorb variations in demand and production;
(c) to provide a buffer between production processes;
(d) to take advantage of bulk purchase discounts;
(e) to meet possible shortages in the future;
(0 to absorb seasonal fluctuations in usage or demand;
(g) to enable production processes to flow smoothly and efficiently;
(h) as a necessary part of the production process;
(i) as a deliberate investment policy particularly during inflation or
possible shortage.
3.3 Inventory Management Techniques

In managing inventory, the firm's objective should be in consonance
with the shareholder wealth maximization principle. To achieve this,
the firm should determine the optimum level of inventory. Efficiently
controlled inventories make the firm flexible. Inefficient inventory
control results in unbalanced inventory and inflexibility. The firm may
sometimes run out of stock and sometimes may pile up unnecessary
Ill
stocks. This increases the level of investment and makes the firm
unprofitable. To manage inventory efficiently, answer should be sought
to the following two questions:
(1) H ow m u c h s h o u ld b e o r d e r e d ?
(2) W h e n s h o u ld it b e o r d e r e d ?
The first question relates to the problem of determining economic order

quantity (EOQ), and is answered with an analys is of costs of
maintaining certain level of inventories. The second question arises
because of uncertainty and is a problem of determining the re -order
point. All these will be discussed as the unit develops and the next unit.
3.4 Stock Costs
Stock represents an investment by the organisation. As with any other

investment, the cost of holding must be related to the benefits to be
gained. This is done through effective identification of the cost. There
are four categories:
 Cost of hold ing ;

 Cos t of ob ta in ing s t oc k
 St oc k- ou t c os t, a nd
 The cost of stock itself
3.4.1 Cost of Holding Stock
These costs, are also known as carrying costs including the following:
(a) interest on cap ital invested in the stock:

(b) storage charge(warehouse — rent, lighting, heating, refrigeration,
air conditioning, etc.);
(c) stores staffing, equipment maintenance and running costs;
(d) handling costs;
(c) audit, stocktaking or perpetual inventory costs;
(0 insurance, security;
(g) d e te r io r a t io n a n d o b s o le s ce ;
( h) pilferage, verm in damage.
3.4.2 Costs of Obtaining Stock
Otherwise known as ordering or procurement costs, include the

following:
( i) The clerical and administrative costs associated with the

purchasing, accounting, and goods received departments;
112
N1131 839
,
QUANTITATIVE ma INIQUES FOR BANKING AND FINAN CE
(ii) Transp ort cos ts;

(iii) Where goods are manufactured internally, the set-up and tooling
costs associated with each production run.
3.4.3 Stock-Out Costs
These are the cost associated with running out of stock. The avoidance
of this cost is the basic reason why this cost is held in the first instance.
These costs include the following:
 Lost contribution through the lost sale caused by the stock-out;

 Lost of future sales because customers go elsewhere;
 Lost of customer goodwill;
 Cost of production stoppages caused by stock-outs of W.I.P. or
raw materials;
 Lab ou r fr us tr ation o ver s top p ag es ;
 Extra cost associated w ith urgent often small quantity,
replenishment purposes.
3.4.4 Costs of Stock
These costs are the buying in prices or the direct cost of production.
You consider this cost when:
(a) Discounts are availab le for bulk purchases;

(b) Savings in production costs arc possible with longer batch runs.
Example 1
The following diagram —figure 9.1, shows a stock situation simplified by

the following assumptions: regular rates of demand, a fixed lead time,
and replenishment in one batch
113
MBF 839 QUANTITATIVE TECIINIQUES FOR BANKING AND FINANCE
Stock level
1000 Re-order level (700 units)
Slopes indicate
900 anticipated rates of demand
800
Reorder Maxi
700 quantity
600 (EOQ)
500
400
300
200
Lead time
safety stock 5 weeks
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14
15 16
Time, days or weeks

Stock terminology illustrated, Figure 1
Notes:
 It could be observed from the above diagram 1 that the safety
stock is needed to cope with periods of maximum demand during
the lead time.
 The lead time is shown as 5 weeks, the safety stock 200 units,
and the re-order quantity 600 units.
 With constant rates of demand, as shown, the average stock is the
safety stock plus 1/2 re-order quantity. For example in the figure
1, the average stock is 200 + 'A (600) = 500 units.
Like most other quantitative techniques, two types of models can be

used for inventor y control — deterministic and stochastic (or
probabilistic) models.
 A deterministic model is one which assumes complete certainty.

The values of all factors (e.g. demand, usage, lead-time, carrying
costs, etc.) are known exactly and there is no element of risk and
uncertainty.
 A stochastic model exists where some or all of the factors are not
known with certainty and can only be expressed in probabilistic
or statistical terms. E.g. if the usage rate or lead -time was
114
specified as a probability distribution, this would be a stochastic

or probabilistic model.
Both types of model will be explained in subsequent units.
3.5 Types of Control System
They include:
(1) Re-order level system

( 2) P er io d ic r e vie w s yst e m.
3.5.1 Re order Level System
-
Ot herwise known as t he t wo -bin syst em wit h t he following

characteristics:
(a) A predetermined re-order level is set for each item.

(b) When the stock level falls to the re-order level, a replenishment
order is issued.
(c) The replenishment order quantity is invariably the EOQ.
(d) The name `two-bin system' comes from the simplest method of
operating the system whereby the stock is segregated into two
bins.
(e) Most organizations operating the re-order level system maintain
stock records with calculated re-order levels which trigger off the
required replenishment order.
Note: Example 1 is a simple re-order level system which is widely used

in practice.
Example 2:
The following data relate to a particular stock item:
Normal usage 110 per day

Minimum usage 50 per day
Maximum usage 140 per day
Lead time 25 — 20 days
EOQ (previously calculated) 5,000

Using this data, calculate the various control levels.
115
MBE 839 QIIANTITATIVE TECIINIQUES FOR BANKING AND FINANCE
Solution:
Re-order Level Maximum Usage x Maximum Lead time
140 x 30
4,200 units
Minimum Level R e- or der Level — Aver age Us ag e for

Average
Lead Time
4.200 —(110 x 27.5)

1,175 units
Maximum level = Re-order Level F.OQ — Minimum

Anticipated
in Lead Time
4.200 + 5,000 —(50 x 25)

7,950 units
Note: In a manual system, the three levels would be entered on a stock

record card and comparisons made between the actual stock and the
control levels each time an entry was made on the card.
Re-order level system
Advantages:
(a) L ower st oc ks on a ver age :

(b) Items ordered in economic quantities via the FOQ calculation:
(c) Somewhat more responsive to fluctuations in demand:
(d) Automatic generation of a replenishment order at the appropriate
time by comparison of stock level against re-order level:
(e) Appropriate for widely differing types of inventory within the
same firm.
Disadvantages:
(a) Many items may reach re-order level at the same time. thus
overloading the re-ordering system:
116
(b) Items come up for re-ordering in a random fashion so that there is

no set sequence;
(c) In certain circumstances (e.g. variable demand, ordering costs
etc.), the EOQ calculation may not be accurate.
3.5.2 Periodic Review System

This system is sometimes called the constant cycle system. The system
has the following characteristics:
(a) Stock levels for all parts are reviewed at fixed intervals, e.g.
every fortnight.
(b) Where necessary, a replenishment order is issued.
(c) The quantity of the replenishment order is not a previously
calculated EOQ, but is based upon; the likely demand until the
next review, the present stock level and the lead time.
(d) The replenishment order quantity seeks to bring stocks up to a
predetermined level.
(c) The effect of the system is to order variable quantities at fixed
intervals as compared with the re-order level system where fixed
quantities are ordered at variable intervals.
Advantages:
(a) All stock items are reviewed periodically so that there is more
chance of obsolete items being eliminated;
(b) Economies in placing orders may be gained by spreading the
purchasing office load more evenly;
(c) Larger quantity discount may be obtained when a range of stock
items are ordered at the same time from a supplier;
(d) Because orders will always be in the same sequence, there may
be production economies due to more efficient production
planning being possible and lower set up costs.
Disadvantages:
(a) In general, larger stocks are required, as re-order quantities must
take account of the period between reviews as well as lead times;
(b) Re-order quantities are not at the optimum level of a correctly
calculated EOQ;
(c) Less responsive to changes in consumption. If the rate of usage
change shortly after a review, a stock-out will occur before the
next review;
(d) Unless demands are reasonably consistent, it is difficult to set
appropriate periods for review.
117
.i What are the advantages and disadvantages of the Periodic

review system of Inventory control?
How may inventory be classified?
Example 3
A simple manual periodic review system control of the 500 piece parts
used in the assembly of the finished products by the periodic review
system. The stock levels of al 500 parts are reviewed every 4 weeks and
a replenishment order issued to bring the stock of each part up to a
previously calculated level. This level is calculated at six-monthly
intervals and is based on the anticipated demand for each part.
Based on the above system, the following graph shows the situation for
one of the piece parts, parts No. 1101x, over a period of 16 weeks.
Stock Level
3,600 Predetermined Forecast (forecast level)
3,000
EOQ I 0Q2 EOQ 3
2,500
2,000
1,500
1,000
500
R e v i e w R e vi e Revi w
Point 1 Point 2 Poin 3
12 3 4 5 6 7 8 9 10 11 12 13 14 15
16
Time (weeks)
Stock levels of Part No. 1101 Figure 2
4.0 CONCLUSION
Inventory management techniques have been demonstrated as a tool for
reducing cost and expressing inventory as an investment which is used
to minimize cost and maximize profit.
118
5.0 SUMMARY
In this unit, stocks have been classified in raw material, work-in-

progress, and finished goods.
 The overall objective of inventory control is to maintain stock
at a level which minimizes total stock-out.
 There are two basic inventory control systems — the re-order
level or two-bin system and the periodic review system.
 In the re-order level system, the usual replenishment order
quantity is the EOQ.

-
The following data relate to a given stock item:
Normal usage 1300 per day

Minimum usage 900 per day
Maximum usage 2000 per day
EOQ 30,000
Calculate the various controls levels.


Fandey, I.M. (2005). Financial Management, UBS Publisher's

119
UNIT 5 INVENTORY CONTROL II

CONTENTS
1.0 Introduction
2.0 Objectives
3.0 Main Content
3.1 Economic Order Quantity (EOQ)
3.2 EOQ assump tion
3.2.1 EOQ formula
3.2.2 EOQ with gradual replenishment
3.2.3 EOQ where stock-outs are permitted
3.2.4 EOQ with discounts
3.2.5 Financial consequences of discounts
3.2.6 Marginal costs and EOQ calculations
4.0 Conclusion
5.0 Summary
1.0 INTRODUCTION
From Inventor y Contr ol I unit ... 3.3 (inventor y management

techniques), the importance of Economic Order Quantity in stock order
arrangement was emphasized. In this unit, an attempt will be made to
explain the different facet of EOQ in inventory control.
2.0 OBJECTIVES
 explain the basic Economic Order Quantity (EOQ);

 calculate order quantity when price discounts are possible
3.0 MAIN CONTENT
3.1 Economic Order Quantity (E0Q)

One of the major inventory management problems to be resolved is how
much inventory should be added when inventory is to be replenished. If
the firm is buying raw materials, it has to decide lots in which it has to
be purchased on replenishment. If the firm is planning a production run,
the issue is how much production is to schedule (or how much to make).
These problems are called Economic Order Quantity problems the task
of the firm is to determine the optimum or economic order quantity (or
economic lot size). The determination of an optimum inventory level
120
involves two types of costs — ordering cost and carrying cost. The
economic order quantity is that inventory level that minimizes the total
of ordering and carrying costs.
3.2 EOQ Assumption

To be able to calculate a basic EOQ, certain assumptions are necessary:
(a) that there is a known, constant stockholding cost;

(b) that there is a known, constant ordering cost;
(c) that rates of demand are known;
(d) that there is a known, constant price per unit;
(e) that replenishment is made instantaneously;
(f) no stock-outs allowed.
It is worthy of note that:
(i) it will be apparent that the above assumptions are somewhat

sweeping and there are good reasons for treating any EOQ
calculation with caution
(ii) Some of the above assumptions are relaxed later in the study
(unit);
(iii) The rationale of EOQ ignores buffer stocks which are maintained
to cater for variations in lead time and demand.
Example 1 A graphical EOQ

—
The following data will be used to develop a graphical solution to the

EOQ problem:
A company uses 50,000 gadgets per annum which are N10 each to
purchase. The ordering and handling costs are N150 per order and
carrying costs are 15% of purchase price per annum, i.e. it costs N1.50
per annum in carrying a gadget in stock (N10 x 15%).
To graph the various costs involved, the following: (pages 240 and 241)
3.2.1 The EOQ Formula
It is possible, and more usual, to calculate the EOQ using a formula.

The formula method gives an exact answer, but do not be misled into
placing undue reliance upon the precise figure. The calculation is based
on estimates of costs, demands etc., which are, of course, subject to
error. The EOQ formula is given below and you should learn it:
121
Basic EOQ formula
E C K ) =
Co
Where Co = Ordering cost per order

D = Demand per annum
C, = Carrying cost per item per annum
Using the data from Example 1, the EOQ can be calculated:
= N150
50,000 gadgets
Co = N10 x 15% or N1.50 per gadget
EOQ = V 2. Co. D
Co
= V 2 x 150 x 50,000
1.5
10,000,000
3,162 gadgets
Notes:
(a) The closest value obtainable from the graph was approximately
3,200 which is very close to the exact figure.
(b) Always take care that demand and carrying costs are expressed
for the same time period. A year is the usual period used.
(c) In some problems, the carrying cost is expressed as a percentage
of the value whereas in others it is expressed directly as a cost per
item. Both ways have been used in this example to provide a
comparison.
The mathematical derivation is expressed below:
Derivation of basic EOQ formula:
Let D = annual demand

0 = order quantity
Co = cost of ordering for one order
C, carrying cost for one item per annum
122
Average Stock = Q/2
Total annual stockholding , QC,/2
Number of orders per annum = D/Q
Annual ordering costs , DCo/Q
Total Cost Qc_c+D.00

2Q
The order quantity which makes the total cost (Tc) at a minimum is
obtained by differentiating with respect to Q and equating the derivative
to zero.
dTc = —
dQ 2 Q2
and when dTc = 0 costs are at a minimum

dQ
i.e. 0 = Cc—Dc
2 Q2
and to find Q
DCQ = Cc
Q2 2
2DC0 = Q22Cc
Q
2DC,2 =
Cc
Q (i.e. the EOQ) = 2 . Co D

Co
3.2.2 EOQ with Gradual Replenishment

In example earlier shown above, the assumption was that the widgets
were ordered externally and that the order quantity was received as one
batch, i.e. instantaneous replenishment as shown in the following
diagram figure 10.1
123
Stock vertical lines indicate instantaneous replenishment

level
Xi and X, are
replenishment points
Time
XI x,
Stock levels showing instantaneous replenishment

Figure 1
If however, the widgets were manufactured internally, they would
probably be placed into stock over a period of time resulting in the
following pattern:
Stock sloping lines indicate net rate of replenishment

Level
X1 and X2 are the periods over which

replenishments are made
Time
x,
x2
Stock levels showing Non-instantaneous replenishment

Figure 2
The net rate of replenishment is determined by the rate of replenishment
and the rate of usage during the replenishment period. To cope with
such situations, the basic EOQ formula needs modification thus:
EOQ with gradual replenishment = V 2 . C2. D

Cc (1 — D)
Where R = production rate per annum, i.e. the quantity that

would
be produced if production of the item was carried
on the whole year.
124
All other elements in the formula have meanings as previously defined

in paragraph 4. Note: The derivation of the above formula is given
Example of EOQ with Gradual Replenishment
Example 1
Assume that the firm described in Example 1 has decided to make
the widgets in its own factory. The necessary machinery has been
purchased which has a capacity of 250,000 widgets per annum. All
other data are assumed to be the same.
EOQ with gradual replenishment = 2 x 150 x 50,000

1.5 (1 —50,000)
250,000
'115,000,000
1.5 (0.8)
3,535 widgets
Notes:
(a) The value obtained above is larger than the basic EOQ because
the usage during the replenishment period has the effect of
lowering the average stock holding cost.
(b) As pointed out earlier on, the ordering costs for internal ordering
usually include set-up and tooling costs as well as paper work and
administration costs.
3.2.3 EOQ Where Stock-Outs Are Permitted
It will be recalled that the overall objective of stock control is to

minimize the balance of the three main areas of cost, i.e. holding costs,
ordering costs and stock-out costs.
Stock-out costs are difficult to quantify but nevertheless may be

significant and the avoidance of these costs is the main reason why
stocks are held in the first place. Where stock-out costs are known then
they can be incorporated into the EOQ formula which thus becomes:
EOQ (where stock-outs are permitted and stock-out costs are known)
=
Cc Cs
125
Where Cs = stock-out costs per item and the other symbols have the
meanings previously given.
Note: It will be seen that the formula is the basic EOQ formula
multiplied by a new expression containing the stock-out cost. The
derivation of the EOQ formula where stock-outs are permitted is given
below.
Example
Assume the same data as in .... Except that stock -outs are now
permitted. When a stock-out occurs and an order is received for widgets
the firm has agreed to retain the order and, when replenishments are
received, to use express courier service for the delivery at a cost of
NO.75 per widget. Other administrative costs associated with stock-outs
are estimated at NO.25 per unit. What is the EOQ?
C„ = N150
50,000
Cc= N1.50
C, = NO.75 + NO.25 NI.00
Thus, EOQ (with stockouts) Ai 2 x 150 x 50,000 i11.5 + 1

1.5 1
5,000
3.2.4 EOQ With Discounts
A particularly unrealistic assumption with the basic EOQ calculation is

that the price per item remains constant. Usually some form of discount
can be obtained by ordering increased quantities. Such price discounts
can be incorporated into the EOQ formula, but it becomes much more
complicated. A simpler approach is to consider the costs associated
with the normal EOQ and compare these costs with the costs at each
succeeding discount point and so find the best quantity to order.
3.2.5 Financial Consequences of Discounts
Price discounts for quantity purchases have three financial effects, two
of which are beneficial and one adverse.
126
Beneficial Effects:
Savings come from:
(a) Lower price per item

(b) The larger order quantity means that fewer orders need to be
placed so that total ordering costs are reduced.
Adverse Effect:
Increased costs arise from the extra stockholding costs caused by the
average stock level being higher due to the larger order quantity.
Example 2 of EOQ with
Discounts
A company uses a special bracket in the manufacture of its products
which it orders from outside suppliers. The appropriate data are;
Demand = 2,000 per annum

Order cost N20 per order
Carrying cost = 20% of item price
Basic item price = NI 0 per bracket
The company is offered the following discounts on the basic
price:
For order quantities 400 — 799 less 2%
800 — 1,599 less 4%
1,600 and over less 5%
It is required to establish the most economical quantity to
order.
This problem can be answered using the following procedure:
(a) Calculate the EOQ using the basic price.

(b) Compare the savings from the lower price and ordering costs and
the extra stock-holding costs at each discount point (i.e. 400, 800
and 1,600) with the costs associated with the basic EOQ, thus:
Basic EOQ = 2 x 2,000 x 20

10 x 0.2 200
brackets
Based on this EOQ, the various costs and savings comparisons are given
in the following Table: 10.1
127
O rd er Q uant i t y 200 (EOQ) 400 800 1,600 Line

No.
Discount 2% 4% 5% 1
Average No. of 10 5 21/2 11/2 2
Orders p.a.
Average No. of 5 71/2 8% 3
Orders saved p.a.
Ordering cost savings (5 x 20)=44100 (71/2 x 20)= N150 (8% x 20)=14175 4
P.a.
Price saving per item - 20p 40p 50p 5
per annum (2000, x 20p) = (2,000 x 40p)— 800 (2,000 x 50p) =
400 1,000
Total gains N500 N950 141,175 6
Stockholding cost p. (100 x 10 x 2) 200 x 9.8 x 21= 400 x 9.6 x 2) = (600 x 9.5 x 2) = 7
a. =N200 N392 N768 441,520
Additional costs - (N392 - N200) = (N768 - 14200) = (141,520-14200) 8
incurred by increased N192 14568 141,320
order
Net gain / (loss) 14308 N382 (N145) 9
Cost/Savings Comparisons EOQ to Discount Points

From the above table, it will be seen that the most economical order
quantity is 800 brackets, thereby gaining the 4% discount.
Notes:
a) Line 2 is Demandof2,000
Order quantity
b) Line 7 is the cost of carrying the average stock, i.e.
Order quantity x cost per item x carrying cost percentage

2
c) L i n e 9 is L i n e 6 m i n u s L i n e 8 .
3.2.6 Marginal Costs and EOQ Calculations
It cannot be emphasized too strongly that the costs to be used in EOQ

calculations must be true marginal costs, i.e. the costs that alter as a
result of a further order or carrying another item in stock. It follows
therefore that fixed costs should not be used in the calculations. In the
examples used in this chapter, the costs have been clearly and simply
stated. In examination questions, this is not always the case and
considerable care is necessary to ensure that the appropriate costs are
used.
4.0 CONCLUSION
In this unit, you have been inculcated with the different applications of
inventory control through economic order quantity (E0Q).
128
MBF 839 QUANT ITA TIVE TECH NIQUES FOR BANKI NG A ND FI NAN CE
5.0 SUMMARY
 The EOQ is the order quantity which minimizes the total costs
involved which include holding and order costs.
 The basic EOQ calculation is based on constant ordering and
holding costs, constant demand and instantaneous replacement.
 The basic ECQ formula is: '12. C 0 .D
Cc
where larger quantities are ordered to take advantage of price discounts,

stock-holding costs increase, but savings are made in the price
reductions and reduced ordering costs.
.i What assumptions are usually made in EOQ calculations?
A customer has been ordering 5,000 units at the rate of 1,000unit

per order during last year. The production cost is N12 per unit —
N8 for materials and labour and N4 overhead cost. It costs
N1,500 to set up for one run of 1,000 units and inventory
carrying cost is 20% of the production cost. Since this customer
may buy at least 5,000 units this year, the company would like to
avoid making five different production runs. Determine the most
economic production run.
1) A company uses 100,000 units per year which cost N3 each.

Carrying costs are 1% per month and ordering costs are N250 per
order. What is the E0Q?
2) What would be the EOQ if the company made the items
themselves on a machine with a potential capacity of 600,000
units per year?



129
MODULE 3
Unit 1 Inventory Control III

Unit 2 Time Series Analysis
Unit 3 Correlation
Unit 4 Regression Analysis
Unit 5 Project Planning
UNIT 1 INVENTORY CONTROL III

CONTENTS
1.0 Introduction
2.0 Objectives
3.0 Main Content
3.1 The Re-order Point
3.1.1 Re-order levels in conditions of certainty
3.1.2 Re-order level and safety stock and holding 3.2
Safety Stock
3.2.1 Safety stock calculation by cost tabulation
3.2.2 Safety stock calculation by statistical met hods
3.3 Inventory Control and Sensitivity Analysis
4.0 Conclusion
5.0 Summary
7.0 References /Further Readings
1.0 INTRODUCTION
In this unit, a cursory study of the problem, how much to order, will be
solved by determining the economic order quantity. The answer should
be sought to the second problem, when to order. This is a problem of
determining the re-order point.
2.0 OBJECTIVES
 plan safety or buffer stocks necessary to cope with variations in

demand and/or lead time
 calculate sensitivity analysis as in inventory control
130
MB,: 839 QUANTITATIVE TECHNIQUES FOR BANKING AND FINANCE
3.0 MAIN C ONTENT
3.1 The Re-order Point
The re-order point is that inventory level at which an order should be

placed to replenish the inventory to determine the re-order point under
certainty, you should know:
(a) lead time
(b) average usage, and
(c) economic order quantity (EOQ)
Lead time is the time normally taken in replenishing inventory after the
order has been placed. By certainty, we mean that usage and lead time
do not fluctuate. Under such situation, re-order point is simply that
inventory level which will be maintained for consumption during the
lead time. That is:
Re-order point Lead x Average usage
3.1.1 Re order levels in conditions of certainty

-
Where the rate of demand and the lead time is known with certainty, the
re-order level is the rate of demand times the lead time. This means
that, regardless of the length of the lead time or of the rate of demand,
no buffer stock is necessary. This results in a situation as follows:
Stock
Level
o = Reorder level
Lead Time
Time
Reorder level in conditions of certainty
Figure 1
131
3.1.2 Re-order Level and Safety Stock Relationship
It will be seen from the above figure that, in conditions of certainty, the
re-order level can be set so that stock just reaches zero and is then
replenished. When demand and/or lead time vary, the re-order level
must be set so that, on average, some safety stock is available to absorb
variations in demand and/or lead time. In such circumstances, the re-
order level calculation can be conveniently considered in two parts:
a) the normal or average rate of usage times the normal or average

lead time (i.e. as the re-order level calculation in conditions of
certainty) plus
b) the safety stock.
3 . 2 S afet y S t oc k
The re-order point in any typical example is computed under the

assumption of certainty. It is always difficult to predict usage and lead
time accurately. The demand for material may fluctuate from day-to-
day or from week-to-week. Similarly, the actual delivery time may be
difficult from the normal lead time. If the actual usage increases or the
delivery of inventory is delayed, the firm can face a problem of stock-
out which can prove to be costly for the firm. Therefore, in order to
guard against stock-out, the firm may maintain a safety-stock — some
minimum or buffer inventory as cushion against expected increased
usage and/or delay in delivery time.
3.2.1 Safety Stock Calcu lation b y Cost Tabulation
The amount of safety stock is the level where the total costs associated
with safety stock are at a minimum. That is, where the safety stock
holding cost plus the stock out cost is lowest. (it will be noted that this
is a similar cost position to that described in the EOQ derivation
described in earlier unit).
132
The appropriate calculations are given below based on the following

illustration:
Example 1
An electrical company uses a particular type of thermostat which costs $5. The
demand averages 800 p.a. and the EOQ has been calculated at 200. Holding costs
are 20% p.a. and stock out costs have been estimated at $2 per item that is
unavailable. Demand and lead times vary, but fortunately, the company has kept
records of usage over 50 lead times as follows:
(a) (b) (c)
Usage in lead time Number of times recorded Probability b/50
25 29 units
— 1 0.02
30 — 34 units 8 0.16
35 —39 units 10 0.20
40 — 44 units 12 0.24
45 — 49 units 9 0.18
50 — 54 units 5 0.10
55 — 59 units 5 0.10
Total 50 1.00
Fro m the abo ve the re -o rder level and s af ety s to ck s ho uld he

calculated.
Solution:
Step 1 — Using the mid-point of each group, calculate the average

usage in the lead time.
X T tx
27 1 27
32 8 256
37 10 370
42 12 504
47 9 423
52 5 260
57 5 285
Total 50 2,125
Table 2
Average usage = 2,125 42.5
50
133
MBF 839 Q U ANT I T ATI VE T ECI I NIQ U ES FO R BANKING AND FINANCE
Step 2 - Find the holding and stock out costs for various re-order
levels.
A B C D E F C II
Ilolding Possible No. of Shortage
Re- Safety cost shortages Probability orders p.a. cost Total
order s t o c k ( A- (B x N (mid-points) (from (8001 (D x E x I: x cost
level 42.5) ) T abl e 2- A) Table 1) 200 N 2) (C +
N N G)
45 2.5 2.5 2 0.18 4 2.88
7 0.10 4 5.60
12 0.10 4 9.60 20.58
50 • 7.5 7.5 2 0.10 4 1.60
7 0.10 4 5.60 14.70
55 12.5 12.5 2 0.10 4 1.60 14.10
60 17.5 17.5 17.50
From the Table (I) it will be seen that the most economical re -order
level is 55 units. This re-order level, with the average demand in the
lead time of 42.5, gives a safety stock of 12.5.
3.2.2 Safety Stock Calculation b y Statistical Methods
The previous method of calculating safety stock was based on relative

holding and stock out costs; but on occasions these costs, particularly
the effects of stock outs, are not known. In such circumstances
management may decided upon a particular risk level they are prepared
to accept and the safety stock and re-order level are based upon this risk
level. For example, management may have decided that they are
prepared to accept a 5% possibility of a stock out.
Illustration of a safety stock calculation by statistical methods
Solution:
a) Safety stock given variable demand and constant lead time:
From normal area tables, it will be found that 5% of the area lies
above the mean +1.64a.
Safety stock = 1.64 x standard deviation of demand for 13.44

days
= 1.64 x (0.4 x .4 13.44)
= 2.40
Note: The standard deviation of daily demand, 0.4 is multiplied

by \)13.44 because standard deviations are not added.
b) Safety stock given variable lead time and constant demand:
Safety stock = 1.64 x standard deviation of lead time for a

demand of 3,162
134
= 1.64 x(0.75)
= 1.23
Example 2
Using the data from example 1, it was found that the average demand during
the lead time was 42% units. The company has carried out further analysis
and has found that this average lead time demand is made up of an average-
demand (D) of 3,162 units per day over an average lead time (L) of 13.44
days. Both demand and lead time may vary and the company has estimated
that the standard deviation of demand (o.D) is 0.4 units and the standard
deviation of lead time (aL) is 0.75 days. The company is prep ared to accept
a 5% risk of a stock out and wish to know the safety stock required in the
following three circumstances:
(i) Whe r e d e m a nd v a r i e s a nd l e a d t i m e i s c o ns t a nt ;
(ii) Whe r e t he l ea d t im e v ar i e s a nd d e m a nd i s co ns t ant ;
(iii) Where both demand and lead time vary.
c) Thisacombinationofthetwopreviousectionsandisthesum
of the separate safety stocks already calculated:
= 2.40 + 1.23
= 3.63
Note:
(a) Safety stock calculations based on risk levels are

commonly used.
(b) Where lead time and demand vary it can be expensive to

maintain sufficient stocks for low stock out risks.
The above example uses for properties of the Normal Distribution

and values obtained from Normal Area tables. It is but one
further application of the statistical principles covered earlier in
the book using continuous probability distributions. Discrete
probabilities and expected values can also be used for incorporate
variability.
3.3 Inventory Control and Sensitivity Analysis

When an inventory control value has been calculated, for example, the
EOQ, the re-order level, the total stockholding cost etc. management
stay wish to know how sensitive the value is to changes in the factors
used to calculate it.
135
For example, is the EOQ greatly affected when ordering cost changes?
The process by which this is done is known as sensitivity analysis. The
following example shows a typical analysis:
Example: 3
Henderson Ltd. made the following estimates for a component they use:
Annual usage 1,125
Ordering costs N50 per order
Carrying costs per year N5 per component
Based on these estimates, an EOQ of 150 and expected total stock costs of N750 p.a.
were calculated, as follows:
EOQ = 42 x 50 x 1,125 = 150

5
Expected total stock costs Ordering cost p.a. + Holding Cost p.a.
(1 125) x N50 +145 x 150 = N750

150 2
During a year, the EOQ of 150 was used for re-ordering, but the actual usage of
components turned out to be 20% higher at 1,350.
(a) Calculate the actual t otal st ock cost s.
(b) Calculate what the total stock costs would have been if the correct EOQ had
been used.
(c) Find out how sensitive total costs are to errors in the usage estimates.
Solution:
a) Actual total stock costs = (1,350) x N50 + 5 x 150 = N825

150 2
b) Correct EOQ = if 2 x 50 x 1,350 = 164

3
Total costs if an EOQ of 164 was used
= (1,350) x N50 +145 x 164 = N822

164 2
c) Thus N3 (i.e. N825 — 822) extra costs were incurred by using the
150 EOQ based on the incorrect usage.
This means that a 20% increase in demand has caused only
approximately {1/2%} change in total costs. We conclude that
total costs are insensitive to errors in demand estimates.
136
From management's viewpoint, this is good news. It means that

even if errors are made in the estimates, which are all too likely,
it doesn't make much difference to the final result.
SELF-ASSESSED EXERCISES
(0 How is the re-order level set when demand and lead times are
known with certainty?
(i) Why is the establishment of safety stocks essentially a cost

balancing exercise?
4.0 CONCLUSION
At the end of this unit, you have been taught that safety or buffer stocks
are needed to cope with variations in demand and/or lead time. This is
essential in the skill of managing inventory in an organisation.
5.0 SUMMARY
That safety stocks are necessary because of demand and/or lead time
variation. Re-order level is the average demand over the average lead
time plus safety stock.
The safety stock level can be established by comparing the safety stock
holding cost and the stock out cost at various re-order levels.

-
The demand for an item is normally distributed with a mean of 50 per

day and a standard deviation of 5 units. Given a lead time of 20 days,
what is the re-order level and safety stock to meet 90% of demands.



137
IvIBF 839 QUANTITATIVE TECHNIQUES FOR BANKING AND FINANCE
UNIT 2 TIME SER IES AN ALYSIS
CONTENTS
1.0 Introduction
2.0 Objectives
3.0 Main Content
3.1 Time Series
3.1.1 The Trend
3.1.2 Seasonal Variation
3.1.3 Residual
3.2 Time Series Analysis
3.2.1 De-seasoning the Time Series
3.2.2 Fitting a Trend Line
3.2.3 Identifying Cyclical Variations
3.3 Forecasting with the Decomposed Time Series
4.0 Conclusion
5.0 Summary
1 . 0 IN TR O D U C T IO N
Time series analysis is one of the most powerful methods in use,

especially for short-term forecasting purposes. From the historical data,
one attempts to obtain the underlying pattern so that a suitable model of
the process can be developed, which is then used for purposes of
forecasting or studying the internal structure of the process as a whole.
Time series analysis looks for the dependence between values in a time
series (a set of values recorded at equal time intervals) with a view to
accurately identify the underlying paths of the data.
In the case of quantitative methods of forecasting, each technique makes

explicit assumptions about the underlying pattern. For instance, in using
regression models you had first to make a guess on whether a linear
model should be chosen and only then could you proceed with the
estimation of parameters and model development. We could rely on
mere visual inspection of the data or its graphical plot to make the best
choice of the underlying model. However, such guess work though not
uncommon, is unlikely to yield very accurate or reliable results. In time
series analysis, a systematic attempt is made to identify and isolate
different kinds of patterns in the data. The four kinds of patterns that are
most frequently encountered are horizontal, non-stationary, seasonal and
cyclical. Generally, a random component is also superimposed.
138
The question of the choice of a forecasting method is taken up. The

characteristics of various methods are summarized along with likely
situations where these may be applied. Of course, considerations of cost
and accuracy desired in the forecast play a very important role in the
choice.
2.0 OBJECTIVES
After reading this unit, you should be able to:
 discuss the role of time series analysis in short-term forecasting;

 decompose time series into its various components;
 identify the underlying patterns of a time series;
3.0 MAIN CONTENT
3.1 Ti me Series
The time series takes a variable and examines the way in which its
magnitude fluctuates over a period of time. For example, consider the
way in which output of a particular firm has varied year by year since
1985. Why should you do this?
Whenever we are interested in planning, you must make forecasts of
what is likely to happen in the future. Despite complicated techniques
applied, the inference will depend, to a large extent, on what has
happened in the past. Thus, a reliable analysis of what has happened in
relevant fields in recent years is the first step in obtaining a reliable
estimate of what is likely to happen in the future.
Any housewife will tell you that egg prices are likely to rise during
particular months of the year, and to fall during other months. How
does she arrive at this?
On the basis of past experience, you can undertake a simple analysis of
more complex problems to predict the future more reliably and more
precisely.
3.1.1 The Trend

Let's consider the following series:
139
Table 1 Output of ABC Limited 1985 — 2002 Million Tons

Year Output
1985 68
1986 100
1987 120
1988 177
1989 100
1990 80
1991 140
1992 200
1993 230
1994 180
1995 280
1996 200
1997 250
1998 170
1999 320
2000 230
2001 210
2002 330
Seeing the figures presented, information reveals two things — figures of

output show that although there have been very marked fluctuations year
by year, there is a general upward rise in the figures. Moreover,
certainly from 1993 to 2002, the figures tend to show a peak output
every second year, followed by a fall in the intervening years.
This becomes much more apparent if you look at the graph below of this
time series in Figure 1:
Output
(M tons)
Trend 300
200
100
1985 1990 1995 2000 2995
140
The general rise in output can be clearly seen. This general tendency of
figures to move in a given direction is known as TREND. In figure 1
above, the trend has been shown as a linear or straight line trend. Other
series to be examined later in the trend will vary its direction, at first
rising and then falling or vice versa. Eg: curvi-linear. An important
characteristic of such trends is that they change direction slowly over
time, and so, baring catastrophes, the continuation can be sketched in
from a careful examination of the existing trend line.
Much more obvious in figure 1 is the succession of rises and falls in

output occurring at regular intervals. Looking at this graph suggests a
certain pattern of behaviour in output in 2003. What is it?
Formal analysis of such a series can reveal much more useful data about
what may happen in the future.
3.1.2 Seasonal Variation

Many series are of such a nature that figures become of much significant
when they are analysed on a monthly or quarterly basis rather than
annually. Most economic series are like this — sales, for example, show
marked variations throughout the year.
Sales tend to rise at Christmas. But if such fluctuations occur, are they
sufficiently regular to be predictable? If they are, then you must try to
ascertain the direction and probable magnitude of the saving. Figure 2
below shows hypothetical sales data plotted quarterly over a period of
years.
141
M131,839 QUANTITATIVE TECIINIQUF,S FOR BANKING AND FINANCE
Sales Seasonal fluctuation

(N'000)
400—
Tren
300—
Seasonal fluctuation
100—
I I I I
Quarter I 234 12314 1234 123I4 12314 1234 1 2 34
Year 1996 1997 1998 1999 2000 2001 2002
Seasonal Variation
Figure 2
The data show a remarkable regularity — every year, there are two sales
peaks, in the second and fourth quarter, while every year, sales slump in
the third quarter. The pattern is regular enough for us to predict that the
same thing will happen in future years.
If y take the trend as the norm, such regular fluctuations around t trend
are given the name of SEASONAL VARIATIONS. It should now be
apparent that if you can obtain a reliable estimate of the trend, and of the
magnitude of the seasonal variation from the trend, you have gone some
way towards obtaining a reliable basis for forecasting the future sales
figures.
142
MISF 839 QUANTITATIVE TECHNIQUES FOR BANKING AND FINANCE
Figure 3
(a)
Output
Year
Linear Trend
(b)
Output
f
Year
Curvilinear Trend
3.1.3 Residuals
Unfortunately for the planner, the data he has is a compound of many
influences, most of which cannot be foreseen and may not occur again in
a similar form. Exports will affect the dock strike, production by a
shortage of some vital raw material, sales by a sudden change in
taxation. Thus, the figures gotten can be broken down into TREND
AND SEASONAL VARIATION AND RESIDUALS.
143
By their very nature residual influences are chance happenings and are
not amenable to analysis. This does not mean that residual influences
can be ignored by any planner of repute for it could make nonsense of
his forecast.
The importance of the residuals lies on this fact. If over the years
residual influences have had a minimal effect on the figures, one can use
ones forecasts with some degree of confidence that will not be unduly
affected by external events. But if the data has been affected regularly
and, to a large extent, by residual factors, then we must use our forecasts
with caution.
3.2 Time Series Analysis
Taking a look at some simple time series and seeing how the trend could
be obtained and if seasonal variation can be calculated, you will begin as
statisticians have done by considering the trend and its nature and scope.
Firstly, take the case of a series with obvious linear trend i.e. which
shows a constant tendency, in spite of fluctuations, for all figures to
move in a direction — to rise or to fall, but not both.
Economic or business oriented time series are made up of four

components — trend, seasonality, cycle and randomness. Further, it is
usually assumed that the relationship between these four components is
multiplicative as shown below:
where Xt
x
=
t Tt S t C t Rt
the observed value of the time series
(1)
T = trend
seasonality
C, = cycle
and R = randomness
Alternatively, one could assume an additive relationship of the form X, =

T, +SL + Rt
But additive models are not commonly encountered in practice. Focus

will be to work with a model of the form (1) and shall systematically try
to identify the individual components.
144
If the time series represents a seasonal pattern of L period of L periods,

then by taking a moving average of L periods, we would get the mean
value for the year. Such a value will obviously be free of seasonal
effects, since high months will be offset by low ones. If 1\4, denotes the
moving average of equation (1), it will be free of seasonality and will
contain little randomness (owing to the averaging effect) which is
expressed thus:
Tt Ct (2)
The trend and cycle components in equation (2) can be further
decomposed by assuming some form of trend. One could assume
different kinds of trends such as:
 Linear trend — implies a constant rate of change (Figure I)

 Parabolic trend — implies a varying rate of change (Figure II)
 Exponential or Logarithmic trend — implies a constant percentage
rate of change (Figure III)
 Curvi-linear trend (an S curve) — implies slow initial growth, with
increasing rate of growth followed by a declining growth rate and
eventually saturated (Figure IV).
Figure I — Linear Trend

•
= a + bt
Increasing
Trend "Th
'6)
Decreasing
Trend
b<0
___ 
Time, t
145
MBI, 839 QU AN TI TA rwE TE CH NI QU E S FO R BA N KIN G AN D FI NAN CE
Figure II — Parabolic Trend
Xt = a + b ± ct (c > 0)
(c > 0)
Time, t
 Figure III — Exponential Trend
Xt = a bt
(a)
•
Log Xt = Log a + Log bt
146
MBE, 839 QUANTITATIVE TECIINIQUES FOR BANKING AND FINANCE
(b)
Figure 1 V — A typical S Curve

Xt = Ca + bit
i
t
Knowing the pattern of the trend, the appropriate mathematical function

could be determined from data by using the methods of regression. This
would be established by the values of parameters of the chosen trend
model. Assuming a linear trend gives:
Tt = a + b, (3)
The cycle component C, can now be isolated from the trend T, in

equation ... (2) by the use of equation .... (3) as follows:
(4)
Tt a + b,
As already indicated, if a linear trend is not adequate, one may wish to

specify a non-linear one. Any pattern for the trend can be used to
separate it from the cycle. In practice, however, it is often difficult to
separate the two; one may prefer to work with the trend cycle figures of
equation .... (2)
The isolation of the trend will ad little to the overall ability to forecast.
This will become clear when we take up an example problem for
solution.
To isolate seasonality, one could simply divide the original series

{equation ... (I)) by the moving average {equation ... (2)} to obtain:
X, = T,______ C, R, = S1R, .......................... (5)
Mt TCt
Finally, randomness can be eliminated by averaging the different values

of equation (....5). The average is done on the same months or seasons
147
MBF 839 QUANTITATIVE TECI INIQUES FOR BANKING AND FINANCE
of different years (for example, the average of all Januaries, all

Februaries .... all Decembers). The result is a set of seasonal values free
of randomness, called seasonal indices, which are widely used in
practiced.
In order to forecast, one must reconstruct each of the components of

equation ... (1). The seasonality is known through averaging the values
in equation ... (5) and the trend through equation .... (3). the cycle of
equation ... (4) must be estimated by the users and the randomness
cannot be predicted.
To illustrate the application of this procedure to a dual forecasting of a

time series, below is an example to consider using decomposition
(adapted by Srivastava et. al, 1978).
A firm producing farm equipment wants to predict future sales based on

the analysis of its past sales pattern. The sales of the company for the
last five years arc given in the table below:
Table 12.2 Ouarterlv sales of a firm during 2003 — 2007

Quarter
Year I II III IV
2003 5.5 5.4 7.2 6.0
2004 4.8 5.6 - 6.3 5.6
2005 4.0 6.3 7.0 6.5
2006 5.2 6.5 7.5 7.7
2007 6.0 7.0 8.4 7.7
The procedure involved in the study consists of:
(a) de-scasonalising the times series which is done by constructing a

moving average M and taking the ratio Xt /Mt which we know
from equation ... (5) represents the seasonality and randomness.
(b) fitting a trend line of the type Tt = a -I- N to the de-seasonalisecl
time series.
(c) identifying the cyclical variation around the trend line.
(d) using the above information for forecasting sales for the next
year.
3.2.1 De Scasonalising The Time Series

The moving averages and the ratios of the original variable to the
moving average have first to be computed — as in table 2 below:
148
Table 3 Computation of Moving Averages M and the ratios X /M

Actual 4 Quarter Centre Centre Moving
Year Quarter sales Moving Moving Average Total XI
Total Tot al (Mil MI
I 5.5
2003 II 5.4
III 7.2 23.8 6.0 1.200
6.0 24.1 23.5 5.9 1.017
IV _____ (Y)
I 4.8 23.4 23.2 5.8 0.828
2004 II 5.6 23.6 22.5 5.6 1.000
III 6.3 22.7 21.9 5.5 1.145
IV 5.6 22.3 21.9 5.5 1.018
I 21.5 22.6 5.7 0.702
O en 0 In
2005 II 22.2 23.4 5.9 1.068
1/46 r1/4-: .6
III 22.9 24.4 6.1 1.148

IV 23.8 25.1 6.3 1.032
I 5.2 25.0 25.5 6.4 0.813
2006 II 6.5 25.2 26.1 6.5 1.000
111 7.5 25.7 26.8 6.7 1.119
IV 7.2 26.4 27.5 6.9 1.043
I 6.0 27.2 28.2 7.1 0.845
2007 II 7.0 27.7 28.9 7.2 0.972
III 8.4 28.6
IV 7.7 29.1
It should be noted that the 4 Quarter Moving Totals pertain to the middle
of two successive periods. Thus, the value (1) computed at the end of
Quarter IV, 1983 refers to middle of Quarter II, III, 1988 and the next
moving total of 23.4 refers to the middle of Quarter III and IV, 1983.
Thus, by taking their average, we obtain the centre moving total of
(23.1 -I- 23.4) = 23.75 = 23.8

2
to be placed for Quarter III, 1983.
Similarly, for the other values in case the number of periods in the
moving totals or average is odd, centering will not be required.
The seasonal indices for the quarterly sales data can now be computed
by taking averages of the X t /Mt ratios of the respective quarters for
different years as shown in Table 3 below:
149
MI31, 839 QUANTITATIVE TECHNIQUES FOR BANKING AND FINANCE
Table 4 Computation of Seasonal Indices

Quarters
Year I H III IV
2003 - - 1.200 1.017
2004 0.828 1.000 1.145 1.018
2005 0.702 1.068 1.148 1.032
2006 0.813 1.000 1.119 1.043
2007 0.845 0.972 -
Mean 0.797 1.010 1.153 1.028
Seasonal Index 0.799 1.013 1.156 1.032
The seasonal indices are computed from the quarter means by adjusting
these values of means so that the average over the year is unity. Thus,
the sum of means in Table 3 above is 3.988 and since there are four
quarters, each mean is adjusted by multiplying it with the constant figure
of 4/3.988 to obtain the indicated seasonal indices. These seasonal
indices can now be used to obtain the de-seasonalised sales of the firm
by dividing the actual sales by the corresponding index as shown in
Table 4 below:
Table 5 De-Scasonalised Sales

Actual Seasonal Deceseasonalised
Year Quar t er sales (Y) Index sales
1 5.5 0.799 6.9
2003 II 5.4 1.013 5.3
III 7.2 1.156 6.2
IV 6.0 1.032 5.8
I 4.8 0.799 6.0
2004 11 5.6 1.013 5.5
III 6.3 1.156 5.4
IV 5.6 1.032 5.4
I 4.0 0.799 5.0
2005 II 6.3 1.013 6.2
III 7.0 1.156 6.0
IV 6.5 1.032 6.3
1 5.2 0.799 6.5
2006 11 6.5 1.013 6.4
III 7.5 1.156 6.5
IV 7.2 1.032 7.0
I 6.0 0.799 7.5
2007 11 7.0 1.013 6.9
III 8.4 1.156 7.3
IV 7.7 1.032 7.5
3.2.2 Fitting a Trend Line
The next step after de-seasonalising the data is to develop the trend line.
We shall here use the method of least squares (trend line). Choice of the
origin in the middle of the data with a suitable scaling simplifies
computations considerably.
150
To fit a straight line of the form Y = a + bX to the de-easonalised

sales, we proceed as shown in Table 5 below. Example:
Table 6 Computation of Trend
Year Quarter Deseasonalised X X2 XY
sales (Y)
2003 I 6.9 -19 361 -131.1
II 5.3 -17 289 -90.1
III 6.9 -15 225 -93.0
IV 5.8 -13 169 -75.4
2004 I 6.0 121 -66.0
II 5.5 81 -49.5
Cis
I
III 5.4 49 -37.8
IV 5.4 25 -27.0
2005 I 5.0 9 -15.0
II 6.2 1 -6.2
III 6.0 1 6.0
....I.
IV 6.3 9 18.9
2006 I 6.5 25 32.5
In r--
II 6.4 49 44.8
III 6.5 81 58.5
cr, —
IV 7.0 121 77.0

2007 I 7.5 13 169 97.5
II 6.9 15 225 103.5
111 7.3 17 289 124.1
IV 7.5 19 361 142.5
TOTAL 125.6 0 2660 114.2
the trend line is a = SY = 125.6 6.3

20
 114.2 = 0.04X
EX2 2660
 6.3 + 0.04Xb 3.2.3

Identifying Cyclical Variation
The cyclical component is identified by measuring de-seasonalised

variation around the trend line, as the ratio of the actual de-seasonalised
sales to the value predicted by the trend line. The computations are
shown in Table 6 below:
151
Table 7 Computation of Cyclical Variation

De-scasonalised Trend Y
Year Quarter Sales (Y) a + bX a + bX
2003 I 6.9 5.54 1.245
II 5.3 5.62 0.943
III 6.2 5.70 1.088
IV 5.8 5.78 1.003
2004 I 6.0 5.86 1.024
11 5.5 5.94 0.926
111 5.4 6.02 0.897
IV 5.4 6.10 0.885
2005 I 5.0 6.18 0.809
II 6.2 6.26 0.990
III 6.0 6.34 0.946
IV 6.3 6.42 0.981
2006 I 6.5 6.50 1.000
II 6.4 6.58 0.973
III 6.5 6.66 0.976
IV 7.0 6.74 1.039
2007 I 7.5 6.82 1.110
II 6.9 6.90 1.000
III 7.3 6.98 1.046
IV 7.5 7.06 1.062
The r and om or . rr egu lar var iation is assu med to b e r elatively

insignificant. Attempts have been made to describe the time series in
this problem using the trend, cyclical and seasonal components. Figure
V represents the original time series, its four quarter moving average
(containing the trend and cycle components) and the trend line.
152
MBE 839 QUANTI TATIV E TECHNIQ UES FO R BAN KING AND FINAN CE
Figure 5 — Time Series with Trend and Moving Averages

- ORIGINAL TIME SERIES
FOUR - QUA RTER MOVING AV ERAG E
FOUR - QUA RTER MOVING AV ERAG E
LT]
3.0
2003 2004 2005 2006 2007

YEAR
-
10
3.3 Forecasting with the Decomposed Components of the

Time Series
Suppose that the management of the firm is interested in estimating the

sales for the second and third quarters of 2008. The estimates of the
deseasonalised sales can be obtained by using the trend line:
6.3 + 0.04 (23)
7.22 (2nd Quarter 2008)
and Y = 6.3 + 0.04 (25)
7.30 (3rd Quarter 2008)
These estimates will now have to be seasonalised for the second and
third quarters respectively. This can be done as follows:
For 2008 2nd quarter scasonalised sales estimate = 7.22 x 1.013

= 7.31
For 2008 3rd quarter seasonalised sales estimate = 7.30 x 1.56
= 8.44
153
MBE 839 QUANTITATIVE TEO INIQUES FOR BAN KING AND FINAN CE
Thus, on the basis of the above analysis, the sales estimates of the firm
for the second and third quarters of 1988 are N7.31k and N8.44k
respectively.
These estimates have been obtained by taking the trend and seasonal
variations into account. Cyclical and irregular components have not
been taken into account. The procedure for cyclical variations only
helps to study past behaviour and does not help in predicting the future
behaviour.
Moreover, random or irregular variations are difficult to quantify.
SELF ASSESSMENT EXERCISES
(i) What do you understand by time series analysis? How would

you go about conducting such an analysis for forecasting the
sales of a product in your firm?
(ii) Find the 4 — quarter moving average of the following time series
representing the quarterly production of Tea in Adamawa State.
Quarters
Year I II III IV
2003 5 1 10 17
2004 7 1 10 16
2005 9 3 8 18
2006 5 2 15 19
2007 8 4 14 20
4.0 CONCLUSION
In this unit, you have been shown the role of time series in business
forecasting and how to apply it in the different scope in decision
making.
5.0 SUMMARY
Some procedures for time series analysis have been described in this
unit with a view to making more accurate and reliable forecasts of the
future.
In this unit, time series models of historical data on demand and variable
interest were discussed, expressing that one is dealing with projecting
into the future from the past. It is mainly short-term forecasting models.
154
MBE 839 QUANTITATIVE TED INIQUES FOR BANKING AND FINANCE
The decomposition method has been discussed — with time series broken
up into seasonal, trend, cycle and random components from the given
data and reconstructed for forecasting purposes — with detailed example
to illustrate the procedure.

-
Choice of the origin in the data with suitable scaling simplifies

computations considerably. Find the least squares line of best fit from
the following variables:
X 1 2 3 4 5 6 7
1 2 6 7 10 16 21
Indira Gandhi National Open University (February, 2001). Young

Printing Press.

155
MBE' 839 QUANTITATIVE TECHNIQUES FOR BANKING AND FINANCE
UNIT 3 CORR EL ATION
CONTENTS
1.0 Introduction
2.0 Objective
3.0 Main Content
3.1 The Correlation Co-efficient (R)
3.2 The significant of the Correlation Coefficient
3.3 Rank Correlation
4.0 Conclusion
5.0 Summary
6.0 Tutor Marked Assignment
1.0 INTRODUCTION
We could quote cases where it has been useful to establish a relationship

between two sets of variables. In fact, the establishment of such
relationship has, in some cases, changed our whole way of life. A good
example of this is the relationship between the numbers of cigarettes
smoked and the incidence of lung cancer. Another example is the
relationship between the probabilities of infection and the degree of
exposure to a certain disease. However, it can be just as useful to
establish that a relationship does not exist — the statistician has been the
scourge of the purveyors of quack medicine! A further use of
establishing relationships is to make predictions about a variable that is
difficult or costly to measure directly. For example, if we can establish a
relationship between the destructive power of a bomb and the amount of
explosive it contains, we can now examine the form that such
relationships take.
2.0 O B J EC TI VE S
After completion of this unit, you should be able to:
 explain the meaning of correlation and its role in decision

making;
 compute the correlation coefficient between two variables from
sample observations;
 test for the significance of the correlation coefficient in business
decision;
156
3.0 MAIN C ONTENT
3.1 The Correlation Co-Efficient (R)

Let us suppose that we take two variables, x and y. we suspect that there
is a relationship between the variables and we plot the results on a
graph. The two variables which we suspect are related form a bi-variate
distribution, and the graph drawn is called a scatter diagram. The scatter
diagram might look like this
X
Figure 13.1
We notice that large values of y are associated with large values of x,

and small values of y are associated with small values of x. we would
say that in this case the variables are positively correlated. In other
words, positive correlation exists when both variables increase together.
Were we to plot the number of motor vehicles registered against the
number of road accidents we would expect there to be a positive
correlation. Can you think of any other examples yourself'?
Now it is possible that the scatter diagram might look like this:
X
Figure 13.2
157
MB!: 839 QUANTITATIVE TECIINIQUES FOR BANKING AND FINANCE
In this case small values of y are associated with large values of x and
large values of y arc associated with small values of x. here, the
variables are negatively correlated, which exists when an increase in one
variable is associated with a decrease in the other. An example of
negative correlation is the association between the number of T.V.
licenses issued and cinema admissions. You should try to compile a list
of such examples.
A third case is where the scatter diagram looks like this:
X
Figure 13.3
These points are scattered at random and here we can say that there is an
absence of correlation between the variables. This of course will be the
most common of the three cases if we randomly select bi -variate
distributions.
Consider now the three diagrams below:
X
Figure 13.4(a)
158
X
Figure 13.4(b)
Figure 13.4(c) X
In each case we see that the variables are positively correlated. The
diagrams clearly show that association is strongest in case (c) and
weakest in case (a). in case (c) all the points are on a straight line and for
any value of one variable we could predict precisely the corresponding
value of the other variable. Case (c) shows perfect positive correlation.
Now in practice we will not meet perfect positive correlation, but we
will meet examples that approach it quite closely. With some
justification we can regard such examples as showing perfect positive
correlation and any deviations of points from a straight line as resulting
from experimental error.
Clearly, we need some measure of correlation, and our measure must
exist between positive and negative correlation. How can we derive such
a measure? Well, you are probably thinking back to the chapter on the
Time Series and releasing that some connection must exist between
correlation and the 'least squares' method used in that chapter. We saw
that if a number of points were plotted on a graph the least squares
line of best fit could be given as y = mx + c, where,
159
Exy a E y
Ex2 — (D)2
and
1 (Ey — Ma)
If this is not clear in your mind you should read again the appropriate
section.
Examine carefully how we obtained this line. We selected a point, say

F,, with the coordinates (XnY„) and said that the value of y on this line
when X = Xn was X„M + C; and the deviation of the point (Xn l',,) from
the point on the line was
Y„— (XnM + C) = Y,,— X,,M C
Now what are the implications of this? We are calculating the value of y
on the line for a given value of x — which implies that we are using the
values of x to predict the values of y. we are supposing that all
deviations from the line occurs as a result of errors in the value y, and so
we have been measuring the vertical deviations of the points from the
line. Now must we find the regression line this way? We could have
considered instead the horizontal deviations of the points from the line.
To do this we would have to consider the line
X= M1y + C
And for any value y„ the value of x on the line would be
Y„MI + C
Hence the deviation of the point (x,,y,) from this line is
xn— (y,,m1 +c = xn—ynmi— ci
160
You can see this in the diagram below:
Figure 13.5
(You were asked to consider this idea in the tutorial following the
analysis of least squares.)
We can use the same method to find m1 and c1 that we used to find m
and c. this would give:
Exy — ExEy
Ex2 — (£,42
and
Ci= 1 (Ex — MIEy)
To find the regression line y = mx +c, we took the value of x as given

and calculated the deviations from y this line is called the regression
of y on x, and is used to estimated a value of y when we are given a
value of x. to the regression line of x = miy +cl, we took the value of y
as given and calculated the deviations from x. this is the regression line
of x on y and is used to estimate a value of x given a value of y. in the
first case we assume that all deviations from a linear relationship are
caused by differences between the observed values of y, and the values
of y calculated from the regression equation
Example 1
When we considered the least squares method we found that for the bi-
variate distribution
161
MBF 839 QUANTI TATIVE TECHNI QUES FOR BAN KING AND F INAN CE
1 3 4 5 6 7
7
1 2 6 7 10 16 21
the least squares line of best fit was

y = 3.29x — 4.16
This is the regression line of y on x. let us now find the regression line
of x on y.
X y2 xy
1 1 1 1
2 9 4 4
3 6 36 18
4 7 49 28
5 10 100 50
6 16 256 96
7 21 441 147
78 63 887 344
a =28 Ey = 63 Ey2 = 887 Exy = 344
E,xy — ExEy
Ex2 — (Ex)2
2y2 — (E1)2
344 — 28 x 7
887— (63)2
7
0.288
= 1 (Ex — MlEy)
C' 1 (28 — 0.28 x 63)

7
1.43
162
The equation is x = 0.288y + 1.43
3.2 The Significant of the Correlation Coefficient
Now let us see if we can obtain a measure of correlation, and all this the
correlation coefficient (r). We will define r as the product of the
coefficients m and mi in the regression equations, i.e.
MMI
We shall consider first the case of perfect, positive correlation: the

points would all lie on a straight line with a positive slope. The
regressive lines would coincide and the two equations y = mx + c and x
= m1y +c, would be identical. Suppose we re-arrange the second
equation:
mly ci
So, y = x - ci
mi
The gradient of this line is 1. As the regression lines coincide, the

gradient of this
m
is also m. Hence,
and the product of the coefficients will be
1 x mi = 1
ml
Thus, with perfect, positive correlation, r = 1: a sensible value. Now let

us consider the case where no correlation exists. The regression lines
will be parallel to the axes and look like the diagram on the facing page.
You may be wondering why the regression lines are parallel to the axes.
If no correlation exists, then y does not depend upon x, and the
regression equation is y = c. you will remember from the introductory
section that the line y = c is parallel to the x axis. Likewise, the
regression
163
x on y
y on x
X
Figure 13.6
Equation x on y must be x = c l, which is parallel to the y axis, in this

case m = 0 and m1= 0, so the correlation coefficient is 0 x 0 = 0
When correlation is absent the value of the correlation coefficient is

zero, again a sensible value.
Finally let us consider perfect negative correlation. Again the regressive

lines will coincide, but this time they will both have a negative slope.
The regression lines will have the equations
mx + c andx = + ci
Earlier we found that m =' so in this case — m = -1 and the product

mi mi
of the coefficients will be — ml x — 1 = 1. Hence if perfect negative

correlation exists the mi
value of our correlation coefficient will be one — the same as for perfect
positive correlation. Now this will just not do! Our correlation
coefficient must be able to distinguish between positive and negative
correlation, but our coefficient r = mm, will not do this! What shall we
do? Let us try making r the squares root of the product of the coefficient
s, i.e.
7' 2
Now when correlation is absent is r 2 = 0 and r = 0 which is quite

satisfactory. When correlation is perfect r 2 = 1 and r = ±1. This is much
better.
The positive root can signify perfect positive correlation and the
negative root perfect negative correlation. But how will we know which
164
MBE 839
QUANTITATIVE TECHNIQUES FOR BANKING AND FINANCE
root to take? Well, if correlation is positive, the gradient of the

regression lines will be positive. Likewise if correlation is negative both
m and m, will be negative. So if m and m, are positive take the positive
root of their product and if m and m1 are negative take the negative root.
measure of correlation, then
Let us summaries what we have learned so far. If we take ? = mm, as a
— 1 < r2 Cl
If 0
correlation is absent
1
correlation is perfect and positive
—1 correlation is perfect and positive
so it follows that if 0 < r Cl, correlation is positive but not perfect and if
— I < r 2 <1, correlation is negative but not perfect The closer r
between the two variables.
approaches to its limits (-1 and +1) the stronger is the association
Now let us consider the correlation coefficient for the example we
3.29 x 0.288 and r r +0.97 -
considered earlier. We found that m = 3.29 and that ml = 0.288 so r 2= A

high degree of positive correlation exists. This leads us to expect that this.
both regression lines would almost coincide. The diagram clearly shows
Now we have defined r 2 as mm i , and we have seen that it is a
satisfactory measure of correlation. It must follow that:
Sty - E._gy"
7'2
E x 2 .. E L I)!
Now suppose that we divide the top andEbottom
y 2 of each
- fraction by n the
number of pairs in the bi-variate distribution. Then
I T.ay - ixLy
1_ lxy -
1 a y
165
You should recognize that the denomination of the first fraction is the
variation of x and that the denomination of the second fraction is the
variance of y. so we can write
L FY-xY
n n [ L
var (x)
var (5r)
C
LSt Y x LL ExY - D-4:Y9
var (x) var (5')
We call the numerator of this last fraction the Covariance so now we can
say
R= COI____Lgig---(1Ce?--VIC
Gx csr
Writing the correlation coefficient in this way we have overcome the
problem of the sign of the coefficient: it is the same as the sign of the
covariance.
Example 2
Index of Earnings and prices (1996 – 100)
1999 2000 2001
Quarter3 4 1 2 3 4 1 2 3 4
Earnings 121 125 130 135 138 145 150 154 158 162
Prices 113 115 118 120 121 12 4 127 1 31 133 1 35
We wish to find the correlation coefficient between earnings and

prices. It will simplify the arithmetic if we take deviations from 140
on the
earnings index and deviations from 125 on the prices index.
166
x2 2
x Y Y Xy
-19 -12 361 144 228
-15 -10 225 100 150
-10 -7 100 49 70
-5 -5 25 25 25
-2 -4 4 16 8
5 -1 25 1 -5
10 2 100 4 20
14 6 196 36 84
18 8 324 64 144
22 10 484 100 220
18 -13 1844 539 944
Table 1
Covariance (xy) = 1 [ Exy — Laid
= 1 [ 944-18x-13]
10 10
= 1 [ 944 + 23.4] =96.74

10
844 — 182]
10 10
=J 1 [ 1844 — 32.4] = '181. 16 13.46

10
1[ — (Ev)2] = \I 1 [ 539 — (-13)2 ]
10 10
= \I 1 [ 539 — 16.9] = \I 52.21

7 23
10
r = Covariance (xv) = 96.74 = 0.994

ax ay = 13.46 X 7.23
3.3 The Significance of the Correlation
Calculating the correlation coefficient has little meaning unless we can
test its significance. Suppose we have a bi-variate population with a zero
correlation coefficient, and we draw samples of n pairs from this
population. Of course, it would be unreasonable to expect all our
167
MBE 839 QUANTITATIVE TECI INIQUES FOR BANKING AND FINANCE
samples to have a zero correlation coefficient: some will be greater than

zero and some less, but we would expect the average value of
correlation coefficients to be zero. This give us some idea as to how we
can assess the significance of r. suppose a sample of n pairs gives a
correlation coefficient of r, and suppose we adopt the Null Hypothesis
that the sample is drawn from an uncorrelated population then if we can
find the standard error we should be able to calculate the probability of
obtaining our observed value of r.
Now we have already stated that correlation coefficients of samples all

drawn from the same population are liable to sampling fluctuations. If
the sample size is fairly large, then the distribution of r will be Normal.
However, unrepresentative pairs will have a significant effect on the
correlation coefficient of the sample, and the range of the correlation
coefficient will be greater (the effect of an unrepresentative pair will be
lost in a large sample). The distribution of the correlation coefficient for
small values of n will not be Normal. This is illustrated in the diagram
below
Large values of n
Ar , Small values of
The distribution of the correlation coefficient

Figure 7
When n is small, the distribution of the correlation coefficient forms the

so-called 'T' distribution. The deviation and use of this distribution is
beyond the scope of this book, and unfortunately most calculations of
the correlation coefficient are from small samples. However, if you turn
to the appendix, you will find that values of r at the 5% and I% levels of
significance are tabulated against differing sizes of n. consulting this
table for n = 10 ( the number of pairs in the last example). We can state
that if no correlation exists there is a 5% chance of r exceeding 0.63 and
a 1% chance of r exceeding 0.76. Now the value of r we obtained in the
168
last example was 0.994. What can we conclude? The probability that
correlation is absent is certainly very much less than 1% and we must
state that positive correlation seems highly likely.
This section on the correlation coefficient will end with a word of

warning. When dealing with correlation coefficient one must never
forget common sense. A significant correlation coefficient does not
necessarily imply cause and effect. Many highly significant correlations
are quite non-sensical. Statisticians in the past pointed to the highly
significant correlation between the annual issues of broadcasting
licenses and the annual rate of admissions into mental institutions. The
only logical conclusion you can draw from this is that quite by chance
both statistics were increasing at the same rate. Even when a significant
correlation seems reasonable you should not assume that cause and
effect is established, as a third variable may be influencing both the
others and thus explain the correlation. The high degree of positive
correlation between overcrowding causes a high infant mortality rate;
through in fact both are indicative of income levels.
3.4 Rank Correlation
Example 3
The correlation coefficient r is rather awkward to calculate and you may

prefer to use instead the Rank Correlation coefficient. This is not as
accurate as the correlation coefficient r but does still give a fairly
reliable indication as to the degree of correlation between two variables.
Let us examine the following vicariate distribution.
2001 Production of Private Value of Exports
Months Cars (thousands) (N million)
February 134 627
March 142 771
April 138 787
May 156 755
June 176 799
August 119 782
September 159 834
October 151 810
November 145 759
December 180 832
Table 2
If we wish to calculate the rank correlation coefficient we must rank the

values in each column. Let us consider the production of private cars
over the ten months. The highest output was in December so we give
169
MBE 839 QUANTITATIVE TECIINIQUF,S FOR BANKING AND FINANCE
this month a rank of one. The second greater output was in June and we
give this month a rank of two. We continue in this way ranking each
month according to its output of private cars. When we have done this
we rank each month according to the value of exports. Our bi-viate
distribution would now look like this:
Months Output (x) Rank of Exports (y) Rank of

Output Export (y)
February 134 9 627 10
March 142 7 771 7
April 138 8 787
May 156 4 755
June 176 2 799 4
Au:1.1st 119 10 782 6
September 159 3 834 1
October 151 5 810 3
November 145 6 759 8
December 180 1 832 2
Table 3
We can now calculate the correlation coefficient using the ranks instead
of the actual values. This will considerably simplify the arithmetic.
Suppose we write p for the rank correlation coefficient. Then
Co-variance (XY)
(Tx cry
before we actually calculate p for this bi-variate distribution let us

consider some. If preliminaries. The ranks (X) are identical to the ranks
(Y) although they will not necessarily appear in the same order. It
follows from this that Ex = Ey and that a, = a y. If you cannot see that
this must be so verify these facts for yourself using the ranks in the table
above. Now we can write
Co-variance (XY)
a2 (x)
so we need calculate only one variance that two standard deviations. Let
us now consider the bi-variate distribution with n pairs. If we calculate p
for this distribution we may well obtain an expression for p which is
easier to manage than the one above. The ranks (A') in this distribution
will be
1, 2, 3 ....................... , n
170
though not necessarily in that order. We will need to calculate the

variance of the ranks (X).
Variance (X) = [Ex2 (a)2}
Now E x = 1 + 2 + 3 + ....................... +n i.e. the sum of the first n

integers. When you worked the tutorial following geometric progression
(it was in the chapter on Discounted Cash Flow) you found the sum of
this series and you were asked to keep the result. Find this result and
check that it is n2 + n. We can now write EX = n2 + n. in the same
Ex2 = 12 + 22 +32 +.................................. + n2
Now it can be shown that 12+22 + 32 + ............... + n2 = n(n +1)(2n + 1)

6
If you wish to verify this; an exercise in the next tutorial will help you to
do so.
We can now say
Variance (X) =- 1 [n(n + 1) (2n + 1) — (n2 + n)2 X 1]

n 6 2 n
 1 [n(n + 1) 2n + n)— n4 + 2n3 + n2]

6 4n
 1 [n(n +1) (2n + 1) — n3 + 2n2 + n

6 4
 1(n+ 1)(2n + 1) — n2 + 2n + 11
6 4

1(n+ 1)(2 +1)—(n+ 1)2]
6 4
Well, this does not seem to be much of a simplification, but if we now
consider the covariance the simplification will eventually occur.
Covariance (XY) - 1 [ EXY — EXEY ]
 1 [ EXY — (EX)2 ]
171
Now we have already found that the last term in the expression can be
written as
(n + 1)2.
4
We shall now see if we can find another way of writing EXY. To do this
we shall use the fact that
0( )02 , 2xy +y2
So E(X — Y)2 = EX2 - 2EXY + EY2
The quantity (X — Y)2 could be found by taking the difference between

the ranks, squaring them and summing the squares. If we write D 2for
this quantity
ED2 = X2 — 2XY + Y2
So EXY = EX2 + EY2 — ED2

2
As EX2 = EY2,
EXY = 2EX2 — DE 2 Ex2 ED2
2 2
We have already found an expression for EX2, so

EXY = n(n + 1)(2n + 1) — ED2
6 2
We can now write the Covariance like this:
Covariance (XY) n(n + 1)(2n + 1) — ED2 — (n +1)2

6 2n 4
(n + 1)(2n + 1) ED2 — (n + 1)2

6 2n 4
and the rank correlation coefficient like this:
172
(n 1) (2n + 1) - ED2— (n + 1)2

6 2n 4
(n + 1) (2n + 1) (n + 1)2
6 4
ED2
2n
=1—
Or 4- 1) (2n 1) _ (n + 1)2
6 4
6ED2
=1—
2n (n -1- 1) (2n -1- I) _ 3n(n + 1)2
= 1 — 6ED2
n(n2 —1)
So the attempt at simplification has proved to be worthwhile as the rank
correlation coefficient will now be simple to calculate.
X Y D D2
9 10 -1 1
7 7 0 0
8 5 3 9
4 9 -5 25
, 4 -2 4
Table 13.4a
X Y D D2
10 6 4 16
3 1 2 4
5 3 2 4
6 8 -2 4
1 2 -1 1
0 68
Table 13.4b
Notice that ED = 0. This must be so (why?) and is a useful check on the

calculations performed.
1— 6 X 68 =0.588
10002 -
173
4.0 CONCLUSION
In this unit, you have been taken through the concept of correlation
coefficient, its computation between two variables from sample
observations and how to apply correlation in decision-making.
5.0 SUMMARY
In this unit, you have been taken through to-
understand the meaning of correlation;

compute the correlation coefficient between two variables from
sample observations;
test for the significance of the correlation coefficient in business
decision;
compute the rank correlation coefficient when rankings rather
than actual values for variables are known;

-
1) What do you understand by the term correlation?
2) Explain how the study of correlation helps in forecasting demand

of a product.

Pandey, I.M. (2005). Financial Management, 9th Edition Vikas

Publishing House PVT LTD. New Delhi, India
174
U N I T 4 R E G R E S S I O N AN A L Y S I S
CONTENTS
1.0 Introduction
2.0 Objectives
3.0 Main Cont ent
3.1 The Need for Statistical Analysis
3.1.1 Linear Regression
3.2 Fitting a Straight Line
3.2.1 Examining the Fitted Straight Line
3.3 Variety of Regression Models.
4.0 Conclusion
5.0 Summary
1.0 INTRODUCTION
In industry and business today, large amounts of data are continuously

being generated. This data may be pertaining, for instance, to a
company's annual sales, capacity utilisation, turnover, profits,
manpower levels, absenteeism or some other variable of direct interest
to management.
Or there might be a technical data regarding a process such as

temperature or pressure at certain crucial points, concentration of a
certain chemical in the product or the breaking strength of the sample
produced or one of a large quality attributes.
The accumulated data may be used to get information about the system
(as for instance, what happens to the output of the plant when
temperature is reduced by half) or to visually depict the past pattern of
behaviour (as often happens in company's annual meetings where
records of company progress are projected) as simply used for control
purposes to check if the process or system is operating as designed (as
for instance of quality control).
The interest in regression is primarily for the first purpose, mainly to

extract the main features of the relationships hidden in or implied by the
mass of data.
175
2.0 OBJECTIVES
After completion of this unit, you should be able to:
 explain the role of regression in establishing mathematical

relationships between dependent and independent variables from
given data
 use the least square criterion to estimate the model parameters;
 determine the stand or the errors of estimate of the forecast and
estimated parameters
 make meaninLi ul forecasts from given data by fitting any
function, linear in unknown parameters.
3.0 MAIN CONTENT
3.1 The Need for Statistical Analysis
For the system understudy there may be variables and it is of interest to

examine the effects that some variables exert (or appear to exert) on
others. The exact functional relationship between variables may be so
complex but we may wish to approximate to this functional relationship
by some simple mathematical function such as straight line or a
polynomial which approximates to the true function over certain limited
ranges of the variables involved.
Broadly speaking, one would have to undergo the following sequence of

steps in determining the relationship between variables, assuming we
have data points already:
(1) Identify the independent and response variables;

(2) Make a guess form of the relation (linear, quadratic, cycle etc.)
between the dependent and independent variables (graphically
plot systematic tabulation to suggest trends or patterns);
(3) Estimate the parameters of the tentative entertained model in (2)
above. E.g. if a straight line was to be fitted — what is the slope
and intercept of this line?
(4) Having obtained the mathematical model, conduct an error
analysis to see how good the model fits into the actual data;
(5) Stop, if satisfied with model otherwise repeat (2) to (4) for
another choice of the model form in (2).
What is Regression?
Suppose one considers the height and weight of adult males for some
given population. If one plots the pair (X 1 , X2 ) = (height, weight).
diagram like figure 1 will result. This is conventional scattered diagram.
176
Note that for any given height, there is a range of observed weights and
vice versa. The variations will be partially due to variation errors but
primarily due to variations between individuals. Thus, no unique
relationship between height and weight can be expected. But one can
note that average observed weight for a given observed height increases
as height increases.
The locus of average of observed weight for given observed height (as
height varies) is called the Regression Curve of weight on height. Let
us denote it by X2 = f (xi). There also exists a regression curve of height
on weight similarly defined which we can denote by X1 = g (X2).
Let us assume that these two "curves" are both straight lines (which in
general may not be). In general, these two curves are not the same as
indicated by the two lines in figure 14.1 below:
•
Figure I
90 Xi = g (X2 )
80 X (X1 )
Weight in kg (X2)
70
— 60
164 168 172 176 180 184 188 192

Height in cms
A pair of random variables such as (height, weight) follows some sort of

bi-variate probability distribution. When we are concerned with the
dependence of a random variable but not a random variable, an equation
that relates Y to X is usually called a regression equation. Similarly,
when more than one independent variable is involved, you may examine
the way in which a response Y depends on variables X I, X2 ... Xn. The
determination of a regression equation from data which cover certain
areas of the X — space as Y = f (Xi, X2 ... Xn).
177
MBE 839 QUANTITATIVE TE,CI INIQUES FOR BANKING AND FINANCE
3.1.1 Linear Regression
The simplest and most commonly used relationship between two

variables is that of a straight line — written in first order model as Y = pa
+ p, x + c .... (1). (That is, for a given X, a corresponding observation
Y consist of the value of 130 + p, X plus an amount €, the increment by
which an individual Y may fall off the regression line. Equation (1) is
the model of what is believed. pa, 130are called the parameters of the
model whose values have been obtained from the actual data. When a
model is linear or non-linear, it is referring to linearity or non-linearity
in the parameters. The value of the highest power of independent
variable in the model is called the order of the model.
For example:
= 13o ± x + 132 X2 -I- €
is a second order (in X) linear (in the P s) regression model. Now in the
model of equation (1 ) 130, 13, and € are unknown and in fact would be
difficult to discover from observation. However, pa, p, remain fixed,
and although we cannot find them exactly without examining all
possible occurrences of Y and X, we can use the information provided
by the actual data to give us estimate b0 and 131 of 130 and PI.
Thus, we can write:
7(
1 b0 + biX (2)
where Sr (Y hat) denotes the predicted value of Y for a given X, when b 0

and b1 are determined. Equation (2) could then be used as a predictive
equation; substitution of a value of X would provide a prediction of the
true mean value of Y for that of X.
3.2 Fitting a Straight Line
Least Squares Criterion
In fitting a straight line (or any other function) to a set of data points, we
would expect some points to fall above or below the line resulting in
both positive and negative error terms as shown in figure 14.2 below:
178
MEW 839 QUANTITATIVE TECHNIQUES FOR BANKING AND FINANCE
•
Figure II — The Least Squares
Crite ion
x, = g (X2)
Response Variable
—ye
error The fitted function attempts
to minimise the sum of
Squares of errors
Independent Variable
Figure 14.2
It is true that we would like the overall error to be as small as possible.

The most common criterion in the determination of model parameters is
to minimise the sum of squares of errors or residuals as they are often
called. This is known as the least square criterion, and is the one most
commonly used in Regression analysis (as well as Time Series
Analysis).
This is, however, not the only criterion available. One may, for
instance, minimise the sum of absolute deviations, which is equivalent
to minimising the mean absolute deviation (M.A.D).
The least squares criterion, however, has the following main advantages:
(i) It is simple and intuit ively appealing.

(ii) It results in linear equations (normal equations) for solution of
parameters which are easy to solve.
(iii) It results in estimates of quality of fit and intervals of confidence
of predicted values rather easily.
For other discussions and equation development refer to Time Series

Unit on least square criterion.
179
An Example:
Data on the annual sales of a company in Naira over the past eleven
years is shown in the table below. Determine a suitable straight line
regression model, Y = (3 0 -I- x C for the data in the table:
Year Annual Sales in Naira
1988 1
1989 5
1990 4
1991 7
1992 10
1993 8
1994 9
1995 13
1996 14
1997 13
1998 18
Solution:
The independent variable in this problem is the year whereas the

response variable is the annual sales. Although you could take the
actual year as the independent variable itself, a judicious choice of the
origin at the middle year 1993 with the corresponding X values for other
years as —5, —4, —3, —2, —1, 0, 1, 2, 3, 4, 5 would simplify calculations.
(From b EX,Y, —2 (YX1/,',Y1) / n (3)

EX, — (EX)2 In
(Ref: Least Squares Criterion in Time Series Unit)
we see that to estimate the parameter b, we require the four summations:

2
EX, Ey, E x, , EX,Y,
Thus, calculations can be shown below where the totals of the four
columns yield the four desired summations:
180
S/N X1 Yi X,2 X,Y,

1 —5 1 25 —5
2 —4 5 16 —20
3 —3 4 9 —12
4 —2 7 4 —14
5 —1 10 1 —10
6 0 8 0 0
7 1 9 1 9
8 2 13 4 26
9 3 14 9 42
10 4 13 16 52
11 5 18 25 90
EX, = 0 EY, = 102 EX12= 110 'CY ; = 158
.
Table14.1
We find that n = 11
Xxi = 0 0/11 = 0
EXY = 102 Y = 102/11
X2 = 110 = 158
=
b, EXLEY,
2
EN — (EXY /n
158 = 1.44
110
The fitted equation is:
Y + b, (X —7)
Or
9.27 + 1.44X
Thus, the parameters 130 and 131 of the model Y = 130 + 131x + E are
estimated b0 and bl which in this case are 9.27 and 1.44 respectively.
Now that the model is completely specified, we can obtain the predicted
values cr, and the errors as residuals Y . Y .
- — - I respectively to the eleven
observations enough, the ANOVA table 14.2 below:
181
S/N X, Yi Y1—Y1
1 —5 1 2.07 —1.07
2 —4 5 3.51 1.49
3 —3 4 4.95 —0.95
4 —2 7 6.39 —0.61
5 —1 10 7.83 2.17
6 0 8 9.27 —1.27
7 1 9 10.71 —1.71
8 2 13 12.15 0.85
9 3 14 13.59 0.41
10 4 13 15.03 —2.0
11 5 18 16.47 1.53
Table 14.2
To determine whether the fit is good enough, the Analysis of Variance
(ANOVA) table can be constructed:

(Sum of squares - SS due to regression
= b, ,
(Associated degrees of
freedom = 1) = [ I,X,Y, — (y,x,zy,) / n 1
E x, 2 (E xi ) 2 i n
= (158)2 = 226.95
110
The total (corrected) SS I x ( y) ( 1 . 1

?
(Associated degrees of
freedom = 11 —1 = 10) 1194 — (102)2/11
 1194 — 945.82
 248.18
The value R2 - SS due to regression

SS about mean
 226.05 = 0.9145
248.18
indicating that regression line explains 91.45% of the total variation

about the mean.
182
Note:
I -- M —' e.--
sum of squares sum of squares + sum of squares

about the mean about regression due to regression
Any sum of squares is associated with it a number called its degree of

freedom. This number indicates how many independent pieces of
information involving the n independent numbers Y Y
-1, - Yn are2, • • • •,
needed to compile the sum of squares.

3 2 1 Examining the fitted Straight Line
In fitting the linear model Y =h o + p, x + E using the least square
criterion as indicated earlier in 3.2, no assumption was made about
probability distribution. The method of estimating the parameters 130
and Pi tried only to minimise the sum of squares of the errors or
residuals, and that simply involved the solution of simultaneous linear
equations (Refer to Time Series Chart). However, in order to be able to
evaluate the precision of the estimated parameters to provide confidence
intervals for forecast values, it is necessary to make the following
assumptions in the basic model Y, =130 + JbiXi +x + E, i = 1,2 ... n.
With some mathematical assumptions, the following could be
determined — (Ref. Time Series and Correlation units).
(1) Standard error of the slope b, and confidence interval of 131.

(2) Standard error of the intercept b0 and confidence interval for 130.
(3) Standard error or Y, the predicted value.
(4) Significance of regression.
(5) Percentage variation explained.
Note:
Errors that occur in many real life situations tend to be normally

distributed due to the Central Limit Theorem. In practice, an error term
such as is a sum of errors from several sources. Then no matter what
the probability distribution of the separate errors may be, their sum will
have a distribution that will tend more and more to the normal
distribution as the number of components increases, by the Central Limit
Theorem.
The studies in Time Series and correlation have established these facts.
You are encouraged to refer to them.
183
An Example of the Calculations:

The various computations stated for a straight line regression situation in
Section 3.2.1 will now be illustrated for the example of annual sales data
for a company that was considered earlier in Section 3.2. Recall that the
fitted regression equation was
t = 9.27 + 1.44X
3.3 Variety of Regression Models

The methods of regression analysis have been illustrated in this unit for
the case of fitting a straight line to a given set of data points. However,
the same principles are applicable to the fitting of a variety of other
functions which may relevant in certain situations like seasonal, which
leads to cyclical forecast and Trend, Polynominal of various order,
Multiplicative Models, Linear and Non-linear Regression (as found in
the Time Series and Correlation Units).
4.0 CONCLUSION
In this unit, fundamentals of linear regression have been highlighted.
Broadly, the fitting of any chosen mathematical function to given data is
known as regression analysis. It has been established that the estimation
of the parameters of this model is accomplished by the least squares
criterion which tries to minimise the sum of squares of the errors for all
the data points.
5.0 SUMMARY
We have touched on the following salient issues:
 How the parameters of a fitted straight line model are estimated,

has been illustrated.
 After the model is fitted to the data, the next logical question is to
find out how good the quality of fit is?
 Making of qualitative statements regarding confidence limits for
estimates of the parameters as well as the forecast values.
 Comparison of alternative regression models in hypothesis.
 Emphasis that the method of least squares used in linear
regression is applicable to a wide class of models.
 Regression is a potent device for establishing relationships
between variables from the given data and the discovered
relationship can be used for predictive purposes. Some of the
models used in forecasting of demand rely heavily on regression
analysis.
184

1) Explain the need for linear regression in statistical analysis.
2) In establishing a relationship between variables from a given

data, discuss briefly highlighting the fundamental steps to be
considered.
7.0 REFERENCES AND FURTHER READING

Indira Gandhi National Open University (2000). Young Printing Press,
Delhi.

185
U N IT 5 PROJ ECT P LANN IN G
CONTENTS
1.0 Introduction
2.0 Objectives
3.0 Main Content
3.1 Nature of Project
3.1.1 Project Planning / Management
3.1.2 Project Life Cycle
3.2 The Network
3.2.1 Network — Symbol, Diagram / Convention
3.3 Total Project Time — Critical Path Method (CPM)
3.4 Gantt Chart
4.0 Conclusion
5.0 Summary
1 . 0 IN TR O D U C T IO N
Planning for a project is expedient if it is to be undertaken efficiently;

moreover, if it involves a number of people. All too often, people work
in isolation, quite ignorant of what other people working on the same
project are doing. It is the task of management to coordinate efficiently
the efforts of such people. In the past, planning has not been so
important as it is today as projects were less complex, the rule of thumb
method would work quite well. However, as projects have become
more complex, management (banking, finance etc.) researchers have
turned their attention increasingly to systematic planning. This unit will
consider the complexity of this day products (services) and you will
soon recognise the need for systematic planning. Shortly af ter the
Second World War (1945), researchers evolved a method called
Network Analysis. The impact of the method has been quite dramatic,
largely because it is applicable to such a wide variety of projects. It has
been used to plan production projects, servicing projects, research
projects, sales projects, operational projects and military projects etc.
2.0 OBJECTIVES
 d is c u s s p r o je c t p la n n i n g ;
 explain network in project development with deterministic and
probabilistic times
 cons tr u ct s imp le netw or k d iag r ams
186
3.0 MAIN C ONTENT
3.1 Nature of Projects
Projects are found in series of stages, a life cycle, which include:
 Planning
 Execution
 Ph a s e — o ut
At each stage of this life cycle, a variety of skillful requirements are

involved. In effect, project unit human resources and with diverse
knowledge and skills, most of whom remain together for less than full
life of the project. Some go from project to project as their contributions
are needed (full or part-time basis) — consulting firms — architects,
writers, publishers, Accountants, Financial Analysts etc. tend to be
involved with project on a regular basis.
3.1.1 Project Planning / Management
The length, size and scope of projects varies widely according to the
nature and purpose of the project. Nevertheless, all projects have
something in common. They go through a life cycle, which typically
consists of five phases:
(1) Concept — recognizing the need for a project.

(2) Feasibility Analysis — examines the expected costs, benefits and
risk of undertaking the project.
(3) Planning — spells out the details of work and provides estimates
of the necessary resources — human, time and cost.
(4) Execution — ensure project is done.
(5) Termination — closure is achieved.
3.2 The Network Symbols, Diagram / Convention

—
How does the method work? Well the first stage of analysis is to divide
the project into a number of different activities. An activity is merely a
particular piece of work identifiable as an entity within the project. If
for example, the project under consideration is the servicing of a motor
car, then one of the activities would be check the brakes for wear.
Now an activity within a network is represented by an arrow, with the

description of the activity written on it viz:
Check the brakes for war
187
Mal, 839 QUANTITATIVE TECHNIQUES FOR BANKING AND FINANCE
Note: In addition to activities, events are identified too.
Events mark the point in time when activity is completed and the next
activity can be started. Events are normally represented by circles:
Check the brakes wear Figure I
The event 111

have its brakes tested.
represents the point in time when the car is ready to
Network ?
A network is a convenient method of showing the logical sequence of

activities in a project.
Suppose that in a certain project, there are two activities A and B, and
activity B cannot be started until activity A is completed. Using the
network symbols, these activities can be represented like this:
Figure 2
The event represents the point in time when activity A is

completed, but it also represents the point of time when activity B is can
begin. This represents the situation when activity B depends upon
activity A; when two or more activities are dependent upon the same.
activity. The situation when neither activity D nor E can start until
activity C is complete would be represented like this:
188
Figure 3
The event x represents the point of time when activity C is completed,

and also the point of time when activities D and E can start, so the
diagram clearly shows that D and E depend on C.
Also, it is likely that an activity depends upon more than one other
activity. If activity A cannot start until activities G and F are both
complete — represent as follows:
Figure 4
Dependence Tables
From the introduction, the first task of network analysis is to sort out the
logical sequence of act ivit ies. This is done by constructing a
dependency table. List all the activities, next is the list of activities that
they depend upon. E.g. there is one way in which the Financial
Manager can affect the volume of credit sales and collection period and
consequently investment in accounts receivable. This is through a
dependable change in credit policy.
Credit policy is used to refer to the combination of three decision
variables (assuming) logically:
Introduction Preceding activity

Credit Standard (a)
Credit Terms (b)
Collection Efforts, (c) on which t he
financial manager has
influence. Following the preceding activity will enhance proper

decision.
189
Constructing a dependence table (trend) is often the most difficult part

of project analysis. Obviously, the construction of dependence tables of
any action (activity) like that of establishing a credit arrangement is in
essence, a team activity. All the experts engaged on the project have to
be consulted.
Network Diagram
A network diagram below is composed of a number of arrows and

nodes. The arrows represent the flow of project activities. A network
diagram is generally the preferred approach for usual portrayal of
project activities as follows:
Order
Locate
umiture
•
Interview Furniture set-up
• i r e d
Figure 5
Traine. Remote
Move in
The network diagram describes sequential relationship among major

activities on a project. For example, activity 2 —> 4 cannot be started;
according to the network until activity 1 —> 2 has been completed
(figure 4 above).
A path is a sequence of activities that leads from the starting node to the
finishing node. Thus, the sequence 1 2 — 4 —> 5 —> 6 is a path.
There are two other paths in the network namely: 1 —> 2 —> 5 6 and 1
—  3 5 6. The path with the longest time is of particular interest
because it governs project completion time. The project life cycle
equals the expected time of the longest path; the longest path is the
critical path, and its activities are referred to as critical activities. The
allowable slippage for any path is called slack, and it reflects the
difference between the length of a given path and the length of the
critical path.
190
Network Conventions
Developing and interpreting network diagrams requires some familiarity

with networking conventions. Discussion centres only on some of the
most basic and most common feature of network diagram. As above,
figure 1, 2, 3 and 4 provides background for understanding the basic
concepts associated with precedence diagrams and permits us to solve
typical problems. See some more examples like the following:
Figure 6
When multiple activities enter a node, this implies that all those
activities must be completed before any activity that is to begin at that
node can start. Hence figure 5 — activities "a" and "b" must both be
finished before either activity "c" or activity "d" can start.
When two activities both have the same beginning and ending nodes, a
dummy note and activity is used to preserve the separate identify of each
activity. In the diagram below, activities "a" and "b" must be completed
before activity "c" can commence. See figure t below:
Figure 15.7
Note: The 'if' Event Numbering Rule

If the network above consisted of real rather than imaginary activities,
then description of each activity would be written above each arrow. It
is convenient to use a coding system to describe a particular activity, and
we do this by numbering the events according to the 'if' rule. The rule
states that the event at the end of an activity must be assigned a greater
number than the event at the beginning of an activity. There is no single
way of numbering the events, for the 'if rule' allows considerable
latitude. One numbering system that obeys the rule is shown in figure 7
below but of course there are many others.
191
We can now describe activity A by its 'if numbers 1 —2.
Example:
Activity Preceding
B and C A
B and C
D
Figure 9 _______4
This network does show a logical sequence of activities, but we cannot

accept it, as activities B and C both have the same 'if number. This
problem is overcome by introducing a dummy activity — represented by
a broken line.
192
The dummy activity does not represent any activity as such — it is

inserted to preserve the sequential numbering system of events. Such a
dummy is called an identity dumni) . Sometimes it is necessary to insert
a dummy to preserve not the sequential numbering system, but the
logical sequence of events.
3.3 Total Project Time: Critical Path Method (CPM)
The main determinant of the PERT and CPM networks are analysed and
interpreted is whether activity time estimates are probabilistic or
deterministic. If time estimates can be made with a high degree of
confidence, the actual times will not differ significantly, we say the
estimates are deterministic.
If estimated times are subject to variation, we say the estimates are

probabilistic. Probabilistic time estimates must include an indication of
the extent of probable variation.
Rules for when and where logical dummies would be needed for critical
path method. The rules are most convenient to use as stated below:
(1) If an activity occurs in the right-hand column of a dependence

table but not in the left-hand column, it cannot depend upon
another activity having been completed. Hence, it must be a
'starting activity'. If there is more than one start activity, you
may need dummies,
(2) If any activity occurs more than once in the right-hand column
then you will need to introduce dummies. If the activity occurs n
times then (n — 1) dummies will have to be drawn from its end
event.
Example:
The dependence table of a certain project looks like the following:
Activity Preceding Activity
B, C and D A
E and F
G and H
193
G, H, C and D
I, J and K
Firstly, note that there is only one starting activity. Both G and H occur
twice, so the dummy will be needed from the end event of both these
activities. The network will look like the following figure 10:
Figure II
Let us sup p os e that the netw or k des cr ib es the activ ities in a

manufacturing process. The network clearly shows that activities B, C
and D can all be started together as long as we have sufficient resources
to do so. Let us assume that we can easily obtain all the resources that
we need.
Assume the project is such that the only resource needed is labour, and
that all the labour available is equally capable of performing any
activity. The time taken to complete each activity is known thus.
Activity
A B C D E F G H I J K L
Time (hrs)
34 5 6 2 1 74 35 6 2
How long will the project take?
Answer: Examine all routes through the network. See there are seven
possible routes! See the list and find the time taken.
194
MB!: 839 QUANTITATIVE TECHNIQUES FOR BANKING AND FINANCE
Route Time Hours

3+4+2-FE7+3+2 =
ABEGIL 21
A B E G Dummy J L 314+2+7+0+5+2 = 23
ABFHIL 3+4+1+4+3+2 = 17
A B E H Dummy J L 3+4+1+4+0+5+2 = 19
AC J L 3+5+5+2 = 15
A D Dummy I L 3+6+0+5+2 = 16
ADKL 3+6+6+2 = 17
The project cannot be completed in less than 23 hours. This is

determined by the longest route through the network — called the critical
path. Activities on this route must be completed on time otherwise the
total project time will lengthen: (i.e. the activities have critical times).
You should realise that it will be very difficult to calculate the total
project time without first drawing the network. Try to find the total
project time just using the dependence table, and you will confirm how
true it is?
Note.
Finding the earliest event times has certainly helped to determine the
total project time, but it has helped in isolating the critical path. To do
this, use latest event times i.e. the latest time that each event can occur if
the network is to be completed on time. Think back to critical activities
— they must be started on time, otherwise the total project time will
lengthen. What does this imply?
Each event on the critical path must have the same earliest and latest
times. Using this fact, one can easily identify the critical path in the
network.
195
MBE 839 QUANTITATIVE TECIINIQUES FOR BANKING AND FINAN CE
Draw the network of the following 'dependable table'.

Activity Preceding Activity
B and C A
C and D
E and F
Remember that there must be one start event and one end event.
3.4 Gantt Chart
This is a popular tool for planning and scheduling simple projects. It

enables a manager to initially schedule project activities and then to
monitor progress over time by comparing planned progress to actual
progress. A Gantt chart for a bank's plan to establish a new direct
marketing department is shown below:
Activity start 0 2 4 6 8 10 12 14 16 18 20
Local new facilities

Interview
sros I ective staff
Flire and train staff
Select and order
furniture
Remodel and install
•hones
Furniture received
and set up
More in start up
To prepare the chart, the personne in charge of the project identifies the
major activities first that would be required. Next, time estimates for
each activity is made, and the sequence of activities will occur, their
planned duration and when they will take place. Then, as the project
progresses and the manager would be able to see which activities were
ahead of schedule and which were delaying project. This enables the
manager to direct attention where it was needed most to hasten the
project in order to finish on schedule.
The advantage of a Gantt chart is its simplicity, and this accounts for its
popularity. However, Gantt charts fail to reveal certain relationships
among activities that can be crucial to effective project management.
196
For instance, if one of the early activities in a project suffers a delay, it

would be important for the manager to be able to easily determine which
later activities would result in a delay.
It assists in allocating resources to the project known as 'loading the

network'. On a Gantt chart, the activities are represented by lines
having lengths proportionate to the duration of each activity.
Each activity on the Gantt chart is identified by its number. It helps to

dictate the project that is overloaded or under-loaded. The Gantt chart
can be used effectively to show how a project is actually progressing
and this is done by drawing a line above the activity symbol which is
proportional to the amount of work completed. Gantt chart is an
important element of project control.
(1) Suppose that the labour available to work on a project was as

follows:
2 men up to the end of the 11th hour
3 men from the 1 1 th hour to the 23" hour
Draw the simple Gantt chart to take account of the supply of labour.
Pro ress Chart
1 2 3 4 5 6 7 8 9
1
2 3
2
4
3
3 4
4
5
3 5
The progress chart shows the state of a project at the end
197
MBE 839 QUANTITATIVE TEO INIQUES FOR BANKING AND FINANCE
4.0 CONCLUSION
You have been able to see the need of a manager to be vested with
creativity and imagination in the smooth management of business
through proper project planning with aid of network studies. This helps
in evaluating the job/duty and giving it proper direction.
5.0 SUMMARY
Projects consist of activities with set objectives. From the study, you
realise that projects are to be properly planned, the analysis understood
through network symbols, convention and diagrams. With simple
network presentation, a manager can identify problems of the project on
time and determine the probable time needed to achieve the objective of
the project and how the resources are allocated and managed.

-
Given the following information:
Determine.
(a) T he le ngt h of e ac h pa t h;
(b) T he cr it ic a l pa t h;
(c) The expected length of the p r oject;
(d) Amou nt of s lack time for each p ath.

National Open University of Nigeria (2006). MBA 701: Production and

Operations Management.
198

MBF 839

Uploaded by

Copyright:

Available Formats

MBF 839

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

MBF 839

Uploaded by

Copyright:

Available Formats

NATIONAL OPEN UNIVERSITY OF NIGERIA

SCHOOL OF MANAGEMENT SCIENCES

COURSE CODE: MBF 839

COURSE TITLE: QUANTITATIVE TECHNIQUES FOR BANKING

Course Developer/Writer Mr. E. U. Abianga

Course Editor Mr. S.O. Israel-Cookey

Course Co-Ordinator Mr. E. U. Abianga

Programme Leader Dr. 0.J Onwe

NATIONAL OPEN UNIVERSITY OF NIGERIA

National Open University of Nigeria

First Printed 2008

All Rights Reserved

Printed in Nigeria by:

Unit 1 Introduction to Basic Mathematics 1 — 18

Unit 1 Valuation of Bonds and Shares .............. 59 — 71

Unit 1 Inventory Control III ........................... 130— 137

UNIT 1 INTRODUCTION TO BASIC MATHEMATICS

As mathematics is full of symbols e.g. x, y, z and expressions involving

their simplest forms. There will be combination of symbols to obtain

In this units too, ratio, proportion as well as variation will be revised.

On completion of this unit, you should be able to:

 perform basic arithmetic operations on numbers like conversion

3.0 MAIN C ONTENT

3.1 Numbers and Basic Arithmetic Operations

Numbering in word and expressing them correctly in figures is

There are ten digits (base 10 or decimal system) namely 0, 1, 2, 3, 4, 5,

 Billions—. Hundred Millions—. Ten Millions—> Millions—>

3.1.2 Basic Arithmetic Operations

The basic arithmetic operations are addition (+), subtraction (—),

(i) perform all operations within brackets (parentheses);

Step I: open the brackets:

Step III: multiply and divide in the order they appear:

5 + 8 — 32 (Add and subtract as they appear)

Fractions and Percentages

Proper and Improper Fractions

Fractions are common phenomenon when working with numbers.

In m/n fraction, m is the numerator while the number n is denominator.

7/3 may be written as 2% (mixed number).

Note: Every mixed number is the sum of an integer and a proper

Finally, note that the reciprocal of a fraction is obtained by simply

Every answer in mathematics in the form of fraction is presumed to be

is the lowest form.

SELF ASSESSMENT EXERCISE 1

( i) Convert 9/4 to a mixed number.

( ii) Convert 4 3 to an improper fraction.

(iii) A store carries 22 different styles of girl's shoes, 16 different

Arithmetical Operations with Fractions

Addition and Subtraction

A similar rule applies for subtraction

a—b = a—b e.g. 5— 3 = 5-3 = 2 = 1

10 + 9 (since LCM of 6 and 4 is 12)

5+3 20 + 18 (24 is the product of 6 and 4)

Step I: Deal first with whole numbers:

Step II: 2A + % = 8+3= 8+3= 11

Multiplication and Division

Irrespective of whether the fractions are proper or improper, it is always

To divide one fraction by another, multiply the first fraction by the

Another common expression of fraction is in decimal form like:

Percent means of one hundred and is denoted by the symbol %. For