Вы находитесь на странице: 1из 38

\

CHAPTER

3

TRANSVERSALITY CONDITI

ONS

OINT

FOR

VARIABLE-

ENDP

PROBLEMS

The

calculus

differential

Euler

equation-the

of

basic

necessary

normally

two

and

condition

a

in the

varia-tions-is

equation

with

fixed

second-order

constants.

points,

the

containing

initial

arbi-trary

terminal

For

problems

two

given

boundary

conditions

provide

sufficient

information

to

definitize

the

two

arbitrary

constants.

But

to

the

discretionary

if

initial

or

terminal

then

the

the

at time

of

point

is

variable

(subject

will

the

no

PT

In

choice),

a boundary

case,

for

preceding

condition

if

faces

the

problem.

be missing.

Such

will

be

dynamic

monopolist

of

externally

imposed

price

choice

as an integral

part

instance,

chapter

can treat

T ,and

the optimization

that case, the boundary condition P(T) = PT will no longer be available, and the void must be filled by a

transversality condition. In this chapter we shall develop the transversality conditions appropriate to various types of variable termi-nal points.

3.1

THE

GENERAL

CONDITION

For expository

convenience,

TRANSVERSALITY

we shall assume that only

the terminal point is variable. Once we learn to deal with that, the technique is easily extended to the case of a variable initial point.

The

Variable-Terminal-Point

objective

is

to

Problem

Our new

Maximize

or minimize

Yy] = /*~(t,y, yf) dt

0

(3.1)

y(0)

= A

(A given)

subject

to

and

Y ( ~)=Y T

( T ,~~f r

This

that

"free"

choice

although

to the problem.

optimal

differs

the

from

the

time

previous

and

version

terminal

of

the

state

problem

a

y,

part

in

terminal

T

are now

of

the

that

in the sense that

T

is

process.

free, only

they

It

have

is

to

become

be

understood

of

T

positive

values

are relevant

To

develop

the

necessary

conditions

for

an

extremal,

we

shall,

as before,

employ

a perturbing

curve

p(t),

and

use

the

variable

E

to

generate

neighboring

paths for comparison

with

the

extremal.

First,

suppose

that

T *

is

the

known

optimal

terminal

time.

Then

any

value

of

T

in the immediate

neighborhood

of

T*

can be

expressed

as

where

E

represents

a

small

number,

and

AT

represents

an

arbitrarily

chosen

(and

fixed)

small

change

in T. Note that since

T* is known

and

AT

is a prechosen

magnitude,

T

can be

considered

as a function

of

E ,T(E)

with derivative

 
 

7 r n

 
 

The

same

E

is used

in conjunction

with

the

perturbing

curve

p(t) to generate

neighboring

paths

of

the extremal

y*(t):

 

However, although

the

p(t) curve

must

still

satisfy

the

condition

p(0)

=

0

[see

(2.2)1

to

force

the

neighboring

paths

to

pass

through

the

fixed

initial

point,

the

other

condition-p(t)

I = 0-should

now

be

dropped,

because

y,

is free.

By

substituting

(3.4)

into

the

functional

V[y],

we

get

a

function

V(E)

akin

to

(2.121,

but

since

T

is

a

function

of

E

by

(3.2),

the

upper

limit

of

integration

in

the V function

will also vary with

E:

(3.5)

~( c=

kT'"~[t, y*(t)

+ rp(t), y*'(t)

+ rpf(t)] dt

 

Y(t>

~'( t

Our problem

is to optimize

this V

function

with

respect

to

E.

definite

integral

(3.5) falls into

the general

form

of

(2.10).

The V/d~ is therefore,

by

(2.111,

on the

right closely resembles the one encountered in (2.13) arlier development of the Euler equation. In fact, much of the rocess leading to (2.17) still applies here, with one exception.

[Fyrp(t)];f

in (2.16) does not vanish

since

in the present

we have

problem,

that

in (2.171,

value

p(T) z 0. With this amendment

= [F,,],=,p(T),

[Fy,p(t)],=,

assumed

result

to the earlier

term in (3.6)

=

[ [

p(t)

1

Fy - -F'!

dt + [FY*],=,p(T

can also

write

Second

term in (3.6)

= [Fit=, AT

these

into

(3.6),

and

setting

d V / d ~= 0,

we

obtain

the

 

new

p(t)[Fy

- :FYr]

dt + [Ff]t=T~(T)+ [F]~=TIT = 0

three

terms

on the left-hand

its own arbitrary

element:

side

of

(3.71, each

contains

p(t) (the

entire

perturbing

curve)

in the (T) (the terminal value on the perturbing

curve)

in the second

arbitrarily

chosen

change

in

T) in

the third. Thus we cannot offsetting or cancellation of terms. Consequently, in order tocondition (3.7), each of

the three

terms

must

individually

be set o

 

T FIGURE FIGURE

3.1 3.1

Step

ii

To

this

end,

we

first

get

rid

of

the

arbitrary

quantity

p(T) by

transforming

it into terms

of

AT

and

AyT-the

changes

in

T

and

y,,

the

two

principal

variables

in

the

variable-terminal-point

problem.

This

can

be

done

with

the

help

of Fig.

3.1. The

AZ'

curve

represents

a neighboring

path

obtained

by

perturbing

the

AZ

path

by

~p(t

),with

E

set

equal

to

one

for

convenience.

Note

that

while

the

two

curves

share

the

same

initial

point

because

p(0)

= 0, they

have

different

heights

perturbing

the vertical

at

t

y,

can

11,

resulting

be

also

the

AZ'

= T

because

The

p(T) #

0 by

construction

of

the

But

p(T),

direct

of

the

by

in

T

=

E

curve.'

distance

from

altered

curve

magnitude

measures

perturbation.

the

amount

if

AT

shown

change

as

ZZ',

the

by

should,

inasmuch

since

out

EAT (= AT

> 0, be

extended

to

2".2 As

a result,

yT

is

further

pushed

up by the vertical

distance

between

Z'

and

2". For

a small

AT, this

second

change

the total

in

y,

can be

change

in y,

approximated

from

point

Z

by

to point

y'(T) AT. Hence,

Z"

is

Rearranging

this

relation

allows

us to express

p(T) in

terms

of

AyT

and

 

AT:3

'~echnically,

the

point

T

on

the

horizontal

axis

in

Fig.

3.1

should

be

labeled

T*. We

are omitting

the

*

for

simplicity.

This

is

justifiable

because

 

the

result

 

that

this

discussion

is

leading

to-(3.9)-is

a

transversality

 

condition,

which,

 

as

a

standard

practice,

is

stated

in terms

of

T (without

the *) anyway.

'~lthough

we are initially

interested

only

in the solid portion

of

the

AZ

path, the equation

for

that

path

should

enable

us

also

to

plot

the

broken

portion

as

an

extension

of

AZ.

The

perturbing

curve is then

applied

to the extended

version

of

AZ.

 

~ h resulte

3

3.1.

in (3.8) is valid

even

if

E

is

not

set equal

to one

as

in

Step iii

dropping the first term in (3.7), we finally arrive at the

desired general transversality condition

Using (3.8) to eliminate

p(T) in (3.7), and

Fig.

This

condition,

unlike

the

Euler

equation,

is

relevant

only

to

one point

of

time, T .Its role

is

to take

the place

of

the

missing

terminal

condition

in the

present

problem.

Depending

on the

exact

specification

of

the

terminal

line

or

curve,

however,

the

general

condition

(3.9) can

be

written

in various

specialized

forms.

 

3.2

SPECIALIZED

TRANSVERSALITY

 

CONDITIONS

 

In

this

section

we

consider

five

types

of

variable

terminal

points:

vertical

terminal

line,

horizontal

terminal

line,

terminal

curve,

truncated

 

vertical

terminal

line, and

truncated

horizontal

terminal

line.

 

Vertical

Terminal

Line

(Fixed-Time-Horizon

Problem)

 

The

vertical-terminal-line

case,

as

illustrated

in

Fig.

1.5a,

involves

g

fixed

T .Thus

AT

=

0, and

the

first

term

in

(3.9 ) drops

out.

But

since

Ay,

is

arbitrary

and

can

take

either

sign,

the

only

way

to

make

the

second

term

in (3.9) vanish

for

sure is

to

have

Fy r= 0

at

t =

T .This

gives

rise

to

the transversality

condition

 

which

is sometimes

referred

to as the

natural

boundary

condition

 

Horizontal

Terminal

Line

 

(Fixed-Endpoint Problem)

For the horizontal-terminal-line case ,as illustrated in

have

Ay, = 0 but AT is arbitrary. So

automatically drops out, but the first does not. Since AT

is arbitrary, the only way to make the first term vanish for sure is to have the bracketed expression equal to zero Thus the transversality condition is

Fig. 1.5b ,the situa-tion is exactly reversed;

we

term

now

the second

in (3.9

It

might

F(t

capital

be

useful

to

give

an

economic

let

,where

interpretation

interpret

represents

to

(3.10 ) and

as

a

(3.11) .To f ix ideas,

profit

function

taking

us

y

net

,y ,y')

stock, and y' entails

represents

resource

away from the current profit-making busines

operation, so as to build up capital which will enhance future profit. Hence there exists a tradeoff between current profit and future profit. At any time t, with a

given capital stock y, a specific investmen decision-say, a decision to select the investment rate yl,--will result in the current profi F(t, y, y',). As to the effect of the investment decision on future profits, i enters through the intermediary of capital. The rate of

capital accumulation

value

F(t, y, y',) to see the overall (current as well as future

profit implication of the investment decision. The

imputed (or shadow) value to the firm of a unit of

is measured by the derivative Fy,.This means that if we

is y',;

if we can convert

that into a

profi

measure,

then we can add it to the current

capita

decide

to leave

(not

use

up) a unit

of

capital

at the

terminal

time

 

it will entail

a

negative

value

equal

to

-Fyg.Thus,

at

t

= T,

the

value

measure

of

y',

is

-y',

Fyto.Accordingly,

the overall profit

implication

of

the

decision

to

choose

the

investment

rate y',

is

F(t, y, y',)

- yb Fyg.The gen

eral expression

Now

for this

is

we can interpret

F - y'Fy,, as in (3.11).

the transversality

condition

(3.11) to mean that

in a problem with a free terminal time, the firm should

select

capital

(current and future) profit. In other words, all the profit opportunities should have been fully taken advantage of by the optimally chosen terminal time. In addition, (3.10)-which can equivalently be writ-ten as [ - FY,],=, = O-instructs the firm to avoid any sacrifice of profit that will be incurred by leaving a positive

terminal capital. In other words, in a free-terminal-state problem, in order to maximize profit in the interva [0, T ] but not beyond, the firm should, at time T, us up all

the capital it ever accumulated.

a

T

such that

a decision

t

=

T, no

longer

to

invest

yield

any

and

accumulate

overal

will, at

Terminal

Curve

With

1.5c,

a terminal

neither

curve

AyT nor

yT = +(T),as illustrated

AT

is

assigned

a

zero

in Fig.

value,

so

neither

term

in

(3.9)

drops

out. However,

for a small

arbitrary

AT, the terminal

curve

implies

that

Ay, qS=AT.

So it is possible to eliminate Ay, in (3.9) and combine the

two

terms

Since

AT

into the form

[F -y1FYp + F,vqS]t-TAT

= 0

is arbitrary,

we can deduce

the transversality

condition

 
 

Unlike

the

last

two

cases,

which

involve

a

single

unknown

in

the terminal

point

(either

y,

or

TI,

the

terminal-curve

 

case

requires

us

to determine

both

y,

and

T. Thus two relations

are needed

for this purpose.

The transversality condition (3.12) only provides one; the

other is supplied by the equation y, = +(T 1.

Truncated

Vertical

Terminal

Line

The usual

case of vertical

terminal

line, with

AT

= 0,

specidixes

(3.9) to

When the line is truncated-restricted by the terminal condition yT 2 ymi where yminis a minimum permissible

level

types

possible

> ymi nor y,*

of

of

y-the

outcome:

optimal

y,*

solution

can

have

two

y,*

= ymin .If

> ymin

the terminal restriction

is

automatically

satisfied;

that

is,

it

is

nonbinding.

Thus,

the

transversality

condition

is

in that event

the same as (3.10):

 

The

basic

reason

for

this

is

that,

under

this

outcome,

there

are

admissible

neighboring

paths

with terminal

values

 

both

above

and

below

y,*,

so that

Ay,

= y,

- yT*

can

take

either

sign.

Therefore,

the

only

way

to

satisfy

(3.13) is to have

The

other

outcome,

Fy,= 0 at

t = T

y,*

= y,,,,

on the

other

hand,

only

admits

neigh-boring

paths

with

terminal

values

y,

2 y,*.

This

means

that

Ay,

is

no longer

completely

arbitrary

(positive

or

negative) ,but

is

restricted

to

be

nonnegative.

terminal

value

Assuming

the

p(T )

> 0,as

perturbing

curve

in

Fig .3.1,Ay,

r

to

0

have

would

mean

that

E

2

0

( E

=

1 in

Fig .3.1) .The

nonnegativity

of E , in turn ,means

condition

(3 .13)-which

has

that

the

transversality

its

roots

in the

first-order

condition

d V / d

~=

O-

must

be

changed

to

an

inequality

as

in

the

Kuhn-Tucker

conditions.*

For

a

maximization

called

for, and

problem

(3.13

,the

should

I

type

become

of

inequality

is

 

(3.15)

[ F y l] t =T A I~~0

And

since

Ay,

2

0, (3.15) implies

condition

(3.16)

[Fy,]t=TsO

for~,*=~,

 

Combining

(3.14)

and

(3.16),

we

may

write

the

following summary statement of the transversality condition for a maximization problem

(3.17)

[F~)],=,5 0

YT* 2 Ymi

 

(YT* - ~rnln)[~Y']t=T=

 
 

[for maximization

of

V

4 ~ o anr

explanation

of the

Kuhn-Tucker

conditions,

see

Alpha

C

Chiang,

Fundamenta

Methods

of

Mathematical

Economzcs,

3d ed

McGraw-Hill,

New

York, 1984, See. 21.2

is instead

inequality sign in (3.15) must be reversed, and the transversality condition becomes

If

the

problem

to

minimize

V, then

(3.17')

[FY*]t=T2 0

(YT* -~rnin)[Fy#]~== 0

YT* 2Ymin

[for minimization

The

tripartite

complicated,

but its

condition

(3.17 )or (3.17' ) may

the

of

V]

seem

practical problem solving, we can try the [Fy#1,=, = 0

condition in (3.14) first, and check the resulting yT* value. If yT* 2 ymin then the terminal restriction is satisfied, and the problem solved. If y,* <

application is not. In

y,,,

on

the

other

hand,

then

the

optimal

y,

lies

below

the

permissible

range,

and

[F,,],,,

fails

to

reach

zero on the truncated

terminal

line. So, in order

to

satisfy

the

complementary-slackness

condition

in

(3.17

) or

(3.17')

we

just

set

y,*

= y,,,

treating

the

problem

as

one with

a fixed

terminal

Truncated

Horizontal

point

Cl", ymin

Terminal

Line

The

horizontal

terminal

line

may

be

truncated

by

the

restriction

T

5

T,,,

where

T,,

represents

a

maximum

permissible

time

for

completing

a

task-a

deadline.

The

analysis

of

such

a

situation

is

very

similar

to

the

truncated

vertical

terminal

line

gust

discussed.

By

analogous

reasoning,

we

can

derive

the

following

transversality

condition

for

a

maximization

prob-lem:

(3.18)

[ F -~ ~ F ~ ~ ] ~T*IT,,,~,L O

( T * - T,,)

maximization

of

V]

[ F - y1FY,]t =,= 0

If

in

the

problem

(3.18) must

is

to

minimize

be changed,

and

V, the

the transversality

first

inequality

[for

condition

is

(3.18')

[ F - Y 'F ~~]I~0,, T* I T,,

 

(T* - T,,)

[F - ytFY.],=, = 0

[for minimization

of

V]

EXAMPLE

1

Find

the

curve with

the

shortest

distance

between

the

point

(0 ,l ) and

the

line

y

=

2

-

t.

Referring

to

Fig.

3.2,

we

see

that

this

is

a

problem

with

similar

a terminal

to

the

curve y(T) = 2 - T ,but otherwise

Example

shortest-distance

problem

in

Sec.

2.2. Here,

the problem

is

to

Minimize

ny]= /*(I

0

+ y12)1'2

dt

subject

to

y(0)

= 1

and

y(T)

= 2

- T

Y

it

4

is

of

1 2

It has previously

been

t

FIGURE

3.2

shown that with

the given

integrand,

the Euler

equation

yields

a straight-line

extremal,

say,

with

condition

two

y*

=a t + b

arbitrary

constants

a

and

b.

From

y(0) =

1, we

can readily

determine

the

that

initial

b

=

1.

To determine

the

other

constant,

a, we

resort

to

the

transversality

condition

(3.12). Since we

have

 

the transversality

condition

can be written

as

 

(1 +y")1/2

+

(- 1 - yl)rl(l

+yr2)-l"

=

o

(at

t = T

Multiplying

through

by

(1 + y'2)1/

2and

simplifying

we can reduce this equation

to

the

form

yt

=

1 (at

t =

T).But

the

extremal

actually

has

a constant

slope,

y*'

= a,at all values of

t.Thus, we must have

 

a

=

1. And

the extremal

is therefore

y*(t)

= t + 1

 

As

shown

in

Fig.

3.2,

the

extremal

is

a

straight

line that meets the terminal line at the point

line

and the extremal are, respectively+ ,- 1 and +1.Thus the

two lines are orthogonal (perpendicular) to each other. What the transversality condition does in this case is to

require

Moreover ,we note that the slopes

(;,

1;).

of

the

terminal

the extremal

to be orthogonal

to the terminal

line

EXAMPLE

2

Find

the extremal

of

the functional

FIGURE

3.3

with

free. Here we have a horizontal terminal line as depicted

in Fig. 3.3.

boundary

conditions

y(0) =

1, y,

=

10, and

T

The general solution

of

the

Euler

equation

for

this problem has previously been found to be a quadratic

path (Sec. 2.2, Example 1)

Since

y(0) =

1, we

have

c, =

1. But

the

other

constant

c, must

condition

be

defini-tized

(3.11)+ .Since

with

the help

F

= ty'

+ y'2+

of

the transversality

so that

F,,

= t

+ 2y1

that

condition

becomes

which

is,

terminal

reduces

extremal

time.

the

to

is

To

-y'2

=

0

required

meet

this

or

to

y'

=

have

0

at

a zero

condition,

we

t

=

T .That

slope

differentiate

at the

the

general

solution

to

get

y*'(t), then

set

t

=

T ,and

let

y*'(T)

=

0.

The

result

+ is

the

equation

- T/2

+ c, = 0; it

follows

that c, = T/2 .However,

we

still

cannot

determine

 

the

specific

value

of

cl without

knowing

the value

of

T

 

To

find

make

use

of

the

information

that

y,

=

10

T ,we

(horizontal

terminal

line).

Substituting

 

and

letting

c2

=

the

1 into

the

resulting

general

expression

solution,

take

t

c, = T/2 T ,and

=

value

setting

the

stipulated

10, we obtain

the equation

 

The

negative value, we end up with T* = 6. Then it follows

that c, = 3, and the extremal is

solution

values

of

T

are therefore

rt 6. Rejecting

the

 

This

time

path

is

shown

in Fig.

3.3. As

required

by

the transversality

condition,

 

the

extremal

indeed

attains

a

zero

slope

at

the

terminal

point

(6, 101.Unlike

in

Example

1, where

the transversality

condition

translates

into

an

orthogonality

requirement,

the

transversality

condition

in

the

pre-sent

example

dictates

that

the

extremal

share

a

common

slope

with

the

given

horizontal

terminal

line

at

t

= T.

 

Application

to the Dynamic

Monopoly

Model

Let

us

now

consider

the

Evans

monopolist

a

variable

as a problem

with

terminal

state

P,

a fixed

2

P,,.

model

of

terminal

With

a

dynamic

time

the

T ,but

<