Mathematical Economics

PART I: COMPARATIVE STATICS ANALYSIS

Hà Văn Hiếu

University of Economics and Law

17th September 2023

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 1 / 138

 Introduction

Lecturer
Dr. Hà Văn Hiếu,
Faculty of Economic Mathematics, UEL.
Email: [email protected].

Students
This module is taken by 413, 413C and 413CA.

Given your very varied backgrounds, you will need to

stay focused,
work with your team,
become practiced at communicating your own insights and challenges...

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 2 / 138

 All about this module

1 Part I-1: What is Mathematical Economics [Chapters 1, 2 (textbook)]

The nature of Mathematical Economics.
Mathematical models in Economics.
2 Part I-2: Comparative Static Analysis [Part 2, 3, 4 - Chapters 3-13 (textbook)]
A review on MEB (aka Toán Kinh tế 1): Matrices, Differentiation, Integration,
Optimization and applications in Economics.
More applications and more accurately.
3 Part II: Dynamic Analysis [Part 5 - chapter 14-19 (textbook)]
Differential Equation
Difference Equation
4 Part III: Linear Programming Problem [Chapters 3-5 (Introduction to Operation Research)]
The simplex method.
Solving Linear Programming Problem by Excel solver.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 3 / 138

 How to pass?

1 Minimum attendance: 80%.

2 No smartphone, no cigarette, etc.
3 Speaking: English, Vietnamese.
4 Writing: English.

Forms of Assessments Type of Collaboration Weight

Attendance, assignments Individual, team 30%
Midterm Individual 20%
Final exam Individual 50%
Table 1: Super important table

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 4 / 138

 Resources

The main recommended text is

Alpha C. Chiang, Fundamental Methods of Mathematical
Economics, 3rd edition, McGraw-Hill, Inc., 2005.
https://lms.uel.edu.vn/

Other resources:
1 Raymond A. Barnett; Calculus for Business, Economics, Life Sciences and Social
Sciences, 13th edition.
2 Frederick S. Hillier, Introduction to Operations Research, 10th edition, McGraw-Hill,
2015.
3 PGS.TS. Hoàng Đình Tuấn; Giáo trình Mô hình Toán Kinh tế; NXB Thống kê,
Trường Đại học Kinh tế Quốc dân.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 5 / 138

 Create your team!

413, 413C, 413CA https://lms.uel.edu.vn

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 6 / 138

 The nature of Mathematical Economics

Definition
Mathematical economics is the application of mathematical methods to represent theories
and analyze problems in economics. Often, these applied methods are beyond simple
geometry, and may include differential and integral calculus, difference and differential
equations, matrix algebra, mathematical programming, or other computational methods.
[Wikipedia]

1 What is the difference between mathematical economics and literary economics?

2 What is the advantage of mathematical economics?
3 What is the difference between mathematical economics and econometrics?
Ref: p2-4

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 7 / 138

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 8 / 138
 Ingredients of a Mathematical model

Variables: something that can take on different values, example: price, profit, revenue,
etc. Each variable can assume various values and presented by a symbol: Endogenous
- value is determined by the model and exogenous - value is determined outside the
model. An economic variable can be exogenous in some models and endogenous in
others. Symbols: x, y, z, t. dependent variables and independent variables.
Constants: a magnitude that does not change. Variables frequently appear in
combination with fixed number or constants. Symbols: 1, 2, 3, a, b.
Parametric constants or parameters: a, b - supposed to represent a given constant.
However, it can take virtually any value. A parameter is nevertheless to be regarded as
a variable of a model.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 9 / 138

 Ingredients of a Mathematical model - 2

Equations: something that express the relation between variables: definitional

equations - set up identity between two alternate expressions that have exactly the
same meaning, behavioral equations - specify the manner in which a variable behaves
in response to changes in other variables, conditional equation - state requirements to
be satisfied.
Example

π =R−C (Definitional equation)

C = 75 + 10Q (Behavioral equation)
Qd = Qs (Conditional equation)

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 10 / 138

 Examples - Partial market equilibrium

Law of supply and demand

supply and demand pull against each other until the market finds an equilibrium price.

P = f (Qd )

P = g(Qs )

Qs = Qd


In the case where the

supply and demand are
modelled as linear
functions:

P = −aQd + b

P = cQs + d

Qs = Qd


Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 11 / 138

 Matrices - Definitions

Definition
"In mathematics, a matrix (plural matrices), is a rectangular array or table of numbers,
symbols, or expressions, arranged in rows and columns." - Wikipedia.

Let  
7 3 4
A=
1 5 6
be a matrix.
1 How many rows and columns the matrix A has?
2 What is the dimension (size) of A?
3 We denote matrices by capital letters (that is, A, B, C, . . . ) and their entries (or
elements) by the corresponding lower-case letter in ordinary type. For example,
a12 = 3. Determine a13 , a22 , a24 , a34 .

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 12 / 138

 Practice Problem

Let  
  1 0 2 3 1
1 2  
A= B = 1 −1 0 6 2 C = 5 7 9 0 2 D = [6]
3 4
3 4 6 7 8

a. Determine the dimensions of the matrices A, B, C and D.

b. Write down (if possible) the values of

a11 , a22 , b14 , c25 , c33 , c43 , d11

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 13 / 138

 Transposition

Definition
The transpose of a matrix is found by replacing rows by columns, so that the first row
becomes the first column, the second row becomes the second column, and so on.

Example
Write down the transpose of the following matrices:
 
1 4 0 1 2  
 3
  1 2 3
7 6 1 4 
, B = 1 5 ,
A= C= 2 4 5 
 2 1 3 5 −1  7 9
3 5 6
2 −5 1 8 0

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 14 / 138

 Row matrices and column matrices

A row vector (or row matrix) is a matrix that has only one row. For example,
B = [1 5 7] is a row vector.
A column vector is a matrix that has only one column. For example,
 
1
 2 
C=  −3 


Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 15 / 138

 Addition and Scalar multiplication
Let         
7 5 5 2 −6 2 0
A= , B= , C= , D= , O= .
2 1 4 2 1 −9 0
Find

a. A+D b. A+C c. B−C d. C −O e. D−D

Definition

A + B = [entries of A] + [entries of B]
α · A = [α · entries of A]

Remark: To add (or subtract) two matrices of the same size, we simply add (or subtract)
their corresponding elements.
Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 16 / 138
 Matrix multiplication

If A is a row vector  
A = a11 a12 a13 . . . a1n
and B is the column vector  
b11
 b21 
 
B =  b31 
 
 .. 
 . 
bn1
Then we define the matrix product to be the 1 × 1 matrix

AB = [a11 b11 + a12 b21 + a13 b31 + · · · + a1n bn1 ]

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 17 / 138

 Matrix multiplication

Definition
 
columns
AB = [Rows of A] ·  of 
B

Practice Exercise: Let

     
2 1 1 1   1 1  
1 2 1 2 3
A =  5 1 0 B = 2 C= D = −1 1 and E =
 
3 1 4 5 6
−1 1 4 1 2 1

Find (where possible)

a. AB c. CD e. AE g. DE
b. BA d. DC f. EA h. ED
Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 18 / 138
 Summary on Operations on matrices

Definition

A + B = [entries of A] + [entries of B] (Addition)

αA = [α · every entries of A] (Scalar multiplication)
 
columns
AB = [Rows of A] ·  of  (Matrix multiplication)
B

Properties
Associativity: (A + B) + C = A + (B + C), (AB)C = A(BC), (αβ)A = α(βA)
Commutativity: A + B = B + A

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 19 / 138

 Zero matrices and Identity matrices
 
a b
Let A = be an arbitrary 2 × 2 matrix, we know that A + O2×2 = O + A = A.
c d

Question
Is there a matrix, says I, such that

A · I = I · A = A?

Identity matrix
The n × n identity matrix is the n × n matrix in which the entries in the main diagonal
equal to 1 while the other entries equal to 0.
 
  1 0 0
1 0
Example: I2 = and I3 = 0 1 0 are identity matrices.
0 1
0 0 1

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 20 / 138

 Matrix inversion

 
a b
Let A = be an arbitrary 2 × 2 matrix, we know that A + (−A) = (−A) + A = O.
c d

Question
Is there a matrix, denoted by A−1 for example, such that

A · A−1 = A−1 · A = I ?

Definition
The inverse of an n × n matrix A (if exists) is the n × n matrix A−1 such that

A · A−1 = A−1 · A = In×n

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 21 / 138

 Practice Problem - Teamwork

1 Find (where possible) the inverse of the following matrices.

 
  1 2 0
6 4
A= B = 0 1 2
1 2
0 0 1

2 Is the following matrix invertible? Why?

 
1 2
C=
2 4

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 22 / 138

 Determinant

Theorem

a b
The matrix A = is invertible if and only if ad − bc ̸= 0. Furthermore, the inverse of
c d
A is  
1 d −b
ad − bc −c a

The number ad − bc is called the determinant of A and is written as

a b
det(A) or |A| or
c d

If the matrix has a non-zero determinant, it is said to be non-singular or invertible;

otherwise it is said to be singular or non-invertible.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 23 / 138

 Determinant - Definition

If A = (a)1×1 then det A = |A| = a.

 
a11 a12 a a
If A = then det A = 11 12 = a11 a22 − a12 a21 .
a21 a22 2×2 a21 a22
 
a11 a12 a13
If A = a21 a22 a23  then
a31 a32 a33 3×3
a a a a a a
det A = a11 22 23 − a12 21 23 + a13 21 22 .
a32 a33 a31 a33 a31 a32
How is about the determinant of 4 × 4 matrices?

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 24 / 138

 Cofactors

If  
a11 a12 a13
A = a21 a22 a23 
a31 a32 a33
then
a22 a23 a a a a
det A = a11 − a12 21 23 + a13 21 22
a32 a33 a31 a33 a31 a32
If we denote
a22 a23 a21 a23 a21 a22
A11 = A12 = − A13 =
a32 a33 a31 a33 a31 a32
Then
det A = a11 A11 + a12 A12 + a13 A13

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 25 / 138

 Cofactors

a22 a23 a a a a
det A = a11 − a12 21 23 + a13 21 22
a32 a33 a31 a33 a31 a32
det A = a11 A11 + a12 A12 + a13 A13

The cofactor of the element, aij , is the determinant of the matrix left when row i and
column j are deleted, multiplied by +1 or 1, depending on whether i + j is even or
odd, respectively.

Definition
For an n × n matrix A, we have
n
X
det(A) = a1j A1j .
j=1

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 26 / 138

 Inversion

Theorem
A square matrix A is invertible if and only if det A ̸= 0
If the dimension of A is 3 × 3 and det A ̸= 0, then
 
A11 A21 A31
1
A−1 = A12 A22 A32 
det A
A13 A23 A33

The adjugate matrix The adjoint matrix

 
A11 A12 A13  
A11 A21 A31
A21 A22 A23 
A12 A22 A32 
A31 A32 A33
A13 A23 A33

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 27 / 138

 Practice Problem

Find (where possible) the inverses of

   
1 3 3 270 −372 0
A = 1 4 3 and B = 552 201 0
1 3 4 999 413 0

Find the determinant of the following matrix:

 
1 70 2
A = 0 3 0
2 300 1

Remark: In order to calculate the determinant, we can expand along any row/column.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 28 / 138

 Matrices and Linear equations

      
 a11 x + a12 y + a13 z = b1 a11 a12 a13 x b1
a21 x + a22 y + a23 z = b2 ⇐⇒ a21 a22 a23  × y  = b2 
a31 x + a32 y + a33 z = b3 a31 a32 a33 z b3


Hence, the system can be written as

Ax = b

The vector of unknowns, x, can be found by

x = A−1 b

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 29 / 138

 When A is non-singular - Cramer’s rule

Given the system

Ax = b
where      
a11 a12 ... a1n x1 b1
 a21 a22 ... a2n   x2   b2 
A= . x= .  b=.
     
.. .. .. 
 .. . . .   ..   .. 
an1 an2 . . . ann xn bn
Then
det Ai
xi =
det A
where Ai is the n × n matrix found by replacing the ith column of A by the
right-hand-size vector b.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 30 / 138

 Practice Problem

1 Solve the system of equations


2x1 + 4x2 = 16
3x1 − 5x2 = −9

using Cramer’s rule to find x2 .

2 Solve the system of equations

 4x1 + x2 + 3x3 = 8

−2x1 + 5x2 + x3 = 4

3x1 + 2x2 + 4x3 = 9


using Cramer’s rule to find x3 .

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 31 / 138

 Rank of a matrix-Elementary row operations

Elementary row operations

1. (Replacement) Replace one row by the sum of itself and a multiple of another row.
2. (Interchange) Interchange two rows.
3. (Scaling) Multiply all entries in a row by a nonzero constant.

Two matrices are called row equivalent if there is a sequence of elementary row operations
that transforms one matrix into the other.
Question
Do these preserve a system of linear equations?

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 32 / 138

 Row echelon forms
A rectangular matrix is in echelon form (or row echelon form) if it has the following three
properties:
1 All nonzero rows are above any rows of all zeros.
2 Each leading entry of a row is in a column to the right of the leading entry of the row
above it.
3 All entries in a column below a leading entry are zeros.
If a matrix in echelon form satisfies the following additional conditions, then it is in
reduced echelon form (or reduced row echelon form):
1 The leading entry in each nonzero row is 1.
2 Each leading 1 is the only nonzero entry in its column.

Uniqueness of the reduced echelon form

Each matrix is row equivalent to one and only one reduced echelon matrix.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 33 / 138

 Gauss-Jordan method
Pivot column
A pivot position in a matrix A is a location in A that corresponds to a leading 1 in the
reduced echelon form of A. A pivot column is a column of A that contains a pivot position.

Gauss-Jordan elimination method

1 Begin with the left-most nonzero the pivot position and cover all rows, if
column. This is a pivot column. any, above it. Apply steps 1–3 to the
2 Select a nonzero entry in the pivot sub-matrix that remains. Repeat the
column as a pivot (interchange rows if process until there are no more nonzero
necessary) rows to modify.
3 Use row replacement operations to 5 Beginning with the rightmost pivot and
create zeros in all positions below the working upward and to the left, create
pivot. zeros above each pivot. If a pivot is not
4 Cover (or ignore) the row containing 1, make it 1 by a scaling operation.
Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 34 / 138
 Reduced row echelon form

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 35 / 138

 Ranks of matrices

Rank of a matrix
The rank of a matrix A is the number of non-zero rows in its echelon form.

Example
Determine the rank of the following matrices:
     
1 3 3 270 −372 0 1 70 2
A = 1 4 3 , B = 552 201 0 , and C = 0 3 0
1 3 4 999 413 0 2 300 1

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 36 / 138

 Kronecker-Capelli Theorem

Theorem
A system of linear equations Ax = b with n variables has a solution if and only if the rank
of its coefficient matrix A is equal to the rank of its augmented matrix [A | b]. In
particular:
if n = rank(A), the solution is unique,
otherwise, there are infinitely many solutions.

Example
Determine if the following is consistent:

x1 − 2x2 + x3 = 0
2x2 − 8x3 = 8
5x1 − 5x3 = 10

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 37 / 138

 Free variables
Example
Suppose, for example, that the augmented matrix of a linear system has been changed
into the equivalent reduce echelon form
 
1 0 2 1
0 1 1 4
0 0 0 0

The variables x1 and x2 corresponding to pivot columns in the matrix are called basic
variables. The other variable x3 is called a free variable.
The (general) solution for this system of equations is

x1 = 1 − 2x3
(
x1 = 1 − 2x3

x2 = 1 − x3 or
 x2 = 1 − x3
x3 is free


Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 38 / 138

 The number of free variables

Theorem
Let Ax = b be a system of linear equations which has n variables.
1 If rankA < rank(A b) then the system has no solution.
2 If rank(A) = rank(A b) = n then the system has a unique solution.
3 If rank(A) = rank(A b) = k < n then the system has infinitely many solutions with
n − k free variables.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 39 / 138

 Rn space

Rn = {(x1 , x2 , . . . , xn ) : x1 , x2 , . . . , xn ∈ R} .
Each element in Rn is called a vector of dimension n.
Vector operations
Let u = (x1 , x2 , . . . , xn ) and v = (y1 , y2 , . . . , yn ) be two vectors in Rn and let α be a real
number then

u + v = (x1 + y1 , x2 + y2 , . . . , xn + yn ) (Addition)
αu = (αx1 , αx2 , . . . , αxn ) (Scalar multiplication)

Remark: For convenience, from now on, we will write vectors as columns.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 40 / 138

 First example

Practice Exercise: Let

       
1 2 −1 5
′
u = 2 , v = 5 , w = −1 , and w = 9  .
      
3 6 −3 12

be the vectors in R3 . Determine a, b, c, d in R such that w = au + bv and w′ = cu + dv.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 41 / 138

 Linear combination

We have w = −3u + v. However, there is no c, d such that w′ = cu + dv. We say that

w is a linear combination of the vectors u and v.
w′ can not be written as a linear combination of the vectors u and v.

Definition
Let {u1 , u2 , . . . , uk } be a family of vectors in Rn .
Any vector of the form α1 u1 + α2 u2 + · · · + αk uk , for some α1 , . . . , αk ∈ R, is called a
linear combination of u1 , u2 , . . . , uk .
Denote Span(u1 , . . . , uk ) := {α1 u1 + α2 u2 + · · · + αk uk : α1 , . . . , αk ∈ R}. Then
Span(u1 , . . . , uk ) is a vector subspace of Rn . This space is called the subspace spanned
by u1 , . . . , uk .

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 42 / 138

 Exercises
   
1 −1
Exercise 1: Let u = 1 and v =  0  be vectors in R3 . Determine Span(u, v) and
0 2
prove that it is a vector subspace of R3 .
Exercise 2: In R3 , given the following vectors:
         
1 0 0 2 0
v1 = 0 , v2 = 1 , v3 = 0 , v4 = 3 , O = 0 .
0 0 1 5 0

Determine a linear combination (if any) of

1 v1 via v2 , v3 , v4 . 4 v4 via v1 , v2 , v3 .
2 v2 via v1 , v3 , v4 .
3 v3 via v1 , v2 , v4 . 5 O via v1 , v2 , v3 .

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 43 / 138

 Linear independence

In R3 , let Definition: A family of vectors

 
1
 
0
 
0 {v1 , v2 , . . . , vm } in Rn is said to be linearly
e1 = 0 , e2 = 1 , e3 = 0 .
     independent if each vi is not a linear
0 0 1 combination of the rest, for each
i = 1, 2, . . . , m.
Then we can verify that
e1 is not a linear combination of Example: {e1 } is independent, {e1 , e2 } is
{e2 , e3 }, independent.
e2 is not a linear combination of Proposition: Let v1 , v2 , . . . , vm be linearly
{e1 , e3 }, and independent vectors. Then {v1 }, {v2 , v3 }, ...
e3 is not a linear combination of are linearly independent. In general, any
{e1 , e2 }. sub-family of {v1 , . . . , vm } is linearly
In this case, we say that {e1 , e2 , e3 } is independent.
linearly independent (in short,
independent).
Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 44 / 138
 A criterion for Linear Independence
           
1 1 0 1 1 0
u1 = 2 , u2 = 1 , u3 = 0 , v1 = 2 , v2 = 3 , v3 = 1
          
0 0 1 0 1 1

1 Prove that u1 , u2 , u3 are linearly 3 Express the zero vector O = (0, 0, 0) as

independent. a linear combination of {u1 , u2 , u3 }.
2 Prove that v1 , v2 , v3 are linearly 4 Write the zero vector O as two linear
dependent. combinations of {v1 , v2 , v3 }.

Proposition
The vectors v1 , v2 , . . . , vm are linearly independent if and only if O has a unique
expression as a linear combination of v1 , v2 , . . . , vm , i.e., the vector equation
x1 v1 + x2 v2 + · · · + xm vm has a unique solution. That unique expression is
O = 0 · v1 + 0 · v2 + · · · + 0 · vm .
Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 45 / 138
 Rank of a system of vectors

Proposition
A system of vectors v1 , v2 , . . . , vm in Rn is linear independent if and only if the rank of the
matrices [v1 v2 · · · vm ] equals to m, the number of vectors.
     
1 0 0
Example: The vectors u1 = 1 , u2 = 1 , and u3 = 0 are linearly independent
1 1 1
because  
1 0 0
rank [u1 u2 u3 ] = rank 1 1 0 = 3
1 1 1
Remark: If m = n then a system of vectors v1 , v2 , . . . , vm in Rn is linearly independent if
and only if the vector equation x1 · v1 + x2 · v2 + · · · + xm vm = 0 has only trivial solution

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 46 / 138

 Demand functions

1 In microeconomics, the quantity demanded (Qd ) of a good

depends on the market price (P ).

Qd = f (P )

Alfred Marshall (1842-1924)

2 However, economists also consider P , the price, as a

function depending on quantity, Qd .

P = g(Qd )
Léon Walras (1834-1910)

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 47 / 138

 The Laws of Demand and Supply

The law of demand

Demand usually falls as the price of a good rises. That is, P is a decreasing function of Qd .

Remark: If the demand function P is modelled as a linear function, i.e., P = aQd + b,

then a < 0 < b.

The law of supply

An increase in price results in an increase in quantity supplied. That is, P is an increasing
function of Qs .

Remark: If the supply function P is modelled as a linear function, i.e., P = aQs + b, then
a > 0 and b > 0. Consequently, if the supply function is expressed as Qs = cP + d, then
c > 0 > d.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 48 / 138

 The law of demand-supply

Law of supply and demand: supply and

demand pull against each other until the
market finds an equilibrium price.
Equilibrium (market): This state occurs
when quantity supplied and quantity
demanded are equal.
Remark: The supply and demand
functions are not necessary to be linear.
However, in this chapter, we assume that
they are linear.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 49 / 138

 The supply-demand model

The supply-demand model

In general case: In the case where the supply and demand
are modelled as linear functions:
 
P = f (Qd )
 P = −aQd + b

P = g(Qs ) P = cQs + d
 
Qs = Qd Qs = Qd
 

Example: The demand and supply functions of a good are given by P = 2Qd + 50 and
P = 1/2Qs + 25, where P, Qd and Qs denote the price, quantity demanded and quantity
supplied, respectively.
a. Determine the equilibrium price and quantity.
b. Determine the effect on the market equilibrium if the government decides to impose a
fixed tax of $5 on each good.
Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 50 / 138
 The market model - Two goods

A general model When Qd , Qs are linear

 
Q = f1 (P1 , P2 ) Qd,1 = a1 P1 + a2 P2 + a3
 d,1





Qd,2 = f2 (P1 , P2 ) Q = α1 P1 + α2 P2 + α3
 d,2

 


 
Q = g (P , P ) Q = b P + b P + b
s,1 1 1 2 s,1 1 1 2 2 3


Qs,2 = g (P
2 1 , P2) 

Qs,2 = β P
1 1 + β P
2 2 + β 3
 



Qd,1 = Qs,1 


Qd,1 = Qs,1
 
Qd,2 = Qs,2 Qd,2 = Qs,2
Example: The demand and supply functions for two interdependent commodities are
given by Qd1 = 10 − 2P1 + P2 , Qd2 = 5 + 2P1 − 2P2 , Qs1 = −3 + 2P1 , Qs2 = −2 + 3P2 ,
where Qdi , Qsi and Pi denote the quantity demanded, quantity supplied and price of good
i, respectively. Determine the equilibrium price and quantity for this two-commodity
model.
Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 51 / 138
 Complementary and substitutable goods

Suppose that the demands and the supplies Question: What can we say about the sign
of two goods are given by of the parameters

Qd,1 = a1 P1 + a2 P2 + a3


 a1 , a 2 , a 3 , α 1 , α 2 , α 3 , b 1 , b 2 , b 3 , β 1 , β 2 , β 3 .
Q = α P + α P + α
d,2 1 1 2 2 3


 Qs,1 = b1 P1 + b2 P2 + b3

Qs,2 = β1 P1 + β2 P2 + β3

1 Two goods are said to be substitutable, if an increase in the price of one good causes
an increase in the demand of the other good.
2 Two goods are said to be complementary, if an increase in the price of one good
causes a decrease in the demand of the other good.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 52 / 138

 National income determination

Macroeconomics is concerned with

the analysis of economic theory and
policy at a national level. We focus
on one particular aspect known as
national income determination. We
describe how to set up simple
models of the national economy
which enable equilibrium levels of
income to be calculated.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 53 / 138

 National income determination - variables

Firms use resources such as land,

capital, labour and raw materials to
produce goods and services. These
resources are known as factors of
production and are taken to belong
to households. National income
represents the flow of income from
firms to households given as
payment for these factors.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 54 / 138

 National income determination - variables

The economy is in equilibrium if


Y = C + I

Y =C +S

I= S


where Y is the national income, C

is the consumption and S is the
savings.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 55 / 138

 Consumption functions

The consumption function is a function of Y : C = f (Y ).

Question
Is f a decreasing function?

If the relationship between C and Y are linear, we may assume C = aY + b, where 0 < b
and 0 < a < 1.
Savings function: S = Y − C.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 56 / 138

 The equilibrium level of income and consumption
If the level of investment that firms plan to inject into the economy is supposed to be
some fixed value, I ∗ , then the equilibrium level of income and consumption can be
calculated via the following system of equations:

 Y = C +I
C = aY + b
I = I∗


Example
Find the equilibrium level of income and consumption if the consumption function is

C = 0.6Y + 10

and the planned investment I = 12.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 57 / 138

 Government Expenditure and Taxation

To make the model more realistic, let us Example: Given that

now include government expenditure, G,
and taxation, T , in the model. If so, the 
income that households have to spend on 
 G= 40
consumer goods is no longer Y but rather I= 55

Y − T , which is called disposable income, 
 C = 0.8Yd + 25
T = 0.1Y + 10.

Yd . 

 Y = C +I +G
C = aYd + b


 T = cY + d Calculate the equilibrium level of national

Yd = Y − T income.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 58 / 138

 IS Schedule
It is more realistic to assume that planned investment depends on the rate of interest, r.
As the interest rate rises, so investment falls. Therefore, we may suppose that I = f (r). If
the relationship between those are linear, we may write

I = cr + d, where c < 0 and d > 0.

From the equations 

 Y = C +I
C = aY + b
I = cr + d,


we may eliminate C and I to set up an equation relating Y and r, which is called IS

schedule.
Question
Determine the IS schedule in the above system.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 59 / 138

 The equilibrium of the money market

To make the model more realistic, we will need some additional information, that is the
equilibrium of the money market.
The equilibrium of the money market
The money market is said to be in equilibrium when the supply of money, Ms , matches
the demand for money, Md : that is, when

Ms = M d .

The level of money supply, Ms , is assumed to be controlled by the central bank and is
taken to be autonomous, so that
Ms = Ms∗
for some fixed value Ms∗ .

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 60 / 138

 Money demand
1 The transactions demand is used for the daily exchange of goods and services, Ms,1 .
2 The precautionary demand is used to fund any emergencies requiring unforeseen
expenditure, Ms,2 .
3 The speculative demand is used as a reserve fund in case individuals or firms decide

to invest in alternative assets, such as government bonds, Ms,3 .

We assume transactions demand and precautionary demand are proportional to national
income. Consequently, Ms,1 + Ms,2 = k1 Y for some positive constant k1 .
As interest rates rise, speculative demand falls. We model this by writing
Ms,3 = k2 r + k3
where k2 is a negative constant and k3 is a positive constant.
The money demanded,
Md = Ms,1 + Ms,2 + Ms,3 = k1 Y + k2 r + k3 .

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 61 / 138

 LM Schedule
If the money market is equilibrium, then Md = Ms , or

k1 Y + k2 r + k3 = Ms∗

LM Schedule
The above equation, relating national income, Y , and interest rate, r, is called the LM
schedule.

Exercise: Determine the equilibrium Sketch the IS and LM curves on the same
income, Y , and interest rate, r, given the diagram. What effect would an increase in
following commodity market C = 0.7Y + 85, the value of autonomous investment have
I = 50r + 1200, and the money market on the equilibrium values of Y and r?
Ms = 300, Md = 0.2Y − 40r + 30.
Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 62 / 138
 Input-output Analysis
The input-output analysis of Prof. Wassily Leontief, a Nobel Prize winner, deals with this
particular question: “What level of output should each of the n industries in an economy
produce, in order that it will just be sufficient to satisfy the total demand for that
product?"
Assumptions:
1 Each industry produces only one homogeneous commodity.
2 Each industry uses a fixed input ratio for the production of its output.
3 Production in every industry is subject to constant returns to scale.
That means:
in order to produce each unit of the j-th commodity, the input need for the i-th
commodity must be a fixed amount, which we shall denote by aij , and refer to as an
input coefficient.
For example, a32 = 0.35 means that 0.35$ worth of the third commodity is required
as an input for producing 1$ worth of the second commodity.
Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 63 / 138
 Example - I/O table of Vietnam in 2007

In order to
produce 1 VND
“Thóc", it is
required the
inputs of 0.08663
VND “Thóc",
0.00477 VND
“Mía cây",
0.01260 VND
“Cây hàng năm
khác", and so on.

Source: IO-Viet-Nam-2007.pdf

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 64 / 138

 Structure of Input-output Model

Coefficient matrix: activities.

Let’s denote by xi the output level of
 
a11 a12 ··· a1n
 a21 a22 ··· a2n  industry i. Then the model is in
A= . equilibrium if
 
.. .. .. 
 .. . . . 
an1 an2 · · · ann x1 = a11 x1 + a12 x2 + · · · + a1n xn + b1
Each column sum represents the partial x2 = a21 x1 + a22 x2 + · · · + a2n xn + b2
input cost and must be less than 1. ···
Final demand is a column vector xn = an1 x1 + an2 x2 + · · · + ann xn + bn
b = [b1 b2 . . . bn ]t , including
consumer demand, government This model is called Leontieff’s
expenditure and export - import input-output model.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 65 / 138

 I/O model in terms of Matrices

x1 = a11 x1 + a12 x2 + · · · + a1n xn + b1




x = a x + a x + · · · + a x + b

2 21 1 22 2 2n n 2
⇔ X = AX+b ⇔ (I−A)X = b ⇔ X = (I−A)−1 b


 · · ·

xn = an1 x1 + an2 x2 + · · · + ann xn + bn

Exercises: Given the input matrix and the 1 Explain the economic meaning of the
final-demand vector entries 0.33, 0, and 200.
   
0.05 0.25 0.34 1800 2 Explain the economic meaning of the
A = 0.33 0.10 0.12 , b =  200  third-column sum.
0.19 0.38 0 900
3 Determine the levels of outputs.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 66 / 138

 Homework (Teamwork)

One-commodity model: Two-commodity model:


Qd =
 Qs Qd,1 = Qs,1 , Qd,2 = Qs,2
Qd = a − bP Qd,1 = ao − a1 P1 + a2 P2


Qs = −c + dP Qs,1 = −bo + b1 P1 + b2 P2
Qd,2 = αo − α1 P1 + α2 P2
Equilibrium solution:
Qs,2 = −βo + β1 P1 + β2 P2
a+c ad − bc
P∗ = , Q∗ =
b+d b+d
Find the equilibrium prices and quantities.
To be economically meaningful, the model Determine conditions on the parameters to
should contain the additional restriction make the model economically meaningful.
that a, b, c, d > 0 and ad > bc. Reference: p.41 of the textbook

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 67 / 138

 Elementary Analysis

Differentiation: the first-order derivative - the slope, the rate of change -

interpretation (marginal values, elasticities), the second-order derivative - the
convexity - interpretation, Rules of differentiation - the sum-difference rule, the
product, quotient, chain rules, the inverse rule.
Partial differentiation: the first, second-order derivatives, partial marginal values,
elasticities, Implicit functions and rate of substitutions.
Integration: (in)definite integral, rules, integral by parts, improper integrals,
economical applications: producer, customer’s surpluses; net investment and capital
formation.
Optimization: constrained, unconstrained optimization, Lagrange multiplier. Utility
maximization and production maximization.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 68 / 138

 Differentiation

The first-order derivative of a function f (x) at x0 is defined as:

f (x0 + ∆x) − f (x0 )

f ′ (x0 ) = lim
∆x→0 ∆x

Example:
1 Given y = u3 + 2u, where u = 5 − x2 ,
find dy/dx by the chain rule.
2 Given y = 4x5 + x3 + 2x. Determine
y ′ (5).

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 69 / 138

 The second-order Derivative, the Convexity

The second-order derivative:

f ′′ (x) = (f ′ )′ (x)

Concave and Convex:

Example.
Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 70 / 138
 Partial Differentiation

Let f (x, y) = x2 y. be a function of two Remark: We also use the notation fx for
variables x and y. Then the derivative of f with respect to x.
Remark: In the case f = f (x1 , . . . , xn ), we
also use the notation fi for the derivative of
f with respect to xi for each i = 1, 2, . . ., n.
Example: Let f (x, y) = x2 y. Then

∂f
= 2xy and fy = x2
∂x

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 71 / 138

 Second-order derivatives
The function obtained by differentiating twice with respect to (w.r.t.) x:
∂2z ∂2f
or or fxx
∂x2 ∂x2
The function obtained by differentiating twice w.r.t. y:
∂2z ∂2f
or or fyy
∂y 2 ∂y 2
The function obtained by differentiating first w.r.t. x and then w.r.t. y:
∂2z ∂2f
or or fyx
∂y∂x ∂y∂x
The function obtained by differentiating first w.r.t. y and then w.r.t. x:
∂2z ∂2f
or or fxy
∂x∂y ∂x∂y
Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 72 / 138
 Examples

Example
Find expressions for the second-order partial derivatives of the functions

a. f (x, y) = 5x4 − y 2 b. f (x, y) = x2 y 3 − 10x

Example
Find expressions for the partial derivatives f1 , f11 and f21 in the case when

f (x1 , x2 , x3 ) = x1 x2 + x51 − x22 x3

Young’s theorem
If fxy and fyx are both continuous then

fxy = fyx

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 73 / 138

 Total Differentials
Let z = z(x, y) be a function of two (independent) variables x and y. Then

∆z = z(x + ∆x, y + ∆y) − z(x, y)

= z(x + ∆x, y + ∆y) − z(x, y + ∆y) + z(x, y + ∆y) − z(x, y)
∂z ∂z
≈ ∆x + ∆y
∂x ∂y
When ∆x, ∆y → 0, we have
∂z ∂z
dz = dx + dy.
∂x ∂y
This is defined as the total differential of z.
Example
Determine the total differential of z = x2 + y 3 and estimate the change on z at
(xo = 50, yo = 50) when both inputs increase by 1 unit.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 74 / 138

 Rules of Total Differential

d(c) = 0 (constant-function rule)

n n−1
d(cu ) = cnu du (power-function rule)
d(u ± v) = du ± dv (sum-difference rule)
d(uv) = udv + vdu (product rule)
u 1
d = 2 (vdu − udv) (quotient rule)
v v

Example
Determine the total differentials of the following functions:

a) z = xy d) z = xex
b) z =x+y e) z = Axα y β
c) z = ex+y f) f = Axα y β + z(M − ax − by)
Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 75 / 138
 Second-order Total Differentials

Given z = f (x, y), then

d2 z = fxx dx2 + 2fxy dxdy + fyy dy 2

Example
Given z = x3 + 5xy − y 2 , then

dz = (3x2 + 5y)dx + (5x − 2y)dy

d2 z = 6xdx2 + 10dxdy − 2dy 2

At the point (x = 1, y = 2), we have

dz = 13dx + dy and d2 z = 6dx2 + 10dxdy − 2dy 2

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 76 / 138

 Total Derivatives - 1 variable

Let y = f (x, w) be a function where x = g(w). Then y is a function of w and

dy dx ∂y dx ∂y
= fx + fw = +
dw dw ∂x dw ∂w
The above equation can be obtained by firstly differentiation the function y totally
and then dividing through by dw.
Let y = f (x1 , x2 , . . . , xn , w) where x1 , . . . , xn are functions of w. Then

dy ∂y dx1 ∂y dx2 ∂y dxn ∂y

= + + ··· + +
dw ∂x1 dw ∂x2 dw ∂xn dw ∂w

Example
Determine the total derivative of z = y ln ex where y = x2 .

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 77 / 138

 Total Derivatives - 2 variables
Let y = f (x1 , x2 , . . . , xn , u, v) where x1 , . . . , xn are functions of u, v. Then

dy ∂y ∂x1 ∂y ∂x2 ∂y ∂xn ∂y

= + + ··· + +
du ∂x1 ∂u ∂x2 ∂u ∂xn ∂u ∂u
Since y can be considered as a function of 2 independent variables u and v, we should
use another notation for the total derivative, instead of dy/du.

ðy ∂y ∂x1 ∂y ∂x2 ∂y ∂xn ∂y

= + + ··· + +
ðu ∂x1 ∂u ∂x2 ∂u ∂xn ∂u ∂u

Example
Determine the total derivatives of z = y1 ex1 + y2 sin(2x2 ) where y1 = x1 + x2 and
y2 = x 1 x 2 .

More exercises: page 193-194.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 78 / 138
 Implicit functions

Explicit function: y = y(x) Implicit function: y = y(x)

y = x1/3 y3 − x = 0

Implicit function
A function y = y(x1 , . . . , xn ) is called an implicit function if y is defined by an (implicit)
equation
F (y, x1 , x2 , . . . , xn ) = 0

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 79 / 138

 Implicit-function Theorem

Implicit-function theorem
A given equation in the form
F (y, x1 , x2 , . . . , xn ) = 0
determines an implicit function y = y(x1 , . . . , xn ) around a point (y ∗ , x∗1 , . . . , x∗n ) if
F has continuous derivatives Fy , F1 , . . . , Fn , and
F (y ∗ , x∗1 , . . . , x∗n ) = 0, and
Fy (y ∗ , x∗1 , . . . , x∗n ) ̸= 0.

Example
The equation F (x, y) = x2 + y 2 − 9 = 0 determines a relation between x and y. Does it
determine an implicit function y = y(x)? Explain.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 80 / 138

 Derivatives of implicit functions
Let y = f (x1 , . . . , xn ) be an implicit function of n independent variables x1 , . . . , xn defined
by the equation
F (y, x1 , . . . , xn ) = 0.
By taking the total differentials for both sides of the equation, we have
Fy dy + F1 dx1 + F2 dx2 + · · · + Fm dxm = 0.
Hence,
(Fy f1 + F1 )dx1 + (Fy f2 + F2 )dx2 + · · · + (Fy fm + Fm )dxm = 0
Since x1 , . . . , xn are independent,
Fy fi + Fi = 0 for every i
Equivalently,
∂y Fi
=− for every i.
∂xi Fy
Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 81 / 138
 Examples
Example
If y = y(x, w) is an implicit function defined by the equation

F (y, x, w) = y 3 x2 + w3 + yxw − 3 = 0,

then
∂y Fx 2y 3 x + yw
=− =− 2 2
∂x Fy 3y x + xw

Example
Assume that the equation F (Q, K, L) = 0 implicitly defines a production function
Q = f (K, L). Then the marginal productivity of capital and labour are

∂Q FK ∂Q FL
MPPK ≡ =− and MPPL ≡ =−
∂K FQ ∂L FQ
Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 82 / 138
 Marginal values and Elasticity

Marginal value:
the absolute change on f
≈ f ′ (x)
the absolute change on x
Interpretation: The marginal value of f at x = xo estimates the absolute change of f
when the variable x increases by 1 unit.
Elasticity:
the percentage change on f x
≈ f ′ (x) ×
the percentage change on x f (x)
Interpretation: The elasticity of f at xo estimates the percentage change of f when
the variable x increases by 1%.
The demand is said to be (unit, in)elastic at xo if |ϵd (xo )|(=, <) > 1.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 83 / 138

 Partial Marginal values and Elasticities

Given an economical function f (x1 , x2 , . . . , xn ),

The partial marginal value of f with respect to the variable (w.r.t.) x1 :

∂f the absolute change on f

≈
∂x1 the absolute change on x1

The partial elasticity of f with respect to the variable (w.r.t.) x1 :

x1 the percentage change on f

fx′ 1 × ≈
f (x1 , . . . , xn ) the percentage change on x1

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 84 / 138

 Substitution Rates

A marginal rate of substitution (MRS) is the amount of an input needed to increase

to maintain the level of the output when another input falls down to 1 unit.
Example. Given a product function Q = Q(K, L), the marginal rate of technical
substitution, denoted by MRST, at (Ko , Lo ) is the amount of capital needed to add to
maintain the output Qo = Q(Ko , Lo ) when the amount of labour is reduced by 1 unit.
Example. Given a utility function U = U (x1 , x2 ), the marginal rate of commodity
substitution, denoted by MRCT, is the amount by which one input needs to increase
to maintain a constant value of utility when the other input decreases by 1 unit.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 85 / 138

Marginal rate of Commodity substitution and Technical Substituation

If y = y(x1 , . . . , xn ) is an implicit function

given by

F (y, x1 , . . . , xn ) = 0

then
∂y F′
= − x′1
∂x1 Fy
Given U = U (x1 , x2 ). Then

dx2 U x1
MRCS = − =
dx1 U x2

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 86 / 138

 Optimization

Stationary points: f ′ (x) = 0

 ′
f (x) = 0
Local maximum:
f ′′ (x) < 0
 ′
f (x) = 0
Local minimum:
f ′′ (x) > 0
f ′ (x) = 0

Stationary point of inflection: ⇒
f ′′ (x) = 0

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 87 / 138

 Higher-order Derivative Test

f ′ (xo ) ̸= 0: xo is not local maximum or minimum.

f ′ (xo ) = 0, f ′′ (xo ) > 0: local minimum.
f ′ (xo ) = 0, f ′′ (xo ) < 0: local maximum.
f ′ (xo ) = 0, f ′′ (xo ) = 0, f ′′′ (xo ) ̸= 0: inflection point.

N th-Derivative test
Let f (x) be a function so that we can its N -th derivative. Assume that there is a point xo
so that f ′ (xo ) = · · · = f (N −1) (xo ) = 0 and f (N ) (xo ) ̸= 0.
1 If N is odd then xo is an inflection point.
2 If N is even and f (N ) (xo ) < 0 then xo is a local maximum point.
3 If N is even and f (N ) (xo ) > 0 then xo is a local minimum point.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 88 / 138

 Higher-order Derivative Test - Practice

Example
Find the stationary values of the following functions:

(a) y = x3 (b) y = −x4 (c) y = x6 + 5

Determine by the N th-derivative test whether they represent local maximum, minimum,
or inflection points.

More exercises: page 254

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 89 / 138

 Unconstrained Optimisation

One variable Two variables

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 90 / 138

 Second-order Derivatives test

1 Solve the simultaneous equations


fx (x, y) = 0
fy (x, y) = 0

to find the stationary points.

2 2 > 0 at (a, b) then the function has a minimum at
If fxx > 0, fyy > 0 and fxx fyy − fxy
(a, b).
3 2 > 0 at (a, b) then the function has a maximum at
If fxx < 0, fyy < 0 and fxx fyy − fxy
(a, b).
4 2 < 0 at (a, b) then the function has a saddle point at (a, b).
If fxx fyy − fxy

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 91 / 138

 Second-order condition in Total Differentials approach

Let z = f (x, y) be a 2-variables function which has a stationary point (xo , yo ), i.e.
fx (xo , yo ) = fy (xo , yo ) = 0 or df (xo , yo ) = 0.
If d2 f (xo , yo ) > 0 for arbitrary values of dx and dy, not both zero, then (xo , yo ) is a
local minimum point of f .
If d2 f (xo , yo ) < 0 for arbitrary values of dx and dy, not both zero, then (xo , yo ) is a
local maximum point of f .
If the sign of d2 f (xo , yo ) is positive for some values of dx and dy, but negative for
others, then (xo , yo ) is a saddle point of f .

Example
Find the extreme value(s) of z = x2 + xy + 2y 2 + 3.

More exercises: p300-301.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 92 / 138

 2-Variable Case

Let
z = f (x1 , x2 , x3 )
Then
dz = f1 dx1 + f2 dx2 + f3 dx3
d2 z = f11 dx21 + f12 dx1 dx2 + f13 dx1 dx3
+f21 dx2 dx1 + f22 dx22 + f23 dx2 dx3
+f31 dx3 dx1 + f32 dx3 dx2 + f33 dx23
Once again,
x∗ = (x∗1 , x∗2 , x∗3 ) is a stationary point if df (x∗ ) = 0.
x∗ is local minimum if d2 z(x∗ ) > 0 for arbitrary dx, dy, dz, not all zero.
x∗ is local maximum if d2 z(x∗ ) < 0 for arbitrary dx, dy, dz, not all zero.
x∗ is saddle if d2 z(x∗ ) changes its sign.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 93 / 138

 Hessian determinant test - Leading principal minors

z = f (x1 , x2 , x3 )
Then
d2 z = f11 dx21 + f12 dx1 dx2 + f13 dx1 dx3
+f21 dx2 dx1 + f22 dx22 + f23 dx2 dx3
+f31 dx3 dx1 + f32 dx3 dx2 + f33 dx23
The symmetric Hessian determinant:
f11 f12 f13
|H| = f21 f22 f23
f31 f32 f33
whose leading principal minors are
f11 f12
|H1 | = f11 |H2 | = |H3 | = |H|
f21 f22
Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 94 / 138
 Hessian Determinant test

Hessian determinant test for 3-variables case

If x∗ is a stationary point of z, i.e. dz(x∗ ) = 0, then the following statement is hold:
 
∗ ∗ maximum
z ≡ z(x ) is a if
minimum
|H1 | < 0, |H2 | > 0, |H3 | < 0 (d2 z negative definite)
 

|H1 | < 0, |H2 | > 0, |H3 | < 0 (d2 z positive definite)

Example
Find the extreme values of the following functions.
z = x21 + 3x22 − 3x1 x2 + 4x2 x3 + 6x23
z = x1 x3 + x21 − x2 + x2 x3 + x22 + 3x23
z = x41 − 16x21 + x22 + x2 x3 + x23

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 95 / 138

 n-Variable Case

z = f (x1 , x2 , . . . , xn )
The total differential will then be
dz = f1 dx1 + f2 dx2 + · · · + fn dxn
The Hessian
f11 f12 ··· f1n
f21 f22 ··· f2n
|H| = . .. .. ..
.. . . .
fn1 fn2 · · · fnn

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 96 / 138

 Constrained Optimization
Let us consider a consumer with the simple utility function
U = x1 x2 + 2x1
To render the optimization problem meaningful, a budget constraint should be
incorporated into the problem. If the consumer intends to spend a given sum, say, $60, on
the two goods and if the current prices are
P1 = 4, and P2 = 2,
then the budget constraint can be expressed by
4x1 + 2x2 = 60
The problem now is to
optimize U = x1 x2 + 2x1
subject to
4x1 + 2x2 = 60
Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 97 / 138
 An illustration by Graph

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 98 / 138

 Example - Practice

Solve
optimize U = x1 x2 + 2x1
subject to
4x1 + 2x2 = 60

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 99 / 138

 Constrained Optimization and Lagrange-multiplier

optimize U = x1 x2 + 2x1
subject to
4x1 + 2x2 = 60
Idea of the method: is to convert a constrained optimization problem into a form of
unconstrained optimization problem.
The (corresponding) Lagrangian function of is of 3-variables:
L(x, y, λ) = x1 x2 + 2x1 + λ(60 − 4x1 − 2x2 )
For then the first-order condition for unconstrained optimization will consist of the
set of simultaneous equations:
Lλ = 60 − 4x1 − 2x2 = 0
L1 = x2 + 2 − 4λ = 0
L2 = x1 − 2λ = 0
Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 100 / 138
 Lagrange-multiplier method

optimize z = f (x1 , x2 , . . . , xn )
subject to
g(x1 , . . . , xn ) = c
It follows that the Lagrangian function will be
L = f (x1 , x2 , . . . , xn ) + λ[c − g(x1 , x2 , . . . , xn )]
for which the first-order condition will consist of the following (n + 1) simultaneous
equations:
Lλ = c − g(x1 , . . . , xn ) = 0
L1 = f1 − λg1 = 0
······
Ln = fn − λgn = 0
If x = (x∗1 , . . . , x∗n ) is a solution of the constrained optimization then there is λ∗ so that
(x∗1 , . . . , x∗n , λ∗ ) is a root of the above system.
Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 101 / 138
 The case of more than one constraint
When there is more than one constraint, the Lagrange-multiplier method is equally
applicable, provided we introduce as many such multipliers as there are constraints in the
Lagrangian function. For example,

optimize z = f (x1 , x2 , . . . , xn )
subject to
g(x1 , . . . , xn ) = c and h(x1 , . . . , xn ) = d

Then we can define:

L = f + λ[c − g(x1 , . . . , xn )] + µ[d − h(x1 , . . . , xn )]
The first-order condition consists of
Lλ = c − g(x1 , x2 , . . . , xn ) = 0
Lµ = d − h(x1 , x2 , . . . , xn ) = 0
Li = fi − λgi − µhi (i = 1, 2, . . . , n)
Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 102 / 138
 Second-order Conditions

optimize z = f (x, y)
subject to
g(x, y) = c

Since g(x, y) = c, we have

dg = gx dx + gy dy = 0
d2 g = gxx dx2 + 2gxy dxdy + gyy dy 2 + gy d2 y = 0
gx
Hence, dy and dx are no longer both arbitrary: dy = − dx. By direct calculations (page
gy
357), we obtain
     
fy fy fy
d z = fxx − gxx dx + 2 fxy − gxy dxdy + fyy − gyy dy 2
2 2
gy gy gy

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 103 / 138
 Second-order Conditions - 2

     
fy fy fy
2
d z = fxx − gxx dx + 2 fxy − gxy dxdy + fyy − gyy dy 2
2
gy gy gy
Remark at a stationary point (x∗ , y ∗ ), we have

d2 z = Lxx dx2 + Lxy dxdy + Lyx dydx + Lyy dy 2

where L is the corresponding Lagragian function:

L(x, y, λ) = f (x, y) + λ [c − g(x, y)]

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 104 / 138
 The Bordered Hessian

optimize z = f (x, y)
subject to
g(x, y) = c

The bordered Hessian:

0 gx gy
|H| = gx Lxx Lxy
gy Lyx Lyy
 
maximum point
(x∗ , y ∗ )
is a of z subject to g(x, y) = c
minimum point
0 gx gy  
>0
if and only if gx Lxx Lxy (ref. p358)
<0
gy Lyx Lyy

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 105 / 138
 n-Variable Case

0 g1 g2 ··· gn
g1 L11 L12 ··· L1n
optimize z = f (x1 , x2 , . . . , xn ) |H| = g2 L21 L22 ··· L2n
subject to .. .. .. .. ..
. . . . .
g(x1 , x2 , . . . , xn ) = c gn Ln1 Ln2 ··· Lnn

L = f (x1 , . . . , xn ) + λ(c − g(x1 , . . . , xn ))

Condition Maximum Minimum

First-order Lλ = L1 = · · · = Ln = 0 Lλ = L1 = · · · = Ln = 0
necessary
Second-order |H2 | > 0, |H3 | < 0,
sufficient |H4 | > 0, . . . , (−1)n |Hn | > 0 |H2 |, |H3 , . . . , |Hn | < 0
Remark. H is an (n + 1) × (n + 1)-matrix, |Hi | is the first (i + 1) × (i + 1) principal minor.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 106 / 138
 Multiconstraint Case

optimize z = f (x1 , x2 , . . . , xn )
subject to
g j (x1 , x2 , . . . , xn ) = cj for j = 1, .., m

0 g1 g2 ··· gn
g1 L11 L12 ··· L1n
|H| = g2 L21 L22 ··· L2n
.. .. .. .. ..
. . . . .
gn Ln1 Ln2 · · · Lnn

L = f (x1 , . . . , xn ) + λ1 [c1 − g 1 (x1 , . . . , xn )]+

+ · · · + λm [cj − g j (x1 , . . . , xn )]

The same second-order sufficient condition is applicable to this case as well.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 107 / 138
 The Effect of Nonnegativity Restrictions

Maximize z = f (x)
subject to
x≥0

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 108 / 138
 The Effect of Nonnegativity Restrictions - 2

Maximize z = f (x1 , . . . , xn )
subject to
xi ≥ 0 (i = 1, 2, . . . , n)

The first-order condition becomes:

fi ≤ 0, xi ≥ 0, xi fi = 0 for all i = 1, 2, . . . , n.

where
∂f
fi =
∂xi

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 109 / 138
 The Effect of Inequality Constraints

Maximize z = f (x1 , x2 , x3 )
subject to
g 1 (x1 , x2 , x3 ) ≤ c1
g 2 (x1 , x2 , x3 ) ≤ c2
and x1 , x2 , x3 ≥ 0
By adding two dummy variables y1 , y2 :
Maximize z = f (x1 , x2 , x3 )
subject to
g 1 (x1 , x2 , x3 ) + y1 = c1
g 2 (x1 , x2 , x3 ) + y2 = c2
and x1 , x2 , x3 , y1 , y2 ≥ 0
The first-order condition on the later problem becomes the first-order condition on the
corresponding Lagrangian function with non-negativity restrictions.
Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 110 / 138
 Kuhn-Tucker condition

Maximize z = f (x1 , x2 , x3 )
subject to
g 1 (x1 , x2 , x3 ) ≤ c1
g 2 (x1 , x2 , x3 ) ≤ c2
and x1 , x2 , x3 ≥ 0
The Lagrangian function
L = f (x1 , x2 , x3 ) + λ1 [c1 − g 1 (x1 , x2 , x3 )] + λ2 [c2 − g 2 (x1 , x2 , x3 )]

Kuhn-Tucker condition
∂L ∂L
= fj − (λ1 gj1 + λ2 gj2 ) ≤ 0 and xj ≥ 0 and xj =0
∂xj ∂xj
∂L ∂L
= c1 − g 1 (x1 , x2 , x3 ) ≥ 0 and λ1 ≥ 0 and λ1 =0
∂λ1 ∂λ1
∂L ∂L
= c2 − g 2 (x1 , x2 , x3 ) ≥ 0 and λ2 ≥ 0 and λ2 =0
∂λ2 ∂λ2
Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 111 / 138
 Kuhn-Tucker Condition for General Case

Maximize z = f (x1 , x2 , . . . , xn )
subject to
g i (x1 , x2 , . . . , xn ) ≤ ci (i = 1, 2, . . . , m)
and x1 , x2 , . . . , xn ≥ 0
The Lagrangian function
L = f (x1 , x2 , . . . , xn ) + λ1 [c1 − g 1 (x1 , x2 , . . . , xn )]
+ · · · + λm [cm − g 2 (x1 , x2 , . . . , xm )]

Kuhn-Tucker condition
∂L ∂L
= fj − (λ1 gj1 + · · · + λm gjm ) ≤ 0 and xj ≥ 0 and xj =0
∂xj ∂xj
∂L ∂L
= ci − g i (x1 , x2 , . . . , xn ) ≥ 0 and λi ≥ 0 and λi =0
∂λi ∂λi
Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 112 / 138
 Examples

Solve the following problem

Maximize f = 2x + 3y
subject to
x2 + y 2 ≤ 2
and x, y ≥ 0

More exercises: page 418.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 113 / 138
 Constrained optimisation on Production functions

Let us suppose that a firm wants to maximise output and that the production function is
of the form
Q = f (K, L)
Let the costs of each unit of capital and labour be PK and PL , respectively. So if the firm
has a fixed amount, M , to spend on these inputs, then

K · PK + L · PL = M

maximise Q = f (K, L)
subject to
K · PK + L · PL = M

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 114 / 138
 Constrained optimisation on Production functions

maximise Q = f (K, L)
subject to
PK K + P L L = M

At the point A, the maximum point,

PL At the maximum point
(based on the line) MRTS =
PK PL MPL MPL MPK
= or =
MPL PK MPK PL PK
(based on the curve) MRTS =
MPK
Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 115 / 138
 Constrained optimisation on Production functions

At the maximum point

PL MPL
=
PK MPK

MPL MPK
=
PL PK

So when output is maximised subject to a cost constant,

the ratio of the input prices is equal to the ratio of their marginal products.
the ratio of marginal product to price is the same for all inputs.
Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 116 / 138
 Constrained optimisation on Utility functions

At the maximum point

M aximise U = U (x1 , x2 ) P1 U′
= x′ 1
subject to P2 U x2
P1 x1 + P2 x2 = M
Ux′ 1 U′
= x2
P1 P2
So when utility is maximised subject to a budgetary constraint,
the ratio of the prices of the goods is equal to the ratio of their marginal utilities.
the ratio of marginal utility to price is the same for all goods consumed.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 117 / 138
 Practice Problem

Example
An individual’s utility function is given by U = x1 x2 where x1 and x2 denote the number of items
of two goods, G1 and G2. The prices of the goods are $2 and $10, respectively. Assuming that the
individual has $400 available to spend on these goods, find the utility-maximising values of x1 and
x2 . Verify that the ratio of marginal utility to price is the same for both goods at the optimum.

Example
A firm’s total cost function is given by

TC = 3x21 + 2x1 x2 + 7x22

where x1 and x2 denote the number of items of goods G1 and G2, respectively, that are produced.
Find the values of x1 and x2 which minimise costs if the firm is committed to providing 40 goods
of either type in total.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 118 / 138
 Practice Problems
Example
A consumer’s utility function is given by

U (x1 , x2 ) = 2x1 x2 + 3x1

where x1 and x2 denote the number of items of two goods G1 and G2 that are bought. Each item
costs $1 for G1 and $2 for G2. Use Lagrange multipliers to find the maximum value of U if the
consumer’s income is $83. Estimate the new optimal utility if the consumer’s income rises by $1.

Example
Use Lagrange multipliers to find expressions for x1 and x2 which maximise the utility function
1/2 1/2
U = x1 + x2

subject to the general budgetary constraint

P1 x 1 + P2 x 2 = M
Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 119 / 138
 Comparative statics analysis on Market model

∂P ∗ 1
=
Q = a − bP (demand) ∂a b+d
∂P ∗ a+c
Q = −c + dP (supply) =−
∂b (b + d)2
∂P ∗ 1
=
Equilibrium: ∂c b+d
∂P ∗ a+c
=−
∂d (b + d)2
a+c ad − bc
P∗ = , Q∗ =
b+d b+d

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 120 / 138
 Comparative statics analysis on Market model

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 121 / 138
 Comparative statics analysis on National-Income model

Y = C + Io + Go
C = α + β(Y − T ) (α < 0; 0 < β < 1)
T = γ + δY (γ > 0; 0 < δ < 1)
Hence,
α − βγ + Io + Go
Y∗ =
1 − β + βδ
It follows that
∂Y ∗ 1
= >0 (government expenditure multiplier)
∂Go 1 − β + βδ
∂Y ∗ β
=− <0 (non-income tax multiplier)
∂γ 1 − β + βδ
∂Y ∗ βY ∗
=− <0 (income tax rate multiplier)
∂δ 1 − β + βδ
Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 122 / 138
 Practices - Ref. p203

Let y1 , y2 be two implicit functions of two (independent) variables x1 , x2 defined by

the following system of equations

F (y1 , y2 , x1 , x2 ) = 0
G(y1 , y2 , x1 , x2 ) = 0

Find the total differentials (i.e. the partial derivatives) of y1 , y2 .

Applying the theory to give the comparative statics analysis on the following
national-income model and compare with the solution solved by Linear Algebra.

Y − C − Io − Go = 0
C − α − β(Y − T ) = 0
T − γ − δY = 0

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 123 / 138
 Integration
Integration: (in)definite integral, rules, integral by parts.

Applications: producer, customer’s surpluses; from a marginal function to a total

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 124 / 138
 Consumer’s surplus

Consumer’s surplus is the excess cost that a person would have been prepared to pay for
goods over and above what is actually paid.

ZQo
CS = −Qo Po + f (Q)dQ
0

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 125 / 138
 Consumer’s surplus - from Investopedia

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 126 / 138
 Producer’s surplus

Producer’s surplus is the excess revenue that a producer has actually received over and
above the lower revenue that it was prepared to accept for the supply of its goods.

ZQo
PS = Qo Po − g(Q)dQ
0

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 127 / 138
 Producer’s surplus - from Investopedia

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 128 / 138
 Investment and Capital formation

Net investment, I, is defined to be the rate of change of capital stock, K, so that

dK
I=
dt
Here, I(t) denotes the flow of money, measured in dollars per year, and K(t) is the
amount of capital accumulated at time t as a result of this investment flow and is
measured in dollars.
If the fund is to provide a continuous revenue stream for n years at an annual rate of
S dollars per year, then the present value can be found by evaluating the definite
integral
Zn
P = Se−rt/100 dt
0

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 129 / 138
 Marginal functions to Original functions

Since MC = (TC)′Q , MR = (TR)′Q , we have

Z Z
TC = MCdQ, TR = MRdQ

Example
a. A firm’s marginal cost function is M C = 2. Find an expression for the total cost
function if the fixed costs are 500. Hence find the total cost of producing 40 goods.
b. The marginal revenue function of a monopolistic producer is M R = 100 − 6Q. Find
the total revenue function and deduce the corresponding demand function.
c. Find an expression for the savings function if the marginal propensity to save is given
by MPS = 0.4 − 0.1Y −1/2 . and savings are zero when income is 100.

More exercises: p470

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 130 / 138
 Improper Integrals - examples

If the fund is to provide a continuous revenue stream for n years at an annual rate of
S dollars per year, then the present value can be found by evaluating the definite
integral
Zn
P = Se−rt/100 dt
0

If the fund is to provide a continuous revenue stream in perpetuity at an annual rate

of S dollars per year, then the present value can be found by evaluating the definite
integral
Z+∞

P = Se−rt/100 dt
0

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 131 / 138
 Improper Integrals
Z ∞ Z b
f (x)dx and f (x)dx
a −∞
Z b
f (x)dx where f is not well-defined on [a, b].
a

Example

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 132 / 138
 Improper Integrals
If f (x) is well-defined on [a, ∞) then
Z∞ Zb
f (x)dx := lim f (x)dx
b→∞
a a

If f (x) is well-defined on (−∞, b] then

Z b Zb
f (x)dx := lim f (x)dx
−∞ a→−∞
a

If f (x) is well-defined on R then

Z ∞ Zb
f (x)dx := lim f (x)dx
−∞ a→−∞,b→∞
a

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 133 / 138
 Improper Integrals - 2
If f (x) is well-defined on [a, b), but not at b then
Z b Zc
f (x)dx := lim f (x)dx
a c→b−
a

If f (x) is well-defined on (a, +∞) but not at a:

Z ∞ Zc
f (x)dx := lim f (x)dx
a b→a+ ,c→∞
b

If f (x) is well-defined on [a, c) ∪ (c, b] but not at c:

Z b Z c Z b
f (x)dx := f (x)dx + f (x)dx
a a c

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 134 / 138
 Practices

Evaluate all the following improper integrals:

a) c)

Z1 Z1
dx x−2/3 dt
√
3
x2 0
−1
d)
b)
Z0
Z∞
ert dt
e−rt dt
−∞
0

More exercises: p464

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 135 / 138
 Dynamic Analysis

Dynamic Analysis is a type of analysis in which the object is either to trace and study the
specific time paths of the variables or to determine whether, given sufficient time, these
variables will tend to converge to certain equilibrium values.
In a dynamic analysis, the question of “attainability" is to be squarely faced, rather
than assumed away.
On salient feature of dynamic analysis is the dating of the variables, which introduces
the explicit consideration of time into the picture. This can be done in two ways: time
can be considered either as a continuous variable or as a discrete variable.

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 136 / 138
 Projects

Team work
(Discrete) Markov chain in Economics and Finance.
Applications of Linear Algebras.

Projects
Leontieff I-O model and applying in Economics and/or Finance: an evidence in Vietnam
(Ref. (1) Using Solow and I-O models to determine the Factors impacting Economic Growth in Ho
Chi Minh City - Nguyen Thi Canh; (2) Vietnam Economic Structure Change Based on Vietnam
Input-Output Tables 2012 and 2016 - Bui Trinh).

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 137 / 138
 Any question?

Thank you!

Hà Văn Hiếu (UEL) Mathematical Economics 17th September 2023 138 / 138