Scribd
Upload
3 views

Concave Convex Functions

Uploaded by

utkarsh.vadhoolan
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
3 views

Concave Convex Functions

Uploaded by

utkarsh.vadhoolan
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 5

Concave and Convex Functions

Sandip Sarkar
BITS Pilani, KK Birla Goa Campus

1 Idea
Concave and convex functions are frequently used in the economics literature.
In this note we shall discuss four definitions of these functions, and then we shall
move on explaining the economic applications. We begin our discussion with
the general definition of concave and convex functions. Here we do not assume
anything related to differentiable/ continuity of the functions.

Definition 1 For any n ∈ N and Dn ⊆ Rn , let f : Dn 7→ R and for any


distinct u, v ∈ Dn , and θ ∈ (0, 1):
1. f is concave ⇐⇒ θf (u) + (1 − θ)f (v) ≤ f (θu + (1 − θ)v)
2. f is strictly concave ⇐⇒ θf (u) + (1 − θ)f (v) < f (θu + (1 − θ)v)

3. f is convex ⇐⇒ θf (u) + (1 − θ)f (v) ≥ f (θu + (1 − θ)v)


4. f is strictly convex ⇐⇒ θf (u) + (1 − θ)f (v) > f (θu + (1 − θ)v)

We now define concave/ convex functions for functions that are differentiable
at least once.

Definition 2 For any n ∈ N and Dn ⊆ Rn , let f : Dn 7→ R, where f (·) is


differentiable, then for any distinct u, v ∈ Dn :
′
1. f is concave ⇐⇒ f (v) ≤ f (u) + f (u)(v − u).
′
2. f is strictly concave ⇐⇒ f (v)) < f (u) + f (u)(v − u).
′
3. f is convex ⇐⇒ f (v) ≥ f (u) + f (u)(v − u).
′
4. f is strictly convex ⇐⇒ f (v) ≤ f (u) + f (u)(v − u).

Before we move to the next definition, we need to understand the concept of


negative (positive) definite and negative (positive) semidefinite. Consider the
function z = f (x1 , x2 , x3 ) as twice differentiable we can write

dz = f1 dx1 + f2 dx2 + f3 dx3 (1)

1
Differentiating 1 once again we get
∂dz ∂dz ∂dz
d2 z = dx1 + dx2 + dx3 (2)
∂x1 ∂x2 ∂x3
One can also write the above equation as follows d2 z = f11 dx21 + f22 dx22 +
f33 dx23 +f21 dx2 dx1 +f12 dx1 dx2 +f31 dx3 dx1 +f13 dx1 dx3 +f23 dx2 dx3 +f32 dx3 dx2 .
Let  
f11 f12 f13
H = f21 f22 f23  ,
f31 f32 f33
and  
dx = dx1 dx1 dx3
One may express 1 in the following quadratic form:

d2 z = dxHdx′ (3)

Notice that d2 z is a scalar. For z = f (x1 , x2 , .., xn ) we can also write d2 z =


dxHdx′ , where
 
f11 f12 · · · f1n
 f21 f22 · · · f2n 
H= , (4)
 
..
 . 
fn1 fn2 · · · fnn
and  
dx = dx1 dx2 ··· dxn ,

Definition 3 Let z = f (x1 , x2 , .., xn ) be a twice differentiable functions, then


1. d2 z is positive definite if for all i ∈ {1, 2, .., n}: we have |Hi | > 0.
2. d2 z is negative definite if ∀i ∈ {1, 2, .., n} such that i is odd: we have
|Hi | < 0, and for all even i we have |Hi | > 0.
3. d2 z is positive semidefinite if for all i ∈ {1, 2, .., n}: we have |Hi | ≥ 0.
4. d2 z is negative semidefinite if ∀i ∈ {1, 2, .., n} such that i is odd: we
have |Hi | ≤ 0, and for all even i we have |Hi | ≥ 0.
where Hi is the ith leading principal minor of H.

Assuming the functions are twice differentiable we now define concavity and
convexity of functions.

Definition 4 Let z = f (x1 , x2 , .., xn ) be a twice differentiable functions, then


1. f is concave ⇐⇒ d2 z is negative semidefinite.
2. f is strictly concave if (but not only if) d2 z is negative definite.

2
3. f is convex ⇐⇒ d2 z is positive semidefinite.
4. f is strictly convex if (but not only if) d2 z is positive definite.

Before we proceed to the fourth definition of convexity and concavity, we


define the notion of convex sets. The basic idea is that a set is said to be
convex, if the convex combination of any two elements within the set is also an
element of the set.

Definition 5 Convex Set: A set S ⊆ Rn is set to be a convex set if for any


u, v ∈ S and for any δ ∈ (0, 1) we have δu + (1 − δ)v ∈ S.

Here are some graphical illustrations of convex sets.


Here are some graphical illustrations of non-convex sets.
We now relate the notion of convex set

Definition 6 For any n ∈ N and Dn ⊆ Rn , let f : Dn 7→ R and for any k ∈ R,


the function f (·) is said to be

1. Concave if S ≥ = {x : f (x) ≥ k} is a convex set.


2. Strictly Concave if S > = {x : f (x) > k} is a convex set.
3. Convex if S ≤ = {x : f (x) ≤ k} is a convex set.

4. Strictly Convex if S < = {x : f (x) < k} is a convex set.

Problem 1 Let f (x) is a convex function, show that −f (x) is a concave func-
tion.

Problem 2 Show that f (x) + g(x) + h(x) is a convex function, if f (), g() and
h() are convex functions.
√ √
Problem 3 Show that f (x) = x1 + x2 is a concave function.

Problem 4 Show that f (x) = x21 + x22 is a convex function.

You might also like