Professional Documents
Culture Documents
Simone Moriconi∗
dy(x) 1
if a = 1/2 → y = x1/2 → dx
= 12 ∗ x 2 −1 = 21 ∗ x−1/2 = 12 x1/2
1
(2)
dy(x)
if a=1 → y=x → dx
= 1 ∗ x1−1 = 1 ∗ x0 = 1 ∗ 1 = 1 (3)
if a = 3 → y = x3 → dy(x) dx
= 3 ∗ x3−1 = 3 ∗ x2 (4)
∗
Associate Professor of Economics, Department of Economics and Quantitative Methods, Iéseg School of
Management E-mail: s.moriconi@ieseg.fr
1
It is can be easily seen when a = 1, the first derivative is a constant, i.e. does not vary
with x itself. In graphical terms this means that the slope is the same for any level of x. This
is indeed the case when the function y(x) is linear in x. (black line in figure 1). Conversely,
whenever a 6= 1 as in examples 3 and 4, the value of the first derivative depends with the value
of x itself. This means that the slope of the function is not constant, which is indeed the case
of the red and blue lines in figure 1. In particular, it can be easily verified that, when a > 1,
the slope of the function is increasing with x. For example, in the case of a = 3 from equation
4 above we obtain:
dy(x)
if x=1 → dx
= 3 ∗ 12 = 3 ∗ 1 = 3,
dy(x)
if x=2 → dx
= 3 ∗ 22 = 3 ∗ 4 = 12,
dy(x)
if x=3 → dx
= 3 ∗ 32 = 3 ∗ 9 = 27,
....
In fact 3 < 12 < 27 < .... This is the interpretation of the blue curve in figure 1 being convex:
whenever a > 1 the slope of the function increases with x. Simmetrically, consider now
the second example, where a = 1/2:
dy(x) 1 1
if x=4 → dx
= 2 41/2
= 12 ∗ 12 = 14 ,
dy(x) 1 1
if x = 16 → dx
= 2 161/2
= 12 ∗ 41 = 18 ,
dy(x) 1 1
if x = 25 → dx
= 2 251/2
= 12 ∗ 15 = 10 1
,
....
The slope of the function is now decreasing when x increases as 1/4 > 1/8 > 1/10 > .... This
is indeed the interpretation of the red curve in figure 1 being concave: whenever a < 1 the
slope of the function y(x) decreases when x increases. Simmetrically, when a < 1, the
slope of the function is decreasing with x.
2
Figure 1: concave and convex functions
All properties already stated hold the z(.) function. In particular the b > 0, will modify
the slope of the function, but will never change its concavity or convexity. The concavity or
convexity of a function will only depend on the exponent. Finally, it is important to recall the
following important properties of derivatives of composite functions:
dh(x) df (x) dg(x)
if h(x) = f (x) + g(x) → = +
dx dx dx
dh(x) df (x) dg(x)
if h(x) = f (x) − g(x) → = −
dx dx dx
These simple properties allow to compute all type of derivatives that combine different types
of exponential function. For example:
dh(x) df (x) dg(x)
if h(x) = f (x) + g(x) = 5x + 2x2 → = + = 5 + 2 ∗ 2x2−1 = 5 + 4x
dx dx dx
dh(x) df (x) dg(x)
if h(x) = f (x) − g(x) = 5x − 2x2 → = − = 5 − 2 ∗ 2x2−1 = 5 − 4x
dx dx dx