Session 2

• Functions of several variables

• Directional derivatives
• 2-variables: 1st order approximation
• 2-variables: Applications in Economy
• 2-variables: Convexity
1
Functions of several variables
In Math1 we considered functions with one variable.
However, many processes in economics and finance are dependent on
several variables, e.g. production functions.
Production functions illustrate the relationship between inputs:
labor (), physical capital (), raw materials, technology, etc.) and output:
goods or services, ….
The most popular one is Cobb-Douglas:
,
where are constants.

2
Graph of functions of several
variables
A function is graphed in as
.

Same for functions of several

variables:
is graphed in as
.

3
Functions of 2 variables
A function of 2 variables can be graphically depicted as a surface in 3
dimensions, with being in a horizontal plane.

Example1:

Example2:
Functions of 2 variables
Example3:

Plane cuts the surface

along a curve called
contour(or level) line
satisying

E.g.

5
Functions of 2 variables
𝑍 fixed, defines a contour(or level) line on the
𝑧 =𝑐
surface which projects onto in plane.
𝑧 = 𝑓 ( 𝑥 , 𝑦 ) is an implicit funtion defining relation
between variables and .
𝑌
𝑐= 𝑓 ( 𝑥 , 𝑦 )
𝑋

fixed, defines as a function of a single variable .

𝑧 = 𝑓 ( 𝑥 , 𝑦 0)
If is differentiable at , its derivative is called
directional (partial) derivative of wrt at .

6
Partial derivatives - 2 variable
examples
Paraboloid:
- constant
𝑦0
𝑦 = 𝑦0
𝜕 𝜕𝑧 𝜕
𝑧= = 𝑓 ( 𝑥 , 𝑦 )=𝑧 𝑥 = 𝑓 𝑥 (𝑥 , 𝑦 )
𝜕𝑥 𝜕𝑥 𝜕 𝑥
Hyperbolic Paraboloid:
- constant

7
Partial differentiation
Directional (partial) derivative of wrt is the derivative of a single variable function
of :
,
where all other variable are considered as constants.

Interpretation through the rate of change of the Cobb-Douglas:

is the rate of change of Production wrt Labor (keeping Capital unchanged)

8
Exercises

1.
Find and . Find and .

2.. Find , , and .

3. . Find , and .

9
Applications in Economy
1. A company manufactures two types of skis, the Lightning and the Alpine
models. Suppose the joint-cost function for producing x pairs of the Lightning
model and y pairs of the Alpine model per week is

where c is expressed in dollars. Determine the marginal costs ∂c/∂x and ∂c/∂y
when and , and interpret the results.

2. Cobb-Douglas function offers one way to express relationship between Labor

and Capital , as inputs, and Production , as output:
,
where are constants.
Find marginal production functions and .

10
Applications in Economy
3. A manufacturer of a popular toy has determined that the production function is ,
where is the number of labor-hours per week and is the capital (expressed in
hundreds of dollars per week) required for a weekly production of units of the toy.
Determine the marginal productivity functions, and evaluate them when and .

11
1 -Order Approximation of :
st

For a single variable function at the point :

1st order approx.
tangent line the at

For a function of 2 variables at the point :

1st order approx.
tangent plane at

𝑀=(𝑎 ,𝑏 , 𝑓 (𝑎,𝑏))
𝑓 ( 𝑎) 𝑀 𝑏

𝑥 𝑎
𝑎 (𝑎 , 𝑏)
12
1st order Approx. of & Gradient
“nabla“
Gradient of

tangent plane at
=

row x column (product of matrices)

as vectors (column matrices)

tangent plane equation

1st order Approximation of

13
Applications in Economy
4. A manufacturer of a popular toy has determined that the production function is ,
where is the number of labor-hours per week and is the capital, expressed in
hundreds of dollars per week, required for a weekly production of units of the toy.
a) Determine the marginal productivity functions, and evaluate them when 0 and .
b) Find the 1st order approximation of at .
c) Using the above approximation, estimate the change of the production if labor and
capital change respectively by and .

14
Higher-Order Partial
Derivatives
Second-order partial derivatives of f(x,y)
fxxmeans (fx ) x and fxy means (fx )y equal if continuous
fyxmeans (fy ) x and fyy means (fy )y (Schwarz‘s theorem)

Example:
Find second-order partial derivatives of f x, y   x 2
y  x 2 2
y .

Solution:

15
Cobb-Douglas Production function
, are constants.

• Marginal product wrt labor

• Marginal product of (wrt) capital

16
Cobb-Douglas Production function
• Law of diminishing returns to capital and labor
𝛼> 𝑜 𝛼< 1
MPL  0 - Q increases when L increases, with capital held constant
QLL  0 - Q increases, but at ever decreasing rate
𝛽 >𝑜 𝛽 <1
MPK  0 - Q increases when K increases, with labor held constant
QKK  0 - Q increases, but at ever decreasing rate

17
Cobb-Douglas Production function
One way to represent production functions in two-dimensional plane is to
keep Q constant (level line) and graph the relationship between K and L
1/ 2 1/ 2
𝑄=5 𝐿 𝐾
In economics this curve is called
an isoquant (meaning equal
quantity).
It shows the various combinations
of L and K for which quantity
production is constant
18
Utility function
Utility function describes the utility (well-being) derived from the
consumption of goods and services
, are constants
U
MU x 
Marginal Utility x
U
MU y 
y
Indifference curve (=isoquant Cobb-Douglas)
combination of goods x and services y
which give the consumer the same utility 19
Graph a function of 2 variables

20
 are projections
Convexity : of onto plane

is (strictly)Convex/Concave at if open segment lies

above/below for any close to M :

𝑧 𝑅
𝑈 =(𝑢 , 𝑓 (𝑢 ))
𝑉 𝑉 = (𝑣 , 𝑓 (𝑣 )) 𝑈 ∙
∙ 𝑈
𝑊
∙
∙𝑀
∙
𝑈
𝑦
∙𝑤
𝑢𝜆1𝑚− 𝜆
∙
∙ 𝑢
∙ 𝑣 𝑅2
∙ 𝑣
𝑥
21
Convexity :
Convex/Concave at may be checked using Hesse-Matrix :

Hesse-Matrix of at
∆1 ∆2

is Convex/Concave at if for of holds

Convex :

Concave :

22
Convexity of :
Exercise 1. Check convexity of the function

Exercise 2. Check convexity of the function

at the point .

Exercise 3. Check convexity of the function

at the point M=.

23