Dr.
Karim
Chapter 3: Optimization Techniques
Optimization
Managerial economics is concerned
with the ways in which managers
should make decisions in order to
optimize the performance of the
organizations they manage.
Optimization Problems
An optimization problem involves the specification
of three things:
1) Objective function to be maximized (e.g. profit
function) or minimized (e.g., cost function).
2) Activities
or
choice
variables
(e.g.,
labor,
quantity produced, price, advertising) that
determine the value of the objective function.
3) Any constraints that may restrict the values of
the choice variables (e.g., maximum total cost).
Optimum Level of Production
Profit ()
Difference between total revenue (TR) and total
cost (TC):
Profit = TR TC
Optimal level of production (Q*) is the level that
maximizes the profit.
Optimal Level of Production
TC
Total revenue and total cost
(dollars)
4,000
D
3,000
B
2,310
2,000
1,085
1,000
C
0
200
TR
* = $1,225
Q
350 = Q
*
600 700
1,000
Production Level
Profit (dollars)
Panel A Total revenue and total cost
curves
1,22
1,00
5
0
600
0
Panel B Profit curve
d
200
350 = Q
*
600
Production Level
1,000
Marginal Benefit & Marginal Cost
Marginal Revenue(MR)
The change in total revenue (TR) associated with a
one unit change in the level of the independent
variable (Q):
MR = /Q = Slope of Total Revenue
Curve
Marginal cost (MC)
The change in total cost (TR) associated with
a one unit change in the level of the
independent variable (Q):
Relating Marginals to Totals
TC
Total revenue and total cost
(dollars)
4,000
100 F
320
3,000
100
520
100
2,000
640
1,000
C
D
D
TR
820
100
520
100
340
100
200
350 = Q
*
800
600
1,000
Production Level
Marginal revenue and
marginal cost (dollars)
Panel A Measuring slopes along TR
and TC
MC (= slope of TC)
8
c (200, $6.40)
6
5.2
0
4
2
d (600, $8.20)
c (200, $3.40)
d (600, $3.20)
MR (= slope of TR)
g
200
350 = Q
*
Panel B Marginals give slopes of
totals
600
Production Level
800
1,000
Using Marginal Analysis to Find the Optimal
Production Level
If marginal benefit > marginal cost
Production should be increased to reach highest
net benefit
If marginal cost > marginal benefit
Production should be decreased to reach highest
net benefit
Optimal level of production
When no further increase in profit is possible
Occurs when MR = MC
Using Marginal Analysis to Find Q*
MR = MC
Profit (dollars)
MR > MC
100
300
c
200
MR < MC
M
100
d
350 = Q *
500
600
Production Level
Q
800
1,00
Profit curve 0
Differential Calculus in Management
A function with one decision variable, X, can be written
as:
Y = f(X)
The marginal value of Y, associated with a small
increase of X, is
My = Y/X
For a very small change in X, the derivative is written:
dY/dX = limit Y/X
X
Quick Differentiation Review
Name
Function
Derivative
Example
Constant Y = c dY/dX = 0Y = 5
Functions dY/dX = 0
A Line
Y = c X dY/dX = c Y = 5X
dY/dX = 5
Power Y = cXb
5X2
Functions
dY/dX = bcX
b-1
dY/dX = 10X
Y=
Quick Differentiation Review
Sum of Y = G(X) + H(X)
dH/dX
Two Functions
example Y = 5X + 5X2
dY/dX = dG/dX +
dY/dX = 5 + 10X
Product of Y = G(X) H(X)
Two Functions
dY/dX = (dG/dX)H +
(dH/dX)G
example Y = (5X)(5X2 )
dY/dX = 5(5X2 ) + (10X)(5X) = 75X2
M a rg i n a l , S l o p e , D e r i v a t i v e
The marginal at point C is Y /X
The slope of the curve at point C
is equal to the slope of the
tangent to the curve at point C =
the rise (Y) over the run (X) =
Y /X.
Or the marginal at a point is
equal to the derivative at this
point The derivative at point C
is also equal to its slope:
Marginal = Slope = Derivative
Max of Y
y
Slope = 0
the function
Y = -50 + 100X - 5X2
value of x
dx
10
20
i.e.,
Value of dy/dx when y is max
0
dY = 100 - 10X
dX
dY = 0
if
dX
X = 10
Value of dy/dx which
Is the slope of y curve
i.e., Y is maximized when
the slope equals zero.
10
20
Note that this is not sufficient for optimization
problems.
In fact, the maximum / minimum points of a
function (e.g., profit function) occur when the
slope of the curve representing the function is
equal to zero Maximum / Minimum profit
occur
when
the
derivative
of
the
curve
representing the profit function is equal to zero.
Max value of y
Since
dY
dX
= 0 at two points, we need another
condition to distinguish between the maximum and
minimum points.
Look at the
Min value of y
dY
dX
curve
* at point 5 the curve is upward, i.e., its slope ( the
second derivative (the derivative of the derivative)) is
positive. Hence
dx
value of dy/dx
d2y/dx2 > 0
d2y/dx2 < 0
d 2Y = > 0
dX 2
( minimum point )
* at point 10 the curve is downward, i.e., its slope is
negative. Hence
d 2Y = < 0
dX 2
( maximum point )
Graphs of an original third-order function and its first and
second derivatives. (what if the second order = 0)
Optimization Rules
Maximization conditions:
12-
dY = 0
dX
d 2Y = < 0
dX 2
Minimization conditions:
12-
dY = 0
dX
d 2Y = > 0
dX 2
Applications of Calculus in Managerial Economics
Profit Maximization Problem
Theprofitfunctionis(=50QQ2).Themaximizationofthefunction
occursif:
1) Firstderivative[d/dQ=502Q]atthatpointisequaltozero.
2) Secondderivative[(d/dQ)=2]is<=0.
Hence,Q=25willmaximizeprofits.
More Applications of Calculus
Minimization problem: Cost Minimization
Supposes that there is a least cost point to produce. An average cost curve
mighthaveaUshape.Attheleastcostpoint,theslopeofthecostfunction
iszero.
1) Thefirstorderconditionforaminimumisthatthederivativeatthatpoint
iszero.
2) Thesecondorderconditionisthatthesecondderivativeis>=0.
TC=5Q260Q,thendC/dQ=10Q60and(dC/dQ)=10.
Hence,Q=6willminimizecost
Where:
10Q60=0.
More Examples - Competitive Firm
Maximize Profits ()
Where = TR - TC = (P Q)- (C Q)
Use our first order condition:
d/dQ = 0 P-C = 0 PRICE = MC
Maximize = 100Q - Q2
First order = 100 -2Q = 0 implies
Q = 50 and;
= 2,500
Second Order Condition: one variable
If the second derivative is negative, then its a maximum
If the second derivative is positive, then its a minimum
Problem 1
Max = 100Q - Q2
First derivative
100 -2Q = 0
second derivative is: -2
implies
Q =50 is a MAX
Problem 2
Max= 50 + 5X2
First derivative
10X = 0
second derivative is:
10 implies
Q = 10 is a MIN
Partial Differentiation
Economic relationships usually involve several
independent variables.
A partial derivative is like a controlled
experiment- it holds the other variables
constant
Suppose price is increased, holding the
disposable income of the economy constant as in
Q = f (P, I )
then Q/P holds income constant.
Problem:
Sales are a function of advertising in
newspapers and magazines ( X, Y)
Max S = 200X + 100Y -10X2 -20Y2 +20XY
Differentiate with respect to X and Y and set equal
to zero.
S/X = 200 - 20X + 20Y= 0
S/Y = 100 - 40Y + 20X = 0
solve for X & Y and Sales
Solution: 2 equations & 2 unknowns
200 - 20X + 20Y= 0
100 - 40Y + 20X = 0
Adding them, the -20X and +20X cancel,
so we get 300 - 20Y = 0, or Y =15
Plug into one of them:
200 - 20X + 300 = 0, hence X = 25
To find Sales, plug into equation:
S = 200X + 100Y -10X2 -20Y2 +20XY
= 3,250
PARTIAL DIFFERENTIATION AND MAXIMIZATION OF
MULTIVARIATE FUNCTIONS.
= f (Q1 , Q2 )
To know the marginal effect of Q 1 on
we hold Q2 constant, and
vice versa.
In order to do that we use partial derivative of
Q1 denoted by
with respect to
( treating Q as constant )
2
Q1
e.g.;
= -20 + 100Q1 + 80Q2 - 10Q12 - 10Q22 - 5Q1Q2;
to find the partial derivative of
with respect to Q1 we treat Q2
as constant; hence
Q1
= 100 - 20Q1 - 5Q2;
Q2
= 80 - 20Q2 - 5Q1;
(1)
therefore
setting
both
simultaneously
partial
(2)
derivatives
equal
to
zero
and
solve
100 - 20Q1 - 5Q2 =0
80 - 20Q2 - 5Q1 =0
multiply by -4 and add
________________
- 220 + 75Q2 = 0
hence
Q2 = 2.933
substitute for Q2 at any of the eq. 1
100 - 20Q1 - 14.665;
hence
Q1 = 4.267.
i.e.,
profit is maximized when the firm produces 4.267 of Q 1 and 2.933 of Q2.
CONSTRAINED OPTIMIZATION
We assume that the firm can freely produce 4.267 of Q1 and 2.933
of Q2. Quite often this may not be the case.
e.g.
Minimize TC = 4Q12 + 5Q22 - Q1Q2;
subject to:
Q1 + Q2 = 30
The constraint function
Solution:
The lagrangian multiplier:
Steps:
1 - set the constraint function to zero
2 - form the lagrangian function by adding the constraint function
after multiplication with an unknown factor to the original
function.
3 - take the partial derivatives and set them equal to zero
4 - solve the resulting equations simultaneously
step 1:
30 - Q1 - Q2 = 0
step 2:
L = 4Q12 + 5Q22 - Q1Q2 + ( 30 - Q1 - Q2)
step 3:
L
Q1
L
Q2
L
= 8Q1 - Q2 -
= -Q1 + 10Q2 -
= -Q1 - Q2 + 30
8Q1 - Q2 - = 0
(1)
-Q1 + 10Q2 - =0
(2)
-Q1 - Q2 + 30
(3)
=0
step 4
multiply eq(2) by -1 and subtract from eq(1)
9Q1 - 11Q2 = 0
(4)
multiply (3) by 9 and add to eq(4)
-9Q1 - 9Q2 + 270 = 0
9Q1 - 11Q2
=0
____________________
-20Q2 +270 = 0
Q2 = 270/20 = 13.5
substituting in eq (3) Q1 = 16.5
the values of Q 1 and Q2 that minimizes TC are 16.5 and 13.5
respectively.
substituting Q1 and Q2 in eq(1) or eq(2) we find that
= 118.5
the interpretation of
measures the change in TC if the constraint is to be relaxed by one
unit.
i.e., TC will increase ( has a positive sign ) by 118.5 if the constraint
becomes 29 or 31.
