Sei sulla pagina 1di 59

Lagos State University

Department of Banking and Finance


Seminar on Application of Goal
Programming Model in an
Organization for Optimum Resource
Allocation using POM-QM(4).

1
Lagos State University
Held on 7th
to 8th November,2016
Speaker:
Dr. A. Arewa

2
Content of the Seminar
Introduction
 What is Goal Program (GP)
Model?
What is it used for?
Definition of Basic Concepts

3
Content of the Seminar cont
 Modeling GP and Definition of Terms

 Types of GP

 Stating the GP Problem

 Arranging the Data in an Excel

 Copying the Data into POM-QM Program

3
Content of the Seminar cont

 Solving the Problem

 Interpreting the Solution

 Marking Recommendation for


Policy Implication
4
Introduction
 It is obvious that in any organization,
management of resources is pivotal.

 Since resources are scarce, an optimum


allocation of them to achieve the desired
organizational goal is inevitable.

 Organizations thrive when their basic goals are


achieved.

5
Introduction Cont
 This why since the World War (11) various
organizations have adopted GP for determining their
optimum production, profit, productive time and
minimum cost to avoid wastage.

 To know the goal(s) that can be achieved or not and


6
their priorities

6
What is Goal Program (GP)
Model?
 GP was first introduced by Charnes, Cooper and Ferguson(1955)
with discrete and continuous variables, linear/non-linear
functions in which case all functions are assumed as goals.

 GP is a realistic mathematical programming model, a


generalization of LP capable of dealing with multi-objectives but
conflicting goals; with the view of selecting optimum decision(s).
7
What is Goal Program (GP) Model?
Cont
 GP is often seen as multi-objective, multi-decision
optimization programming model; inherently used
in solving complex institutional or industrial
problems of choice, preference, allocation,
budgeting and allotment.

 GP is a platform of analyzing multiple measures


of organizational performance in terms of costs,
revenue, productivity and profitability.

8
What is Goal Program (GP)
Model? Cont
 It is an algorithm of modeling
the goals of an organization,
giving priorities and weights,
solving to show whether these
priorities are achieved or not.
What is it used for?

 Minimizing cost, increasing level of productivity,


efficiency, effective usage of resources including
time and attainment of optimum growth,
expansion and healthy industrial or institutional
outlook.

 Thus, GP is operationalized for three purposes:


1. Amidst competing resources, course of actions and
priorities, it gives optimum solution.
10
What is it used for? Cont

2. It shows the extent to which an


organization has attained its goal(s)

3. It gives the level of resources that


is needed to attain desired goals.

11
Definition of Basic Concepts
 Decision marker is the entity who sets off decision
problems.

 Decision variables are the factors that can be


controlled by the decision maker.
 Criterion is a measure of the goodness of a solution to a
decision problem. Thus, decision problem with multiple
criteria is referred to multi-criteria decision analysis or
making.
12
Definition of Basic Concepts cont
 Constraint is the restriction imposed on the decision
variable that had to be met if the solution would be
implemented in real life.

 Two type of constraints: goal constraint and system


or structural constraint.
 Goal constraint is simply the original function with
set of priorities/goals including positive/negative
deviational variables.
13
Definition of Basic Concepts cont
 system constraint is strictly the traditional linear
programming function without the negative-
positive deviational variables.

 Achievement Function measure the extent to


which minimization of unwanted goal deviational
variables have been achieved.

 Multiplex is the multiple phased simplex algorithm


employed in solving LP.
14
Definition of Basic Concepts cont
 Aspiration level referred to target level.

 Positive Deviation is overachievement of aspiration


or target . Or positive deviation of goal from target
level.

 Negative Deviation is underachievement of


aspiration or target . Or negative deviation of goal
from level of aspiration.
Definition of Basic Concepts cont

 Feasible Region is a set of


solutions that fall in the decision
space where all constraints are
duly satisfied.

16
Modeling GP and Definition of
Terms
Note the general GP model was first developed by
Charnes and Cooper (1977) as follows.

m
1. min z   (dev  dev ); i  1,2,..., m

i

i
i 1

st
17
Modeling GP and Definition of
Terms cont
n
2. c
j 1
i, j

i

y j dev  dev  Ai ; j  1,2,..., n
i

n

3. c
j 1
i, j y j  Ai ; i  m  1,..., m  p

18
Modeling GP and Definition of
4. Terms cont
 
d ,d , y  0
i i j
It is important to note that:
1.Equation (1) is referred to objective function which
is the summation of all deviational variables
2. Equations 2 and 3 are called goal and system
constraint functions; and they are both referred to
linear constrain function
3. Equation 4 is non-negativity constraint.
19
Definition of Terms
m is number of goals, p number of structural
constraints and n number of decision variables.

Z is the objective function expressed as the


summation of all the deviational variables.

yj is the jth decision variables.

20
Definition of Terms cont
ci , j is the coefficient of the jth decision variables in
the ith goal.

 
devi and
devi are amount of deviation below and
above aspiration level respectively. That are
accordingly called underachievement and
overachievement variables. Therefore, in typical GP
model, there are two variables: decision and
deviational.

A
i
aspiration level
21
Types of GP
There two types of GP in literature:
 Lexicographic Goal Programming Model or
preemptive or non-Archimedean goal
Programming model.

 Weighted Goal Programming model.

22
Types of GP cont
Iserman (1982) and Sherali (1982) introduced factors
to the general GP objective function that is known
as preemptive priority factors. Thus, their
innovation gave rise to preemptive GP.

These factors are ranked according to priority given to


a goal. In essence preemptive GP is used when
there is clear priority ordering of the desired goals
Types of GP cont
The preemptive factors ( p's ) relationship follows:
p1  p2  ...  pi  pi 1  ...  pm

This indicates that the first priority/goal is much more


important than the second, the second than the
third, the trend continues until the last one.
Therefore, you must achieve the first goal
completely before considering the second and the
trend continues.
24
Types of GP cont
Thus, the preemptive GP can be specified as.
m

5. min z   pi (dev  dev ); i  1,2,..., m



i

i

st
i 1

Linear constraint functions of equations 2 and 3; and


non-negativity constraint of 4

25
Types of GP cont
The second type of GP is weighted one which is
used when the decision maker is interested in
the direct comparison of the goals. This is
done by attaching weight to each of the
deviational
m
variables as follows.
6. min z   wi devi  wi devi ; i  1, 2,..., m
i 1

st Equations 2, 3 and 4

26
Types of GP cont

Where: w and
i
w i
are relative weights attached to
positive and negative deviation variables and they
must be non-negative constants. These weights are
any real number or integers of which the greater the
weight the greater the importance assigned to
minimize the deviation variable.

The difference between the general GP model and the


preemptive GP model is preemptive priority factor.

27
Types of GP cont
The difference between the general GP model and the
weighted GP model is the arbitrary weight assigned
to the deviation variables.

Charnes and Cooper (1977) based on Ijiri (1965)’s


innovation suggested that both the preemptive and
weighted GP models can be mathematically
combined together as follows.

28
Types of GP cont
m ni
7. min z   pi ( w dev  w dev ); i  1,2,..., m

i ,k

i

i ,k

i

st
i 1 k 1
Equations 2, 3 and 4

 
Where: w ;w
i ,k i ,k 0
29
Types of GP cont
Note that overachievement and underachievement of
a goal cannot occur concurrently. Therefore is either
 
devi  0 or devi  0

Or both dev  0 dev  0
i

i

30
Types of GP cont

Therefore, the procedures for achieving a


goal are.
1.Minimize the underachievement
2. Minimize the overachievement
3. Minimize both underachievement and
overachievement.
31
Stating the GP Problem
A typical balance sheet of a bank is given below.

year 2007 2008 2009 2011 2012 total

asset 1191042000 1673333000 1548281000 1617696000 1864457000 7894809000

liability 1142207000 1478052000 1361452000 143827000 1682301000 7102322000

shareholder 48835000 195231000 186829000 179426000 182156000 792471000

earnings 109512000 189506000 246725000 18518600 198725000 929654000

proft 2154000 41239000 2113000 668000 797000 66357000

total 3776985958 3577361000 3345440000 2.1268E+10 3928430000 1373812396

32
Coding
Since the values of the items in the balance sheet are
reported in trillions, there is need to reduce them to
decimals by equally dividing by 1,000,000,000 and
then round them up to three decimal places as
follows:
year 2007 2008 2009 2011 2012total
asset 1.191 1.673 1.548 1.18 1.864 7.894
liability 1.142 1.478 1.361 1.437 1.682 7.1
shareholder 0.048 0.195 0.187 0.179 0.182 0.791
earnings 0.11 0.19 0.247 0.185 0.199 0.931
proft 0.022 0.041 0.002 7E-04 8E-04 0.067
total 2.513 3.577 3.345 3.42 3.928 16.78

33
Goals or Objectives of the Bank
• In formulating GP, the goals of an organization must
be explicitly stated. Let us hypothetically assume that
the goals of this bank are:
1.Reduce liability to exactly 5.7 trillion.
2.Increase asset to at least 9 trillion.
3.Increase shareholder funds to at least 2.1 trillion.
4.Increase earnings to at least 1.2 trillion.
5.Increase profit to at least 0.4 trillion.
6.Increase balance sheet to at least 18 trillion.

34
Setting Priority
In GP you have to prioritize your goals, which
one comes first or second, even last.

In this example, our goals are prioritized as


follows:
Priority 1= liability reduction
Priority 2= asset maximization

35
Setting Priority Cont
Priority 3= maximization of owners’ funds

Priority 4= maximization of earnings

Priority 5= maximization of profit

Priority 6= optimum balance sheet management

36
Variables
There are two sets of variables here: Decision and
deviation.
Decision Variables: x1, x2, x3, x4, x5 and x6
Where:
X1 is the total value of assets in the balance sheet =
7.894 trillion.
X2 is the total value of liability in the balance sheet =
7.1 trillion.
X3 is the total value of shareholders fund in the balance
sheet = 0.791 trillion.
37
Variables cont
X4 is the total value of earnings in the balance sheet =0.931
trillion.
X5 is the total value of profit in the balance sheet = 0.067 trillion.
X6 is the total value of all items in the balance sheet = 16.78
trillion.
Deviation Variables: and are amount by which goal i is
underachieved and overachieved respectively.

d i d i

38
GP Functions
In GP, there are two functions which we have
earlier mentioned- structural constraint functions
and goal functions.
In this example the constraint functions are:
X1 ≤ 7.894 asset constraint
X2 ≤ 7.1 liability constraint
X3 ≤ 0.791 shareholder fund constraint
X4 ≤ 0.931 earning constraint
X5 ≤ 0.067 profit constraint
X6 ≤ 16.78 balance sheet item constraint
39
GP Functions Cont
While the goal functions are:  
X1+d1--d1+= 9 asset maximization goal
X2+d2--d2+= 5.7 liability reduction goal
X3+d3--d3+=2.1 shareholders’ wealth goal
X4+d4--d4+=1.2 earning maximization goal
X5+d5--d5+=0.4 profit maximization goal
X6+d6--d6+=18 optimum balance sheet management goal
X1, X2 , X3, X4, X5 , X6, d1-, d1+, d2-,d2+, d3-,d3+, d4-,d4+ ,d5-,d5+, d6-,d6+ ≥ 0 non-
negativity constraint

40

38
Summary of Functions
Min p1(d1-)+ p2(d1+) + p3(d2-)+ p4(d2+)+ p5( d3-)+ p6(d3+)+
p7(d4-)+ p8(d4+)+ p9(d5-)+ p10(d5+)+ p11( d6-)+ p12(d6+)
st
X1 ≤ 7.894
X2 ≤ 7.1
X3 ≤ 0.791

41
Summary of Functions Cont
X4 ≤ 0.931
X5 ≤ 0.067
X6 ≤ 16.78
X1 +d1--d1+ =9
X2 +d2--d2+ = 5.7
X3 +d3--d3+ =2.1
X4 +d4--d4+ =1.2
X5 +d5--d5+ =0.4
X6 +d6--d6+ =18
• X1, X2 , X3, X4, X5 , X6, d1-, d1+, d2-,d2+, d3-,d3+, d4-,d4+ ,d5-,d5+, d6-,d6+
≥0
42
Arranging the Data in an Excel
  wt(d+) prty(+) wt(d-) prty(-) x1 x2 x3 x4 x5 x6   RHS

Asset 0 0 0 0 7.894 0 0 0 0 0 <= 7.89

Liabilit 0 0 0 0 0 7.1 0 0 0 0 <= 7.1

Shareholder 0 0 0 0 0 0 0.791 0 0 0 <= 0.79

Earnings 0 0 0 0 0 0 0 0.931 0 0 <= 0.93

Proft 0 0 0 0 0 0 0 0 0.067 0 <= 0.07


balance-
sheet 0 0 0 0 0 0 0 0 0 16.783 <= 16.8
reduce
liability 1 1 1 1  1 0 0 0 0  = 5.7
increase
asset 0 0 1 2 1 0 0 0 0 0  = 9
increase
shareholder 0 0 1 3 0 0 1 0 0 0  = 2.1
increase
earnings 0 0 1 4 0 0 0 1 0 0  = 1.2
increase
profit 0 0 1 5 0 0 0 0 1 0  = 0.4
max balance
sheet 0 0 1 6 0 0 0 0 0 1  = 18
Copying the Data into POM-QM
Program
To perform this function follow the following
steps:
1.Right click or double click on the POM-QM
Program.
2. Once it opens click on ok.

3 Array of programs appears select Goal


Programming.
44
Copying the Data into POM-QM
Program Cont
4. Or click on the module to select Goal
Programming.

5. Click on the first icon directly below file.

6. In the work box that opens fill in the number of


constraint and goal together and decision
variables.
45
Copying the Data into POM-QM
Program Cont
7. Copy the data column by column in the excel
file and paste column by column.

46
Solving the Problem
Click on solve option on the second menu to obtain results. Also click on summary.

Click on the grid, then on the copy icon on the second menu to copy entire tableau.

Paste results as follows

47

45
Summary Results One
Item

Decision variable analysis 1

X1 1

X2 1

X3 1

X4 1

X5 0.4

X6 1
Summary Results Two

Priority analysis Nonachievement

Priority 1 4.7

Priority 2 8

Priority 3 1.1

Priority 4 0.2

Priority 5 0

Priority 6 17
Summary Results Three
Constraint Analysis RHS d+ (row i) d- (row i)

asset 7.89 0 0

liabilit 7.1 0 0

shareholder 0.79 0 0

earnings 0.93 0 0

proft 0.07 0 0.04

balance-sheet 16.78 0 0

reduce liability 5.7 0 4.7

increase asset 9 0 8

increase shareholder 2.1 0 1.1

increase earnings 1.2 0 0.2

increase profit 0.4 0 0

max balance sheet 18 0 17


Interpreting the Solution
The summary result one gives the optimum solution
of the GP.
This shows that the optimum profit is 0.4 trillion; the
optimum asset to hold is 1 trillion and the
maximum level of liability is 1 trillion. The optimum
level of earnings is 1 trillion.

51
Interpreting the Solution Cont
The summary result two shows the priority that
is achieved or not. From the solution:

The first priority-liability reduction- is not


achieved. Also, third, fourth and sixth
priorities; but the forth priority which is
earning maximization is nearly achieved.

52
Interpreting the Solution Cont
However, the second priority (asset maximization)
and fifth priority (profit maximization) are achieved
since the non achievements are zeros.

The summary result three shows the constraint and


goal analyses.

.
53
Interpreting the Solution Cont
The constraint analysis indicates that exactly all
the assets in the balance sheet are used, all
funds and earnings are distributed, liability is
exactly used to create asset but 0.04 trillion
amount of profit constraint is not realistic.
Interpreting the Solution Cont
The goal analysis shows that liability reduction
goal is underachieved by 4.7 trillion, wealth
maximization is underachieved by 1.2 trillion
and earnings underachieved by 0.2 trillion.

But however, the goals of asset and profit


maximization are exactly achieved.

55
Reference
• Charnes,  A ., Cooper, W.W., & Ferguson, R
(1955). Optimal estimation of executive
compensation by linear programming,
Management Science, 1, 138-151.
• Charnes, A., & Cooper, W.W. (1977).Goal
programming and multiple objective
optimizations. European Journal of
Operational Research. 1(1);39-54.
Reference Cont

Ijiri, Y. (1965). Management Goals and


Accounting for Control. Rand Mc Nally,
Chicago.

Isermann, R.(1982). Parameter adaptive control


algorithms: a tutorial. Automatical. 18; 513.

57
Reference Cont
Sherali, H. D. (1982). Equivalent weights for
lexicographic multi-objective programs:
characterizations and computations. European
Journal of Operational Research, 11(4);367–
379.

58

Potrebbero piacerti anche