Optimization (Maximum and Minimum)

The equilibrium state is defined as the optimum position for a given economic unit (a
household, a business firm, etc.) in which the economic unit strive to deliberate for
attaining the equilibrium. The attainment of equilibrium is the outcome of the impersonal
balancing of the forces and does not require the conscious effort on the part of anyone to
accomplish a specified goal. That is, the consuming households behind the forces of
demand and the firms behind the forces of supply are each striving for an optimal
position under the given circumstances, but as far as the market itself is concerned, no
one is aiming at any particular equilibrium price or equilibrium quantity. Similarly, in
national-income determination, the impersonal balancing of leakages and injections is
what brings about an equilibrium state, and no conscious effort at reaching any particular
goal (such as an attempt to alter an undesirable income level by means of monetary or
fiscal policies) needs to be involved at all. Therefore, our primary focus is on the classical
techniques for locating optimum positions by using differential calculus.

The most common criterion of choice among alternatives in economics is the goal of
maximizing something (such as maximizing a firm’s profit, a consumer’s utility, or the
rate of growth of a firm or of a country’s economy) or of minimizing something (such as
minimizing the cost of producing a given output). Economically, we may categorize such
maximization and minimization problems under the general heading of optimization,
meaning “the quest for the best”. From a purely mathematical point of view, however, the
terms maximum and minimum do not carry with them any connotation of optimality.
Therefore, the collective term for maximum and minimum, as mathematical concepts, is
the more matter-of-fact designation extreme, meaning an extreme value.

In formulating an optimization problem, the first order of business is to delineate an

objective function in which the dependent variable represents the object of maximization
or minimization and in which the set of independent variables indicates the objects whose
magnitudes the economic unit in question can pick and choose, with a view to
optimizing. We shall therefore refer to the independent variables as choice variables. The
essence of the optimization process is simply to find the set of values of the choice
variables that will lead us to the desired extreme of the objective function.

1
For example, a business firm may seek to maximize profit  , that is, to maximize the
difference between total revenue R and total cost C . Since within the framework of a
given state of technology and a given market demands for the firm’s product, R and C
are both functions of the output level Q , it follows that  is also expressible as a
function of Q :
 (Q)  R(Q)  C (Q) .
This equation constitutes the relevant objective function, with  as the object of
maximization and Q as the (only) choice variable. The optimization problem is then that
of choosing the level of Q that maximizes .

To cast the problem into a more general mold for further discussion, let us consider the
general function
y  f (x)
and attempt to develop a procedure for finding the level of x that will maximize or
minimize the value of y .

 If the object function is a constant function then there exist all values of the choice
variable x will result in the same value of y , and the height of each point on the
graph of the function may be considered a maximum or, for that matter, a minimum
or, indeed neither. In this case, there is in effect no significance choice to be made
regard the value of x for the maximization or minimization of y .
 The function is strictly increasing, and there is no finite maximum if the set of
nonnegative real numbers is taken to be its domain. However, we may consider the
end point on the left (the y intercept) as representing a minimum.
 The points E and F in the following figure are the relative (or local) extreme in the
sense that each of these points represents an extreme in the immediate neighborhood
of the point only. The fact that point F is a relative minimum is, of course, no
guarantee that it is also the global minimum of the function, although this may
happen to be the case. Similarly, a relative maximum point such as E may or may
not be a global maximum. Here it can be noted that a function can very well have
several relative extreme, some of which may be maxima while others are minima.

2
O
Y

y  f (x)

(-)
)_) (+)
(+)

O X

First-Derivative Test
If the first derivative of a function f (x) at x  x0 is f ( x0 )  0 , then the value of the
function at x0 , f ( x0 ) , will be
 A relative maximum if the derivative f (x) changes its sign from positive to
negative from the immediate left of the point x0 to its immediate right.
 A relative minimum if f (x) changes its sign from negative to positive from the
immediate left of x0 to its immediate right.
 Neither a relative maximum nor a relative minimum if f (x) has the same sign
on both the immediate left and the immediate right of point x0 .
Example
Find the relative extreme of the function
y  f ( x)  x 3  12 x 2  36 x  8
First, we find the derivative function to be
f ( x)  3x 2  24 x  36
To get the values of x under condition f ( x0 )  0 then we have following quadratic
equation

3
 3x 2  24 x  36  0

By applying the quadratic formula, we then obtain the following pair of roots (solutions) :
x1  6 [at which we have f (6)  0 and f (6)  8]
x2  2 [at which we have f (2)  0 and f (2)  40]
Therefore, the immediate neighborhood of x  6 , we have f ( x)  0 for x  6 , and
f ( x)  0 for x  6 ; thus the value of the function f (6)  8 is a relative minimum.
Similarly, the immediate neighborhood of x  2 , we find f ( x)  0 for x  2 , and
f ( x)  0 for x  2 , the value of the function f (2)  40 is a relative maximum.

Example
Find the relative extreme of the average-cost function
AC  f (Q)  Q 2  5Q  8
The derivative here is f (Q)  2Q  5 , a linear function. Setting f (Q) equal to zero,
we get the linear equation 2Q  5  0 , which has the single root Q  2.5 . To apply the
first-derivative test, we obtain Q  2.4 and Q  2.6 . Since f (2.4)   0.2  0
whereas f (2.6)  0.2  0 , we can conclude that the stationary value
AC  f (2.5)  1.75 represents a relative minimum.

Second-Derivative Test

If the value of the first derivative of a function f at x  x0 is f ( x0 )  0 , then the

value of the function at x0 , f ( x0 ) , will be
 A relative maximum if the second-derivative value at x0 is f ( x0 )  0 ;
 A relative minimum if the second-derivative value at x0 is f ( x0 )  0 .

4
Example
Find the relative extreme of the function
y  g ( x)  x 3  3 x 2  2
The first two derivatives of this function are
g ( x)  3x 2  6 x and g ( x)  6 x  6
Setting g (x) equal to zero and solving the resulting quadratic equation, 3x  6 x  0
2

, we obtain x1  2 and x2  0 , which in turn yield the two stationary values :

g (2)   2 (a minimum because g (2)  6  0 )
g (0)  2 (a maximum because g (0)   6  0 )

Interpretation of the Second Derivative

 The derivative function f (x) measures the rate of change of the function.
 The second derivative function f (x) is the measure of the rate of change of the
first derivative.
 In other words, the second derivative measures the rate of change of the rate of
change of the original function f (x) .
 If we put x  x0 in f (x) then following explanation may be made:

f ( x0 )  0 increase
 means that the value of the function tends to 
f ( x0 )  0 decrease
 For second derivative we may be explained that
f ( x0 )  0 increase
 means that the slope of the curve tends to 
f ( x0 )  0 decrease

 A positive first derivative coupled with a positive second derivative at x  x0 implies

that the slope of the curve at that point is positive and increasing.
 That is, the value of the function is increasing at an increasing rate.
 A positive first derivative with a negative second derivative indicates that the slope of
the curve is positive but decreasing.
 That is, the value of the function is increasing at a decreasing rate.

5
 Further, when f ( x0 )  0 and f ( x0 )  0 , the slope of the curve is negative and
increasing.
 Lastly, when f ( x0 )  0 and f ( x0 )  0 , the slope of the curve must be negative
and decreasing.
 Therefore, according to first-order and second-order condition, a function must be
satisfied the following conditions for identifying the maximum or minimum value of
the function :

Condition Maximum Minimum

First-order necessary f ( x)  0 f ( x)  0
Second-order sufficient f ( x)  0 f ( x)  0
Second-order necessary f ( x)  0 f ( x)  0
Note: when marginal revenue and marginal cost are equal then f ( x)  0 is applicable.

Conditions for Profit Maximization

In order to maximize profit, a firm must equate marginal cost and marginal revenue. For
this purpose, let us consider the total-revenue function R  R(Q) and total-cost function
C  C (Q) , both of which are functions of a single variable Q . Then the profit function
is defined as
   (Q)  R(Q)  C (Q) (1)
For find the profit-maximizing output level, the first-order necessary condition for
maximization is d dQ  0 . That is
d
  (Q)  R(Q)  C (Q)  0 iff R(Q)  C (Q) (2)
dQ
Thus the optimum output (equilibrium output) Q * must satisfy the equation
R(Q*)  C (Q*), or MR  MC . This condition constitutes the first-order
condition for profit maximization.
To check the second-order condition, we obtain the second derivative by differentiating
the first derivative in (2) with respect to Q :

d 2
  (Q)  R(Q)  C (Q)  0 iff R(Q)  C (Q)
dQ 2

6
This last inequality is the second-order necessary condition for maximization. If it is not
satisfy, then Q * there is not possible to maximize profit. If R(Q*)  C (Q*) , then we
are unable to reach a definite conclusion. If it is observed R(Q*)  C (Q*) , this
satisfies the second-order sufficient condition for a maximum. In that case, we can
*
conclusively take Q to be a profit-maximization output. Therefore, we may conclude
that if the rate of change of MR is less than the rate of change of MC at the output
where MC  MR , then that output will maximize profit.

Example
Let the R(Q) and C (Q) functions be

R(Q)  1200 Q  2Q 2
C (Q)  Q 3  61.25Q 2  1528.5 Q  2000
Then the profit function is
 (Q)   Q3  59.25 Q 2  328.5 Q  2000
where R, C, and  are all in dollar units and Q is in units of tons per week. This
profit function has two critical values, Q  3 and Q  36.5 , because
d 3
  3Q 2  118.5 Q  328.5  0 , when Q
dQ 36.5
But since the second derivative is
d 2 0 when Q  3
  6 Q  118.5 
dQ 2  0 when Q  36.5
the profit-maximizing output is Q *  36.5 (tons per week). By substituting Q * into
the profit function, we can find the maximized profit to be  *   (36.5)  16318.44
(dollars per week).

7
Example 1:
The farmer’s available amount of land is 1000 acres, the unit net profit of crops 1 and 2
are Tk.10 and Tk.8 respectively, and the production functions are
y1  l10.6 and y 2  l 20.8
What is the optimal land allocation?
Solution
The Lagrange function is
L  10l10.6  8l 20.8   1000  l1  l 2 
where  indicates the Lagrangian multiplier. And the first order conditions are
6l10.4    0
6.4l 20.2    0
1000  l1  l 2  0
The first two conditions give
6l10.4  6.4l 20.2
implying that
6l 20.2  6.4l10.4
and so
5
 6.4  0.4 
l 2    l1   1.38 l1
2

 6  
Substituting into the constraint gives the quadratic equation
1.38l12  l1  1000  0
This solves to give
l1*  26.6 and l 2*  973.4
This is the optimum land allocation for y1 and y2 .

8
Example 2:
A publisher pays the author of a book a royalty of 15%. Demand for the book is
x  200  5 p and the production cost is C  10  2 x  x 2 . Find the optimal sales from
both the author’s and the publisher’s perspective.
Solution
The inverse demand function is
p  40  0.2 x
The author’s income is
Y ( x)  0.15 px  0.1540  0.2 x x  6 x  0.03x 2
The optimal level of sales from the author’s viewpoint, x A , leads to a maximum values
for Y x  , and so
Y x A   0  6  0.06 x A  0  x A  100
The profit for the publisher, on the other hand, is
 ( x)  R( x)  C ( x)  Y ( x)
 px  (10  2 x  x 2 )  (6 x  0.03x 2 )
 (40  0.2 x) x  (10  2 x  x 2 )  (6 x  0.03x 2 )
 32 x  1.17 x 2  10
The optimum level of sales from the publisher’s viewpoint, x p , leads to a maximum
value for  (x), and so
32
 ( x p )  32  2.34 x  0  x p 
 13.68  14
2.34
Clearly, the author seeks a level of sales which maximizes sales revenue, while the
publisher takes costs, including royalty costs, into account and so prefers a lower level of
sales.

Example 3:
Find the price and quantity that will maximize total revenue, given the demand function
𝑃 = 12.50𝑒 −0.005 𝑄 . Check the second-order condition.
Solution:
Total revenue function is defined as
𝑇𝑅 = 𝑃𝑄,
where 𝑃 indicates unit price and 𝑄 indicates number of units of the quantity.
⟹ 𝑇𝑅 = 12.50𝑒 −0.005 𝑄 𝑄

9
 𝑑𝑇𝑅
Then = 12.50𝑒 −0.005𝑄 1 + 𝑄 −0.005 12.50𝑒 −0.005 𝑄
𝑑𝑄

= 12.50𝑒 −0.005 𝑄 1 − 0.005𝑄 = 0

⟹ 1 − 0.005𝑄 = 0
∴ 𝑄 = 200
Thus, 𝑃 = 12.50𝑒 −0.005 200
= 12.50𝑒 −1 = 12.50 0.36788 = 4.60

𝑑 2 𝑇𝑅
Again, = 12.50𝑒 −0.005 𝑄 −0.005 + 1 − 0.005𝑄 −0.005 12.50𝑒 −0.005 𝑄
𝑑 𝑄2

= −0.005 12.50𝑒 −0.005 𝑄 2 − 0.005𝑄

= −0.005 12.50𝑒 −1 1 = −0.0625 0.36788 < 0.
That is, the function is satisfied the maximization condition.

Example 4:
Land bought for speculation is increasing in value according to the formula
3
𝑉 = 1000𝑒 𝑡
The discount rate under continuous compounding is 0.09. How long the land should be
held to maximize the present value.
Solution:
The present value 𝑃 is defined as
𝑃 = 𝑉𝑒 −𝑟𝑡
Substituting the value of 𝑉 𝑎𝑛𝑑 𝑟, we have
3 3
𝑡 −0.09𝑡 𝑡 −0.09𝑡
𝑃 = 1000𝑒 𝑒 = 1000𝑒
⟹ ln 𝑃 = ln 1000 + 𝑡 1 3
− 0.09𝑡
𝑑 1 𝑑𝑃 1 −2 3
Then ln 𝑃 = = 𝑡 − 0.09 = 0
𝑑𝑡 𝑃 𝑑𝑡 3
𝑑𝑃 1 −2 3
⟹ =𝑃 𝑡 − 0.09 = 0
𝑑𝑡 3
1 −2 3
⟹ 𝑡 = 0.09
3

∴ 𝑡 = 0.27−3 2
≈ 7.13 years
𝑑2𝑃 2 1 −2 3 𝑑𝑃 2𝑃 𝑑𝑃
Again, = 𝑃 − 9 𝑡 −5 3
+ 𝑡 − 0.09 =− 3 < 0, since = 0.
𝑑𝑡 2 3 𝑑𝑡 9 𝑡5 𝑑𝑡

10
 Exponential Functions

The term exponent means an indicator of the power to which a variable is to be raised. In
power expressions such as x 3 or x 5 , the exponents are constants; but there is no reason
x r
why we cannot also have a variable exponent, such as in 3 or 3 , where the number 3
is to be raised to varying powers (various values of x or t ). Therefore, a function
whose independent variable appears in the role of an exponent is called an exponential
function. For example, the exponential function may be represented in the form
y  f (t )  bt (b  1)
where y and t are the dependent and independent variables, respectively, and b
denotes a fixed base of the exponent. Further, the function in which the variable 𝑥
appears as an exponent then this function is called exponential functions. A simple
example, 𝑓 𝑥 = 2𝑥 , a function whose domain is all the real numbers. That is,
(i) if 𝑥 is a positive integer, 2𝑥 means “multiply 2 by itself 𝑥 times”;
(ii) if 𝑥 = 0, 20 = 1, by definition;
𝑛
(iii) if 𝑥 = 1 𝑛, 21 𝑛
= 2 , the nth root of 2;
𝑛 𝑚
(iv) if 𝑥 = 𝑚 𝑛, 2𝑚 𝑛
= 2 , the mth power of the nth root of 2; and
(v) if 𝑥 is a negative number, 2𝑥 means 1 2 𝑥 , the reciprocal of 2 𝑥

To explain the exponent form promptly, an irrational number may be considered as the
preferred base which is denoted by e , where e  2.71828... . When this base e is used
in an exponential function, it is referred to as a natural exponential function, examples of
which are
y  et  exp(t ), y  e3t  exp(3t ), y  Aer t  A exp(r t ) etc.
t
Then the derivative of e is defined as
d t
e  et .
dt
of y  Ae  Aew ,
rt
Further, the derivative where
w  r t , and A, r are cons tan ts . Then, by the chain rule, we can write
dy dy dw
  A ew. r  r A er t .
dt dw dt
d
That is Aer t  r A e r t .
dt

11
Natural Exponential Functions and the Problem of Growth

 The Number e
Let us consider the following function:
m
 1
f ( m)  1   . (1)
 m
If larger and larger values are assigned to m , then f (m) will also assume larger values;
specifically, we find that
1
 1
f (1)  1    2
 1
2
 1
f (2)  1    2.25
 2
3
 1
f (3)  1    2.37037...
 3
4
 1
f (4)  1    2.44141....
 4
Moreover, if m is increased indefinitely, then f (m) will converge to the number
2.71828.....  e ; thus e may be defined as the limit of (1) as m   :
m
 1
e  lim f (m)  lim 1   . (2)
m m  m
That the approximate value of e is 2.71828 can be verified by finding the Maclaurin
series of the function  ( x)  e x which is expressed as
1 2 1 3 1 4
ex 1  x  x  x  x  ......
2! 3! 4!
As a special case, for x  1 , we find that
1 1 1
e 11    .......
2! 3! 4!
 2.7182819

12
An Economic Interpretation of e
The number e can be interpreted as the result of a special mode of interest compounding.
Suppose that, starting out with a principal (or capital) of $1, we find a hypothetical
banker to offer us the unusual interest rate of 100 percent per annum ($1 interest per
year). If interest is to be compounded once a year, the value of our asset at the end of the
year will be $2; we shall denote this value by V(1), where the number in parentheses
indicates the frequency of compounding within1 year:
V (1)  initial principal (1  int erest rate )
1
 1
 1(1  100%)  1    2
 1
If interest is compounded semiannually, however, an interest amounting to 50 percent
(half of 100 percent) of principal will accrue at the end of 6 months. We shall therefore
have $1.50 as the new principal during the second 6-month period, in which interest will
be calculated at 50 percent of $1.50. Thus our year-end asset value will be 1.50(1+50%);
that is,
2
 1
V (2)  (1  50%)(1  50%)  1  
 2
3 4
 1  1
By analogous reasoning, we can write V (3)  1   , V (4)  1   , etc. ; or,
 3  4
m
 1
in general, V (m)  1   , where m represents the frequency of compounding in 1
 m
year.

In the limiting case, when interest is compounded continuously during the year, i.e., when
m becomes infinite, the value of the asset will grow in a “snowballing” fashion,
becoming at the end of 1 year
m
 1
lim V (m)  lim 1    e (dollars)
m m  m
Thus, the number e  2.71828 can be interpreted as the year-end value to which a
principal of $1 will grow if interest at the rate of 100 percent per annum is compounded
continuously.

13
Here it is noted that the interest rate of 100 percent is only a nominal interest rate, for if
$1 becomes $e = $2.718 after 1 year, the effective interest rate is in this case
approximately 172 percent per annum.

Interest Compounding and the Function Aer t

The continuous interest-compounding process just discussed can be generalized in three

directions, to allow for : (1) more years of compounding, (2) a principal other than $1,
and (3) a nominal interest rate other than 100 percent.
If a principal of $1 becomes $e after 1 year of continuous compounding and if we let $e
be the new principal in the second year (during which every dollar will again grow into
$e), our asset value at the end of 2 years will obviously become $e(e) = $e 2. By the same
token, it will become $e3 at the end of 3 years or, more generally, will become $et after t
years.
Next, let us change the principal from $1 to an unspecified amount, $A. This change is
easily taken care of: if $1 will grow into $et after t years of continuous compounding at
the nominal rate of 100 percent per annum, it stands to reason that $A will grow into
$Aet.
Further, the nominal interest rate of other than 100 percent, i.e., r  0.05 may be
expressed that an initial principal of $A to be invested for t years at a nominal interest
rate r then the compound-interest formula is defined as
mt
 r
V ( m)  A 1   ,
 m
where the coefficient A reflects the change of principal from the previous level of $1.
 The expression r m indicates that, in each of the m compounding periods in a year,
only 1 m of the nominal rate r will actually be applicable.
 The exponent m t indicates that, since interest is to be compounded m times a year,
then there total of m t compounding in t years.

14
The above formula can be transformed into an alternative form
rt
 m

 r r 
V ( m )  A 1  
 m  
  ,
rt
 1 
w
 A 1   
 w  
m
where w . Then the asset value in the generalized continuous compounding
r
process is defined as

V  lim v(m) A e r t .
m

Instantaneous Rate of Growth

rt
For interpretation of natural exponential function Ae , interest compounding merely
exemplifies the general process of exponential growth (here, the growth of a sum of
money capital over time) and apply the function equally well to the growth of population,
rt
wealth, or real capital. Therefore, the coefficient r in Ae no longer denotes the
nominal interest rate. That is, r can be reinterpreted as the instantaneous rate of growth
of the function Ae . Given the function V  Ae , which gives the value of V at
rt rt

each point of time t , the rate of change of V is to be found in the derivative

dV
 r Ae r t  r V .
dt
But the rate of growth of V is simply the rate of change in V expressed in relative
(percentage) terms, i.e., expressed as a ratio to the value of V itself. Thus, for any given
point of time, we have
dV dt r V
Rate of growth of V   r.
V V
Several observations should be made about this rate of growth. With this understanding,
let us make following comments :
 The rate of growth of V is an instantaneous rate of growth. Since the derivative
dV dt  r A e r t takes a different value at a different point of t , as will V  Aer t ,

15
 their ratio must also have reference to a specific point (or instant) of t , in this sense,
the rate of growth is instantaneous.
 In the present case, however, the instantaneous rate of growth happens to be a
constant r , with the rate of growth thus remaining uniform at all points of time. This
may not. of course, be true of all growth situations actually encountered.
 Even through the rate of growth r is measured at a particular point of time, its
magnitude nevertheless has the connotation of so many percent per unit of time, say,
per year (if t is measured in year units). Growths, by its very nature, can occur only
r at the instant
over a time interval. Therefore, V indicates the rate of growth of
t  t0 , by which we may understand that if the rate of change dV dt ( r V )
prevailing at t  t 0 is allowed to continue undisturbed for one whole unit of time
(1 year), then V will have grown by the amount r V at the end of the year.

For the exponential function V  Ae , the percent rate of growth is constant at all
rt

points of t , but the absolute amount of increment of V increases as time goes on,
because the percentage rate will be calculated on larger and larger bases.

Upon interpreting r as the instantaneous rate of growth, it is clear that little effect will
henceforth be required to find the rate of growth of a natural exponential function of the
form y  Ae , provided y  75 e o.o 2 t , for instance, we
rt
r is a constant. Given a function

can immediately read off the rate of growth of y as 0.02 or 2 percent per period.

Discounting and Negative Growth

In a compound-interest problem, we seek to compute the future value V (principal plus
interest) from a given present value A (initial principal). The problem of discounting is
the opposite one of finding the present value A of a given sum V which is to be
available t years from now.

Let us take the discrete case first. If the amount of principal A will grow into the future
value of A (1  i ) after t years of annual compounding at the interest rate i per annum,
t

i.e., if V  A (1  i) t then by dividing both sides of the equation by the nonzero

expression (1  i) , we can get the discounting formula :
t

16
 V
A  V (1  i)  t
(1  i) t

which involves a negative exponent. It should be realized that in this formula the roles of
V and A have been reversed : V is now a given, whereas A is the unknown , to be
computed from i (the rate of discount) and t (the number of years), as well as V .

Similarly, for the continuous case, if the principal A will grow into Aer t after t years
of continuous compounding at the rate r in accordance with the formula
V  Aer t
then we can derive the corresponding continuous-discounting formula simply by dividing
rt
both sides of the above equation by e :
V
A
rt
 Ver t .
e
Here again, we have A (rather than V ) as the unknown, to be computed from the given
future value V , the nominal rate of discount r , and the number of years t . The
r t
expression e is often referred to as the discount factor. The above equation considers
as the exponential growth function, we can immediately read  r as the instantaneous
rate of growth of A . Being negative, this rate is in effect a rate of decay. Just as interest
compounding exemplifies the process of growth, discounting illustrates negative growth.

Example : Find present value of a 5-year bond with a face value of $1000 considering 9
percent interest rate under annual compounding.
Solution :

t
We know A  V 1  r   1000 1  0.09 5
 A  1000 0.64993  649.93
That is, $649.93 at 9 percent interest will grow to $1000 in 5 years.

17
Effective vs. Nominal Rates of Interest

The nominal interest rate will earn different effective rates of interest which depend on
the type of compounding. When compounded annually for 2 years, $100 will be worth
$121 ( V  A 1  r t ); when compounded semiannually, V  $121.55 (
V  A 1  r m
mt
); when compounded continuously, V  $122.14 ( V  A e rt ) under
10 percent interest rate.

To find the effective annual rate of interest re for multiple compounding;

A1  re   A 1  r m
t mt

Dividing by A and taking the t th root of each side

1  re  1  r m
m

re  1  r m  1
m

To find the effective annual rate of interest for continuous compounding:

1  re  e r
re  e r  1

Example 1: Find the effective annual rate of interest for a nominal interest rate of 10
percent when compounded for 2 years (1) semiannually and (2)
continuously.

Solution :
re  1  r m  1  (1.05) 2  1
m
(1) Semiannually,
re  1.1025  1  0.1025  10.25%

(2) Continuously, re  e r  1  e 0.1  1  1.10517  1  0.10517  10.52%

18
Example 2: Determine the interest rate needed to have money double in 6 years when
compounded semiannually.

Solution :
mt
 r
V  A 1   ,
 m
where V is the future value (principal plus interest), A is the principal, m is the number
of compounding at time t and r is the interest rate.

2 ( 6)
 r
Then 2 A  A 1  
 2
2  1 1  0.5 r 
12

 1  0.5 r  12 2

ln 1  0.5 r   0.69315  0.05776

1 1
 ln 2 
12 12
 1  0.5 r  e 0.05776  1.05946
 0.5 r  1.05946  1  0.05946
 r  0.11892  11.89%

Example 3: At what interest rate will money treble if compounded continuously for 8
years ?.

Solution :
V  A e rt
 3 A  A e r (8)
 ln 3  ln e8r
 1.09861  8r
 r  0.1373  13.73%

Example 4: How long will it take money to double at 5 percent interest when
compounded quarterly ?

Solution :
4t
 0.05 
Then V  A 1  
 4 
2  1.0125
4t

 ln 2  4t ln 1.0125
 0.69315  4 0.01242 t
 t  13.95 years .

19
Example 5: Costs C of a government program escalate from 5.39 billion in 1995 to
10.64 billion in 2001. Express costs in terms of an ordinary exponential
function, and find the annual rate of growth.

Solution :
5.39  C0 1  r   C0
0
(i)
10.64  C0 1  r 
6
and (ii)

Substitute C0  5.39 from (i) in (ii) and simplify

10.64  5.391  r 
6

 1.974  1  r 
6

Take the common log of both sides

log 1.974  6 log 1  r 


1
0.29535  log1  r 
6
 log1  r   0.04923
 1  r  anti log 0.04923  1.12
 r  1.12  1  0.12  12%
That is annual growth rate is 12%. For this purpose general formula is defined as
C  5.39 1  0.12 .
t

Example 6: An animal population goes from 3.5 million in 1997 to 4.97 million in 2001.
Express population growth 𝐴 in terms of a natural exponential function and
determine the rate of growth.
Solution : The population growth function is defined as
V  Ae rt
r (0)
 3.50  Ae A (i)
 4.97  Ae 4r (ii)
Substitute 𝑃0 = 3.50 from (i) in (ii) and simplify,
4.97  3.50 e 4r
 1.42  e 4r
Take the natural log of both sides
ln 1.42 = ln e4r = 4r
 0.35066 = 4𝑟
 𝑟 = 0.08767 ≈ 8.8%
Thus, 𝑉 = 3.50𝑒 0.088 𝑡

20
 Logarithms
The Meaning of Logarithm

When we have two numbers such as 4 and 16, which can be related to each other by the
equation 4  16 , we define the exponent 2 to the logarithm of 16 to the base of 4 , and
2

write
log 4 16  2 .
It should be clear from this example that the logarithm is nothing but the power to which
a base ( 4 ) must be raise to attain a particular number ( 16 ). In general, we may state that
y  bt  t  log b y
which indicates that the log of y to the base b (denoted by log b y ) is the power to
which the base b must be raised in order to attain the value y . Given y , the process of
finding its logarithm log b y is referred to as taking the log of y to the base b . The
reverse process, that of finding y from a known value of its logarithm log b y , is
referred to as taking the antilog of log b y .

Common Log and Natural Log

The base of the logarithm, b  1 , does not have to be restricted to any particular number,
but in actual log applications two numbers are widely chosen as bases : the number 10
and the number e . When 10 is the base, the logarithm is known as the common
logarithm, symbolized by log 10 . With e as the base, the logarithm is referred to as the
natural logarithm and is denoted either by log e or by In .

Common logarithms, used frequently in computational work, are exemplified by the

following :
log 10 1000  3 [because 10  1000 ]
3

log 10 100  2 [because 10  100 ]

2

1 [because 10  10 ]
1
log 10 10
0 [because 10  1 ]
0
log 10 1
1
log 10 0.1   1 [because 10  0.1 ]
2
log 10 0.01   2 [because 10  0.01 ]

Observe the close relation between the set of numbers immediately to the left of the
equals signs and the set of numbers immediately to the right. From these, it should be

21
apparent that the common logarithm of a number between 10 and 100 must be between
1 and 2 , and that the common logarithm of a number between 1 and 10 must be a
positive fraction, etc.

In analytical work, however, natural logarithms prove vastly more convenient to use than
common logarithms. Since by the definition of logarithm, we have the relationship
y  et  t  log e y (or t In y)
it is easy to understand that the analytical convenience of e in exponential functions will
automatically extend into the realm of logarithms with e as the base. The following
examples are given for natural logarithms :
In e3  log e e3  3
In e 2  log e e 2  2
In e1  log e e1  1
In 1  log e e 0  0
1
In  log e e 1   1
e
The common log and natural log are convertible into each other; i.e., the base of a
logarithm can be changed, just as the base of an exponential expression can.

Rules of Logarithms

Rule I : (log of a product) In (u v)  In u  In v (u, v  0)

Rule II : (log of a quotient) In (u v)  In u  In v (u, v  0)
Rule III : (log of a power) In u a  a In u (u  0)
Rule IV : (conversion of log base) log b u  (log b e)(log e u) (u  0)
1
Rule V : (inversion of log base) log b e 
log e b

22
Derivatives of Log-Function Rule

The derivative of the log function y  In t is

d 1
In t 
dt t

The Rules Generalized

The log-function and exponential-function rules can be generalized to cases where the
variable t in the expression e t and In t is replaced by some function of t , say, f (t ) .
The generalized versions of the two rules are
d f (t )  d u u du 
e  f (t ) e f (t ) or dt e  e dt 
dt
d f (t )  d 1 dv 
In f (t )  or
 dt In v 
dt f (t ) v dt 
Given a function y  e , we can first let u  f (t ) , so that y  e . Then, by
f (t ) u

d f (t ) d u de u du du
e  e  .  eu  e f (t ) f (t )
dt dt du dt dt
Similarly, given a function y  In f (t ) , we can first let v  f (t ) , so as to form a
chain: y  In v , where v  f (t ) . Then, by the chain rule, we have
d d dIn v dv 1 dv 1
In f (t )  In v  .   f (t )
dt dt dv dt v dt f (t )

 An Application of Logarithms

The rules of logarithms enable us to solve with ease certain simple exponential equations
(exponential functions set equal to zero). For example, if we seek to find the value of x
that satisfies the equation
ab x  c  0, (a, b, c  0)
we can first try to transform this exponential equation, by the use of logarithms, into a
linear equation and then solve it as such. For this purpose, the c term should first be
transposed to the right side:

23
 ab x  c .
This is because there is no simple log expression for the additive expression, but there

exist convenient log expressions for the multiplicative term ab x and for c individually.
Thus, after the transposition of c and upon taking the log of both sides, we have
log a  x log b  log c ,
which is a linear equation in the variable x , with the solution
log c  log a
x .
log b

 Example
Find dy dt from y  t In t .
3 2

2
The given function is a product of two terms t 3 and In t , therefore, we use product
rule of differentiation :
dy 3 d
dt
t
dt
In t 2  In t 2d 3
dt
t 
 2t 

 t 3 2   In t 2 3t 2  
t 
 2t 2  3 t 2 2 In t 
 2t 2 1  3In t 

 Example
Find dy dt from

x2
y .
x  32 x  1
Instead of applying the product and quotient rules, we may first take the natural log of
both sides of the equation to reduce the function to the form
In y  In x 2  In x  3  In 2 x  1
The derivative of the equation is explained as

24
 d
dx

In y   d In x 2  In x  3  In 2 x  1
dx

1 dy 2 x 1 2 7x  6
 2  
y dx x x  3 2 x  1 x x  32 x  1
dy 7x  6
 y
dx x x  32 x  1
7x  6 x2 x 7 x  6
 
x x  32 x  1 x  32 x  1 x  32 2 x  12

 A Problem of Timber Cutting

To identify the best time for timber cutting optimal time for sale, we suppose the value of
timber (already planted on some given land) on the basis of following increasing function
of time :

V 2 t

expressed in units of $1,000. Assuming a discount rate of r and assuming zero upkeep
cost during the period of timber growth. To solve this problem, first we convert V into
its present value :

A(t )  V e r t 2 t
e r t
thus In A  In 2 t
 In e  r t  t In 2  r t  t 1 2 In 2  r t .
To maximize A , we must set dA dt  0 . The first derivative is obtainable by
differentiating In A with respect to t and then multiplying by A :
1 dA 1 1 2
 t In 2  r
A dt 2
dA  In 2 
thus  A   r  .
dt 2 t 
Since A  0 , the condition dA dt  0 can be met if and only if
In 2 In 2
r or t .
2 t 2r
Consequently, the optimum number of years of growth is

25
 2
 In 2 
t  
*
 .
 2 r 
It is evident from this solution that, the higher the rate of discount, the earlier the timber
should be cut.
To make sure that t * is a maximizing solution, we should be needed to checked second
order condition.

In this example, we have abstracted from planting cost by assuming that the trees are
already planted, in which case the (sunk) planting cost is legitimately excludable from
consideration in the optimization decision. If the decision is not one of when to harvest
but one of whether or not to plant at all, then the planting cost (incurred at the present)
must be duly compared with the present value of the timber output, computed with t set
at the optimum value t * . For example, if r  0.05 , then we have
2
 0.6931 
t    6.931  48.0 years.
* 2

 0.10 
A*  26.931 e 0.05 48.0   122.0222 e 2.40
and
 122.0222 0.0907   $11.0674 in thousands 
*
So only a planting cost lower than A will make the venture worthwhile, there also to be
taken nil upkeep cost.

 Finding the Rate of Growth

When we observe a variable y is a function of time, y  f (t ) , then the growth is
defined as
dy dt f (t ) m arg inal function
ry    .
y f (t ) total function
Therefore, it is observed that this ratio is precisely the derivative of In f (t )  In y .
Thus, to find the instantaneous rate of growth of a function of time f (t ) , we can
differentiate it with respect to t , and then dividing by f (t ) .

26
  Rate of Growth of a Combination of Functions

To examine the instantaneous rate of growth of a product of two function of time, let us
consider :
u  f (t )
y uv where 
v  g (t )
Taking the natural log of y , we obtain
In y  In u  In v
Thus the desired rate of growth is
d d d
ry  In y  In u  In v
dt dt dt
But the two terms on the right side are the rates of growth of u and v , respectively. Thus
we have the rule
ru v   ru  rv
Expressed in words, the instantaneous rate of growth of a product is the sum of the
instantaneous rates of growth of the components.

By a similar procedure, the rate of growth of a quotient can be shown to be the difference
between the rates of growth of the components:
ru v   ru  rv .

 Example
If consumption C is growing at the rate  , and if population H is growing at the rate
 , what is the rate of growth of per capita consumption ? Since per capita consumption
is equal to C H , its rate of growth should be
rC H   rC  rH    

Now consider the instantaneous rate of growth of a sum of two functions of time:
 u  f (t )
z u v where 
 v  g (t )

27
This time, the natural log will be
In z  In u  v .

In u  v 
d d
Thus rz  In z 
dt dt
1 d
 u  v 
u  v dt


1
 f (t )  g (t )
uv
But we know ru  f (t ) f (t ) , so that f (t )  f (t ) ru  u ru . Similarly, we have
g (t )  v rv . Then we may write
u v
ru v   ru  rv
uv uv
which shows that the rate of growth of a sum is a weighted average of the rates of growth
of the components.
By the same token, we have
u v
ru v   ru  rv
uv uv

 Example
The exports of goods of a country, G  G(t ) , has a growth rate of  t , and its exports
of services, S  S (t ) , has a growth rate of b t . What is the growth rate of its total
exports ?

Since total exports is X (t )  G(t )  S (t ) , then its rate of growth is defined as

G S
rX  rG  rS
X X
G   S  b  G   S b
     
X  t  X t Xt

28
  Point Elasticity
Let us consider y  f (t ) , the derivative of In y measures the instantaneous rate of
growth of y . Then, for the function y  f (t ) , we differentiate In y with respect to
In x , rather than to x .
To begin with, let us define u  In y and v  In x . Then we can observe a chain of
relationship linking u to y and thence to x and v as follows:

u  In y y  f (x) x  e v  e In x
Then the derivative of In y with respect to In x is
d In y  du du dy dx
 
d In x  dv dy dx dv
d   dy   d  1 dy v 1 dy dy x
  In y     e v   e  x
 dy   dx   dv  y dx y dx dx y
But this expression is precisely that of the point elasticity of the function. Hence we have
established the general principle that, for a function y  f (x) , the point elasticity of y
with respect to x is
d In y 
y x  .
d In x 

Example 1: The price of agricultural goods is going up by 4 percent each year, the
quantity by 2 percent. What is the annual rate of growth of revenue 𝑅 derived from the
agricultural sector?
Solution:
The revenue, 𝑅 = 𝑃𝑄
⇒ ln 𝑅 = ln 𝑃 + ln 𝑄
The derivative of the natural log function equals the instantaneous rate of growth 𝐺 of the
function. Thus,
𝑑 𝑑 𝑑
𝐺 = 𝑑𝑡 𝑙𝑛𝑅 = 𝑑𝑡 𝑙𝑛𝑃 + 𝑑𝑡 𝑙𝑛𝑄
According to question,
𝑑 𝑑
𝑙𝑛𝑃 = 𝑔𝑟𝑜𝑤𝑡𝑕 𝑜𝑓 𝑃 = 4% and 𝑙𝑛𝑄 = 𝑔𝑟𝑜𝑤𝑡𝑕 𝑜𝑓 𝑄 = 2%
𝑑𝑡 𝑑𝑡
𝑑
Thus, 𝐺= 𝑙𝑛𝑅 = 0.04 + 0.02 = 0.06
𝑑𝑡
That is, the annual rate of growth of revenue 𝑅 from agriculture sector is 6%.

29
Example 2: A firm experiences a 10 percent increase in the use of inputs at a time when
input costs are rising by 3 percent. What is the rate of increase in total input costs?
Solution:
The costs is defined as 𝐶 = 𝑃𝑄, where P indicates the price of inputs and Q indicates
the quantity of inputs
⇒ ln 𝐶 = ln 𝑃 + ln 𝑄
The derivative of the natural log function equals the instantaneous rate of growth 𝐺 of the
function. Thus,
𝑑 𝑑 𝑑
𝐺 = 𝑑𝑡 𝑙𝑛𝐶 = 𝑑𝑡 𝑙𝑛𝑃 + 𝑑𝑡 𝑙𝑛𝑄 = 0.03 + 0.10 = 0.13
That is, the rate of increase in total input costs is 13%.

Example 3: Employment opportunities 𝐸 are increasing by 4 percent a year and

population 𝑃 by 2.5 percent. What is the rate of growth of per capita employment PCE?
Solution:
The per capita employment is defined as
𝐸
𝑃𝐶𝐸 = 𝑃
⇒ ln 𝑃𝐶𝐸 = ln 𝐸 − ln 𝑃
Taking the derivative to find the growth rate,
𝑑 𝑑 𝑑
𝐺= 𝑙𝑛𝑃𝐶𝐸 = 𝑑𝑡 𝑙𝑛𝐸 − 𝑑𝑡 𝑙𝑛𝑃 = 0.04 − 0.025 = 0.015 = 1.5%
𝑑𝑡
That is, the rate of growth of per capita employment PCE is 1.5%.

Example 4: National income 𝑌 is increasing by 1.5 percent a year and population 𝑃 by

2.5 percent a year. What is the rate of growth of per capita income PCY?
Solution:
The per capita income is defined as
𝑌
𝑃𝐶𝑌 = 𝑃
⇒ ln 𝑃𝐶𝑌 = ln 𝑌 − ln 𝑃
Taking the derivative to find the growth rate,
𝑑 𝑑 𝑑
𝐺= 𝑙𝑛𝑃𝐶𝑌 = 𝑑𝑡 𝑙𝑛𝑌 − 𝑑𝑡 𝑙𝑛𝑃 = 0.015 − 0.025 = −0.01 = −1%
𝑑𝑡
That is, per capita income is falling by 1percent a year.

30
Example 5: A country exports two goods, copper 𝑐 and bananas 𝑏, where earnings in
terms of million dollars are
𝑐 = 𝑐 𝑡0 = 4 and 𝑏 = 𝑏 𝑡0 = 1
If 𝑐 grows by 10 percent and 𝑏 by 20 percent, what is the rate of growth of export
earnings 𝐸?
Solution:
The export earnings function is defined as
𝐸 =𝑐+𝑏
⟹ ln 𝐸 = ln(𝑐 + 𝑏)
𝑑 𝑑 1
Then 𝐺𝐸 = 𝑑𝑡 ln 𝐸 = 𝑑𝑡 𝑙𝑛 𝑐 + 𝑏 = 𝑐+𝑏 𝑐 𝑡 + 𝑏(𝑡) (i)
1 𝑔 (𝑥)
𝑓 𝑥 = 𝑙𝑛𝑔 𝑥 𝑡𝑕𝑒𝑛 𝑓 𝑥 = 𝑔 𝑥
𝑔 𝑥 = 𝑔(𝑥)
𝑐 (𝑡) 𝑏 (𝑡)
The growth of 𝑐 is defined as 𝐺𝑐 = 𝑐(𝑡) and the growth of 𝑏 is defined as 𝐺𝑏 = 𝑏(𝑡) , then

𝑐 (𝑡) = 𝐺𝑐 𝑐(𝑡) and 𝑏(𝑡) = 𝐺𝑏 𝑏(𝑡). Substituting in (i), we have

1
𝐺𝐸 = 𝐺 𝑐 𝑡 + 𝐺𝑏 𝑏(𝑡)
𝑐+𝑏 𝑐
𝑐(𝑡) 𝑏(𝑡)
⟹ 𝐺𝐸 = 𝑐+𝑏 𝐺𝑐 + 𝑐+𝑏 𝐺𝑏
Then substituting the given values, we have
4 1 4 1
𝐺𝐸 = 4+1 0.10 + 4+1 0.2 = 5 0.10 + 5 0.20 = 0.12 = 12%
That is, the rate of growth of export earnings is 12%.

31
32