Page 1 of 12

MULTIPLE PROJECTS AND CONSTRAINTS STRUCTURE

1 Introduction
2 Constraints
3 Method of ranking
4 Mathematical programming approaches
5 Linear programming models
6 Integer linear programming model
7 Summary
1 INTRODUCTION

Rarely an organization has unlimited resources to go after all available opportunities. In most
cases, they have limited resources and therefore they need to select the best combination of
options that can facilitate accomplishment of organization’s goals and objectives. We can
identify different methods /techniques for project selection from a give set of identified projects.
Put differently we can say that an organization which has many promising proposals at hand but
they cannot undertake all of them due to different significant constraints preventing an
organization from undertaking all the projects at a time. Had it not been these constraints
operating, your organization might have preferred to undertake all the projects at a time and
thereby could have maximized its worth. So operating under existing economic and market
circumstances, and internal shortcomings, an organization may be forced to select only a few
projects from the identified set of projects. As a project manager, you may not have a significant
role to play in the project selection process but you need to know why the projects have been
selected and how they fit into the organization’s strategic objectives. This is because of the fact
that the project selection process has to operate under a number of constraints.

When investment projects are considered individually, any of the discounted cash flow technique
may be applied for acceptance or rejection of that project. However, in an organization, capital
investment projects cannot be considered individually or in isolation. So we need to discuss a
few of constraints which may further restrict our choice. It is obvious that projects under
consideration may or may not be independent; available resources may be limited and partial
acceptance of a project may not be feasible.
These constraints may be categorized and listed as follows:
i. Project Dependence,
ii. Capital Rationing, and
iii. Project Indivisibility

2 CONSTRAINTS
i) Project Dependence: Economic Dependency
a) Mutually Exclusive Projects (Substitute Projects)
Positive Economic Dependency

Page 1 of 12
 Page 2 of 12

Negative Economic Dependency

b) Mutually Not Exclusive projects but Complimentary in Nature
Asymmetric Complementarity and
Symmetric Complementarity.

Project A and B are economically independent if the acceptance or rejection of one does
not change the cash flow stream or does not affect the acceptance or rejection of the
other. On the other hand, Project A and B are economically dependent if the acceptance
or rejection of one changes the cash flow stream of the other or effects the acceptance or
rejection of the other.

What kinds of economic dependency are found in projects?

1) The most conspicuous kind of economic dependency occurs when projects are
mutually exclusive. If two or more projects are mutually exclusive, acceptance of
any one project (out of the set of mutually exclusive projects) automatically
precludes the acceptance of all other projects in the set.
From an economic point of view, mutually exclusive projects are substitutes
for each other. For example, the alternative possible uses of a building represent a
set of mutually exclusive projects. Clearly if the building is put to one use, it
cannot be put to any other use.

2) Economic dependency also exists when projects, even though not mutually
exclusive, negatively influence each other’s cash flows if they are
accepted together.
For example, project for building a toll bridge and a project for
operating a toll ferry. These two projects are such that when they
are undertaken together, the revenues of one will be negatively
influenced by the other.
3) The third kind of economic dependency, which may be referred to as positive
economic dependency, occurs when there is complimentarity between
projects. If undertaking a project influences favorably the cash flows of
another project, the two projects are complimentary projects.
Complementarily can be of two types:
a) Asymmetric Complementarity and
b) Symmetric Complimentarity.

a) In asymmetric complementarity, the favorable effect

extends only in one direction. For example, project
A favorably influences the cash flows project B
whereas project B, in turn, has no influence on the
cash flows of project A.
b) In symmetric complimentarity the favorable effect extends
in both directions: project A favorable influence on
the cash flows of Project B and likewise, project B has
favorable influence on cash flows of project A.

Page 2 of 12
 Page 3 of 12

ii. Capital Rationing:

Capital rationing exists when funds available for investment are inadequate to undertake
all projects which are otherwise acceptable.
What are the sources of capital rationing?
Capital rationing may arise because of an internal limitation or an external constraint.
a. Internal capital rationing is caused by a decision taken by the management to
set a limit to its capital expenditure outlays; or, it may be caused by a choice of
hurdle rate higher than the cost of capital of the firm. Internal capital rationing, in
either case, results in rejection of some investment projects which otherwise are
acceptable.
b. External capital rationing arises out of the inability of the firm to raise sufficient
amounts of funds at a given cost of capital. In a perfect market, a firm can obtain
its entire funds requirement at a given cost of capital. In the real world, however,
the firm can raise only a limited amount of funds at a given cost of capital. With
increased fund requirement, the cost of capital increases and beyond a certain
point, the implication of this rising tendency of capital supply curve is that some
projects which would be acceptable if the cost of capital were constant will have
to be postponed or abandoned.

iii. Project Indivisibility:

Capital projects are considered indivisible, i.e. a capital project has to be accepted
or rejected in toto - a project cannot be accepted partially. Given the indivisibility of
capital projects and the existence of capital rationing, the need arises for comparing
projects. To illustrate this point, consider an example.
A firm is evaluating three projects A, B, and C which involve an outlays of Rs. 0.5
million, Rs. 0.4 million, and Rs. 0.3 million respectively. The net present value of these
projects are Rs. 0.2 million, Rs. 0.15 million, Rs. 0.1 million respectively. The funds
available to the firm for investment are Rs. 0.7 million.
In this situation, acceptance of project A (project with the highest net present value)
which yields a net present value of Rs. 0.2 million results in the rejection of
projects B and C which together yield a combined net present value of Rs. 0.25 million.
Hence, because of the indivisibility of projects, there is a need for the comparison of
projects before the acceptance/rejection decisions are taken.

Project Selection Methods

Organizations invest huge sums of money on new projects. It is important for organizations to
ensure that these investments are safe and will result in good returns and benefits for the
organizations. Hence there is a need for analyze all new project opportunities to justify the
decisions for making the needed monetary investments. An organization may be having multiple

Page 3 of 12
 Page 4 of 12

project opportunities to invest on and reap the benefits. But they cannot invest on all projects,
they have to be selective.

The process of analyzing the new project opportunities to decide which ones will be worthwhile
taking up so that organization gets the most benefits is known as project selection methods.

Project selection is one of the most important steps before taking up any new project investment
decision. Project selection study will be done by professionals with solid knowledge in these
methods. Both quantitative and qualitative methods are used for project selection. Organizations
can define clear guidelines for project selection procedures and methods to be used.

There are two major categories of project selection methods as below:

1. Qualitative Methods :

Many times project selection is done based on non-financial considerations. These may include
political reasons, personal biases, safety and security considerations, investor and customer
requests etc.

2. Quantitative Methods:
Because of economic dependency, capital rationing, and project indivisibility, a need arises for
comparing projects in order to accept some and reject others. What approaches are available for
determining which projects to accept and which projects to reject?

3 METHOD OF RANKING
Two approaches are available for determining which project to accept and which projects to
reject:

i) Method of Ranking Methods (Benefits measurement methods)

ii) Method of Mathematical Programming Approach

Ranking Approach
(A) The Method of Ranking
Here in this case we need to measure benefits and then rank projects accordingly in
ascending or descending order. Benefits Measurement Methods
Some of the most important methods under this category include the following:
 Cost Benefit Analysis – Benefits from the project should be more than the cost invested
in the project for any project to be considered for selection. A ratio of benefits to cost (BCR)
should be more than 1. Higher the BCR, better the opportunity.
 Payback Period – Payback period is the time in which the total investment of a project is
recovered. After the payback, organization truly will start making profits. Payback time of a

Page 4 of 12
 Page 5 of 12

project should ideally be lower. Lower the payback period, the project looks that much more
attractive.
 Net Present Value – Net present value is the difference between the sum of present
value of all revenue and the sum of present value of all the investment. The NPV must be more
than zero for the project to be profitable. Higher the NPV, better is the choice. NPV is the most
practical criteria which discount all future values of revenue and investments to their
corresponding present values to calculate profit.
 Internal Rate of Return (IRR) – IRR is the maximum expected return on money from a
given project. Higher the IRR, better the opportunity. IRR is the rate of interest at which the
NPV becomes equal to zero.
 Scoring Model – Scoring models are used to study different project options against
certain parameters by assigning scores to them for each parameter. The project option with
highest score will be the preferred one.
 Economic Value Added – EVA is to see of the project is profitable after deducting all
cost of capital. Because then only there is economic value add in the project.
 Opportunity cost – Opportunity cost concept is used to study what we lose by selecting
something. This can be used when we have to give up some benefits while we will be selecting
some other benefits. If we have 2 projects A and B. Opportunity of selecting A is what we will
be losing by not selecting B.

i. The method of ranking consists of two steps:

a) Rank all projects in a decreasing order according to their individual NPV’s,
IRR’s or BCR’s.
b) Accept project in that order until the capital budget is exhausted.

The method of ranking is seriously impaired by two problems:

a. Conflict in ranking as per discounted cash flow criteria, and
b. Project indivisibility.

Conflict in Ranking
In a given set of projects, preference ranking tends to differ from one criterion to another. For
example, NPV and IRR criteria may yield different preference rankings. Likewise, there may be
a discrepancy between the preference rankings of NPV and BCR (benefit cost ratio) criteria.
When preference rankings differ, the set of projects selected as per one criterion tends to differ
from the set of projects selected as per some other criterion.
Illustration:
Consider a set of five projects A, B, C, D AND E, for which the investment outlay , expected
annual cash flow , and project life are as shown below:
Project Investment Outlay Expected Annual Cash Flow Project Life
(Rs) (Rs) (Years)
A 10,000 4,000 12
B 25,000 10,000 4
C 30,000 6,000 20
D 38,000 12,000 16
E 35,000 12,000 9

Page 5 of 12
 Page 6 of 12

The NPV, BCR and IRR of the five projects and the ranking along these dimensions are shown
below:

Project NPV (Rs) NPV IRR IRR BCR BCR

Ranking (percentage Ranking Ranking
)
A 14.776 4 39 1 2.48 1
B 5370 5 22 4 1.21 5
C 14,814 3 19 5 1.49 4
D 45,688 1 30 2 2.20 2
E 28,936 2 29 3 1.83 3

It is clear that in the above case the three criteria rank the projects differently. If there is no
capital rationing all the projects would be accepted under all three criteria though internal
ranking differ across criteria.
However, if the funds available are limited, the set of projects accepted would depend on the
criterion adopted.

Project Indivisibility
A problem in choosing the capital budget on the basis of individual ranking arises because of
indivisibility of capital expenditure projects. A problem in choosing the capital budget on the
basis of individual ranking arises because of indivisibility of capital expenditure projects.

To illustrate, consider the following set of projects (ranked according to their NPV)_ being
evaluated by a firm which has a capital budget constraint of Rs2,500,000.
Project Outlay (Rs) NPV (Rs)
A 1,500,000 400,000
B 1,000,000 350,000
C 800,000 300,000
D 700,000 300,000
E 600,000 250,000
If the selection is based on individual NPV ranking, projects A and B would be included in the
capital budget --- these projects exhaust the capital budget.
A cursory examination, however, would suggest that it is more desirable to select projects B, C
and D. These three projects can be accommodated within the capital budget of Rs 2,500,000, and
have a combined NPV of Rs 850,000 for projects A and B.

Page 6 of 12
 Page 7 of 12

Feasible Combinations Approach

The above example suggests that the following procedure may be used for selecting the set of
investments under capital rationing.
1. Define all combinations of projects which are feasible, given the capital budget
restriction and project interdependencies.
2. Choose the feasible combination that has the highest NPV.
Illustration:
Consider the following projects that are being evaluated by a firm which has a capital budget
constraint Rs 3000,000.
Project Outlay (Rs) NPV (Rs)
A 1,800,000 750,000
B 1,500,000 600,000
C 1,2000,000 500,000
D 750,000 360,000
E 600,000 300,000
Project B and c are mutually exclusive. Other projects are independent.

Solution:
Given the above information the feasible combinations and their NPVs are shown below:
Feasible Combinations Outlay NPV
A 1,800,000 750,000
B 1,500,000 600,000
C 1,2000,000 500,000
D 750,000 360,000
E 600,000 300,000
A and C 3,000,000 1,250,000
A and D 2,550,000 1,110,000
A and E 2,400,000 1,050,000
B and D 2,250,000 960,000
B and E 2,100,000 900,000
C and D 1,950,000 860,000
C and E 1,800,000 800,000
B, D and E 2,850,000 1,260,000
C, D and E 2,550,000 1,160,000
The most desirable feasible combination consists of projects B, D and E as it has the highest
NPV.

Page 7 of 12
 Page 8 of 12

4 MATHEMATICAL PROGRAMMING APPROACH

The ranking procedure described above becomes cumbersome as the number of projects
increases and as the number of years in the planning horizon increases. To cope with a problem
of this kind, it is helpful to use Mathematical Programming Models. The advantage of
mathematical programming models is that they help in determining the optimal solution without
explicitly evaluating all feasible combinations.

These methods are mathematical models of project selection which are used for larger projects
that require complex and comprehensive calculations and simulations to study the feasibility of
projects by understanding the uncertainties involved clearly.

A mathematical programming model is formulated in terms of two broad categories of equations:

(i) The Objective Function, and
(ii) The Constraint Equations.
The objective function represents the goal or objective the decision maker seeks to achieve.
Constraint equations represent restrictions-arising out of limitations of resources, environmental
restrictions, and managerial policies-which have to be observed. The mathematical model seeks
to optimize the objective function subject to various constraints.

The objective function and constraint equations are defined in terms of parameters and decision
variables. Parameters represent the characteristics of decxision environment which are given.
Decision variables represent what is amenable to control by the decision maker.

A wide variety of Mathematical Programming Models is available, a few of them are as

mentioned below:
 Linear Programming Model.
 Integer Programming Model.
 Goal Programming Model
 Non-linear programming
 Dynamic programming

1. LINEAR PROGRAMMING MODEL

The linear programming model is based on the following assumptions:-
 The objective functions and the constraint equations are linear.
 All the coefficients in the objective function and constraint equations are defined
with certainty.
 The objective function is one-dimensional.
 The decision variables are considered to be continuous.
 Resources are homogeneous. This means that if 100 hours of direct labour are
available, each of these hours is equally productive.

Page 8 of 12
 Page 9 of 12

1. Linear Programming Model of a Capital Rationing Problem

The general formulation of a linear programming model for a capital rationing problem
is:

Maximize Z = NPVj Xj (1)

Subject to  CFjt Xj ≤ Kt (t = 0, 1, ……….., m) (2)
0 ≤ Xj ≤ 1 (3)

where NPVj = net present value of projects j

Xj = amount of projects j accepted
CFjt = cash outflow required for project j in period t
Kt = capital budget available in period t

The following features of the model may be noted.

1. All the input parameters (NPVj , CFjt , Kt) are assumed to be known with
certainty.
2. The Xj decision variables are assumed to be continuous but limited by a lower
restriction (0) and an upper restriction (1).
3. The NPV calculation is based on a cost of capital figure which is known with
certainty.

2 INTEGER LINEAR PROGRAMMING MODEL

The principal motivation for the use of integer linear programming approach are :
i. It overcomes the problem of partial projects which besets the linear
programming model because it permits only 0 or 1 value for the decision
variables.
ii. It is capable of handling virtually any kind of project interdependency.

The Basic Integer Linear Programming Model for capital budgeting under capital rationing is
as follows:
Maximize  Xj * NPVj (4)
Subject to  CFjt* Xj ≤ Kt (t = o, 1… m) (5)
Xj = (0, 1) (6)
It may be noted that the only difference between this integer linear programming model and the
basic linear programming model discussed earlier is that the integer linear programming model
ensures that a project is either completely accepted (Xj = 1) or completely rejected (Xj = 0).

Page 9 of 12
 Page 10 of 12

Evaluation
The merits of the integer linear programming model are:
1. It overcomes the problem of partial projects which besets the linear
programming model.
2. It is capable of handling virtually any kind of project interdependency.
The main limitations of the integer linear programming model are:
1. The solution of linear programming model takes considerably more time than
the solution of the integer of the linear programming model.
2. Meaningful shadow prices are not available for the integer programming
formulation. This happens because the integer linear programming model
permits only discrete variation, not continuous variation, of the decision
variable. In the integer linear programming model, constraints which are
not binding in the optimal solution are assigned zero shadow prices though
the objective function would decrease when the availability of resources
representing non-binding constraints, is diminished.

3. GOAL PROGRAMMING MODEL

The principal goal of financial management is to maximize the wealth of shareholders, which
under conditions of perfect capital market, can be realized by selecting the set of capital projects
that maximize the net present value.

a) However in the real world, capital market imperfections (like capital rationing,
differences in lending and borrowing rates etc.) exist.
b) Further, imperial observation show that managers pursue a multiple goal structure which
includes, inter alia, the following:
i. Growth in sales and market share
ii. Growth and stability of reported earnings
iii. Growth and stability of dividends

Therefore, a realistic representation of real life situations should reflect in the multiple goals
pursued by management. The Goal Programming approach, a kind of mathematical
programming approach, provides a methodology for solving an optimization problem that
involves multiple goals.

To use the goal programming model, the decision maker must:

1. State an absolute priority other among his goals, and

2. Provide a target value for each of his goals.

Page 10 of 12
 Page 11 of 12

The goal programming methodology seeks to solve the programming problem by minimizing the
absolute deviations from the specific goals in order of priority structure established. Goals at
priority level one are sought to be optimized first. Only when this is done well, the goals at
priority level two will be considered: so on and so forth. At a given priority level, the relative
importance of two or more goals is reflected in the weights assigned to them.

Goal Programming format

In general, the goal programming format is as follows:

Minimize Z = {P1 [f1 (a1, a1)] + P2 [f2 (a2, a2)] + …….. + Pm [fm (dm, dm)] }
Subject to:
Constraints aj1 Xj ≤ C1
aj1 Xj ≤ C1
aj1 Xj ≤ C1
.
.
.
.
Target Levels of various goals bj1 Xj + d1- - d1+ ≤ G1
bj1 Xj + d1- - d1+ ≤ G1
.
.
.
.
bjm Xj + dm- - dm+ = Gm
Non-negativity constraints Xj, , di-, di+ ≥ 0

The goal programming model is formulated in terms of three principal components:

(i) The objective function
(ii) The economic constraints, and
(iii) The goal constraints

The objective function seeks to minimize the weighted deviation from the target levels of various
objectives in accordance with a priority structure. A formal representation of the objective
function calls for specifying the following:
 The priority levels of goals
 The relative weights attached to each goal where there are two or more goals at the same
priority level.
 The relevant deviational variables which should be minimized with respect to each goal.

Page 11 of 12
 Page 12 of 12

Economic Constraints
The economic constraints represent resource limitations or restrictions emanating from the
decision environment which cannot be violated. Hence they are referred to as “hard constraints”
The economic constraints in goal programming model are similar to the constraints in found in a
Linear Programming Model. Hence, such constraints require the usual slacks or surplus
variables.

Goal constraints
The goal constraints represent target levels of various goals that are pursued by the decision
maker. Defined as strict equalities, goal constraints contain two deviational variables denoted by
di- and di+.
di+ indicates that the desired level of goal i has been over-achieved.
di-, indicates that the desired level of goal i has been under-achieved.
When the desired goal level has been over-achieved di+ is non-zero and di- is zero.
When the desired goal level is exactly achieved, both di+, di-, are zero.
The deviational variables tie the goal constraints and objective function.
For each goal, the appropriate deviational variables(s) is (are) placed in the objective function.

4. SUMMARY
In the selection of new projects, these are many constraints which include project dependence,
capital rationing and project indivisibility. Capital rationing exists when funds available for
investment are inadequate to undertake all projects which are otherwise acceptable. Method of
ranking and method of mathematical programming are the two approaches available for
determining the acceptance or rejection of projects. The method of ranking consists of ranking of
all projects in a decreasing order according to their individual NPV or IRR or BCR and
acceptance of projects in that order until the capital budget is exhausted. Mathematical
programming models help in determining the optimal solution without explicitly evaluating all
feasible combinations. The mathematical model seeks to optimize the objective function subject
to various constraints. Linear programming model and integer programming model are the
important types of mathematical programming model.

Page 12 of 12