Chapter11 15

CHAPTER
Gale Tiory
and Strjegiehavior
-AL
U Preview
• !ntroducfion to Game Theory

The Payoff Matrix
Naah Equilibrium
Dominant Strategies
Dominated Strategies
Maxirnin Strategies
Mixed Strategies
IN Genie Theory and Oligopoly
Noncooperative Ganies:ThePrisorier's Dilemma
Cooperative Games: Enforcing a Cartel
Repeated Games: Dealing with Cheaters
Sequential Gaines: The Advantage of Being First
N Strategic Behavior
Present versus Future Profits: Limit Pricing
The Value of a Bad Reputation Price Retaliation
Establishing Commitment: Capacity Expansio;i
Preemptive Action: Market Saturation
N Summary
N Discussion Questions
N Problems
11
370
CHAPTER II Game Theory and Strategic Behavior 371
PREVIEW
The di sti n gu i shing characteristic of oligopolistic markets is that managers must con-
sider the effects of their decisions on other firms and must also anticipate how managers
of those other firms will respond. Basically, life in an oligopoly can be interpreted as a
high-stakes game where the objective is to earn economic profits b y outguessing your
rivals. In fact. important insights into oligopolistic markets have been achieved by using
a method of analysis called game theory.
The first section of this chapter introduces basic concepts of game theory. The sec-
ond illustrates how game theor y can be used to understand and predict the behavior of
managers in oli g opolies. The final Section of the chapter considers strategies that are
used to prevent entry in a market.
INTRODUCTION TO GAME THEORY

(lame theory was developed in the 1950s by mathematician John von Neumann and
economist Oskar Morgenstern. The technique was designed to evaluate situations
where individuals and or g anizations can have conflicting objectives. It can be used to
analyze a broad range of activities, includin g dating and mating strategies parlor games.
legai and political negotiations, and economic behavior. For example, in wage negotia-
tions between unions and firms, a primary objective of the management team is to keep
the total wage bill as small as posible, while union negotiators want to maximize wage
payments. In peace talks between two nations at war, each country wants to arrange a
settlement that is to its advantage. During courtship, both men and women may adopt
complicated and sometimes devious strategies based on their individual objectives. In
all of these situations, game theory could he used to analyze the bargaining process be-
tween the two parties.
Over time, game theory has evolved to become a very broad and complex subject.
Consequently, only the most basic elements can be presented here. However, even these
fundamental concepts can provide valuable insights into business behavior. In this sec-
tion, the nomenclature and structure of game theory are introduced. Although the po-
litical and courtship applications of the techniques probably are more interesting, the
discussion is limited to the use of game theory in an economic setting.
\Y^yoff Matrix
At the heart of game theory are the concepts of strategies and payoffs tA s trategy is a
course of action taken by one of the partiipants in a game, and the payoff is the result
or outcome of the strate\9Consider the simple example of two children engaged in a
penny-flipping game. One child tosses the coin and the other calls it. If a head or tail is
correctly predicted, then the caller gets the penny. If not, the other person wins the coin.
In this situation, the strategy is the caller's choice of a head or a tail, and the payoff or
outcome is that cithcr the caller wins a peitnv or the coin Losser wins the coin.
In this simple game, the coin tosser is a passive participant. and the outcome de-
pends on the caller's choice and the result of a random flip of the coin. In more com-
plicated games, the outcome may depend not onl y on the choice made by one person.
but also on the strategies selected by one or more other participants.
372 PART IV Market Structure
Firm 2
No Price Price
Change Increase
No Price Change I 10,10 100, —30
Price Increase —20,30 140,35
Consider A rnrket with two competing firms whose objective is to increase their
profits by price changes. Assume that each firm has two possible strategies—it can ei-
ther maintain its price at the present level or it can increase its price. In this game, there
are four possible combinations of strategies—both of the firms increase their prices,
neither firm increases its price, firm 1 increases its price but firm 2 does not, and firm 2
increases price but the other firm does not.
The results for each of the four combinations of strategies are shown in Table 11.1.
The first number in each cell is the profit for firm 1, and the second value is the corre-
sponding profit for firm 2. Table 11.1 is referred to as a payoff matrix because it shows
the outcomes or payoffs that result from each combination of strategies adopted by the
two participants in the game.
yash Equilibrium
In discussing models of perfect competition and monopoly in chapter 9, the focus was
on determining equilibrium conditions. These conditions were the rates of output that
allowed the firms to maximize profits. Because the equilibriums represented the best
the firms could do, there was no reason to change. For both of these market structures,
the profit-maximizing equilibrium was the rate of output where marginal revenue
equaled marginal cost. Actions of the otlir firms in the market Were irrelevant because
they had an insignificant impact on the firm (perfect competition) or because there
were no other firms (monopoly).
But the payoffs shown in Table 11.1 depend not only on the pricing alternative cho-
sen by each firm, but also on the strategy selected by the other firm If firm 1 decides
not to raise prices, it will have $10 million in profits if firm .2 also does nrel1ange its
price, and $100 million if firm 2 implements a price increase. If firm 1 raises prices, prof-
its will be $ —20 million if the other firm makes no change and $140 million if firm 2 does
increase its price.
The objective of each firm is to do the best it can based on what the other firm does.
Faced with the payoff matrix of Table 11.1, is there an equilibrium result? Suppose that
firm I does not change its price. In this situation, the optimal strategy for firm 2 is also
-no price change because it will earn $10 million in profit, compared to a loss of $30 mil-
lion if it increases prices. Note that if firm 2 does not change prices, the best strategy for
firm 1 is to hold the line on prices.
Thus, for the payoff matrix of Table 11. 1, no price changes is an equilibrium, be-
cause neither firm can benefit by increasing its price if the other firm donot. This re-
suit is referred to as a Nash equilibrium (for mathematician John Nash j! Nash equi-
librium is defined as a set of strategies such that none of the participants in the game
can improve their payoff, given the strategies of the other participan or the game de-
CHAPTER 11 Game Theory and Strategic Behavior 373
No Price Price
Change Increase
No Price Change 10,10 100,-3_J
Firm 1
Price Increase —20,30 140,25
scribed in Table 11.1, firm l's choice not to raise prices is optimal if firm 2 doesn't raise
prices and vice versa. In chapter 10, the equilibrium for a Cournot duopoly was deter-
mined. That resultsor a Nash equilibrium because each firm was producing the op-
timal rate of output based on the output for the other firm. Consequently, neither firm
had any incentive to change.
A limitation of the concept of Nash equthbrium is that there can be more than one
equilibrium.F5iexámple, in Table 11.1, if, JiL 1 were to increase its price, then firm2
would earn more profit by increasing its price than byteavigji price unchanged. Sim-
ilarly, if gm..2-were to initiate a price increase, firm 1 would earn $40 million more by
matching the increase., Tnus both urms increasin g meir mice is also sn eciuiirnnum)
The actual outcome of the game depends on which action occurs first. Because no price
change is the initial condition, the expected outcome of this game would be that prices
would not change. For some games, there may be no Nash equilibrium. In these cases,
pa pants may continuously switch from one strategy to another.
key Concepts
' A strateg' ira course of action that can:hc used by a player ma game.
The payoff matrix indicates the outcomes of all the possible conibinations of
strategies in a game..
• A Nash equilibrium is a set of Strategies such that notof the players in a game V
can imprdve their payoff given the strategies ofthe 6ther prt;cipantt
Dominant Strategies
For the payoffs inTable 11.1, the optimal strategy for each firm depends on the strategy
selected by the other firm. But in some sjtuation4ne firm's best strategy may not de-
pend on the choice made by other participants in iTegame. In this case. that firm has a
dominant strategyonsider the payoff matrix shown in Table 11.2.
If firm 1 does not change prices, firm 2's best strategy is to also make no price ad-
justment. But, based on the profit numbers frornTable 11.2, if firm. 1 increases its price,
firm 2 is still better off with no price change because profit will be $30 million, compared
to $25 million if it increases price. Hence, firm 2's dominant strategy is to hold prices at
the existing level, regardless of what firm 1 does. When one player has a dominant strat-
egy, the game will always have a Nash equilibrium because that player will use that
strategy and the other will respond with its best alternative. For the payoffs in Table
11.2. firm l's best response to no price change by firm 2 is also not to change its price.
In the analysis of games, the first step is to determine if any participant has a dom-
inant strategy. If such a strategy exists, then the outcome of the game should he easily
determined, because the player will use the dominant strategy, and other participants
will adopt their best response. If there is no dominant strategy, the next step is to search
for other Nash equilibriums. If there are none, it may still he possible to analyze the
game using other techniques.
Case Study
t ndiana Jones and the Holy Grail
Among the most popular movies of all time are the Indiana Jones sagas. In Indiana
Jones and the Last Crusade, it is revealed that the hero is a great adventurer but not an
astute student of game theory. Near the end of the movie. Jones, his father, and the Nazis
are at the site of the Holy Grail. Because the pair refuse to help. the Nazis shoot the fa-
ther, knowing that he can be saved only by taking a drink from the sacred cup.
Indiana Jones makes his way to the Grail but finds that it is located among hun-
dreds of other chalices. Drinking from the right cup brings eternal life, but a sip from
any other causes instant death. The Nazi leader impatiently drinks from the wrong cup
and dies, while Indiana Jones makes a well-reasoned (but risky) choice of a chalice,
which does turn out to be the Holy Grail. His father also drinks from the cup and is
healed of his mortal wound.
Jones should be applauded for his courage, but given a failing mark in economics
for not recognizing he had a dominant choice. The best strategy was to give the drink to
his father without tasting it first, lithe cup was the Holy Grail, his father would be saved.
Hence, he would have been as well off as by tryingit himself and then giving the life-
giving fluid to his father. However, if Jonesdrank first and it was not the Hol y Grail, both
would die—Indiana from the liquid and his father from the wound. But if the wrong cup
were given to his father first, he would die - but Indiana Jones would be spared.
Giving the cup to his father first is no worse than tasting it fit-st if the Holy Grail
is selected and better if it is not. Thus, this was the dominant strategy and should have
been selected.
SOURCE: This example is taken from A. Dixit and B. Nalebuff, Think! ig Strategically
1991). p p. 59-60 (New York: Norton
D o minated Strategies
Many games do not have a dominant strategy for any player. In fact, dominance is the
exception rather than the rule. However, it may be possible to simplify a game by elim-
inating dominated strategieA dominated strategy is
Payoff than some other an alternative that yields a lower
iek*.. no matter what, the other players in the gamc d,
Defensive Strategy
Defense Linebackers
Against Run Back Blitz
Offensive Run 2 6
Strategy Pass 8 7 10
Consider a sin gleplay in a football game in which the goal of the offense is to
gain
as many yards as possible and the goal of the defense is to minimize the yards gained.
Assume there is time for just two more plays, and the team with the ball wants to use
the first play to get as close to the goal line as possible to try a field goal. The offense
has two strategies—to run or to pass. The options for the defense are to defend against
the run, to protect against the pass by dropping back their linebackers, or to defend
against a pass by using a quarterback blitz: The outcomes of these offensive and defen-
sive strategies are shown in Table 11.3. Each of the numbers in the cells shows yards
gained by the offense.
Neither team has a dominant strategy. For the offense, this can he seen by noting
that the run gains more ground against the blitz but less against the other two defenses.
For the defense, there is no strategy that will always give up less yardage than the oth-
ers. However, the defense does have one strategy that is dominated. Table 11.3 shows
that the blitz yields more ground against both the run and the pass than either of the
other two defenses. The implication is that the defense should either counter the run or
drop back its linebackers, but should not blitz.
Whenever there is a dominated strategy, the game can be simplified because the
dominated strategy should always be avoided. Thus, b y reducing the number of viable
options, it may be easier to identif y an equilibrium or to use other techniques to ana-
lyze the outcome.
M.aydmin Strategies
Thus far, the analysis of market structures has assumed that managerial decisions focus
on maximizing profits. But in highly competitive situations, such as an oligopoly, von
Neumann and Morgenstern suggested that decision makers might adopt a risk-averse
strategy of assuring that the worst possible Outcome is as beneficial as possible, regard-
less of what other decision makers do. This decision rule is referred to a maxim in
strategy because it specifies that each player in the game will elect the option that max-
i izes the minimum possible profit (or other desirable outcome).
Consider the fo o g example. The two firms in a duopoly are each thinking about
introducing a new product. The profit outcomes for the four possible combinations of
strategies are shown by the payoff matrix in Table 11.4, If the firms are trying to maximize
profit, the matrix has two Nash equilibriums—the two cases where one firm introduces a
new product but the other does not. For example, if firm 1 markets the new product, firm
2 will earn $2 million if it follows suit, but profit will be $3 million if it does not. If firm 2
does not introduce a new product, the first firm will earn $2 million more if it markets the
new product than if it does not. Hence, the bottom left-hand cell of the matrix is a Nash
equilibrium. Using similar logic, the top right-hand cell is also an equilibrium.
IMAM
Firm 2
/
No New New Firm I
Product Product Minimum
No New 4.14 3,6 3
Producr
Firm 1
New 6.3 2.2 j 2
Product
Firm 3 2
Minimum
But the maximin decision criterion is not a pure profit-naximizing strategy. Rather,
it is designed to avoid highly unfavorable outcomes. In applying this principle, each firm
first determines the minimum profit that could result from each strategy it could
choose. As shown in the table, for firm 1, this is $3 million if the firm does not introduce
the new product and $2 million if it does. The numbers are the same for firm 2.
The second step is to select the maximum of the minimums. The result is that nei-
ther firm should introduce a new product because the y will be guaranteed a profit of at
least $3 million by adopting this strategy. Note that the maxirnin outcome is not one of
the two Nash equilibriums. The reason is that loss avoidance rather than profit maxi-
mization was the criterion used for decision making.
Case Study
Texaco and Pen nzoil
In January 1984, the Pennzoil Corporation offered to buy 40 percent of Getty Oil's
stock for $128.50 per share. Getty's board of directors agreed, but then Texaco stepped
in and offered $128 per share for 100 percent of the (Jetty stock.The (Jetty directors re-
versed their approval of Pennzoil's bid and sold the company to Texaco.
Pennzoil promptly sued Texaco for breach of contract. The case was tried in
Texas, and in 1985, a jury awarded Pennzoil $10 billion in damages. Texaco immedi-
ately appealed the verdict, and by the fall of 1987, the case was on its final appeal be-
fore the U.S. Supreme Court. Before a decision was announced-by the Court, Pennzoil
made an offer, which was accepted by Texaco, and the suit was settled in December
1987 for $3 billion.
With some oversimplification, Pennzoil's offer to settle a $10 billion lawsuit for $3
billion can be viewed as a maximin problem. That is, the company wanted to maximize
the minimum possible payment frOm Texaco. Assume there were just two possible out-
comes from the Supreme Court—either the $10 billion jury award would be affirmed
or it would be overturned and nothing would be awarded. Also assume that Pennzoil's
managers had just two options—to wait for the Court's decision or to make an offer
they believed Texaco would accept.
If Pennzoil waited for the Supreme Court decision, the outcome would be $10 bil-
lion if the jury verdict was affirmed and zero if it was not. By settling out of court, the
minimum payment would be $3 billion, Thus, Pennzoil's maximin decision strategy was
to make the settlement offer.
The use of the maxiniin strategy was a logical choice by Pennzoil. If the Supreme
Court reversed the jury award and the firm received nothing, Pennzoil's managers
would have been criticized for not reaching a settlement. Indeed, they could have been
sued for failing to act in the best interest of the firm's stockholders. By offering to set-
tle, they were g uaranteed at least $3 billion and avoided the potential criticism and/or
litigation. U
Mixed Strategies
In all of the games discussed, it has been assumed that each participant selects one
course of action.This approach is called a pure strategy. But in many games, a pure strat-
egy would be a veiy poor choice. For exarnpk, think about the duel bctwccn a baseball
pitcher and hitter. If the pitcher throws all curves or all fastballs, a major league batter
would have a good chance of getting a hit. To be effective, the pitcher must keep the hit-
ter off balance by throwing a mixture of curves and fastballs. This approach is referred
to as a mixed strategy.
Table 11.5 depicts the payoffs for the pitcher and hitter. Low percentages favor the
pitcher, and high values are the objective of the batter. Note that a batter is more suc-
cessful when he can anticipate the pitch, and the pitcher is more likely to succeed if he
can fool the hitter.
If the pitcher throws all fastballs, the batter will look only for this pitch and will hit .400.
Any pitcher who throws only curve balls will also be facing a .400 hitter. Clearly, either of
these two pure strategies would be a disaster for the pitcher. The best strategy is to throw
a mixture of fastballs and curves. Similarly, the hitter cannot succeed by always anticipat-
ing one pitch. Any batter who always goes to the plate looking for a curve will find his av-
erage hovering around .200, because pitchers will throw him fastbaHs and one who antic-
ipates only fastballs will do no better because he will receive a steady diet of curves.
For the payoff matrix shown in Table 11.5, there are no pure strategies that result
in a Nash equilibrium. However, there are equilibrium mixed strategies. If the hitter
randomly alternates between anticipating fastball and a curve on a 50-50 basis and
Pitcher
Throws Throws
Fast ball Curve
Hitter Anticipates Fastball 40% 20%
Anticipares Curve L20 %40%
the pitcher randomly throws a mixture of 50 percent fastballs and 50 percent curves, the
hitter's batting average will be .300. This is because each of the four payoffs in the ma-
trix will occur 25 percent of the time; Thus, the expected frequency of base hits will be
.25(40%) +.25(20%) + 25(20%) + .25(40%) = 30%.
Suppose the batter adopts a strategy other than anticipating that fastballs and
curves are equally likely. If the pitcher continues to throw a random assortment of 50
percent of each pitch, the hitter will guess wrong more often than right, and his average
will drop below .300. Thus, given the pitcher's strategy, the best approach for the batter
is to anticipate a random 50-50 assortment.
In contrast, if the batter continues to ss11me that fastballs and curves arc equally
likely and the pitcher throws a different mix, the hitter's average will still be .300. For
example, suppose the pitcher throws 60 percent fastballs and 40 percent curves. The
probability of each payoff in Table 11.5 will now be
Expect fastball and fastball thrown: .50 X .60 = .30
Expect fastball and curvebafl thrown: .50 X .40 = .20
Expect curveball and fastball thrown: .50 x .60 = .30
Expect curveball and curveball thrown: .50 X .40 .20
Multi p lyin g these probbilitie by the payoffs for each of the cells in the table, the
expected frequency of hits will be .30(40%) + .20(20%) +30(30%) + .20(40%) =
30%. Surprisingly, this result does. not depend on the mixture of fastballs and curves
used by the pitcher. As long as the battercontinues to anticipate that curves and fast-
balls are equally likely, the hitter's average will be .300 regardless of what strategy the
pitcher selects. Thus the game has many Nash equilibriums. When mixed strategies are
allowed, every game, with a finite number of players and a finite number of strategies
has at least one equilibrium.
Mixed strategies can be important in many settings. Sometimes they are used to re-
duce costs. Consider the Internal Reveiue Service and its tax audit procedure. It would
be extremely expensive to audit-,every income tax return, so the IRS chooses individu-
als at random (after using some preselection criteria) for audits. Although the vast ma-
jority , of returns are not audited, the possibility of an audit encourages tax compliance
among the general population.
K&. Deep
• Ira player has i b' oion rLrc'tescfwh 't otl'erpa' Lrs do this i rt
1h as a dominirtt slratcgy. ...
• A j!t'd tIt•(W k ar i 1hr,' :' ' lw.r • :.v o !hn s-i
orhcrLrtgy
e The maxirnin criterion inês ma*i n the tnininu desirabje ouledme of
• A p ure sIrit4'l inyolvas ilsay, inakrnj tb mt. 'inui c'vhir cI:n J

rcq1rL'. ri-idornL' rwu.,c' diftrct J'traqs.e
Caine Theory and Strategic Behavior 379
Suspect 2
Don't Person I
Confess Confess Maximum
15,5 15
Suspect / Confess
Confess L 515 5
Person 2 15 5
KY AND OLIGOPOLY
Game theory can be used to analyze specific Situations faced by managers in oligopolis-
tic markets. This section considers four pecific applications. The first is noncooperative
games, which is used to illustrate how firms in an oligopoly can find themselves in a situ-
ation that is nonoptimal for all participants. The discussion of cooperative games shows
how businesses can work together to avoid such situations, and the evaluation of rcpated
games presents methods for dealing with cheaters. Finally, sequential games are discussed
to illustrate advantages that can come from being the first to act in a business setting.
Noncooperative Gaines: The Prisoner's Dilemma -

A game is considered noncooperative if it is not possible to nego
tiate with other partic-
ipants and enter into some form of binding agreement. When U.S. firms in oligopolistic
markets use pricing strategies to compete for profits, they usually are engaged in non-
cooperative games because they are legally prohibited by antitrust laws from coordi-
nating their prices.
In some cases, noncooperatives games can result in outcomes that are undesirable
for the participants and also for society. One example is the Prisoner's Dilemma. This
model takes its name from the story of two people who were jailed for a crime they al-
legedly committed. The two suspects are separated and interrogated by the police. Each
is told that if she does not confess and the other person does, she will be convicted and
put in jail for 15 years. But if she does confess and implicates her friend, then the jail
sentence will only be five years. However, because the evidence is circumstantial, if nei-
ther confesses, it will be impossible to get a conviction and neither person will go to jail.
The payoff matrix for this game is showi in Table 11.6.
Because the two prisoners are i nterrogated separately. they have no idea whether
the other person will confess or not. Hence, this is an example of a noncooperative
game. If the suspects are risk averse, they may adopt a minimax decision criterion. 'This
is similar to the maximin strategy discussed earlier, except that now the outcomes are
undesirable. Consequently, the minimax approach involves minimizin g the maximum
jail sentence they could receive.
For suspect, 1, the strategy of not confessing involves a maximum jail term of 15
years, while a confession would result in no more than five years in jail. Thus, the max-
imum sentence is minimized by co nfessing. The same is true for suspect 2. which implies
____
Firm 2
Low-Level High-Level
Advertising Advertising
Low-Level $30,$30 $10,40
Advertising -
Y 1
High-Level $40,$10 20, $20 ) (
Advertising
that the minimax solution is for both people to confess. But note that if neither person
confesses, neither will go to jail. Although not confessing is the best joint strategy, be-
cause they are involved in a noncooperative game and cannot influence what the other
person does, they end up with a decidedly nonoptimal outcome.
The Prisoner's Dilemma model can be used to explain a number of interesting phe-
nomena in business. One example is the resource waste from advertising in oligopolis-
tic markets. For simplicity, assume that there are just two firms and that managers can
choose a high level of advertising or a low level. Also assume that the managers are risk
averse and that their objectivc is to maximize the minimum profit earned by their firms.
That is, they use the maximin rule (maximin because the outcome is profit, which is a
desirable outcome). Should managers select the low or the high level of advertising?
The payoff matrix for the four possible combinations of advertising strategy is
shown in Table 11.7. If firm 1 advertises at a low level, its profit will be $30 million if the
second firm also advertises at a low level and $10 million otherwise. Thus, minimum
profit from low-level advertising is $10 million. Table 11.7 shows that a high level of ad-
vertising will guarantee firm I at least $20 million in profit Hence, the maximin strat-
egy for firm 1 is to advertise extensively. Similar logic suggests that firm 2 will also
choose the high advertising option.
Both flrmswill opt for high-level advertising, and the result is a profit of $20 million
for each company. However, note that a joint decision not to advertise would have been
more profitable for each firm because profit would have been $30 million. But neither
firm dares to select this choice because of the possibility that the other might select the
high-advertising strategy and leave it with only S Wmillion in profit.The result is that both
firms earn less profit and waste resources on mutually offsetting advertising. Although
much simplified, this analysis provides insight into the dilemma faced by oligopolists.
The Prisoner's Dilemma is applicable to many game situations. One of the most im-
portant is the arms race between the Soviet Union and the United States. Each country
spent trillions of dollars on missiles and acquired the capability to destroy the other na-
tion many times over. These expenditures came at the expense of other programs, such
as education, medical research, and housing. But neither country was willing to reduce
its spending for fear the other would gain a military advantage.
Cooperative Games: Enforcing a Cartel

In cooperative games it is possible to negotiate and enforce agreements that bind the
participants in the game to a particular strategy. For example, the two suspects men-
tioned earlier were in a noncooperative game because they had no way of knowing
CHAPTER 11 Game Theory and Strategic Behavior
381
what the other person was going to do: If
. they had been allowed to jointly decide on
their strategies and had some means of assuring that the other did not renege,
would have confessed, and they could have avoided spending any time neither
in jail. In the ad-
vertisng version of the Prisoner's Dilemma, if the firms could have signed a binding
contract pledging that they would use the low advertising strategy, the y
each would
have earned $10 million more in profit. With the arms race, years were spent trying to
negotiate an arms limitation treaty. But, because there was no satisfactory way to ver-
ify compliance by the other side, the buildups Continued.
Theaexperience of the cigarette industry during the 1960s is an interesting example
of how nnn cooperati'e game was lurned; by government policy, into a cooperative
game. Historically, the major cigarette manufacturers had spent large sums on TV ad-
vertising to promote their products. Much of this advertising was mutually offsetting, so
the firms probably were in the Prisoner's Dilemma just described. They would have
been better off if the y had reduced their advertising
ex penditures.but no one firm could
afford to do so unless there was some assurance that their lead would be followed by
the other firms.
In 1968, the federal government banned the advertising of cigarettes on television.
Initially, the companies fought the ban, but it soon became apparent that any lost sales
were more than made up by the savings in advertising. In
fact, there is some evidence
that sales actually mci eased because stations were no longer required by the Federal
Communications Commission to air antismoking ads that warned of the dangers of cig-
arettes. With its 1968 ban, the government did for cigarette manufacturers what they
were unable to do for themselv
es---enforce a reduction in advertising.
(' nepts
•' In ti•niop-,.r ', not pwslje to ncgttiii'tt
fliePrisowe Jand CafOr$ ag;:]cr1%
nina imidel illus{ratcs that the
'lCsirablé ,r. zlt f a game cai im-
IC
can -be en1'rc,.'ecL5 ucli 'njj ay bei y

to' avoid the l
Repeated Games: Dealing with Cheaters

Even if enforced a
greement is not possible, firms may be able to escape the Prisoner's
Dilemma if the action is a repeated game; that is, if
it is played many times. Consider the
advertising example shown in Table 11.7. if the game is played once, neither firm will
adopt a low advertising level because the other could select high advertising, capture
most of the profit, and the game would be over. Even if each firm agrees to hold down
advertising, unless there is some way
to enforce this agreement, the high -advertising
equilibrium is likely to occur.
But if the advertising decision is made repeatedly, the outcome may change. A firm
that reneges on an agreement and heavily advertises the first time the game is played
will find that the other firm will respond by increasing its advertising in the second pe-
riod. Thus, the cheater's advantage will be temporary, and profits for both finns will be
low for the second period. In addition, the fact that cheating occurred once should cause
the other firm to be more cautious in the future. With repeated games, reputations are
important in determining the outcome. S
What is the optimal strategy for firms playing noncooperative repeated games? A
general answer depends on the nature of the game, but an experiment by Robert Axel-
rod provides interesting insights.' He proposed a tournament of two-person repeated
Prisoner's Dilemma games with high and low prices as the two alternatives. He invited
world to submit computer programs that embodied
game theorists from all over the . re paired against each other, and the game was re-
strategies for playing. Program s we
peated a large number of times.
The result was surprising. e winning strategy Th h was an approac
described as"tit-
for-tat." Basically, it was that each firm should mimic its rival's behavior from the pre-
vious period. If one firm cheats and cuts prices, the other firm responds by cutting prices
raises
period. If one firm cooperates by raising prices, the other firm also .
in the next
prices. The experiment was later repeated with a larger number of contestants, but tit-
for-tat again proved to be the optimal strategy.
In the real world, the advantage of tit-for tat is that it embodies four principles
that are important in any good strategy. First, it is simple, reducing the chance that it
will he misunderstood. Second, titfortat, never initiates cheating, which could cause
owing such be
in cooperation Third it never rewdrd; chea'ing by al l
oreakdovvii
ahavior to go unpunished. Finally, tit-for-tat is forgiving because it allows cooperation
to be quickly restored.
Unfortunately, tit-for-tat can break down if the players know for sure how many
times the game will be repeated.The explanation is quite simple. Although cooperation
can increase profits, a firm can gain even more if it cheats the last time the game is
But the same is true for the other
played,beCaUse there is no opportunity for retaliation .
player, so both firms should cheat on the last play. If cheating will occur on the last
round, there is no reason to coopere on the next to the-last round; and so on. The re-
suit may be that each play of the game degenerates into a Prisoner's Dilemma.
Sequenhial Games:
The Advantage of Being First
In the games described so far, it has been implicitly assumed that both players reveal
sequential game, one of the players acts first and
their strategies simultaneously . In a
then the other responds. Entry into a new market is an example of a sequential game.
The new firm decides whether or not to enter, and the existing firm then decides
whether to ignore the new firm or try to prevent entry.
For sequential games, there may be an advantage to the player who acts first. Think
about two firms contemplating the introduction of nearly identical new products. The
first firm to get the product to the market is more likely to be successful because it can
develop brand loyalties and may be able to associate the product with the firm in the
minds of consumers. Also. if consumers invest time in learning to use the product of the
first firm, users wit! be less willing to retool and use a similar product from some other
York: Basic Books. 1984).

1 R. Axelrod, The Evolution of Coo peruiu'n (Ncw
CHAPTER I Game Theory and Strategic Behavior
383
Firm 2
No New Introduce
Product New Product
---------.-.--
No Nets, S2$2
Rrin 1 Product
Introduce I $it, S-S
New product $-7,$-7
supplier. Word processing and spreadsheet software are good examples. Users who are
proficient in one program are unlikely to switch to other ,
nificant advantages. programs unless they offer sig-
The advantage of moving first is shown by Table 11.8. Each firm is faced with the
decision of whether or not to introduce anew product. Assume that the firms use the
maxirnin criterion. If the f
m ultaneously, the maximin irms must announce their decisions independe ntly and Si-
criterion specifies that neither should
uct and that each firm will earn $2 million. introduce a new prod ..
Now assume that firm I has research and development advantages thai give it the
O ption of introducing its product first. With firm 1 already in the market, firm 2's opti-
mal strategy is to stay out because it will lose only $5
it enters. Co million, compared to $7 million if
nsequently, firm I will earn $10 million as the only supplier. Clearly, firm 1
has benefittecj by being the first mover in the market.
STRATEGIC BEHAVIOR
In the previous chapter s
barriers to entry wre identified as probably the most impor-
tant determinant of market structures But the tour traditional b
scribed in that chapter (e.g., e a rriers to entry de-
scarce in conomies of scale, product differentiation, control over
puts, and legal factors) are primarily the result of basic conditions that exist in
each market. As such, the number of firms in an industry would be determined by how
those conditions affect entry in each industry. Over time, markets where entry was not
difficult would have many sellers, and those with significant barriers to entry would be
dominated by a small number of firms.
But this passive view of barriers to entry is too simplistic. Businesses are run by
managers, and these individuals will react a
ggressively if they believe that entry could
significantly affect the profitability of their firms, These
reactions may take the form of
Price, cost per unit

($)
PL
Quality per
period
Qe Q
strategic behavior designed to deter entry. Although it is not possible to describe all of
the options available to managers, this section considers some of the strategic responses
that may be used to thwart entry.
Present versus Future Profits: Limit Pricing

1, it is assumed that long-run profit maximization is
Based on the discussion in chapt er s achieved by maximizing the present value of
the goal of managers This objective i
profits over some planning horizon. To this point, however, the analysis of managerial
pricing decisions has focused on maximizing profit in a single period. But the price that
maximizes profits in one period may not be consistent with long-run profit maximiza-
tion. In this section, long-run pricing strategies are considered.
Baja suggested that setting prices to limit entry describes the pricing practice of
His model assumes that monopolists or firms in an oligopoly pursue a pric-
many firms.2
ing strategy designed to prevent new firms from entering the industry. The approachDD, is
illustrated by Figure 11.1. Consider a monopolist facing the market demand curve.
and assume that increasing returns to scale provide a cost advantage for large firms.ThiS
in Figure 111_ by an average cost curve that is downward
cost advantage, is depicted
sloping to the output rateQ.
To maximize profit in a single periods the monopolist should increase production
until marginal revenue equals marginal cost- This implies an output rate of Q,, and a
ill earn economic profit for
ptheiceperiod. At that price and outp ut rate, the monopolist w
of PrnBut unless there are substaiflial barriers to entry, the lure of pi ofit will cause
J. S. Bain.'A Note on Pricing in Monopoly and Oligopoly." American Economic Review (March 1949).
2
44g-464.
CHAPTER I Game Theory and Strategic Behavior
385
Profit
(S)
Time
new firms to enter the industry. A potential barrier to entry is the cost advantage of the
large firms. That is, because of scale economies, the monopolist producing at Qm will
have lower costs than a new firm that will produce at output rates of less than Qm.
ever, by setting the P r Flow-
ofit-maximizing price, the monopolist makes it possible for new
firms to enter on a relatively small scale and still earn at least a normal profit. For ex-
ample, suppose a potential entrant believes that the monopolist will not reduce its price
if a new firm enters the market. Thus if the price is Pm, a new entrant could earn a nor-
ma] profit by producing at an output rate as low as Q. This
can be seen by observing
that at ;he output rate Qe, the price, P,,,, equals average cost.
Although a single new entrant may have little effect on the rate of profit earned by
the monopolist, if the managers of the monopoly continue to use a pricing strategy that
maximizes short-run profit, it is likely that additional firms will enter the market. Over
time, the increased competition will force the firm to reduce its price and will also re-
duce the market share and economic profit of the firm.
Alternatively, the firm could utilize a pricing strategy designed to prevent entry.
This approach would require that in each period, managers set a price below the level
that maximizes short-run profit. For example, as shown in Figure 11.1, the price might
be Set at PL .
At this price, new firms entering the industry and producing at output rate
less than QL
could not
Tier to small-scale entrearn a normal profit. Thus, the pricing strategy establishes a bar-
yAlthough new firms still may enter the market, the size re-
quirement makes entry more difficult and therefore less likely than if the short-run
Profit-maximizing price, m, had been set.
Using entry -limiting pricing
managers alter the firm's profit stream, as shown in Fig-
ure 11.2. Economic profit resulting from setting the profit-maximizing price in each pe-
riod is depicted by the profit stream labeled I. Note that economic profit is substantial in
the initial period but declines as new fixs enter the market. In contrast, entry-limilitig
386 PAR'r IV Market Structure
Profit Stream I Profit Stream It

Period (millions of dollars) (millions of dollars)
1 80 30
2 40 50
3 20 50
pricing aenerates a relatively constant profit stream over the planning horizon. In fact, if
the market is growing. profit may actually increase over time.This is shown by the profit
stream labeled II.
In contrast to entry-limiting pricing is Stigler's open oligQpoly model- 3 Stigler ob-
served that the objective of managers is to maximize the present value of profit. In some
cases, this may be achieved by setting a price designed to deter entry. But in other cir-
cumstances, profits will be maximized in the long run by setting a high price that pro-
vides substantial economic profit in initial periods and then allowing profit to decline
as new firms enter the market. In its extreme form, the open oligopoly model might in-
volve setting the profit-maximizing price in each period. In this case, the profit stream
would he as shown by i in Figure 11.2.
The optimal long-run pricing strategy is the one that maximizes the present value
of profit. Examining the two profit streams in Figure 11.2, it is not obvious which should
be chosen. Profit stream I has a high initial profit, but earnings decline rapidly over time.
In contrast, profit stream II has less initial profit but shows a slight increase over time.
Basically, the optimal strategy depends on the discount rate used by managers to
determine the present value of profit.A high discount rate gives less weight to profits
in the future. Thus, a high discount rate would cause the present value of profit stream
I to be relatively greater. Conversely, low discount rate would be more favorable to
the entry-limiting approach associated with profit stream II.
The importance of the discount rate in selecting a pricing strategy can be illustrated
by a simple example. Suppose that the two profit streams involve only three periods and
the annual profit is as shown in Table 11.9. The first profit stream shows rapidly de-
creasing profit and is consistent with the open oligopoly model. Profit stream II has con-
stant profit, as might result from entry-limiting pricing.
Assume that the discount rate is 20 percent and profit is received at the end of the
period. In this case, the present value of profit stream I is $106 million, which is greater
than the present value of II, which is $105 million. But if the discount rate is 10 percent,
the result is reversed. Now the present value of II is $124 million, whereas that of I is
only $121 million.
Clearly, the appropriate long-run strategy depends on how managers perceive fu-
ture profits. Managers with a short planning horizon and those who view short-term
profits as paramount would be more likely to behave in a manner consistent with the
open oligopoly model. Conversely, those who have a longer time horizon and use a
lower discount rate would be more likely to pursue entry-limiting pricing.
G. J. Stigler, The Theory of Price (Ncw York; Macmillan, 1987), chapter 13.
CHAPTER 11 Game Theory and Strategic Behuvior
387
Gaje Study
Saudi Arabia and World Oil Prices
The Organization of Petroleum Exporting Countries (OPEC) is a cartel consisting of

13 of the world's most important oil-producing nations. Oil ministers from each mcm-
er of the alliance meet on a regular basis, and their primary objective is to influence
world oil prices by establishing maximum production quotas fo the c.ai ei and foi each
country. When oil prices are low, reductions in OPEC production can significantly in-
crease world oil prices. Conversely, if OPEC members increase their supply, prices tend
to decline.
But the history of OPEC suggests that agreements on production levels have of-
&en been difficult to formulate. Although there are exceptions, countries such as Iraq and
Iran have usually advocated lower prodIction quotas, which would result in high world
oil prices. In contrast, Saudi Arabia has t raditionall y
supported more moderate prices. It
is sometimes alleged that the Saudi position is politically motivated by that nation's ties
to the United Slates, which is a large im p orter of oil. Althou g
may bc a factor, the limit-pricing model can also be used to explainh polilical consicleral infls
Saudi policy.
Saudi Arabia is the world's largest petroleum producer. Its known reserves of 162
billion barrels represent one-fourth of all known, oil resources, and the present rate of
production can be maintained for over 50 y ears. In normal times (the Persian Gulf War
was an exception), Saudi oil revenues are sufficient to sustain a high standard of living
for Saudi citizens. In contrast, some other Middle East oil producers, notably Iran and
Iraq, have an immediate need for increased revenues to rebuild their economies, which
have been ravaged, by war and embargoes.
By supporting the advocates, of high prices, Saudi Arabia could increase its
short-term oil revenues. But Saudi policymakers must also consider the long-term ef -
fects of high prices. Economic theory predicts that higher prices will increase quan-
tity supplied because they will allow high-cost oil to be recovered that was not prof-
itable at lower prices.
The problem with higher prices is that oil is just one part of the global energy mar-
ket. As petroleum prices increase, consumers will shift to other forms of energ
y, such as
nuclear, natural gas, and coal. Higher oil prices may also allow other forms of energy,
such as solar, geothermal, and nuclear fusion, to be economically viable- in addition,
consumers will respond to higher prices by finding ways to conserve on energ y use.
Thus, the long-term effect of higher p etroleum prices ma y
because of s be reduced demand for oil
ubstitution, If large capital expenditures have been made to develop other
energy sources, markets for oil may be permanently lost, even if prices decline.
Saudi Arabia's oil reserves are intended to provide the bulk of that nation's export
revenues for many years. Thus, Saudi decision makers must adopt pricing policies that
will preserve the future demand for petroleum. A limit-pricing approach. which makes
it less profitable to substitute other energy forms for oil, is consistent with this objective.
In contrast, for countries such as Iran and Iraq, which have critical needs for increased
short-term oil r
evenues, the open oligopoly approach is a rational pricing strategy. •
38 PART 1V Market Structure
The Value of a Bad Reputation: Price Retaliation

The purpose of limit pricing is to reduce the likelihood of entry by keeping prices at a
low level over a long period of time. Another strategic response to the threat of entry is
to retaliate b y reducing prices when entry actually does occur or if it appears imminent.
When the perceived danger has diminished, prices can be increased to whatever level
management views as appropriate for market conditions..
An interesting example of price retaliation involved General Foods and Proctor &
Gamble. In the early 1970s, General Foods, with its Maxwell House brand, had a 43 per-
cent market share of the noninstant (ground) coffee market in the eastern United
th 1 A
-.A Proctor Q. '
-
LdLS.
•--
U11U ti aiii -
seller in the West, but was not distributed in most areas of the East.
In 1971, Proctor & Gamble began to advertise and distribute Folgers in selected
eastern cities. This effort was initiated in General Food's YongstoWfl, Ohio, sales dis-
trict. which included the cities of Cleveland and Pittsburgh, General Foods' response to
the new entrant- was to increase its advertising and cut prices for Maxwell House coffee
in this region- At times, the price was actually less than the price of producing the cof-
fee. Profit figures indicate the impact of the price cuts. Profits as a percent of sales
dropped from a preentry level of plus 30 percent in 1971 to a negative 30 percent in
1974. When Proctor & Gamble reduced its promotional activities for Folger's coffee in
rtie area, the price of Maxwell House coffee was increased, and Geucial Foods' profit
rates quickly returned to their previous level.
General Foods also responded to entry in its eastern markets by aggressively re-
ducing prices in midwestern cities where Maxwell House and Folger's were both being
marketed. When Proctor & Gamble moved into the Youngstown region, General Foods
Kansas City. When Proctor & Gamble became
cut prices and increased its advertising in
less aggressive in Youngstown, prices and promotional efforts for Maxwell House were
allowed to return to prior levels in the Midwest.
Reducing prices every time entry dccurs or appears likely to occur would be a costly
proposition for the existing firms in a market. But a few applications of this strong med-
icine can have a significant preventative effect. If a firm establishes a consistent pattern
of reacting to entry by drastically reducing prices, then potential rivals may become con-
vinced that they will face the same response and decide not to compete. Th us, by firmly
establishing a reputation for dealing harshly with all new entrants, the firm rny create
an effective barrier to entry.
Example Evaluating Price Retaliation

Wild Tides is a water park in Southern Georgia.The firm has fixed Costs of $500,000 per
year, a variable cost per customer of $5, receives an average of $15 in revenuc per ad-
mission, and has 100,000 customers each year.
- At present, the firm is the only water park in the area, but Wild Tides management
has learned that another facility is being planned for a nearby cbmmunity. After the com-
peting park opens. management expects that to get 100,000 patrons to Wild Tides each
$12.50.
y ear, prices would have to be reduced and that the average revenue would drop to
Management believes that if it immediately reduces prices and keeps them at a
lower level for 2 years. it can prevent the other water park from ever opening. The nec-
esary price cut would venuereduce
to $8 but wouldaverage
increase attendance
reto
CHAPTER .11 Game TheorY and Strategic Behavior
389
120,000 customers per year. After two years. prices would be raised to their previous
level and revenues would again average $15 for 1 (
)0.000 patrons.
Assume that Wild Tides uses a discount rate of 10 percent and a planning horizon of
10 years and that profits are received at the end of each year. Also assume that if prices are
reduced for 2 y ears, there will be no entry for the
Should the firm cut prices? r emaining 8 ears of the p1annin pe
riod
Solution The optimal strategy can be determined by
c omputing the present value of
Profits for each alternative. If prices are not reduced and entry occurs, revenues will be
$1.250.M) per year and variable COSiS
able cost and the $500,000 will be S00,000 per year. Subtracting out the vari-
count rate and a 1 0-y fixed cost, profit per year will he $250,000. Using a 10 percent dk-
ear planning horizon, the present value of profits would be $1,536.1
With price cutting to prevent entry, revenues will be $960,000 (i.e.) 42
and variable costs will be $600,000 (i.e. $5 X 120.000). 120. 00(
Su btracting the $500,000 fixed
cost, profit in each of the first two years will be $-140.
Planning period, revenue per customer 000 For the next 8 years of the
will be $15 for each of the 100.000 atteiidees and
profits will be $300,000 per year. When these amounts are d
iscounted, the present value
of profits is $1,961,540. Because price retaliation gives a greater present value of p,-
its, it is the optimal strategy. ()f-
-
Key Concepts
• Limit pricing involves setting lower prices, which

• A firm that alway make small-scale en try difficult
s retaliates with price
trant may use its reputation to deter entry. reductions when faced
1 . with anew en-
• in evaluating Limit pricing and price
r ea]iation, the present
suit ing from alternative strategies should be considered. value of profits cc-
Establishing Commitment: Capacity Expansion

The threat of price retaliation against" new entrants ma y
not he credible if existing firms
are unable to produce enough output to meet extra demand resulting from lower prices.
In anew
ing rapidly growing market in particular, a new entrant may be able to survive by serv-
c
ustomers that the existing firms cannot supply with their present production ca-
pacity. A strategic response by established firms to prevent this from occarring would be
to invest in additional capacity Once this investment has been made, it becomes a sunk
cost and places existmg firms in a position to expand their production at relatively low
cost. The
existence of excess capacity provides a strong signal that the established firms
can (and probably will) reduce prices as a strategic response to entr y
Inve in their market.
stment in excess capacity reduces the profits earned by an existing firm. Hence,
this investment will he undertaken Only if
ma nagement believes that the certain and im-
mediate loss of profit from making the
investment is less than the expected future profit
loss resulting from entry. Table 11.10 illustrates this trade-off. Suppose a
monopolist
mustdecide
must choose between building a small plant and a large plant and that a secd firm
wh
ether to enter the market. Profit outcomes for each firm for eac on h of our
Possible scenarios are shown in the table
- - Entrant
New n Pm t t -, monopolist
Small plant
$4 million $4 million
Entry
No
N entry $0 million $12
Large plant
g-4 million $-4 million
Entry
$0 - $8 million
No entry
If the monopolist builds the small plant, the competitor will enter because its profit
will be $4 million, as opposed to zero if entry does not occur. But the larger plant will
l meet total market demand.
give the existing firm the ability to reduce its prices and sti l
Consequently, the new firm will not be able to cover its costs and will experience a loss
of $4 million. Thus. if the large plant is built, the better strategy for the new firm is to
stay out of the market.
For the monopolist the better strategy is to construct the small plant, restrict out-
put, and continue as the only supplier. But this option is not viable because it will inis
duce entry, and the monopolist'S profit will be only $4 million But if management
confident that the large plant will deter entry, its construction is the best strategy for the
monopolist because profit will be $8 million.
Preemptive Action: Market Saturation

Just as the total amount of productive capacity can affect the rate of entry, the geo-
graphic location of that capacity can also cause barriers to entry. When costs of trans-
porting a good are high relative to its value, consumers who are not close to a produc-
tion facility may be required. to pay substantially higher prices to have the good
delivered to their location. Thus, firms that locate closer to those consumers will have a
cost advantage and should be able to attract those customers.
Thissituation is depicted in Figure 11.3a. A monopolist has a production facility at
as shown. Although the monopolist may be able to reduce prices and prevent
point E, may allow new entrants located at point
entry near point E. high transportation c osts
7.
CHAPTER 11 Game Theory and Strategic Behavior
391
Ni to underprice the m
near N1 . Having S onopolist and capture the demand of cu
uccessfully
s upplier may, over rime, be ableentered on the geographic fringestomers who are located
to e of the market, the new
in othel- areas, such as N2. xpand i position and challenge the existing firm
One entry-deterring strategy
for
the existing firm would be to disperse its prod
tion facilities, as shown in Figure 11.3b. By the existing firm
out the market area, the o spreading its plants through.
uc-
tr ansportation costs is greatlpportunity
y for the new entrant to take advantage of high
benefits of eco nomies of scale reduced. Although the m
from this dis onopolist may lose some of the
likelthood of entry may ca use ma persion, the ability to prevent or reduce 4he
is a viable long-run strategy, nagement to decide that geographic mark et saturation
The analysis of geographic saturation can also be applied to product characteristics.

Now consider the circles of Figure 11.3 as rep
produci such as aut resenting possible char acteristics of a
omobiles Cars may be small, medi
hardtop; loaded with acc um-sized, or large; co nvertible or
essories or spatan; high Performance or fuel-effj
Point E in Figure 11.3a represent am cjet Let
onopolist Producing a single type of car. Pui ex-
ample, in the early days of the U.S. automobile i ndustr y, Henry Ford only produceci the
Model T and allowed his Cu
wanted black. Initially. Ford experienced great any color they wanted as long as they
stomers to select
com pan y SUCCeSS in the market, hut
eventually the
's market share
tures that were not a v declined as its competitors introduced models with new fea-
ailable on the Model T. This entry is shown in
Points N1 , and N2 on the fringes df the diagram. Figure 1 L3a, by the
Ford's d
ominance of the market might have been extended if it had opted for the
strategy illustrated by Figure 11.3b of filling the market characteristic space with a va-
riety of models. If Ford had provided more options at the outset, new firms would have
had more difficulty carving out their positions in the market.
H owever, another mariu-
facturer and
mobiles first recognized the potential of producing many different varieties of auto-
r
esponded to saturate the product space. Cadillacs were targeted for the
luxury market, Chevrolets for the low-price buyer, and Buicks,
Old smobiles and Ponti-
acs for the middle-price range. Each make was available in several models, and options
could be selected to meet the specific wants of individual Customers. As a result.
lost market share and General Motors Ford
man ufacturer in the United States. for many years was the d ominant automobile
Case Std
Brand Pr oliferation in the Cereal Industry
A perusal of grocery store shelves indicates the large number of brands that are avail-
able from each of the three major cereal
Foods, and Kellogg's. For e manufact urers__General Mills, General
logg's includes xample, just a partial list of the cereals produced by Kel-
392 PART IV Marker Structure
Kenmei Rice Bran
Corn Pops Frosted Bran
Double Dip Crunch
Frosted Flakes All Bran
Nut and Honey Os Low Fat Granola
Mini Buns Smacks
Just Right Apple Jacks
Mueslix Bran Flakes
Corn flakes Raisin Squares
Fruit Loops Raisin Bran
Rice Krispies Cracklin' Oat Bran
Crispix Frosted Mini-Wheats
Spec - Product 19
The cereal industry is of particular interest because an antitrust case was filed by
the Federal Trade Commission alleging that the major manufacturers had attenpted to
prevent entry into the market by engaging in the "proliferation of brands." In this case,
the issue involved not only filling the product market space with cereals brands of every
conceivable type, but also saturating shelf space.
In modern supermarkets there is competition for scarce shelf space. Only a cer-
tain amount of area in a store is allocated for cereals. Thus, if more of that space is used
for the brands of the major manufacturers, then less will he available for cereals from
smaller companies. Because they had filled the shelf space with their products, General
Foods, Kellogg's. and General Mills were accused of saturating the market and creating
different brands, the FTC also argued that
a harrier to entry. Beca'use they had so many
product market saturation tituted a barriercons
to entry.
The case attracted considerable attention from politicians, who felt that pursuing
this issue was a waste of money. Ultimately, the FTC dropped the case. Still, the episode
illustrates the concept of market saturation. U
KEY Concept5
as harner to entry because it 9110ws
• 1iiestment In additkSflal capacity may'act
to iteaSe produCtlon at low variable, cost-
a firm macke with pro-
entryby atnxatiflg the geographic
a Firms rnaye able to deler
duction facilities orth product utt ith braids model-s.
SUMMARY
Game theory is a technique used to analyze situations where individuals or organizations
possible com-
have conflicting objectives. Payoff matrices indicate the outcomes of all the
binations of strategies in a game. A pure strategy involves always selecting one course of
action, while a mixed strategy involves one participant randomly using different strategies.
A Nash equilibrium is defined as a situation where none of the players can improve their
payoff, given the strategies of the other players. A game may have more than one Nash
equilibrium. If mixed strategies are allowed, every game has at least one equilibrium.
If a player has a strategy that is best, regardless of 'what other players do, this is called
a dominant strategy, and it should always be used. Dominated strategies always yield a
implified by elimi-
lower payoff than some other strategy. The analysis of games can be s
CHAPTER ii Game Theory and Strategic Behavior
393
rtating dominated strategies. Risk-averse decision makers may use the maxirnin criterio
which involves maximizing the minimum profit or other desirable outcome. If the outcomen,
is undesirable, the minimax strategy would be the corresponding choice. This criterion spec-
ifies th'at the strategy to be chosen. should minimize the maximum undesirable outcome.
A game is no ncooperative if it is not possible to negotiate with other players and en-
ter into a binding agreement on the strategies to be employed. Noncooperative games may
result in undesirable equilibriums. One example. the Prisoner's Dilemma, can be used to
explain many phenomena, such as over advertising by oligopolists. The Prisoner's Dilemma
outcome may be avoided in cooperative games where agreements can be enforced.
A study by Axeb-od deteiuiinej t1litt1he most effective strategy in repeated games
may be tit-for-tat, which involves always selecting the choice used by the other player
in the previous round. In sequa,al games, where one player acts first and then the
other responds, there may be an advantage to being first.
Firms QAca engage in strategic behavior to deter new entry. One approach is to
keep prices at a low level, which makes small-scale entry more difficult. Price retalia-
tion when entry occurs or if it appears likely is another strategy. A firm that gains a rep-
utation for aggressive price cutting ma y be able to effectively discourage entry. In eval-
uating both of these strategies compared to a policy of setting the monopoly price and
letting entry occur, the present value of profits for each option should be comparcd.
The threat to cut prices must be credible to be effective. A firm that invests in ca-
pacity expansion signals potential entrants that it is able to meet the demand that would
result from lower prices. Once the capacity is in place, variable costs of increasing pro-
duction are low. Firms may also be able to deter entr y by situating their production fa-
cilities to saturate the geographic space. Market saturation may also involve filling the
product space with many different brands or models.
Discu&sion Questions
11-1. Consider the cournot model discussed in cpter 10. is the solution to that prob-
lem a Nash equilibrium? Why or Why not?
11-2. If one particjpjg in a two-person game has a dominant strategy, will that person
always receive a greater payoff from the game than the player who does not have
a dominant strategy? Explain.
11-3. Reread
the Indiana Jones case study. How might ethical or personal values have
affected the choice made by Jones?
11-4. Under what conditions might a maximin strategy be a rational criterion for a
manager to use in decision making?
11-5. Suppose a baseball pitcher adopted a strategy of throwin g a fastball, then a curve
, then
a fastball, and so on. Is this mixed strategy? Would it be effective? Wh y or why not?
11-6. How could the Prisoner's Dilemma model be used to explain a price war between
duopolists?
11-7. Would a manager who was two years from retirement be more likely to favor a
limit pricing or an open oligopoly model? Explain.
11-8. How could the retaliation model be used to explain the tactics used by Israel and
other nations in dealing with terrorists?
11-9. In general, how could a firm evaluate whether an investment in excess capacity
to deter entry would be profitable? -
11-10. How would brand loyalty for a product affect the success of a product market
saturation strategy?
Problems
can either reduce their prices or keep them at the present 1vel. If firm
11-1. Pvo firms B also cuts prices, and $20
A cuts prices, it will earn $10 million in profit if firm
makes no price change, it will
million if firm B does not change prices. If firm A makes no price change.
B
earn $0 if firm B reduces prices and $5 million if firm
The outcomes for firm B are the same as for firm A.
a, Develop the payoff matrix for this game.
b. Does the game have a Nash equilibrium?
c. Does either firm have a dominant strategy?. Explain,
11-2. Two firms produce a homogeneous product for which variable costs are zero. The
total out-
market demand for the product is given by P = 100 - 4Qr, where QT'S
put of both firms. Determine the Nash equilibrium output and price.
11-3. Consider the mixed strategy game involving the pitcher and batter described on
pages 377-378. If the hitter considers a curve and a fastball to be equally likely,
prove that any mixed stxategy is optimal for the pitcher. Hint: Set up the problem
in terms of the probability, P, that thc pitcher will throw a fastbafl.
Two manufacturers of the same product must independently decide whether to
The profit payoff matrix is shown here.
\2 7/ build a new production facility.
a. If they both use the maximin decision criterion, what will be the outcome?
/7 b. Does either firm have a dominant strategy? Explain.
Firm 2
Don't
Build Build
Don't
Build 4ii0
Firm
Build 10,0
11-5. A single decision maker has four strategieS, each with four undesirable outcomes
(i.e., bigger numbers are worse), as shown here.
a. If he is highly risk averse, which strategywill he choose? Explain.
If he is risk neutral and knows that each outcome is equally likely, which strat-
b.
egy will he choose? Explain.
Outcome!
1 2 3 .4
A 10 20 30 30
100 8
60 40 40
D 200
11-6. The manager of a firm has four strategies with four possible desirable outcomes
for each, as shown here.
a. If there are any dominated strategies, rewrite the matrix to show the relevant
choices faced by the decision maker.
b Which strategy would a maximin decision maker select? Explain.
Outcomes
1 2 3 4
A
B
L-'° 30
LJ10
C 60 40 40 40 I
D L5 j 1 560 J 25]
11-7. A firm has three strategies and is subject to four states of nature as showxi fol-
lowing. The outcomes are profits earned by the firm.
States of Nature
A B C D
2 74_2 141J
-. I
6 4 iJ
4 [11_20J
a, Does the firm have a dominant or a dominated strategy? If so, which strategy
is dominant or dominated?. Explain.
b. Which of the four strategies is the firm's maximin strategy? Explain.
c. If the outcomes were undesirable, what would be the minimax strategy? Explain.
If the four states of nature are equally likely, what is the expected profit of
each strategy?
If Joe does not confess, he will not go to prison if Sally does not confess, but he
will go to prison for 15 years if she does confess. If J4does confess, he will go to
prison for 3 years regardless of what Sall y does. The same outcomes hold for
Sally depending on what she does and what Joe does.
Write the payoff matrix for this problem.
Does either person have a dominated strategy? Explain.
c. What is the minimax solution to this problem?
If firm 1 does not advertise, it earns a profit of $10 million if firm does not ad-
vertise and 'a profit of $4 mil]on iffjrm 2 does advertise. If firm 1 does advertise,
- it earns $20 million if firm 2 does not adèrtjse. and $6 million if firm 2 does ad-
vertise. 'The same outcomes hold for firm 2 depending on what firm 1 does.
a. Write the payoff matrix for th is prohIen.
h. If manager?e r
isk averse and have no information about the strategy of the
other firm, what will be the strategy for each firm?
11-10. Phyzz, Inc. is considering introduction of a new soft drink. Sparkle Corporation
may in troduce a similar drink. If Phyzz does not market the drink, its profit will
be $100 million if Sparkle does not enter the market, and $80 million if Sparkle
does introduce a drink, If Phyzz enters, profits will be $120 million if Sparkle
does not, and $90 million if Sparkle introduces a drink.
396 PART IV Marker Structure
a. Set up the pay all matrix. If the choices are made independently and simulta-
neously, what should Phyzz's highly risk-averse managers do?
b. If Sparkle moves first to introduce the drink, what will be Phyzz's optimal
strategy? Explain.
11-11. Green Vista golf course charges $20 per round and has 200,000 golfers play the
course each year. Fixed costs for the enterprise are $1 million, and the variable
cost is $5 per round. if a competing course opens in another part of the city, the
price would have to be reduced to $16 to keep the number of rounds at 200,000.
Management believes that entry could be permanently prevented if the price
were reduced to $12 for 3 years. At this price, joutids played would increase to
220,000. Management uses a 6-year planning horizon, for decision making. As-
sume all amounts are received or incurred at the end dfthe year.
a. If the discount rate is 5 percent, what is the optimal strategy for the firm?
b. How would a higher discount rate affect the optimal pricing strategy? What
about a longer planning horizon?
11-12. The manager of the McRod Corporation is considering pricing strategies for its
fishing equipment. One option is to make entry difficult by charging relatively
low prices. A profit stream for the next 5 years using this option is shown below.
The second option is to set higher prices and allow entry to occur The 5-year
profit stream associated with this pricing strategy is also shown. Assume that
profits are received at the end of the year and that the objective of the firm is
profit maximization over the planning horizon.
Profit
Year Limit Entry Allow Entry
1 100 130
2 100 110
3 100 90
4 100 70
5 100 50
Which pricing strategies should be selected if the manager uses
a. A discount rate of 5, percent and a 5-year planning horizon?
b. A discount rate of 10 percent and a 5-year planning horizon?
c. A discount rate of 5 percent and a 3-year planning horizon?
d. A discount rate of 10 percent and a 3-year planning horizon?
11-13. A manager must select between an open oligopoly and an entry limiting strat-
egy. The profit streams for the two strategies are as shown below. Assume prof-
its are received at the end of the period.
Year Limit Entry Open Entry
100 120
100 100
100 70
a. If the manager uses a discount rate of 10 percent, which is the optimal strat-
egy? Explain.
b. What discount rate would make the manager indifferent between the two
strategies?
Pacific Copper
Pacific Cüpper, a family-owned business, produces copper that is purchased by other

firms to make wire, tubing, and sheets. The Copper is produced in
and is i 1 ,000-pound
dentifiable as having been produced by Pacific Copper only by the firm's ingots
name
Stamped on each ingot.
PART I
Pacific operates the only copper mine and smelter in the South Pacific region. Be-
cause imports are limited by high transportation costs, the firm is essentially a mo-
nopoly with respect to the sale of Copper ingots. The only real source of competition
comes from scrap copper that has been melted back into ingot form. However, this
scrap copper is Considered inferior by buyers and sells for
price. / a su bstantially lower
Pacific Copper sells to a
pproximately 200 firms in the region. Individual pur-
chases are typically made by experienced buyers, but orders tend to be small and fre-
quent to allow buyers to keep their inventory costs down. Although Pacific maintains a
published list of prices, it is not uncommon for preferred customers to be secretly
quoted a lower price or better credit terms.
M
anagement estimates that demand for the firm's product is given by the equation
P=3,000—01OQ
where P
is the price per ton and Q is the number of tons sold per year. Regression analy-
sis suggests that the firm's average and marginal cost equations are
AC
2,400 - 0.10 Q + 0.000002 (22
and
MC=2,400_020Q+ 0.000006 Q2
where AC and MC
are average and marginal costs per ton.
Pe
riodically, Pacific advertises in local business publications. Advertising costs
$1,000isper unit, andtothe
period estimated be marginal effect of additional units of advertising on profit per
AIT
= 5,000 - 1,000A
where A is units of advertising per period.
Requirements
1. If the objective of Pacific management is short-run profit maximization, what will
be the optimal price and rate of output? How much profit is being earned?
2. Suppose that copper ore deposits are discovered on nearby islands, and various
area entrepreneurs are considering the establishment of competing smelters to
produce ingots. What factors should be considered by Pacific's managers in adopt-
ing a long-run pricing strategy? Be specific.
PART II
After 2 years eight new firms enter the market. At the present time, Pacific has a 50 per-
cent share of copper ingot sales and the rest of the market is shared equally by the eight
new firms. During a recreational outing in Alaska, the managets of the nine copper-pro-
ducing firms decide to collude aHd 6eL the price of ingot at the monopoly level.
Requirement
Is the price-fixing plan likely to be successful? Why or why not? Be specific.
ilulawl
PricinV
icisions
CHAPTER 12
Pricing of Goods and Services
CHAPTER 13
Pricing and E,np/oynient of Inputs
INTEGRATING CASE
Study 17: Northern Lumber Products, Inc
CHAPTER
• Pric $ods
anjSeps
2 Preview
• Pricing of Multiple Products
Products with Interdependent Demands
Joint Products
Fully Distributed Versus Inctemental Cost Pricing
Ramsey Pricing
Intermediate Products (fransfer Pricing)
• Price Discrimination
- Necessary Conditions for Price Discrimination
Types of Price Discrimination
• Product Bundling
• Peak-Load Pricing
• Cost-Plus or Markup Pricing
Mechanics of Cost-Plus Pricing
Evaluation of Cost-Plus Pricing
Cost-Plus Pricing and Economic Theory
• Summary
U Discussion Questions
• Problems
401
402 PART V Pricing Decisions
PREVIEW
Of the decisions made by a manager, none is more critical to the success of a fin-n than
setting the price of output. The immediate and obvious effect of pricing choices is re-
flected in short-run profits. But prices set today can also have an important impact on
future profits. indeed, pricing decisions frequently are a major factor in determining a
firm's long-term success or failure.
The material presented in the previous chapters makes the pricing decision appear
deceptively simple. if the firm has some control over price, the rule is to produ:e until
marginal re'.enue equals marginal cosi and charge the price indicated by the cemand
curve for that quantity. As always, economic theory involves simplifications of re lity. A
large corporation might produce several hundred products that are sold in many dif-
ferent markets. Sometimes, the price set for one of these products can affect the drnand
for other products sold by the firm. For example, the price the Gillette Corporation sets
for razors may affect the quantity of blades demanded that fit these razors, and vice
versa. Similarly, production decisions relating to one product may affect the manufac-
turing or marketing costs of other products.
This chapter provides a broader perspective for pricing decisions. The first sction
discusses firms with multi ple products and considers both dcmand and Production in-
terdependencies. The second and third sections examine price discrimination and prod-
uct bundling. Section Four considers the advantages of peak load pricing. Finally, the
fifth section introduces cost-plus, or markup, pricing and demonstrates that this prac-
tice is consistent with profit maximization.
PRICING OF MULTIPLE PRODUCTS

Procter & Gamble began in 1837 as a partnership selling soap to residents of Ci ncin-
nati, Ohio. Today, the firm has annual sates of over $30 billion and sells hundres of
products throughout the world. Procter & Gamble's most popular brands are sho' vn in
1'ahle 12.1.
2j
10011 N
Laundry and Cleaning Personal Care
Products Products Food Products
Cascade Bounty Crisco
Cheer Camay Crush
Comet Charmin Duncan Hines
Dash Crest Folger's
Downy Head & Shoulders Hires
Ivory Liquid Ivory Soap Jif
Mr. Clean Luvs rringe s
Spic and Span Pampers
Tide Pepto-Bismol
Scope
Sure
CHAPTER 12 Pricing of Goods and Services
403
Some of these products are unrelated. For example, the demand for
Potato Chips is unlikely to be affected by the price of Tide. Similarly, productionPringle's
costs
of Prigle's are in
dependent of the amount of Tide produced However, demand and
production of other Procter & Gamble brands are in
terrelated. Clearly, Luvs and Pam .
pers would be considered Substitutes by consumers of disposable diapers. As such, the
price of Luvs affects the demand for Pampers and vice versa. Also, the two competing
brands share the same production facilities. Thus, if pricing decisions are based partially
on costs, prices will be dependent on how costs are allocated between the two products.
When firms produce several products, managers must consider the
ships between those interreltinn
in this section, products. Pricing techniques for multiproduct firms are considered
Products with
In terdependent Demands
Products with inte
rdependent demands are either substitutes or co
mplements For sub-
stitutes, such as Luvs and Pampers,a price increase for one good tends to increase the
demand for the other. However, the magnitude of the increase also depends on the
number of substitutes a v ailable f
rom other suppliers. For goods that are complements,
aadded
price increase tends to reduce the demand for the other good. Options that can he
to new cars are a good exanipie of co
mplements Sales of antilock brakes, power
windows, and stereo systems by an automobile manufacturer are dependent on the
number of vehicles sold by the, company. If the price of the basic a
u tomobile is in-
creased, vehicle sales will decline, and the demand for options will also decrease.
Correct pricing decisions require that demand interde
pendencies be taken into ac-
count. Instead of setting each price in isolation, the impact of each price on the demand
for other products produced by the firm must be considered. The basic objective of the
manager should be to determine prices that maximize total profit for the firm rather
than only profit earned by individual products.
When demands are in t
errelated, insight into managerial decisions can be gained by
considering the marginal revenue equations for the products. Consider a firm that pro-
duces only two goods,X and y Assume that sales
o fXhave an impact on the demand
for Y and
ematically asvice versa. In terms of marginal revenue, this assumption can be stated math-
dTR. +
d Qx dQ (12-1)
and
MR. ± g2i
dQ.. dQ (12-2)
Equation (12-1) indicates that marginal revenue associated with changes in the
quantity of X can be separated into two com p
onents, The first, d TR/dQ . represents
the change in revenues for good X resulting from a one-unit increase in sales of
X The second, dTR IdQX reflects the demand good
change in r int erdependency, It indicates the
evenue from the sale of good Ycaused by a one-unit increase in sales of good
X Equation (12-2) has a similar in
of good Y terpretation in terms of a one-unit increase in sales
depend on the
The signs of the interdependency terms, dTR y/dQx and dTRxIdQy,
nature of the relationship between X and Y. If the two goods are complements, both
terms will be positive, and increased sales of one good will stimulate sales for the other.
Conversely, if the two goods are substitutes, the two terms will be negative because ad-
ditional sales of one good reduce sales of the other.
Clearly, the firm must consider demand interdependencies in order to make opti-
mal pricing and output decisions. Assume that goods X and Y are complements. In de-
termining the profit-maximizing rate of output for good X, if the effect of sales of X on
the demand for Y is not considered, output of XPut would be increased only until
.,o man ha eaan from anncstan
no m" r,f nrnrinrinn V -' "
IA/i an,ink tl-, ,nornIet
AP P'X '-A I''
(121), dTR/dQx understates the actual incremental revenue generated which by selling an
is posi-
additional unit of X. Specifically, revenue also is affected by d TR 1dQ,
tive if the two goods are complements. Thus, when demand interdependence is taken
X. In fact,
into accou'ii, profit maximization requires a greater rate of output for good
output of X should be increased until
dTR +
= MC
where MC, is the additional cost incurred by the firm in producing an additional unit
of good X. Similarly, for goods that are substitutes, it can easily be shown that ignoring
the demand interdependency will cause too many units of output to be produced.
Case Study
Turkey Prices at Thanksgiving
For most American families, roast turkey is an important part of their Thanksgiving fes-
tivities. Although pumpkin pie, potatoes and gravy, and cranberry sauce are a tradi-
tional part of the meal, it is the size and taste of the turkey that determine the quality
of the dining experience.
For many years, most families ate turkey only at Thanksgiving and Christmas. To-
day, turkey is recognized as a highly nutritious and relatively inexpensive meat and is
eaten throughout the year. Turkey consumption has, risen from 5.5 pounds per capita
during the 1950s to about 20 pounds per pdron in the late 1990s. However, demand for
turkey still increases dramatically in November and December as families prepare for
the holidays.
Economic theory predicts that when the demand for a product increases, the price
should also increase. This is exactly what happens at the wholesale level. In early No-
vember, turkey producers raise their prices in anticipation of increased purchases by
grocery stores and restaurants. For example, in 1997, the average wholesale price of
turkey was 72 cents per pound. But during the fourth quarter of that year, the whole-
sale price averaged 75 cents.
Higher wholesale prices for turkey should result in a higher price at the retail
level, but this is not the case. Typically. the price that shoppers pay for turke y is actually
405
less around Thanksgiving and Christmas than during the rest of the year. The reason is
that turkey is often used as a loss leader to entice customers into the store. Retailers as-
sume tijat any losses they experience on turkey can be more than recouped as shoppers
buy all of the other items that they need for the holiday season. Basically, the store own-
ers recognize the int
erdependence
ucts they sell. Because between the demand for turkey and the other prod-
of this comple mentary re lationship, turkey prices are kept low.
Joint Products
Products can be related in production as well as demand. One type of production in-
ter
dependency exists when goods are jointly produced in fixed proportions The process
in productionbeef
of producing Eachand hides in a slaughtefhouse is a good example of fixed proportions
carcass provides a certain amount of meat and one hide. There is
little that the sl
aughterhouse can do to alter the proportions of the two products.
cc
product p a produced in fixed proportions they should be thought of as a
goods are
ckage." Because there is no way to produce one pan of this package with
out also producing the other part, there is no conceptual basis for allocating total pro--
duction costs between the two goods. These costs have meaning only in terms of the
product package. This idea is shown by Figure 12. la. Note that the figure identifies two
demand curves, one for hides and one for beef Although the twogoods are
together, their demands are independent However, there is a single produced
for both products. This marginal cost curve
reflects the fixed proportions of production, that is, the marginal
cost is the cost of supplying one more unit of the product package.
Price, cost
per unit ($) Price, cost
per unit ($)
PB
PB
PH
PH 18
uantity
per Quantity
period per
MRE -' period
() No excess production of h(i MR MRB
(ii) Excess P roduction of hides
Where goods are jointly produced, pricing decisions should take this interdependency
into account. Figure 12.1a indicates how profit-maximizing prices and quantities are de-
terrained. MR B and MR11 are the marginal revenue curves for beef and hides. But when an
additional animal is processed at a slaughterhouse, both the beef and the hide become
available for sale. Hence, the marginal revenue associated with sale of a unit of the prod-
uct package is the sum of the marginal revenues. This sum is represented by the line MR 2-
in Figure 12.1a. JVR7, is determined by adding MR11 and MR for each rate of output.
Graphically, it is the vertical sum of the marginal revenue curves of the two products.
The profit-maximizing rate of output, Q,, is determined by the intersection of MRT
and MC The profit -maximizing prices, P,1 and PB, are specified by the demand curves
for each good at output rate Q0.
In Figure 12.1a, note that both MR H and MR B are positiveat Q0 . In contrast, Fig-
ure 12.1b was drawn so that MR, is negative at the profit-maximizing quantity.The im-
plication is that sale of an extra hide reduces the revenues received from that product.
Clearly, in this situation, Q0 cannot be the optimal rate of output for hides. A firm should
never produce at an output rate where marginal revenue is negative. But a problem
arises because the products are jointly produced. Thus, cutting back on production of
hides would mean reduced supplies of beef for which marginal revenue is positive.
In cases such as that shown in Figure 12.1b. the profit-maximizin g choice for beef is
to sell Q11 at a price of F,,. Although Q0 hides will be produced, sales should not be
made beyond the point where marginal revenue is negative. Thus, only Q11 should be
sold and the price set at P11. Unless costs of disposal or storage are high, the excess hides
should be withheld from the market.
Example Calculating the Profit-Maximizing Prices for Joint Products

A rancher sells hides and beef. The two goods are assumed to be jointly produced in fixed
proportions. The marginal cost equation for the beef-hide product package is given by
- MC=30+5Q -
The demand and marginal revenue equations for the two products are
BEEF HIDES
P=60 --1Q P=80_2Q
MR= 60-2Q MR=80-4Q
What prices should be charged for beef and hides? How many units of the product
package should be produced?
Solution Summing the two marginal revenue equations gives
MRTl40-6Q
The optimal quantity is determined by equating MR T and MC and solving for Q Thus
.
- 140-6Q--'30 1 5Q
and, hence, Q 10.
Substituting Q = 10 into the demand curves yields a price of $50 for beef and $60
for hides. However, before concluding that these prices maximize profits, the marginal
revenue at this output rate should he computed for each product to assure that neither
CHAPTER 12 Pricing of Goods and Services 407
is negative. Substituting Q = 10 into the two marginal revenue equations gives $40 for
each good. Because both marginal revenues are positive, the prices just given maximize
profits. Ifmarginal revenue for either product is negative, the quantity sold of that prod.
uct should be reduced to the point where marginal revenue equals zero.
Keyco
ncept s . .::... . .. ..
• If a firm pioduces two goods that are substitutes, the optimal rate of output for
cwi good is less than the rate that would maximize profit if there were no de-
• mand interdependence.
* If a frm produces complementan' goods, the optimal rates of output are greater
than if there were no interdependence.
For products produced jointly in fixed pro portions, output be increased
until the sum of the marginal revenuesequals the marginal cost of the product
package. . ,.
Fu.11y Distributed versus Incremeittal Clost Pricing

Sonic costs are clearly related to the provision of a particular product or service. For ex-
ample, meters on homes are there for the sole purpose of measuring the amount of elec-
tricity used in the house. No one could seriously challenge an accounting system that as-
signed the costs of those meters to residential electricity customers. However, other
costs may not be clearly attributable to a particular product or service. High-voltage
transmission lines fit into this category. Electricity for residential users is transmitted to
urban areas over such lines. These same facilities are also used to serve industrial and
commercial customers. If any two of the three classes of customers stopped using elec-
tricity7 the same high-voltage transmission lines would still be required to continue serv-
ing the remaining customers.
For an electric utility, expenses associated with high-voltage transmission are re-
.erred to as common costs. Because the facilities are necessary to provide service to
ach class of customers, any allocation of these common costs is essentially arbitrary. In
1 act, the concept of allocating common costs is really a contradiction in terms. It in-
olves allocating costs that already have been determined to be conceptually unassign-
a ble to any specific product or service.
Despite the problems, many businesses make extensive use of a practice called fully
dtributed cost pricing.This approach alloctes a portion of the firm's common costs to
each product or Service. That is, all common costs are distributed among the products
and services of the firm.Then the price of each is set so that it covers the design ated por-
tion of common costs plus costs that are directly related to the provision of the product
or service.
As mentioned previousl y, any assignment of common costs must. be arbitrary. The
real problem with this method of pricing is that the choice of allocation scheme may have
an important effect on the price set and, hence, the quantity demanded of the goods and
services provided by the firm.A scheme that allocates a small portion of common Costs
to a product will result in a lower price and greater quantity demanded fo r that product
than will a method that apportions a larger fraction of such costs to the product.
Consider the following example. A firm provides two services, temporary secretar-
ial help and data processing. The firm has $10 million of common Costs that must be paid
even if neither service is provided. Provision of the first service is very labor intensive,
and 80 percent of all labor costs involve the secretarial workers. In cont!:ast, data pro-
cessing is capital intensive, and 80 percent of all capital costs involve this service.
The firm's management decides to base prices on fully distributed costs. Two allo-
cation schemes are proposed. The first is to apportion common costs on the basis of la-
bor costs resulting from each service. The second approach allocates common costs in
proportion to the amount of capital investment that can be attributed 'directly to sup-
plying each service.
Note how the alternative methods would affect the price set for eaci i service. An al-
location based on labor costs would result in 80 percent, or $8 inhl lion of common costs,
being assigned to temporary secretarial help. Thus, the price set for this service would
have to be relatively high to cover the common costs. But the data processing service,
with its relatively smaller labor expense, would be priced lower because only 20 percent,
or $2 million of common cost, would be allocated to that service. Conve:rsely, an appor-
tionment based on investment would have the opposite effect. Because data processing
is capital intensive, a large fraction of common costs wçuld be assigned to that service,
and hence the price set for data processing would be higher and that of temporary sec-
retarial assistance lower than with the first allocation scheme.
Although a firm must recover its common costs, it is not necessary that the price of
each product be high enough to cover an arbitrarily apportioned share of common
costs. Proper pricing does require, however, that prices at least cover the incremental
cost of producing each good. Incremental costs are additional costs that would not be
incurred if the product were not produced. As long as the price Of a product exceeds its
incremental costs, the firm can increase total profit by supplying that product. Hence
decisions should be based on an evaluation of incremental costs.
The contrast between fully distributed andincremental costs in pricing can be il-
lustrated by considering railroad passenger service. Suppose a railroad has a route that
carries passengers between San Francisco and San Diego. The managers are consider-
ing an intermediate stop in Los Angeles. The rails, locomotives, and passenger cars al-
ready exist, so the primary additional expense would be the energy cost of transporting
extra passengers and the establishment of terminal facilities in Los Angeles.
Assume that competition from bus travel between San Franciscà and Los Angeles lim-
its the fare that can be charged by the railroad for the trip. Specifically, the managers of the
railroad believe that the fare cannot be greater than $30. Should the new service be offered?
One member of the management team argues that the decision should be based on
fully distributed costs and that the San Francisco—Los Angeles service should not be of-
fered unless fares will cover direct costs plus a share of the common costs. For a rail-
road, the primary common costs would include the rails between the two cities and the
engines and cars. Based on the common cost allocation method used by the railroad, it
is determined that the fare would have to be at least $40: Thus, with fully distributed
costs as a standard for pricing, the service would not be offered because the fare would
be higher than for bus transportation.
Using incremental costs to evaluate and price the service is advocated by another mem-
ber of the management team. This person's analysis suggests that with the rail lines in place
and trains already operating between San Francisco and San Diego, the additional expense
CHA RYE R 12 Pricing of Goods and Services 409
of' transportm a passenger from San Francisco to Los Angeles would he onl y $15. The
implication is that the service should he offered and priced between $15 and $30 per trip.
clearl y, the decision regarding the service should he based on incremental analysis.
Both the railroad and its patrons could benefit from adding the stop in Los An geles. If
the ticket price is set higher than $15. the total profits of the firm would increase. Trav-
elers between San Francisco and Los Angeles would benefit because an alternative
form of transportation becomes available. Even the customers going from San Fran-
cisco to San Diego could benefit. Although the San Francisco-Los Angeles fare would
not be high enough to cover its full share of common costs, an y price in excess of V5
could provide a contribution to common costs. Thus, the price of traveling from San
Francisco to San Diego could he reduced.
Where common costs are involved, all of a firm's products and services cannot he
priced at their incremental cost. In aggreg1e, prices must be set high enough to allow
the firm to recover its common costs. However, the example demonstrates that it is not
necessary for each product to cover a arbitrarily determined share of those costs. As
long as the price of a new product exceeds its incremental cost, total profit can be in-
creased by providing the product.
The choice between incremental and fully distributed costs is far from academic.
Frequently, poor choices are iriade by managers who insist that every price murt cover
full y distributed costs. Clearly, the proper approach by managers attempting to maxi-
mize profit is to make decisions based on incremental costs.
Case Study
What Constitutes Unfair Competition?
The Louisville and Nashville Railroad filed a petition with the Interstate Commerce
Commission requesting permission to lower its freight rates for a particular route from
Si 1.86 to $5.11 per ton. The firm's objective was to be able to meet competition from a
company hauling freight by truck and barge. The railroad had substantial costs that
could not be clearly allocated to specific routes. Fully distributed costs for the route in
question were $7.59 per ton, but incremental costs were only $4.69. Because the
truck-barge operation had few fixed costs, its fully distributed and incremental costs
were nearly equal—about $5.19 per ton.
The railroad argued that a rate of $5.11 should be allowed because the $4.69 in in-
cremental costs would be cover e d. The competing truck-barge company contended that
the proposed rail rates were unfair because they were less than the $7.59 per ton that
represented the firm's fully distributed costs. In this case, the Interstate Commerce Com-
mission based its decision on fully distributed costs and rejected the railroad's proposal
for a rate rcduction. For oiaiiv years, such decisions by the ICC made it difficult for rail-
roads to compete with other modes of transportation. But since the early 1980s, railroad
shipping rates have been dere g ulated, giving the firms more latitude in price setting.
*ingot Molds, Pa.. to Steelton. Ky.. 326 ICC 77 (196).
Price, cost
per unit ($) Price, cost
per unit ($)
p.r
Px P UCY
I MC
Quantity
Quantity
per period (Q) per period (Q)
Q IQ
Ramsey Pricing
No product should be supplied by a firm unless its incremental revenues are expected
to exceed its incremental cost. If there are common costs, managers must also decide
which products will be priced above incremental cost and how much above. For unreg-
ulated, profit-maximizing firms, the rules'presented in this and other chapters can be
used to make such decisions. For regulated firms limited to a maximum rate of profit
and for nonprofit enterprises expected to just cover their costs, some other approach
may be necessary. One such method is Ramsey pricing.'
Assume that an enterprise produces two products, X and Y, that the demand for
good Xis more elastic than demand for good Y, and that marginal costs are constant.
If X and Y are priced at marginal cost (P and P.,
respectively), then, as shown in
Figure 12.2, quantity demanded will be Q for X and Q y for Y However, because
none of the enterprise's common costs are included in marginal costs, total revenue
will be less than total cost.
For the enterprise to cover its total costs, at least one of the two goods must be
priced above marginal cost. It was shown in chapter 9 that any deviation from marginal
pricing results in allocative inefficiency. If efficiency is the objective, there is a need
for a "second-best" pricing scheme that allows the enterprise to at least break even
while minimizing the adverse effect on resource allocation.
'The name is in recognition of pioneering work by economist Frank Ramsey. See F Ramsey, "A
Contribution to the Theory of Taxation," The Economic Journal 37 (March 1927):27-61.
If an enterprise is providing several goods, Ramsey pricing suggests guidelines for

the price that should be charged for each good, Because all prices cannot equal mar-
ginal costs, the question is how far to set the price of each good above or below mar-
ginal cost* for that good. In its most simple form, Ramsey pricing requires that price de-
viations from marginal costs he inversely related to the elasticity of demand. That is, for
goods with very elastic demand, the price should he set close to marginal cost. Con-
versely, for goods with relatively inelastic demand, the price should deviate more from
marginal cost. In terms of Figure 12.2, the price of X should be closer to its marginal cost
than the price of good Y For example, a price of P. results in quantity demanded of Q.
and P, corceS j ,oDds to Q,. The shaded area in each panel is the amount that each good
contributes to the recovery of the enterprise's common costs. Note that good Y makes
a much greater contribution than good X
The rationale for the Ramsey rule is easy to understand. If demand is elastic, in-
creasing the price causes a substantial reduction in quantity demanded. But if demand
is highly inelastic, large changes in price will result in little change in the quantity de-
manded. In the extreme, if the demand were totally inelastic (a vertical demand curve),
there would be no change in quantity demanded as price increased. Hence, if deviations
from marginal cost pricing are greatest for those goods with inelastic demand, the re-
source misallocation will be minimized.
Ramsey pricing is sometimes criticized because the largest deviations from mar-
ginal cost pricing are imposed on those with the fewest alternatives (i.e., least elastic de-
mand). Although this is true, the niore relevant point is that there, really is no alterna-
tive to using some variant of Ramsey pricing. There is a limit to the contribution to
common costs that can be obtained from the sale of goods with elastic demand. If large
price increases are imposed on such goods, consumption will decline substantially as
buyers shift to alternative goods. The net result is that the firm will obtain little contri-
bution to fixed costs from goods with elastic demand.
Many nonprofit enterprises receive subsidies from tax revenues or charitable con-
tributions. Where such subsidies are provided, revenues from sales do not need to cover
the firm's total costs. However, the concept of Ramsey pricing may still be relevant for
such enterprises. Even with a subsidy, prices set equal to marginal costs may not allow
the enterprise to cover its total costs. However, by using the Ramsey principle, man-
agers could set prices that would recover the necessary amount while minimizing the
adverse effect on resource allocation.
In some situations, if prices are equated to marginal costs, the revenues of an en-
lerprise (sales plus subsidies) might exceed its total costs.This could occur if the subsidy
were very large or if decreasing returns to scale caused marginal costs to be greater than
average total costs. In this case, the organization would be earning economic profit.
However, such an outcome is not generally consistent with the notion of a nonprofit en-
terprise. Thus, there would be a need to reduce prices until total revenue just equals to-
tal cost. As before, this should be done in a manner that minimizes changes in con-
sumption patterns in comparison to mar ginal cost pricing.
The Ramsey pricing rule can be used to achieve this result. The deviations of prices
below marginal costs should he inversely proportional to the elasticity of demand. Thus
prices of goods for which demand is elastic should be priced near their marginal cost. Con-
versely, where demand is inelastic, prices should be lower in relation to marginal cost.
Key Concepts -
• A firm's common costs are those that cannot be assigned to any single product
or service
• The use of fully distributed costs can lead to poor pricing decisionsA product
can be profitably. produced if its price exceeds incremental cost cif . supplying
the product. : :.. 1.. . ....
• Ramsey pricing is a second-best alternative that can be used when marginall cost
pricing is not feasible
i. Acrninle vercinn nf 1imcev nncin p snFclfies that nnce deviaturrns fi-in mr
-' ginal cost should be inversely related to the elastiuty of demand
Yxainple Using Ramsey Pricing

Consider an enterprise that supplies two goods, X and Y For ease of exposition, assume
that the marginal cost of providing each is constant and equal to $10. However, also as-
sume that the firm has common costs of $99 per period and that these costs must be re-
covered. Further, suppose that demand elasticities are —0.1 for Y and —1.0 for X, and
that if prices are equated to marginal costs, 10 units of each product Will be sold each
period. However, note that if prices equal marginal costs, the firm will incur a loss of $99
per period. What prices for X and Y wouldallow the firm to recover its fixed and mar-
ginal costs while minimizing the adverse effect on resource allocation?
Solution The pricing problem is how to increase prices to recover the $99 in com-
mon costs while minimizing the changes in consumption patterns in comparison to
those with marginal cost pricing. The Ramsey principle suggests that since demand is
more inelastic, product Y should be pried higher in relation to marginal cost than
should product X. One simple formulation of Ramsey pricing uses the inverse elastic-
ity rule. This specifies that departures from marginal cost should be inversely propor-
tional to elasticity of demand. In this example, since elasticity for Y is one-tenth that of
X, the deviation of the price of Y from its marginal cost should be 10 times the devia-
tion of the price of X from its marginal cost.
Using the inverse elasticity rule, the solution is to price product Y at $20 and X at
$11. Note that a 100 percent increase in the price of Ywill decrease quantity demanded
by only 10 percent (because the demand elasticity equals —0.1), to nine units. The 10
percent increase in the price of X will decrease quantity demanded by 10 percent (be-
cause elasticity equals —1.0), also to nine units. Now each unit of Y sold makes a $110
contribution to common costs for a total of $90. Each unit of X sold makes a $1 contri-
bution for a total of $9. Together, this pricing approach allows the enterprise to recover
its common costs of $99. This objective is achieved with minimal impact on the pattern
of consumer demand. For both goods, the reduction in quantity demanded is only one
unit as compared to marginal cost pricing.
Intermediate Products (Transfer Pricing)

Vertical integration is common in modern economic systems.A firm is said to be verti-
cally integrated when it operates at more than one stage of the production process. For
example, some steel producers mine coal and iron, transport the ore on boats owned by
the firm, use the coal as an energy source to transform the iron into steel ingots, shape
the steel ipgots into finished products, and distribute those finished products to con-
sumers.The fabricated steel products received by the consumers are final goods. In con-
trast, the coal, iron ore, steel ingots, and undelivered steel products are referred to as tn-
termediaze goods. That is, they are goods or materials that will he needed as inputs at a
later stage of the firm's operations.
By taking advantage of scale economies, avoiding possible supply disruptions, and
bringing complementary aspects of the production process together. a verticall y inte-
grated company may be more efficient than several small firms, each operating at a sin-
gle stage of the productionprocess. However. greater size resulting from vertical inte-
gration can cause control problems. Top management ma y find it difficult to become
familiar with each stage of the operation or a cumbersome bureaucrac y may develop
that makes it difficult to implement decisions.
One method of dealing with such problems is to organize vertically integrated firms
into semiautonomous divisions. Each of these divisions has its own function and its own
management. Each management team is rewarded based on the performance of the di-
vision. In some cases, the evaluation is based on the profit earned by the unit. In an in-
tegrated firm, determining the amount of profit that should be credited to a division
producing an intermediate product is a difficult task. The problem is that if the unit's
function is to provide an input for the next stage of the production process, revenue will
depend on the price that is charged for the intermediate good. A high price will increase
profits of the unit at the earlier stage of production, whereas a low price will make the
later production stage appear more profitable.
A more serious problem is that an incorrect price set for an intermediate good can
affect the total profit earned by the firm. Specifically, if the decision makers in each di-
vision attempt to maximize profits for their units, the total profit of the firm might be
reduced. Thus, it is important that prices of intermediate products be set lo maximize
overall profits rather than division profit.This objective may require thai top-level man-
agement be involved in pricing intermediate goods.
The following discussion provides guidelines for the pricing of intermediate prod-
ucts, sometimes referred to as transfer pricing. For simplicity, it is assumed that there are
only two stages of production. In the first, rolls of paper are manufactured as an inter-
mediate product. In the second, paper is cut into writing tablets and sold to final con-
surners. Two alternative situations are considered. The first occurs when an external
market exists for the intermediate good. That is, the division making rolls of paper can
sell its output to buyers outside the firm, and the unit requiring the paper has alterna-
tiv e sources of supply. The second case is when there is no external market and paper
can be bought and sold only between the two divisions of the firm. In this case. there is
no market-determined price for the rolls of paper.
External Market Assume that the writing tablet division of the inTegrated firm
has the option of obtaining paper from the paper manufacturing division or from in-
dependent suppliers. Similarly, the paper manufacturing unit can sell to the writing
tablet division or to other buyers. Also assume that the external market is perfectly
competitive. This implies that the two units can buy or sell as much as they want at the
market-determined price.
Price, cost
per unit ($)
sir . Mr
Pp
Quantity per
Q1, period
J
7 '
With a perfectly competitive external market, there really is no price decision to be

made.The paper manufacturing Unit, like a competitive firm, faces a horizontal demand
curvd at the market-determined price. If the managers of that unit attempt to maximize
profits, they will increase production until price equals marginal cost, if the paper man-
ufacturing division tries to charge a price in excess of the market price, the writing tablet
Unit should buy paper from independent suppliers. If the final product division is un-
willing to pay the market price, the paper rolls can be sold on the open market.
Where an external market exists, it is not necessary that the output of the paper
manufacturing division equal the input demand for the tablet unit. If there is an excess
supply of paper rolls, it can be sold to other firms. Similarly, if the firm's internal supply
of paper is insufficient, the tablet division can buy from other producers.
No External Market If no external market exists or if the divisions are not allowed
to trade with other firms, a conflict may arise regarding the proper price to be charged
for paper. The paper manufacturing unit may benefit from a higher price, while the di-
vision that makes tablets may benefit from a lower price. However, the goal of top man-
agement is to determine the price for paper that results in maximum profit for the com-
bined firm.
The optimal price of both the intermediate and the final good can be determined
using Figure 12.3. The demand and marginal revenue curves for writing tablets are D
arid MR, respectively. The marginal cost of producing the paper necessary to make a
writing tablet is M, while the marginal cost involved in transforming the paper into a
tablet is MCI. Hence, from the perspective of the firm, the marginal cost of each addi-
tional tablet is the sum of MC,, and MCI, which is designated as MC. Thus, for the com-
bined firm, the profit-maximizing choice is to produce where MR MC., or at an out-
put rate of Q. per period. The corresponding price would be Pa..
Because Q. is the profit-maximizing output of writing tablets for the combined

firm, the price set for the intermediate product (paper) must cause the managers of the
writing tablet division to produce Q tablets and the managers of the paper division to
supply ttri amount of paper consistent with producing Q tablets.
The solution is for top management to require the paper unit to set its price equal to
the marginal cost of producing paper. This directive will cause the tablet division to view
MC = MC,, + MC as its marginal cost curve and select Q,. as the profit-maximizing
quantity and P,.,, as the price. At the same time, by setting a price of F,, for paper, the pa-
per division will supply the exact amount of product necessary to produce Q, tablets.
Key Concepts
• LI the re
is a pen ectly competitive external niarkt for an interrnedia good,
L1!C priCe 01 the stood to approach maruii:l cost, and
&UIi.S W ii cau. te
no maxarial p icing decision will he necessary.
• In the sence f an eternal markei, profit maximi:'ation requrcs that the- !:
pric f .acti 1n -i-mediate go ( &: he set equal to its marginal cost.
PRICE DISCRIMINATION -
Frequently, the same book is sold at a much lower price in South America and Europe
than in the United States. This practice is an example of price discrimination and is de-
signed to increase the total profit of the book publisher. From an economic perspective,
price discrimination occurs when price differences between consumers or markets do
not reflect variations in the cost of supplying the product.2
Sometimes price discrimination involves charging a uniform price when costs dif -
fer. Consider the firm that advertises an all-you-can-eat buffet for $9.95. The first cus-
tomer is a jockey who does little more than pick at the salads. The second is an NFL de-
fensive tackle, who requires 10,000 calories a day just to maintain his body weight. The
uniform price of $9.95 constitutes price discrimination because the cost of serving the
two customers differs markedly.
More commonly, price discrimination occurs when prices differ even though costs
are essentially the same. Physicians' services are an example. For a given treatment,
there is no reason to believe that costs depend on the income of the recipient. Yet high-
income patients are sometimes charged more than the poor for the same services.
Necessary Conditions for Price Discrimination

Three conditions must he met before a firm can successfully practice price discrimina-
tion. First, the firm must have at least some control over price. Obviously, a price taker
in a perfectly competitive market is not in a position to engage in price discrimination.
Second, it must be possible to group different markets in terms of the price elasticity of
demand in each. It will be shown later in the section that firms can increase total profit
by charging relatively higher prices in markets where demand is less elastic.
21he legal interpretation of price discrimination differs somewhat from the economists definition. The
legal status of price discrimination is discussed in chapter 19
416 PART N' Pricing Decisions
Finally, the firm's markets must be separable, meaning that products cannot he pur-
chased in one market and then resold in another. Suppose that a firm has identified two
markets and charges a high price in the first and a lower price in the second. If the two
markets are not separable, price discrimination cannot be successful. Either consumers
will go to the low-priced market and make their purchases, or enterprising individuals
will buy in the low-priced market and resell at a price below that established by the firm
in The high-priced market. In either case, the price differential between the markets will
disappear as prices decline in the first market and increase in the second.
TI
Lae otuvy
Price Discrimination and the Airlines
Finding the. lowest airfare can be a bewildering experience. On any given day, there are
tens of thousands of different fares available. With ' a full 150-seat aircraft flying between
two U.S. cities, it would not be uncommoti for the passengers to have paid 20 different
fares for their seats. In some cases, these differences at least partially reflectamenities as-
sociated with higher price tickets. For exam p le, first-class passengers have more leg room
and better-meals. But in othercases, different prices are charged for the same travel ex-
perience. During the summer of 1998, a regular coach round-trip ticket from Chicago to
San Francisco cost as much as $900, but promotional pricing by the airlines allowed some
passengers to make the same journey for only $300. Once on board the plane, service was
identical—same crampedseats, same bland meals, and same inconvenient rest rooms.
For many years, the airlines have used what they call yield management to in-
crease their profits.This practice involves both price discrimination and marketing.The
price discrimination component is baseo on variations lii price elasticities for different
types of customers. , Typically, business flyers have less elastic demands because they
must meet with suppliers and customers at specific times and in specific locations, Of-
ten, these trips are made on relatively short notice. Airlines take advantage of this situ-
ation by setting higher prices for tickets that do not require advance purchase. In con-
trast, vacation travelers often choose between man y destinations (including some that
do not involve air travel) and plan their trips far in advance. Because these discretionary
travel demands are more price sensitive, airlines advertise, some seats at lower prices if
passengers are willing to buy their tickets 7 to 30 days in advance.
The marketing aspect of yield management strategies involves determining how
many low-priced seats to offer. Although airlines are required to set aside at least some
seats at the promotional price, they have considerable latitude in determining exactly how
many they will all to each flight. Flights that usually are full will not have many mw-
c st seats, while on those that have a history of excess capacity, airlines will offer mans'
o
such Seats in an attempt to draw additional customers. Determining the most profitable
mix of seac prices is a complex and ongoing process for the airlines. Computers are used
to continuously reevaluate' and alter the optimum composition of prices based on the lat-
est information. U is possible for a potential customer to call a travel agent on a Tuesday
and be told that there are no promotional fares available on a flight and for another per
son to call the agent on Wednesday and obtain the low-cost fare on the same flight.
Although airline pricing practices have the appearance of price discrimination,

there are other factors that should be considered. Many low-cost fares involve restric-
tions. The ticket may have to be purchased in advance, a Saturday night stay at the lo-
cation may be required, and the ticket could be nonrefundable. A purchaser of a regu-
lar coach-fare ticket does not face these constraints. Thus, it could he argued that the
tickets represent different services and that the higher price for the regular fare reflects
the additional convenience associated with that ticket.0
pes of Price Discrimination

There are many forms of price discrimination, but the standard method of classification
identifies three types or degrees of discrimination. Their common characteristic is that
they allow the firm to capture part of the consumer surplus that would have resulted
from uniform pricing.
First-Degree Discrimination Figure 12.4a shows the demand curve faced by a mo-
nopolist. The curve indicates the maximum price that can be obtained for successive
units of output. For example, the first unit, Q, could command a maximum price of P,
the second could be sold for a maxtmum of 2, and so on. To simpl i fy the discussion, it
is assumed that marginal cost is constant and equal to average cost.
First-degree price discrimintion involves charging the maximum price possible for
each unit of output. Thus, the consumer who attaches the greatest value to the product
is identified and charged a price of P 1 . Similarly, the consumers willing to pay P2 for the
second unit and P3 for the third are identified and required to pay P2 and P3, respectively.
Price; cost Price per Unit

per Unit ($) ($)
P1
P2
P1
P2
=AC
'P3
f-i
LI
/I\ I
QD
Quantity per
period Qi
Quantity per
period
Qi Q2 @3
() First degree price (b) Second degree

discrimination discrimination
--i-
With first-degree price discrimination, the profit-maximizing output rate is where the
marginal cost and demand curves intersect. In Figure 12.4a, this occurs at QD. At this
point, the maximum price that can be obtained for the product is just equal to the mar-
ginal cost of production. Any attempt to sell more than QD units would reduce profits
because price would have to be less than marginal cost. Conversely, any rate of output
less than QD would not maximize profits because the additional units could be sold (as
shown by the demand curve) at prices greater than the marginal cost.
First-degree discrimination is the most extreme form of price discrimination and
the most profitable pricing scheme for the firm. Because buyers are charged the maxi-
mum pric& for each unit of output, no consumer urp1Iis remains. As defined in chapter
9, consumer surplus is the difference between the price a consumer is willing to pay and
the actual price charged for the good or service.The maximum consumer surplus results
when there is no price discrimination, and price is set equal tomarginaI cost. In Figure
124a, this maximum consumer surplus is shown as the arca of the triangle AP CB. In con-
trast with first-degree price discrimination, there is no consumer surplus because APB
is captured by the firm as economic profit.
First-degree discrimination is not common because it requires that the seller have
complete knowledge of the market demand curve and also of the willingness of individ-
ual consumers to pay for the product. In addition, the seller must be able to segment the
market so that resale between consumers cannot take place-These requirements are sel-
dom met in actual market situations. However, one possible case involves the sale of Trea-
sury bonds by the federal government. In selling these bonds, the government requires
each prospective buyer to submit a sealed bid. Those conducting the auction determine a
minimum bid. All the bids that exceed the minimum are accepted and the bidders are ob-
ligated to buy at the price they indicated in their bid.Thus, through this process, the gov -
ernment attempts to extract the maximum price that each buyer is willing to pay.
Second-Degree Discrimination Second-degree price discrimination is an imperfect
form of first-degree discrimination. Instead of setting different prices for each Unit, it
involves pricing based on the quantities of output purchased by individual consumers.
This is illustrated by Figure 12.4b. For each buyer, the first Q1 units purchased are priced
at P1 .the next Q2-Q1 units are priced at P2 , and all additional units are priced at P3.
In most cases, second-degree price discrimination involves goods and services
whose consumption is metered. Electricity is an example. Many electric utilities in the
United States use a declining-block tariff in pricing electricity A typical tariff might
specify the following monthly rates for blocks of usage:
First 100 kilowatt-hours $0.12 per kwh
Next 300 kilowatt-hours $0.10 per kwh
All additional kilowatt-hours $0.08 per kwh
It should be observed that just because different prices are charged for different blocks
of consumption, it does not necessarily imply second-degree price discrimination. The
high rate for the first 100 kilowatt-hours may be intended to recover the fixed costs
of serving a customer, such as billing and metering. As such, the $0.12 for the first 100
kilowatt-hours may not involve price discrimination. But if kilowatt-hours beyond 300
cost the utility the same amount to provide as those in the second block, then price dis-
crimination is being practiced.
CHAPTER 12 Pricing of Goods and Sertices 419
In addition to electricity, second-degree price discrimination is often used in setting

rates for water, gas, and time-share computer usage. it is also practiced by fast-food es-
tablishments. For example, a seller may offer soft drinks for $1.00, with refills available
at $0.50. This pricing policy reflects the fact that the second drink is less valuable to the
customer and will be purchased only at a lower price.
Third-Degree Discrimination The most common t y pe of price discrimination is
third-degree discrimination. It involves separating consumers or markets in terms of
their price elasticity of demand. This segmentation can be based on several factors.
Often, third-degree price discrimination occurs in markets that are gcographieally
arated.The practice of selling books at a lower price outside the United States is ancx-
ample. Evidently, buyers in other countries have greater elasticities of demand than: do
U.S. buyers. At the same time, Costs of collecting and shipping books make it unprof-
itable for other firms to buy in foreign countries and resell in the United States.
Discrimination can also be based on the nature of use. Telephone customers are
classified as either residential or business customers. The monthly charge for ;phone
located in a business usually is somewhat higher than for a telephone used in a home.
The explanation is that business demand is less elastic than residential cicmand. An in-
dividual without a t elephone may he able to use a pay phone or go to a neighbor's house
to make a call, but for wan y businesses a telephone ts an absolute necessity.
Finally, markets can be segmented based on personal characteristics of consumers.
Age is a common basis for price discrimination. Most movie theaters charge a lower
price for children than they do or adults. But there is no difference in ithe cost of pro-
viding service to the two groups—one seat is required for each patron regardless of age.
The reduced price for children is based on differing demand elasticities. The assump-
tion is that, with less mone y to spend, a child's demand for movies is more price sensi-
tive than an adult's.
Figure 12.5 is used to show how profit-maximizing prices and quantities iire de-
termined with third-degree price discrimination. Consider a firm selling in two mar-
kets, I and 1. To simplify the discussion, it is assumed that marginal costs are equal
and constant in both markets. Dmand is assumed to be less elastic in market I than
in market II. the marginal revenue curves for the individual markets are AIR, and
MR 11 . respectively.
The combined demand curve for the two markets is also shown in Figure 12.5. it
represents the sum of the demands in each of the markets at each price. The combined
marginal revenue curve is also shown and was computed as the sum of marginal rev-
enues in each market. For the firm, the optimal total output is at Q.,-, the point where
MR .,- = MC. Once the profit-maximizing total output has been determined, the next
• step is to determinc the output rates and prices in each market.
Because marginal costs are constant, the decision rule for allocating output is that
the marginal revenue should be equal in the two markets. That is, the extra revenue ob-
tained from selling an additional unit in the first market should be the same as that re-
ceived from selling one more unit in the second market. If the two are not euai, the
firm could increase its revenue and profit b y allocating additional output to the market
with greater marginal revenue. In Figure 12.5, equating marginal revenue to marginal
cost means that Qi units of output should be sold at a price of P1 in market I and Q11
units at a price of P11 in market H.
Combined
Market II Markets
Market I
Price, cost Price, cost Price cost

per unit (S) per unit ($) per unit ($)
At
P11
DI D DT
MR
1M MRII
_
- -
Note that a higher price is charged in market I, wheie demand is relatively less alas-
tic.This relationship is true in general and is easily explained. Where demand isles' elas-
tic, consumers are less sensitive to price, meaning that relatively high prices n be
charged. Conversely, in markets where demand is more elastic, the quantity derr nded
is more sensitive to price. Hence the profit-maximizing price is lower.
eyConceptL : :...
• PrIL ,, $ on oi1rs icn vanatiofl in tX .tor a prO4Ucti
markets does aot refltct van ation in -. .................
• The4fre criteria for sucs.(u1 prc cr1D1rnO1) re mti powcrpaia-
• .jk' fiiarkes, and variatiL$.P densd e1aticiy amçng markets.: -
• .kgx4 p1CC tLou man* foc
iSi by .CU bflC4..

ird-degre tscwnijiati r price in
ie. J.stc dnand. 1
Example Third-Degree Price Discrimination and Profits
A firm sells in two markets and has constant marginal costs of production equal to $2 per
unit. The demand and marginal revenue equations for the two markets are as follows:
MARKET! MARKET 11
P1 = 14 - 2Q1 P11 = 10
MR, = 14 -. 4Q 1MR11 = 10 —2Q11
Using third-degree price discrimination what are the protit-n-Iaxh1iLjng prices and
quantities in each market? Show that greater profits result from price discrimination
than would be obtained if a uniform price were used.
Solution With price discrimination, the condition for profit maximization is
MR, = MR,, = MC
Because marginal cost is equal to $2, the optimal quantities are the solutions to the
equations:
MR, =14— 4Q 1 = 2 which implies that 0 1 = 3
and
MR11 = 10 - Qn = 2 which implies that Q11 = 4
Optimal prices are obtained by substituting the profit-maximizing quantities into the
demand equations. Thus P1 = 8 and P11 = 6.
Profits in each market are equal to total revenue (P Q) minus total costs (MC . Q).
Hence
profit1 = $ 24 - $6 = $18 and profit11 = $ 24 - $8 = $16
Hence combined profit for the two markets is $34.
To compute profits in the absence of price discrimination, the combined demand
and marginal revenue equations must be computed. The first step is to express the de-
mand equations in terms of quantities. Thus
Q1 =7— and Q11=10—P
Note that the subscript has been dropped from price because the same price is to be
charged in each market. Adding the two demand curves gives
= 17 -
The corresponding marginal revenue equation is calculated by solving for P and using
the principle that the marginal revenue function for a linear demand curve has the same
intercept and twice the slope. Thus
P = 11- - -QT
and
422 PARE V Pricing Decisions
MRT = 11•- --
Equating marginal revenue to marginal cost gives
11- - =2
which implies that Q = 7. Substituting Q = 7 into the combined demand equation yields
P = 62 . Hence profit without price discrimination is
P Q - MC Q = $ 44 —$14 = $32
But price discrimination resulted in total profit of $34. Thds it has been shown that
profit can be increased by the use of price discrimination.
PRODUCT BUNDLING
In the movie Five Easy Pieces, actor Jack Nicholson enters a diner and asks for toast
and coffee. The waitress curtly informs him that toast is not available, even though the
restaurant has both bread and a toaster. To obtain his toast. Nicholson is forced to or-
der a chicken salad sandwich without the chicken, mayonnaise, and lettuce. Although a
little unusual. this is a form of product bundling. Bundling is the practice of selling two
or more products together for a single price. When the products are only available as a
package, the pricing strategy is referred to as pure bundling. If at least some products
can also be purchased separately, then the firm is using mired bundling.
Bundling is a common practice. College and professional sports teams offer season
ticket packages that include seats for popular games that are likely to be sold out, to-
gether with .seats for other games that have less fan appeal. Cultural arts series are of-
ten marketed in the same way. Many restaurants offer complete meals that include ap-
petizers and dessert. Car manufacturers provide vehicles with features such as air
conditioning, antilock brakes, cassette decks, and airbags as standard equipment at "no
extra" price. Computer companies often include certain software, such as an operating
system and a word processor, with the machines that they sell.
Why is bundling such a common strategy? One reason is that firms can reduce their
production and marketing costs by packaging goods and services in this way. For example,
General Motors can customize its facilities to manufacture automobiles with a limited
number of option packages. Successful college football teams such as Notre Dame can re-
duce their ticket sale costs because they have many season ticket holders. The waiters and
waitresses at restaurants that offer only complete meals are spared much of the time re-
quird for customers to make decisions about each individual component of the meal.
But product bundling can be profitable even where there are no cost savings. Like price
discrimination, bundling allows firms to increase their profits by extracting additional con-
sumer surplus. However, in some situations, bundling may be preferable to price discnmi-
nation because it requires less information about tastes and preferences of consumers
A simple example will be used to illustrate how bundling can increase profits. Be-
cause of their extremely high production costs, television series rarely make money for
their producers when they are first seen on TV. The real payoff comes if the program
lasts two or three years so that there are enough episodes produced to allow triem to be
sold to individual stations, which rely on reruns to meet their programming needs,
Consider a firm that has acquired the rights to 50 episodes each of two popular pro-
grams—Seinfeld and Star Trek Suppose stations in two different cities of similar size
are contemplating the distributor's offerings and that the maximum prices they will pay,
often referred to as the reservation price, for a 50-episode series are as shown here:
Seinfeld Star Trek
Memphis, Tenn. $500,000 $300,000
Seattle, Wash. $300,000 $500,000
The numbers were set up so that preferences of the two stations would be negatively
correlated. Presumably because of local tastes, the station in Memphis has a higher
reservation price for Seinfeld than for Star Tre
k. In contrast, the Seattle station attaches
more value to Star Trek. Assume that the same prices must be offered in both cities.
Because preferences of the two stations are negatively correlated, the distributor can
increase its profit by bundling. If the firm prices the two series separately, the maximum
amount it could charge (and still sell to both stations) would be $300,000 for each, and to-
tal revenue would be $1200,000. Now suppose the firm bundles the series and sets a price
Of
$800,000 for the two. iach station will buy the package because the price does not ex-
ceed the sum of the reservation prices of the stations for the two series. But total revenue
to the distributor will be $1,600,000. By bundling, the firm earns an additional $400,000.
If demands for the series had not been negatively correlated, bundling would not
have been advantageous. For example, assume that the reservation prices for the two
stations were as follows:
Seinfeld Star Trek
Memphis, Tenn. $300,000 $400,000
Seattle, Wash. $400,000 $500,000
In this case, the firm could charge a maximum of $300,000 for Seinfeld and $400,000 for
Star Trek and eam a total of $1,400,000. But the highest package price that could be
charged is $700,000 per station, and this would generate the same amount of revenue.
For bundling to be profitable, there must be at least some consumers whose preferences
are negatively correlated with others. If there are consumers who have a high reserva-
tion price for one good but place a low value on the second, there must be other con-
sumers whose reservation price for the second good is higher than for the first.
Key Ccepts.
• Pure bundling mv more p li'pethcrQcthe same pncc'
• If reser'cr&,fl prk es are tegti
)ri elatod, btg c nácreae prufts by
capturing pan pf the consuruc; us... -
PEAK-LOAD PRICING
A firm selling in many markets at the same time can increase its profit by using price
discrimination. Similarly, a firm that uses the same facility to supply several markets at
different points in time can increase total profits by the use of peak-load pricing. Basi-
cally, peak-load pricing involves charging a higher price for consumers who require ser-
vice during periods of peak demand and a lower price for those who consume during
low or off-peak periods.
Pricing of long-distance telephone calls is a good example. Most long-distance calls are
for business purposes and are placed on weekday afternoons. As a result, the switching fa-
cilities and lines provided by telephone companies are designed to meet demand during
this peak period. In contrast, fewer calls are made late at night and on weekends. But these
off-peak calls use the same facilities necessary for peak-period calls. To induce consumers
to shift their calling patterns to periods of "d phone companies offer substactial
discounts for late-night and weekend long-distance calls. If successful, costly additions to
capacity can be postponed and existing facilities used more efficiently. As a result, both the
company and its customers can benefit. The firm will have rediiced costs and hence in-
creased profit. Off-peak customers will pay lower rates. Even peak-period customers may
benefit in the long run because the firm's facilities will be more efficiently utilized.
The fundamental principle of peak-load pricing is that those who impose the great-
est demand on a firm for production capacity should be those who pay for most of that
capacity. The traditional theory of peak-load pricing is discussed here.
Peak-load pricing may be appropriate if three conditions are met in producing a
good or service. First, the product cannot be storable. For example, in the case of long-
distance telephone calls, 'the service involves direct communication between two or
more people. Calls cannot be stored for use at a later time. A busy executive in Portland
would not consider a recorded message from an associate in New York to be an ac-
ceptable substitute for a telephone call received during business hours. Second, the
same facilities must be used to provide the service during different periods of time.
Again using long-distance telephone service as an example, this condition is met be-
cause calls placed at different times use the same lines and switching equipment. The
third condition is that there must be variation in demand characteristics at different pe-
riods of time. Long-distance calls also meet this requirement. Demand is greater during
business hours than at other times. In addition, demand for business calls at any given
time is usually less elastic than is the demand for personal calls.
To illustrate the concept of peak-load pricing, assume that demand for telephone
calls during a day can be divided into two periods of equal length. Period 1 is a time of
low demand and extends from 7P.M. to 7 A.M. Period 2 is from 7 A.M. to 7 P.M. and is
the time of peak demand. Let demand for telephone calls in period 2 be greater at all
prices than demand in period 1. That is, at any given price the quantity demanded is
greater in period 2 than in period 1.
Also assume that only labor and capacity costs are incurred in providing telephone
service. Let b represent labor cost per unit of service,-where b is assumed not to vary by
the period or by the demand for service. Similarly, let B represent the rental cost of a unit
of capacity, with B assumed constant with respect , to the amount of capacity purchased.
That is, the first unit of capacity costs $B, and all additional units cost the firm $B.
Let the price paid for telephone service be thesum of the labor cost and the ca-
pacity cost (which includes a reasonable return on capital). Because the labor cost per
unit of service is assumed to be constant, it can be subtracted from the price in each pe-
riod without altering the results of the analysis. Figure 12.6 depicts the two demand
curves, D 1 and D.,, after the labor cost has been subtracted. As a result, the vertical scale
Price per unit
($)
PT
P2
P1
Capacity
-
starts with b (the labor cost) instead of zero. Because labor costs have been netted out,
Figure 12.6 can be considered a graph of the demand for telephone capacity. Hence.
each point on the two demand curves indicates the amount of capacity that the firm's
customers demand at a given price of capacity.
The D T curve in Figure 12.6 is the result of vertically summing the two individual
demand curves (i.e., adding the vertical distances above the horizontal axis at each point
along the horizontal axis). Remember that the demands are noncompeting and hence
the same capacity is used to provide service in each of the two periods. The individual
demand curves can be interpreted as indicating the willingness of consumers to pay for
capacity during each period. For example, consumers in period 1 will pay P 1 per unit for
Xunits of capacity and consumers in period 2 will pay P2 for X units, Because the ca-
pieity is usable to serve consumers in each period, the total value or demand for X units
of capacity is PT = P1 + 2•
Efficient resource allocation suggests that capacity should be added until the value
of the last unit is just equal to the cost of obtaining it. The cost of each unit of capacity is
B. The total value of capacity is read from the DTCUrVC and indicates that at a cost of B,
X@ units of capacity should be employed. The peak-load pricing problem is to determine
who should pay for the X* units of capacity. Notice that even if the cost of capacity were
zero, less than X* units of capacity would be required to meet the demand in period 1.
Thus, the selection of X* units provides no benefits to consumers in period 1. The impli-
cation is that they should not be required to pay for capacity. Users in that period would
properly be charged only the labor cost, b, per unit of telephone service consumed.
In contrast, because period 2 users attach a positive value to X units of capacit y, they
should he assessed the cost. Specifically, the value of capacity in period 2 is P' = B. By as-
signing this cost to period 2 users, the marginal cost of capacity can be recovered. Thus. the
total price 01 telephone calls to period 2 users should be P*, ± h and to period I users. h
Notice the basic principles involved in this scheme of peak-load pricing. Because
the off-peak users place no value on the marginal units of capacity, they pay only the la-
bor cost; the capacity charge is paid by the users who require the capacity. In general,
peak-load pricing charges all or most capacity costs to peak-period users and charges
off-peak users a price based on the noncapacity costs of serving them.
..-.r- ".-.:-:.'-$,'-................?. . ........ ..
Key Conctpls
• Peak lo d pncutg can be used to reduce costs and mereas6 profits if
J. The 'mt. nciliftes are used to pi ,vxcle a produLt or service at different pe-
riods of tlmc.
2, The product or set-vice is not storthl
Demand characteristics vary from Period to period
• Me theorv of neak-cad nriin '.. iget thf n p'r'l Ic'rs chonlcl p
most capacity costs,-While- off-peak users may be r equired to pay only variable
COStS- . . ., . .. ...-. . .........
Case Study
Peak-Load Pricing of Computer Time
The central processing unit (CPU) is the heart and brain of a computer. Input/output
devices transmit information to and from the computer, and storage devices such as on-
line disks maintain files of information, but it is the compUter's CPU that manages the
ciitire operation and processes the data.
The three criteria for successful peak-load pricing are met by CPU time at large
computational facilities. First, usually a computer has only a single CPU, but this equip-
ment is in use constantly. Thus, the same facility is used to provide the service at differ-
ent periods of time. Second, CPU time not used is lost forever—clearly, the service is
not storable. Finally, although a computational center may be able to provide the same
service at 3 A.M. as at 3 P.M., late-night service is often not as desirable to the customer.
As a result, demand for CPU time is usually much greater during the day than it is at
night. The figure. depicts demand for CPU time at a hypothetical computer facilit y dur-
ing a 24-hour period- The solid line illustrates the usage pattern when the price per sec-
ond of CPU time does not change with the time of day. Note that demand is very low
during the early morning hours, begins to increase around 7A.M., peaks in the after-
noon, and declines in the evening. The result is that the facility is very busy at some
times and virtually idle during other periods.
To encourage the use of the computer during off-peak hours, many computational
facilities have implemented peak-load pricing. Although the specifics differ among cen-
ters, these pricing policies specify relatively high prices during periods of peak demand
and much lower prices during off-peak early-morning hours. For example, at a univer-
sity computer facility, CPU time is priced as follows:
CHAPTER 12 Pricing of Goods 4nd Services 427
CPU
usage
Midnight Time of day

6 AM Noon 6 PM Midnight
LJTL& .. _I_1
The weekend rates also apply to holidays. Note that CPU time at peak hours
(weekdays from 1 to 5 P.M.) is 30 times more costly than during off-peak hours (week-
day evenings from I to 8 A.M. and weekends and holidays from 5 PM. to 8 AM). AS a
result, computer usage has declined during peak periods and is much greater during the
off-peak hours. This change, resulting from peak-load pricing, is depicted by the dashed
line in the figure. By reducing demand at peak periods, the facility may be able to post-
pone purchases of additional computing capacity. U
Cosi per
Day Time Period Second
Monday—Friday 8 A.M.-1 P.M. $0.03
Monday—Friday 1 P.M.-.5 P.M. 0.06
Monday-Friday 5 .-4 A.M. 0.01
Monday-Friday I A-MA A.M. 0.002
Saturday-Sunday 8 A.M.-5 P.M. 0.01
Saturda y-Sunday 5 P.M.-8 A.M. 0.002
COST-PLUS OR MARKUP PRICING

The traditional assumption of economic theor y is that firms attempt to maximize prof-
its and that this objective is accomplished by increasing production until marginal rev-
enue equals marginal cost and then charging a price determined by the demand curve.
In actual practice, however, many firms use cost- plus pricing. Basically, this approach in-
volves setting prices that cover the cost of purchasing or producing a product plus
enough profit to allow the firm to earn its target rate of return.
The remainder of this section considers why cost-plus pricing is so popular and how
it is implemented The section also examines the apparent conflict between the cost-
plus and marginal revenue equals marginal cost pricing rules.
Mechanics of Cost-Plus Pricing

Two steps are involved in cost-plus pricing. First, the cost of acquiring or producing the
good or service must be determined. The total cost has a variable and a fixed compo-
nent. In either case, costs are computed on an average basis. That is.
AC = AVC+AFC (12-3)
where
AVC— and AFC=

Q
and AC is average total costs, AVC is average variable cost, TVC is total variable cost,A FC
is average fixed cost, TFC is total fixed cost, and Q is the number of units produced. But
there is a problem in making the computations. With cost-plus pricing, quantity is used to
calculate the price, but quantity is determined by price. This problem is avoided by using
an assumed quantity. Typically, this rate of output is based on some percentage of the firm's
capacity. For example, for many years General Motors used cost-plus pricing and com-
puted average costs on the assumption that sales would be 80 percent of capacity.
The second step in cost-plus pricing is to determine the marku p over costs. As men-
tioned previously, the overall objective is to set prices thai allow the firm to earn its tar-
geted rate of return. Thus if that return requires $XOf total profit, the markup over costs
on each unit of output will be XIQ. Hence, the price will be
P='AVC+ AFC ±- (12-4)
Example Cost-Plus Pricing of Automobiles

An automobile manufacturer estimates that total variable costs will he $500 million and
total fixed Costs will be $1 billion in the next year. In setting prices, it is assumed that sales
will be 80 percent of the firm's 125,000-vehicle-per-year capacity, or 100,000 units. The
target rate of return is 10 percent, which is to be earned on an investment of $2 billion.
If prices are set on a cost-plus basis, what puce should be charged for each automobile?
Solution A target return of 10 percent on $2 billion requires that the firm earn a to-
tal profit of $200 million. Computations are based on assumed sales of 100,000 vehicles.
Thus, average fixed cost, average variable cost, and profit per unit are
AFC = !on = $10,000
$500 million =
AVC
100,000
• X $200 million - $2000
Q 100,000 -
Hence, if profit of $2,000 per vehicle is added to the $15,000 in average fixed and vari -
able costs, the price will be $17.000 per car.
Evaluation of Cost-Plus Pricing

Cost-plus pricing has some important advantages that help explain its appeal. First, it
may contribute to price stability. This is a desirable result because price changes can be
expensive and may provoke undesirable reactions by competitors. Second, the formula
used in cost-plus pricing is simple and easy to use. As will be discussed later, the neces-
sary information is less than for marginal revenue equals marginal cost pricing. Finally,
cost-plus pricing provides a clear justification for price changes. A firm desiring to in-
crease its price can point to cost increases as the reason.
cost-plus pricing has been criticized on a number of points. One alleged
problem is that it is based on costs and does not take demand conditions into account.
This shortcoming is compounded by the fact that the cost data used may the wrong
costs. Instead of using incremental or opportunity costs, cost-plus prices often rely on
historical or accounting data. in addition, most applicationsdf the procedure are based
on fully distributing common costs to the various goods produced by the firm. Problems
with prices based on fully distributed costs were discussed earlier in the chapter.
Cost-Plus Pricing and Economic Theory

At fitst glance. cost-plus pricing woffid appear to be inconsistent with economic theory
that assumes profit maximization. Moreover, the wide use of cost-plus pricing seems to
make analysis based on the marinal revenue equals marginal cost decision rule largely
irrelevant. However, the conflict may be more apparent than reaLmere is reason to be-
lieve that use of cost-plus pricing is simply a tool used by businesses in pursuing the goal
of long-run profit maximization. As such, cost-plus prices can be shown to be related to,
although not identical to, prices based on marginal revenues andnsainal costs.
A comparison of the two approaches to pricing starts with a.eonsideration of
costs. Although it is true that cost-plus pricing is based on average; rather than mar-
ginal costs, frequently long-run marginal and average costsare not greatly different.
This is especially true in many retailing activities. Thus,, use of average costs as a basis
for price determination may be. considered a reasonable approximation of marginal
cost decision making.
llie secc:d.tep : tht cmtpaison involves the target rate of return and the re-
sulting markup. How does a manager determine whether the target should be 10 per-
cent or 20 percent? Basically, the decision involves management's perception of de-
mand elasticity and competitive conditions. An example would be grocery stores. The
intense competition among these firms holds down profits. As a result, the typical
markup for most food items is only about 12 percent over cost.
If the markup over cost is based oll demand conditions, cost-plus pricing may not
be inconsistent with pout maximization. This can be shown mathematically. Marginal
revenue is the derivative of total revenue with respect to quantity. Thus
MR=TdD=P+dQ
dQ dQ dQ
But P + Q can be rewritten as -1- Note that (dP/dQ)(Q/P) is 1/EAr,,

where E,, is price elasticity of demand. Thus
MR= P 1E (12-5)
Profit maximizatithi requires that MR = MC. As a simplifying assumption, let MC =

AC. Thus, the profit-maximizing price is the solution to
P(l+*)=AC
which can be written as
p(E± 1) =
AC
Solving for P yields
P = A C _E ) (12-6)
+I
Equation (12-6) can be interpreted as a cost-plus or markup pricing scheme. That
is, the price is based on a markup over average costs. The markup, E/(1 + Er), is a func-
tion of the price elasticity of demandAs demand becomes more elastic, the markup be-
comes smaller.-3 For example, if E = — 1.5, the markup is 3.0. But for E = —4.0, the
markup is only 1.33 times average cost.
Thus, cost-plus pricing may simply be the mechanism by which managers pursue
profit maximization. Most managers have limited information on demand and costs.
Obtaining the additional information necessary to generate accurate estimates of mar-
ginal costs and revenues may be prohibitively expensive. Hence, cost-plus pricing may
be the most rational approach in maximizing profits.
Kiy CEnteepts
• Cusl-plus pricing is widely used by managers and tnvchcs a markup over the
average cost of acquiring or producing a product.
• Cost-plus pricing may stabilize prices and provide ajusiiiication for price changes.
• The markup used in cost-plus pricin g is determincd by demand elasticities and
competition. Markups are lower where demand is more elastic and competition -
IS Intense.
• Cost-plus pricing may simply represent the decision rule used by managers in
purswt of pr ft rnixirniz.ati&tu.
SUMMARY -
Modern corporations may produce hundreds of different products. When the demand
for products is interrelated, the firm should take this interdependence into account.
3 1f E = —1.0. the markup is undefined, and if —1 <E <0, the markup becomes negative. However, these
cases are irrelevant because the profit-maximizing firm will never operate on the inelastic portion of its
demand curve. The reason is clear. As long as marginal revenue is positive, demand is elastic. But the profit-
maximizing firm produces where marginal revenue equals marginal cost. Because marginal cost is positive,
marginal revenue is also positive, and demand is elastic at the profit-maximizing rate of Output.
When a firm produces complementary goods, the output rate of each good should be
greater than if no demand interrelationship existed. For substitutes, output rates should
he less than if the goods were independent.
Some goods are jointly produced in fixed proportions. Profit maximization requires
-that the two goods be considered as a product package. Thus, the rate of output should
he increased until mar g inal cost equals the sum of marginal revenues obtained from
selling an additional unit of the product package.
Not all costs are clearly attributable to a particular good or service. Attempts to at-
locate such common costs are arbitrary. Decisions based on fully distributed costs can re-
suit in inefficient resource allocation; incremental costs are a better guide. A good or ser-
vice can be profitably produced if its price exceeds the incremental cost of supplying it.
If an enterprise has common costs, marginal cost pricing may not he feasible. Ram-
sey pricing is a second-best alternative that allows the firm to recover its costs while
minimizing adverse effects on allocative efficiency. A simple version of Ramsey pricing
specifies that deviations from marginal costs should be inversely proportional to the de-
mand elasticities of the goods or services.
Vertical integration causes intermediate goods to be transferred from one division
Of firm to another. However, profit maximization for the combined firm requires that
a
those intermediate products be correctly priced. When a perfectly competitive external
market exists, market forces will cause price to equal marginal cost. Hence, the inter-
mediate product will be appropriately priced. If there is no external market, manage-
ment should set the price of the intermediate good at marginal cost.
Price discrimination occurs when variation in price for a product sold in different
markets does not correspond to differences in costs. The three conditions for successful
price discrimination are market power, variation in the elasticity of demand, and sepa-
rable markets. First-degree price discrimination involves charging each consumer the
maximum amount that he or she is willing to pay. Second-degree discrimination occurs
when prices vary depending on the amount purchased. Third-degree discrimination
separates markets in terms of elasticity of demand.The segmentation may he based on
location, use, Or personal characteristics. Usually, higher prices are charged when de-
mand is less elastic.
Product bundling is a pricing scheme that allows firms to capture part of the con-
sumer surplus. Pure bundling involves the selling of two or more products only as a
package. If the reservation prices of consumers are negatively correlated, firms can use
bundling to increase their profits.
Peak-load pricing can reduce costs and increase profits. The practice is appropriate
when three conditions are met. First, the same facilities must be used to provide a prod-
uct or service at different tLrnes. Second, the product or service must not be storable.
Third, demand characteristics of consumers must vary from period to period. With
peak-load pricing, peak-period consumers will pay most of the capacity costs, while off-
peak users will be charged a price based on variable costs.
Cost-plus, or markup, pricing is widely used by firms. This practice involves set-
ting price at average cost plus a markup designed to provide a target rate of return.
Advantages of cost-plus pricing include stable prices, a simple formula for pricing,
and a clear justification for price changes. Problems include the use of historical and
fully distributed costs rather than marginal costs and an apparent failure to take de-
mand conditions into account in price setting. -
On closer investigation, it can be shown that markups are related to demand elas-
ticity and competition. Less elastic demand and lack of competition are associated with
high markups ove: cost, while the reverse is true for markets characterized by more
elastic demand and intense competition. Faced with a lack of information about de-
mand and costs, it may be that cost-plus pricing is simply the method by which managers
pursue profit maximization.
Discussion Questions
124. Macmillan Manufacturing produces razor blades and razors. Propose a pricing
strategy that would allow the firm to maximize its profit on the two goods. Explain.
12-2. Why should goods produced in fixed proportions be regarded as a product pack-
age in developing production and pricing strategies?
12-3- Should a sheep-ranching operation consider Iamb and wool as a product pack-
age? Why or why not?
12-4. What is meant by the statement that "the assignment of common costs must, by
definition, be arbitrary"?
12-5. Generating equipment is used to provide electric power to both residential and
industrial customers. Assume that total consumption by residential users is
twice that of industrial consumers. in setting iriees, would it be ap propriate to
assign two-thirds of the cost of the generating equipment to the residential
users? Why or why not?
12-6. The managers of a firm are considering offering a new service. The proposed
price of the new service would be greater than its incremental cost but would
not cover fully distributed costs. As a user of services already provided by the
firm, should you favor or oppose the new service? Explain.
12-7. How does the presence or absence of external markets affect the role played by
top management in pricing inte'-mediate products produced by a vertically in-
tegrated firm?-
12-8. If an intermediate product is available from a perfectly competitive industry,
why would a vertically integrated firm produce the product internally? That is,
what is the advantage of vertical integration in this case?
12-9. A city has only one furniture Store. Is it likely that the store could successfully
practice price discrimination? Why or why not? -
12-10. How is bundling similar to price discrimination? Which requires more informa -
tion about consumer preferences?
12-11. Are the three conditions necessary for peak-load pricing met in the case of
movies shown in theaters? Explain.
1212. How can peak-load pricing improve resource allocation?
12-13. How can cost-plus pricing be reconciled to the "marginal revenue equals mar-
ginal cost" rule of economic theory?
rob1ems
12-1. A small firm traps rabbits for their fur and feet. Each rabbit yields one pelt and
two feet (only the hind feet are used to make good-luck charms). The demand
for pelts is given by
Pp = 2.00 - 0.001Q
and the demand for rabbit's feet is given by

1.60 -
The marginal cost of trapping and processing eaLh rabbit is $0.60.
a. What are the profit-maximizing prices and quantities of pelts and rabbits feet?
b. If the demand for rabbit's feet is P F = 1.00 - 0.00I QF, what are the profit-
maximizing prices and rates of output?
12-2. Mike's Shear Shop provides 4,000 haircuts each month at an average price of $10
Per haircut. The common eoss of operating the store are $24,000 per month. The
business is considering hiring a photographer who would take pictures of cus-
tomers after they had their hair cut. The price of the photographs would be $5.00,
and it is estimated that 2,000 customers would purchase the service each month.
The total extra cost of the photographic service is 2,000 ± 2Q per month, where
Q is the number of photographs sold.
a. If the decision is to be based on incremental revenue and incremental cost,
should the service be offered? Explain.
b. If the decision is to be made on the basis of fully distributed costs and if the
$24,000 in monthly common costs are to be apportioned based on the rev-
en-nes from haircuts and photos sold each month, should the new service he of-
fered? Why or why not?
12-3. Write-Right, a vertically integrated firm, produces both paper and writing tablets.
The demand for tablets is given by
P T = 1.00 - 0.001Q
where Q is the quantity of tablets. The marginal cost of producing the paper nec-
essary for each tablet is
MC = 0.20 + 0.001Q
It costs the firm $0.10 to make the paper into a writing tablet. If there is no external
market for the paper, what transfer price should top management set for the paper?
12-4. A firm has found a way of using first-degree price discrimination. Demand for its
product is given by
P = 20 - 2Q
Marginal cost is constant and equal to $6.
a. With first-degree discrimination, what will be the pofii-maximizing rate of
output? How much economic profit will the firm earn?
b. What will be the profit-maximizing rate of output if the firm does not dis-
criminate and sets one price for all customers? How much economic profit will
the firm earn in this case?
12-5. Smith Distributing sells videocassettes in two separable markets. The marginal
cost of each cassette is $2. For the first market, demand is given by
= 20 - 5P1
The demand equation for the second market is
20 - 2P2
a. If the firm uses third-degree price discrimination, what will be the profit-
maximizing price and quantity in each market? How much economic profit
will the firm earn?
b. If the firm charges the same price in both markets, what will be the profit-
maximizing price and total quantity? How much economic profit. will the
firm earn?
12-6. Global motors sells its automobiles in both the United States and Japan. Due to
trade restrictions, a vehicle sold in one country cannot be resold in the other. The
demand functions for the two countries are
US. P = 30,000 - 0.40Q
Japan P = 20,000 - 0.20Q
The firm's total cost funôtion is TC = 10,000,000 + 12,000Q. What price should
Global charge in each country in order to maximize profit? What will be the to-
tal profit?
12-7. The demand for a firm's product is given by P = 200 - 10Q. Marginal cost is con-
stant and equal to $10.
a. Using first-degree price discrimination, what is the profit-maximizing output
and price? How much profit will the firm earn?
b. What is the profit-maximizing output and price if the firm sets a uniform price
to all buyers? How much profit wifi the firm earn?
12-8. A monopolist sells in two markets and the marginal cost is $2 in each market.
a. If demand is given by P = 21 - 3Q in the first market, what is the profit-
maximizing price and output?
b. If demand is given by Q = 21 - P in the second market, what is the profit-
,
maximizing price and output?
c. Which market has the more elastic demand? Explain.
12-9. A firm produces two products,A and B. Demands are independent and marginal
costs are zero. The products are sold to three consumers, and each must be
charged the same price.
a. For the following reservation prices, can profits be increased by bundling? Ex-
plain. What is the maximum profit?
GoodA Good B
Consumer I $50 $40
Consumer 2 $60 $35
Consumer 3 $70 $50
b. For the reservation prices shown here, what is the p rof
it-maximizing price
strategy? Explain. What is the maximum profit?
GoodA Good B
Consumer 1 $50 $60
Consumer 2 $60 $60
Consumer 3 $70 $50
435
12-10. The manager of a Sporting goods store uses c
profit -maximizing price of bicycles. The cost ofost-plus
a bic y
pricing to d etermine the
cle to the store is $80, The
manager estimates that the price elasticity of demand is —3.0. What is the profit-
maximizing price?
12-11. Grasscutter Inc: makes a product used to trim lawns. The firm has fixed costs of
$100,000 per year. Management expects to sell 2,000 units per year, and at that
rate of output, total variable costs will be $50,000. The firm uses cost-plus pric-
ing to earn a target rate of return on an investment of $200.000. If the price is set
at $100, what is the target rate of return?
- Problems Requiring Calculus

12-12. The
House of Music sells low-cost CD players and speakers. The total re v
equation for sales of the two products is given by enue
TR = 200 QC - 6Q + 100Q5
- 4Q * QQ
where Qc and Qs are quantities of CD players and speakers, respectivel
y. The
marginal cost of CD players is $20 and the marginal cost of speakers is $10.
a. Are the two goods substitutes or
complements?
b. What is the profit-maximizing rate of output for each good?
c. What would be the P ro fit-maximizing rate o
f
int erdependence betveen the two goods? output if there were no demand
12-13.
Culture Extravaganza produces ballets in Boston and New
revenues aregiven by York. Monthly total
TRB 1,000Q5
and
TRN 2,000Q,5
where QB is the monthly number of Boston patrons and
QA, is the monthly num-
ber of New York ballet attendees. Salaries of the performers are based on atten
-
dance, and the firm estimates that the marginal cost is $10 per attendee in each cicy.
a.
If Culture Extravaganza attempts to practice third-degree price discrimina-
tion, what will be the P ro
b. fit-maximizing prices and rates of output in each city?
Will the firm earn more profit using price discrimination than if a uniform
price is set? Explain.
12-14, A firm produces two types of c
are shown below with n alculators,x and y. The revenue and cost equations
and Q measured in thousands of calculators per year.
Total revenue 2 Qr + 3 Qy
Total cost = Q 2QXQY - 2Q + 6Q + 14Q, + 5
a. To maximize profit h
produce? ow many of each type of calculator should the firm
b. What is the maximum profit the firm can earn?
CHAPTER
Pricingf3oment
• Preview
• Input Pricing and Employment
Market Structure I: Monopolist—Perfect Competitor
Market Structure II: Monopolist—Monopsonist
• The Correspondence between Output and Input Decisions
• Economic Rent
• Wage and Income Differentials
Dcmand . Side Considerations
Supply-Side Considerations
• Labor Unions
Labor Unions and Collective Bargaining
Union Objectives
• Minimum Wage Laws
• Summary
• Discussion Questions
• Problems
436
CHAPTER 13 Pricing and Employment of Inputs 437
PREVIEW
Individuals earn income by selling resources such as capital and labor to firms. The firm
uses these resources to produce the goods and services demanded by consumers. Of
critical importance to both the individual and the firm is the price of a unit of the re-
source. The analysis presumes that the distribution of resources among individuals is a
given. That is, the ownership of land, capital resources, and human resources (i.e., the
education, training, and experience embodied in labor) has already been determined.
Given this initial distribution, the level and distribution of income are determined by
the price of each resource unit: 'fliee include the wage rate per hour of labor, the rent
per acre of land, and the price per unit of capital. Clearly, input prices are also impor-
tant to the firm because they are a critical factor in determining the amount and mix of
resources employed. This chapter focuses on the determination of those input prices
and how fj ñs decide how much of an input to employ.
The price of a productive input is determined by supply and demand, just as is the
price of output. However, the firm's demand for an input is a derived demand. Firms de-
mand capital, labor, and land not because they have value as such, but because those re-
sources can he used to produce goods and services that have value to consumers. Thus,
the demand for inputs is dependent on the demand for the goods arid services those in-
puts are used to produce; hence the use of the term derived demand.
The first part of this chapter focuses on the determination of input prices and the
amount of an input to be employed. Two different market structures are considered. In
the second section, the correspondence between output and input decisions is outlined.
The next two sections include an introduction to the concept of economic rent and a
discussion of the reasons for significant income differentials among workers. The role
of labor unions in input markets is analyzed in the fifth section. Both the economics of
collective bargaining and the objectives of labor unions are considered. Minimum wage
laws are analyzed in the final section,
INPUT PRICING AND EMPLOYMENT

The analytical framework for studying the demand for productive inputs is based on the
theory of production developed in chapter 6. Recall that efficient production requires
that the ratio of marginal product to input price be equal for all inputs. For example, if
capital and labor are the only inputs, the efficiency condition is
= (13-1)
where MPK and MP1 are the marginal products of capital and labor and rand w are the
prices of those inputs.
If one input, say capital, is held constant, it can be shown that the ratio of input price
to marginal product is equal to marginal cost. Thus for labor
= MC (13-2)
MPL
Wage rate,
Marginal revenue product
($/unit)
30
20
10
Marginal
revenue
product
Units of
25 40 50 l abor (L
Ive
Recall that the profit-maximizing firm must increase oulput, and therefore the em-
ployment of inputs, to the point where marginal cost equals marginal revenue. Thus, in
equilibrium, the ratio of input price to marginal product (i.e.,rnarginal cost) must equal
marginal revenue. That is,
(13-3)
MPL
Multiplying both sides of equation (14-3) by MPL yields
W = MRMPL (13-4)
AN
= MRPJ (13-5)
which indicates that the profit-maximizing firms should hire an input until the price of
the input equals the marginal revenue product (MRP) of that input.
The MRP function is the firm's input demand function. Consider the MRP function
shown in Figure 13.1. It can be used to determine the rate of labor input that will be
hired at any wage rate. For example, if the wage is $10 per unit, 50 units of labor are em-
ployed; at $30 per unit, only 25 Units will be employed.
The input demand function (i.e.. the marginal revenue product or MRP function)
is determined by multiplying marginal product by marginal revenue. Because of the law
CHAPTER 13 Pricing and Employment of Inputs 439
of diminishing marginal returns, the marginal product function will be decreasing, at
least in the range that is relevant for the firm. Further, the firm's marginal revenue func-
tion will either be horizontal, as in the case of the perfectl y competitive firm, or down-
ward sloping. Thus, because the input demand function is found by multiplying the mar-
ginal revenue function by the marginal product function, the input demand function
must be downward sloping.
The firm sells goods and/or services in the product market and buys inputs in the fac-
tor market. The structure of both of these markets will influence the price of the input
and the amount employed. The circumstance of the firm on the output side of the mar-
ket (i.e., whether the firm is a perfect competitor, oligopolist, or monopolist) will affect
the firm's input demand function, whereas the market structure on the supply side will
influence the input supply curve facing the firm. In this section, input pricing and em-
ployment decisions are considered for the following types of firms: (1) a firm th at is a
monopolist in the product market and a perfect competitor in the input market and (2) a
firm that is a monopolist in the product market and a monopsonist in the input market.
The key difference in analyzing the effect of differing conditions in the product
market is that price and marginal revenue are constant for the perfect competitor. In
contrast, the demand curve facing the monopolist is downward sloping, which implies
that the marginal revenue function also has a negative slope.
A perfect competitor in an input market is one of many buyers of the input. No one
buyer has any influence on the price of the input.This implies that the input supply func-
tion is horizontal (i.e., perfectly lastic). meaning that any firm can buy as much of that
input as desired without influencing the price, Thus, the input price facing a perfect com-
petitor in an input market is constant. In contrast, the input supply function facing the
monopsonist or single buyer of an input is upward sloping. Additional inputs are ob-
tainable only at higher prices. For example, a large firm in a small town may employ a
substantial share of the local labor force. Thus, as a near-monopsonist, any significant
increase in the firm's employment of labor would require an increase in the wage rate.
Examples of the supply curves for a perfect competitor and a monopsonist are shown
in Figure 13,2.
Key Concep ': ' . -........
• The demand for a productive input is derived from the demand for the goods
/ and services that input is used to produce.
• The demand function for an input is deri.ed from the condition for efficient
production. In genera!, an input should he employed until the input price equals
the marginal revenue product of theinput.
• Firms that are perfectl y competitive, in input markets face a horizontal input
supply function. Monopsonists face an upward_sloping input supply function.
Market Structure I: Monopolist (Product

Market)— Perfect Competitor (Input Market)
If the firm is a monopolist, or at least has some degree of market power, it faces a
downward-sloping demand function for its output. Thus, the marginal revenue curve
lies below the demand function. In this case, the firms demand for labor, as shown by
Price of input
(S,'unit) Supply (monopsOfliSt)
supply (perfecc
competitor)
Quantify of
n
the marginal revenue product functionwill decline more rapidly than if price equaled
marginal revenue. Because the firth in this example is a perfect competitor in the in-
put market, the supply function is horizontal.
Table 13.1 shows hypothetical values for a firm's total product, marginal product,
and marginal revenue. The resulting marginal revenue product or input demand func-
tion is shown in the right-hand column of the table. This demand is computed by multi-
1 OeiIJt -. 14 .
Ru Marginal
of Labor Totat Marginul Outpid Total - Margina! Revenue
Product Product Price Revenue Revenue Product
Input
0 - - - - -
1 10 10 $5.00 $ 50.00 $5.00 $50.00
2 19 9 4,50 85.50 394 35.50
3 26 7 4.00 104.00 2.64 18.50
4 30 4 3.75 112.50 2.13 8.50
5 32 2 3.60 115.20 1.35 2.70
6 33 1 3.50 115.50 0.30 0.30
CHAPTER 13 Pricing and Employment of Inputs
441
Wate rate,
marginal revenue product
(SIL)
50
4(
30
20
10
Rate of
labor input
plying each entry for marginal product by the corresponding marginal revenue per unit
entry.' As MRPL
is the net addition to the firm's total revenue associated with an addi-
tional unit of labor, MRPL also could be determined as the change in total revenue [in
column (5)] associated with a one-unit change in labor.
The input demand function and the suppl y function are shown in Figure 13.3.1f the
market price of input is $8.50, the firm will employ four units of labor because at that
input rate, the marginal revenue product is just equal to the wage rate. If the price of la-
bor increased to $3530, the firm would employ only two units.
Market Structure II: Monopolist (Product

Market) -_Monopsonist (Input Market)
Now consider a hypothetical firm that not onl y is a monopolist in the product market
but also is the only buyer of labor in the area. An electric utility company with a large
generating facility near a small town could be an example of such a firm. The company
may have a franchise as the only seller of electricity in the region and could also be the
only important employer of labor. Therefore, the firm is both a monopolist and a
monopsonist. if this firm's output rate is to increase, it mus hire more labor. But be-
cause the firm faces an upward-sloping supply curve, hiring more of an input will re-
quire that the price of that input be increased for all its workers.
'Recall that m
arginal revenue is defined as the change in total revenue divided by a one-unit change in
output. When the rate of labor input increases from 0 to 1. output increases from 0 to 10. and total revenue
increases from 0 to S(L Therefore, marginal revenue per Unit over that interval averages $50/10 or $5.00.
E 12
onLborfr*ranst -
Rate Labor Total Marginal
of Labor Price Expenditure Expenditure
Input Per Unit on Labor on Labor
o -- $0 -
1 $ 5.50 5.50 $ 5.50
2 8.00 16.00 10,50
3 11.50 34.50 18.50
I C nfl n nfl nc cn
4 VV.VV
-t
5 18.50 92.50 32.50
6 22.00 132.00 39.50
A labor supply schedule facing the firm is shown in the first two columns of Table
13.2. Also shown is the firm's marginal expenditure on labor function, which is defined
as the change in total expenditure on labor associated with a one-unit change in the rate
of labor input. Note that to hire more labor, the firm must offer a higher wage rate, and
this higher rate must be paid to all inputs hired. Thus, the marginal expenditure on in-
put function lies above the supply curve, reflecting both the higher price necessary to
attract new workers and the higher price that must be paid to those already working.
The relevant functions for determining the optimal labor input rate are the mar-
ginal revenue product function (from Table 13.1 on page 438) and the marginal expen-
diture on labor function (from Table 13.2). These data are summarized in Table 13.3.
To maximize profit, the firm will employ labor until the marginal revenue product
equals the marginal expenditure on labor. In Table 13,3 the marginal revenue product
equals marginal expenditure when thç rate of labor inputs is three, which is the profit-
maximizing input rate. This result is also shown in Figure 134, where the MRPL func-
tion intersects the marginal expenditure function at point c, which corresponds to
L 3. Note that the wage rate paid labor, $11.50, which is determined by the labor
supply function, is less than the marginal revenue product of Labor. A monopsonist is
able to hire workers at a wage less than the value of their contribution to output. This
has led to claims that firms with monopsony power exploit labor.
Rate Marginal Labor MarginaL

of Labor Revenue Price Expenditure
Input Product Per Unit on Labor
0 - - -
1 $50.00 $ 5.50 $,5.50
OAA 1flfl
L. .I.).JU Q.IiIJ
3 18.50 11.50 18.50

4 8.50 15.00 25.50
5 2.70 18.50 3250
6 0.30 22.00 39.50
CHAPTER 13 Pricing and Emplovm p ,a, of Inputs
443
Wage rate,
rnargina rexenue product
S/L)
Maigioat revenue produc:

50 L_
40
)C tidi tare
30 or
2(1
io
U 1 Rate of
23 4 5 6 labor input
I (L)
rk
Case Study
Economic Profits, Monop50,
and the Professional Sports Industry
Admission prices to most p r ofessional sports events are relatively expensive. One rea-
son is that most major-league teams have monopoly power in the market they serve. For
example, in baseball, onl y in the metropolitan areas of Chicago. New York, San Fran-
cisco, and Los Angeles are there two tems; in all other cities there is only one provider
Ofa m jor-league baseball entertainment While there may he substitutes, such as high
school, college, and se
miprofessional teams, judging by attendance and ticket prices,
these products are considered inferior to the major-league brand of baseball.
Not only are most major-league Sports franchises monopoly sellers of baseball en -
tertainment they have traditionally been
inonopsony buyers of the services of the play-
ers. By using a system of drafting new pla y ers, each team was given the exclusive right
to negotiate with the players it selected. The draft reduced the competition for the
player's services and held wages down. To maintain this monopsony right, the league
also used a reservc clause. Ever y player was required to sign a standard
c ontract that
contained a provision reserving the team's right to that player's services during the next
L
•1
Year Average Salary

1970 $ 29,307
1975 44,676
1980 143,751
1985 371,157
1990 597,357
1995 1,110,1
Source: Bureau of Labor Statistics, Compensation

and Working Condition., 2, No.5 (Washington,
D.C.: U.S. Government Printing Office, Fall 1997).
season. The combination of the draft and the reserve clause effectively tied a player to
one team forever. He could be traded to another team but was not free to shop around
among other teams for a belier salary.
In baseball, this system existed for about 70 years, but has now been substantially
weakened, largely as the result of one owner failing to pay a bonus to a player. In 1974,
Jim "Catfish" Hunter, an outstanding pitcher for the Oakland Athletics, claimed that he
had not been paid one-half of the $100,000 specified by his contract. He took the mat-
ter to arbitration, as required under baseball rules, and argued that he should be paid
the remaining $50,000 and that he should be declared a free agent, thus allowing him to
negotiate with any and all of the other teams. He won on both issues and subsequently
signed a long-term contract with the NW York Yankees for $750,000 per year—more
than seven times his 1974 salary. The increase indicates that Oakland had been paying
him substantially less than his marginal revenue product!
This development was not lost on other players. In 1975, Andy Messersmith of
the Los Angeles Dodgers and Dave McNally of the Baltimore Orioles both played
without signing a contract. Then they went to arbitration, arguing that the reserve
clause in their contract for the prior year was nonbinding, as it only applied to the 1975
season. The arbitrator agreed, and both were declared free agents who could negoti-
ate with any team.
After Messersmith and McNally signed contracts similar to that obtained by
Hunter, other players realized the effect of the reserve clause in holding down salaries
and demanded that a new contract system be developed. Negotiations between the
players' association and management broke down over this issue, and a 17-day strike
occurred during spring training in 1977. Both sides finally agreed to an arrangement
wherein a player was tied to a team for six years (later reduced to five years), after
which he automatically became a free agent. As shown in the table, the result was a
tremendous escalation in players' salaries; by 1992, the average exceeded $1,000,000 per
year. Baseball owners were still charging monopoly prices, but they were now sharing
the economic profits with the players. 0
C HAPTER 13 Pricing and Employmen
j of Inputs 445
Key Concepts
• A firm that''j a perfct competitor
in an input market faces
Jpply fun a1orjjóntai input
ction. Profit maximization for such firms requires that
at an input be in-
creased until its marginal revenue product equal the price of the input
'. A firm that has market power in input markets fae gs
supply function. The determination an upward -sloping input
c ombination for the input is made by of the Profit-maximizing price — quantity
equating the firm's
Product and marginal e rn rgina1 revenue
xpenditure on input functions.
THE
The managerial decision about how much to produce is made

d si muJtaneous1, with the
ecision about how much input to employ. Suppose that the capital stock is fixed
(K)
and the firm is jointly deciding the rate of output (Q) and the amount of labor to em-
ploy (L). Given the production function
Q=f(RL) (13-6)
if either Q or L
the firm has onlyisone
specified,
decisionthetoother
make.variable is determined Thus, in a technical sense,
This can be shown in another way. Suppose that the firm is a perfect competitor
in both the output and input markets. Profit maximization" requires that the firm in-
crease output until the price of output (P)
equals marginal cost (MC), and efficient
input resource
product Utilization requires that labor be hired until the marginal revenue
(MRP) equals the unit price of labor (w).
ficiency conditions, It can be shown that these two cf-
P = MC
and (13-7)
MRPJ w
(13-8)
are equivalent; that is, if one is met, so is the other.
Recall that for the perfect competitor, the marginal revenue product is equal to mar-
ginal product multiplied by output price (F). Thus equation (13-8) can be rewritten as
MP) .p=w
Now divide both sides of this e
quation by the marginal product of labor (MPL ),
yielding
MP L (13-9)
The ratio of the wage rate to marginal product is simply marginal cost, so the condi-
P = specified
tion by equation (13-9) is identical to that in equation (13-7). Thus, the rule
MC implies that MRP L = 1V L,
here, the profit maxi and vice versa. Although it is not demonstrated
also equivalent mization rules for a monopolist MR MC
and MR!-'J = w, are
Price of input
Supply of input
(S/unit
P1
Rate of input use
Key Cnncept .. -
• The decision tc produce a given rate of output implies a decision to hire a eer-. -
tam rate of capital and lab r and vice
The. condition for a profit .. maximiziflg output rate (Le. MR MC) is equivalent
to the endi1ion for a profit-maximizing input ra ti (ic.. MRP - input price). -
ECONOMIC RENT
Rent is defined as a payment to any factor of production that has a relatively fixed sup-
p1'.. The term is often used to describe a return to land, and reflects the notion that there
is a fixed amount of land available. 2 As shown in Figure 13.5, when the supply of an in-
put is fixed, the supply function is a vertical line. As a result, the price of that input or
its rent is entirely determined by demand. If the demand curve is D1 , the price will be
P2. Note that regard-
F1 . If the demand curve increases to D2, the price will increase to
less of the change in price, the amount of the input supplied is unchanged.
Economic rem is a payment to a factor of production in excess of the minimum
amount necessary to induce that factor into employment. Suppose that Q in Figure
13.5 represents the number of undeveloped acres in a particular tract of land. Assume
that the cost of developing and marketing this land is equal to P1 . If the demand tune-
2 Although in a strict sense the total amount of land in the world is fixed, from an economic perspective
land is not fixed in supply. The relevant dimensions of the land input are the amount and quality of land in
use. Both of these dimensions of the land input are variable. There is much land that is not used for any
ecosornic purposc. Periodically, some of this land is developed for agricultural or other uses, and thus the
supply of land 'actually is increased in an economic sense.
CHAPTER 13 Pricing and Employment of Input
447
tion is below I), the land will not be used because the market price would be less than
the cost of supplying it. Only if the price is at least P1 will the land be offered for use.
Thus, P1 is said to be the reservation price
for this input. At this price, the land is em-
ploye but there is no economic rent. If demand increased to D2 , the price would in-
crease to P2 , and the rent would be the shaded rectangle defined by
P1P2ha.
Only when the market price of an input is above the reservation price is there an
economic rent paid. Because this return is over and above the amount necessary to em-
ploy the input, economic rent is sometimes said to be unearned. If part or all of this rent
were taxed away, it would not change the resulting resource allocation.
Case Study
Economic Rent iii Professional Baseball
The previous case study discussed how the loss of monopsonv power by the owners of
major-league baseball teams allowed player salaries to increase.
- The All-Salary Teatis (i.e.. the highest paid players) for 1997 included the following:
Player Team
Mm
Contract Years - Average per year
Albert Belle Chicago White Sox 1997-01 $11,000,000
Ken Griffey, Jr. Seattle 199740 8,500,000
Roger Clemens Toronto 1997-99 8,250.000
John Smoltz Atlanta 1997-00 7,750.000
Mike Piazza Los Angeles 1997-98 7,500,000
Barry Bonds San Francisco 1993-98 7,291,667
Source: Bureau of Labor Statistics, Compensation and Working Conditions 2, No.5
(Washington, DC: U.S. Government Printing Office, Fall 1997).
Certainly, a large part of these salaries represent economic rent. The following con-
ceptual model explains how that rent would be determined. Suppose that there are a
fixed number of athletes capable of playing major-league baseball and that each will do
something other than play baseball unless the salary is at least $30,000 per year. Hypo-
thetical de mand-and-supply functions for these athletes are shown in the case figure.
Note the unusual shape of the supply function. For these athletes, the supply function is
horizontal at a price of $30,000 and then is Vertical. This reflects the requirement that the
typical athlete must have this much in salary per year to attract him to play baseball; oth-
erwise, he would take another job. The $30,000 represents the opportunit y cost of play-
ing baseball and is the player's reservation price. Any return above that level is an eco-
nomic rent. The intersection of demand and supply occurs at $1,000,000, and the athletes
are each paid this wage rate. Of this amount, $970,000 is economic rent and $30,000 rep-
resents the payment necessary to attract these workers away from other jobs.0
Annual salary Supply
$ 1.000,000
id
$30,000
Input quantity
S.'
WAGE AND INCOME DIFFERENTIALS

There are large variations among the wage rates paid. to different kinds of labor. Many
unskilled workers are paid the federal minimum wage rate and others apparently have
so little to offer employers that they are unable to find employment even at this low
wage rate. in contrast, it is not unusual for senior lawyers to be paid $200 or more per
hour, skilled surgeons may be paid several thousand dollars for an operation that may
take less than an hour, and the salaries of many professional athletes are thousands of
dollars per game. Why are there such large differentials in wage rates? Are they the re-
suit of market imperfections, such as a lack of information about prices and availability
of competing services, or. are there economic reasons that explain this phenomenon?
Actually, there are important forces on both the demand and supply sides of the input
market that explain must of the differences. Those economic forces are considered in
this section.
Demand-Side Considerations .
All persons are not created equaL.Differeaces in aptitude, intellectual ability, strength.
and other , individual attributes mean that some workers are more productive than oth-
ers. The more productive workers add more to output and to the revenue of the firm.
That is, their marginal revenue product is higher than for other workers,'fl-, ere fore, they
are able to command higher wage rates.
CHAPTER 13 Pricing and Employment of Inpus 419
Some workers, such as salespersons and assembly line workers, are paid in direct
proportion to their output.Thus, their compensation is tied directly to their marginal
revenue product, and the more-productive workers will automatically receive higher
wages than will their less-productive counterparts. Even where pa y is not tied directly
to output in a formal sense, firms generally pay higher wages and salaries to their more
productive workers.
Of course, some firms have plantwide or even industr ywide wage rate schcdules
that prescribe that all workers in a certain job category he paid the same rate. Thus,
one might find differences in productivit y among workers who are paid the same
wage rate. Eve1i in these cases, it is expected that the more productive workers will
earn more in the long run. These workers will he the first to he asked to work over-
time at premium rates and the better workers will tend to he promoted to higher-
paying supervisory positions. Furthermore. the less-productive workers wifl he termi-
nated if their contribution to output is significantl y below that of their counterparts.
Therefore, even in firms where there might be a fixed wage- rate for all workers in a
category, productive workers generally will earn more than sk ill their less-productive
associates.
Supply-Side Considerations
Adjustments in the supply of workers will also result in differential wage tates among
jobs. Consider two jobs, A and B. where the Skill and trainin g requirements are much
different. For example,job A may be a clerk in a retail stoic that requires a high school
education, whereas job B requires a college de gree in electrica.i engined ing. Clearly, it
will he more costly for a person to prepare for job B. The cost of attending college for
four or more years. including the wportunitv cost of the lost iiccme during that period,
is substantial. That investment in education will not be made unless there is an expec-
tation of higher earnings than can be earned in job A.
The wage adjustment for this required training is made on the supply side of the
market. Considerably fewer high school graduates would attend college if they did not
expect to increase their earning power. That is, fewer workers will make the necessary
investment unless there is the prospect of higher earnings than can he earned in com-
parable jobs that do not require this trainingmus, the supply function will be such that
for any given wage, far fewer workers will offer their labor for sale for job B than tor
job A.
Other factors that result in supply-side adjustments in wage rates include risk of
death or injury, working conditions, job content, and hours of work. Jobs that in'.ohe
health risks, poor working conditions, unusual working hours, and!or tedious or strenu-
ous work generally must pay a higher wage rate than jobs with more desir:hk' charac-
teristics. For example, holding other factors constant.a job that required hea' lifting in
a very dirty place on the graveyard shift (midnight to 8 A.M.) would probabl y have to
pay a premium wage rate to attract applicants. As in the preceding example. the wage
adjustment for these job characteristics is made on the supply side of the oarkci At any
given wage rate, more labor will be supplied to a firm offering ajob with desirable char
acteristics than to a firm offering a less-desirable job. Managers are asare of thiesi. fac-
tors and offer hi gh ei wages for less-desirabie jobs.
Key Concepts
• Economic rent isdefmed as the return to a factor of production ui excess of the
payment necessary to keep the factor in its current employment
* wage differentials among workers rerlect demand .-side constderations, such as
greater productivity, and supply-side force including training requirements,
health risks, and working conditions.
Case Study
The Premium for Accepting Risk
The typical answer to the question "What value do you place on your life?" would be
that it is of infinite value. Few people would offer to give up their life in return for a
monetary gain except under the most extraordinary. circumstances. Indeed, one of the
arguments for a military draft is that military service is so risky that it is not possible to
attract adequate volunteers by offering higher pay and benefits. However, the success
of the all-volunteer military in the United States has disproved that notion.
Reflection on individual behavior, particularly with respect to job selection, sug-
gests that people will take all kinds of risks to obtain economic and other kinds of re-
wards. Some recreational thrills, such .s skydiving, hang-gliding, and skiing, can only
be obtained by assuming a significant risk of injury. Many drivers exceed the speed
limit to conserve on time, in fact, even driving at the speed limit is risky. To eliminate
all deaths from vehicle accidents might require speed limits below 10 mph and vehi-
cles built like Army tanks. For most people, the opportunity cost of eliminating traf-
fic deaths is simply too high, and they are willing to take their chances at or above cur-
rent speed limits. Furthermore, people regularly accept work in occupations that
involve significant risk of injury. For, example, those who repair the steam systems of
nuclear power plants often are paid as much as $100 for a few minutes of work. The
reason is because of the exposure to radiation that increases the risk of cancer and
other diseases.
Many civilian occupations are more risky than being in military service during a
war. During the Vietnam War, an average of about 7,000 U.S. servicemen were killed
each year. At that tune, there were some 3 million men and women in the military
forces. Thus, the death rate was about 2.33 per 1,000 employed in that field. Thaler and
Rosen studied death rates and income differentials for a number of high-risk jobs. Sev-
era] of these jobs, including guards and lumberjacks, had annual death rates higher than
that for a military force at war! Jobs such as firefighter and police officer, where risk is
often publicized, are actually less risky than a number of other jobs. Annual death rates
per 1,000 employ ees for some of these occupations are shown in the following table.
CHAPTER 13 Pricing and Emplovnzent of Inputs
431
Annual Annual
Death Rate Death Rate
Per Thousand Per Thousand
Occupation Workers Occupation Workers
Fire fighters 0.44 Mine workers 1.76
Police officers 0.78 Lumberjacks 2.56
Electricians 0.93 Guards 2.67
Crane operators 1.47
SOURCE: R. Thaler and S. Rosen, "The Value of Saving
a Life: Evidence from
the Labor Market; to Household Production and Consumption,
Terleckv (New York Columbia University Press—National Bureau ed, N B.
of
Economic Research, 1975),265-298.
Thaler and Rosen also analyzed thc i

ncremental annual wage associated with
these jobs. This increment measures the trade-off people are willing to make between
risk and income. On average, it was found that about $700 in additional annual wages
are needed to offset the additional risk associated with the incidence of one more death
per 1,000 workers per year. The death rate for lumberjacks (2.56) suggests that the av-
erage worker would have to be paid about $1,800 dollars per year in additional annual
wdes 6. e.. 2.56 x $7001 to accept the additional risk compared to a job where the risk
of death is near zero.0
LA BOR UNIONS
About tO percent of nonagricultural workers in the United States are members of la-
bor unions. Where workers are organized, the nature of management—worker negotia-
tions is different than for firms without labor unions. The union may act as a monopo-
list, that is, as a single seller of labor to the firm.
In this section two topics related to labor unions are considered. First, the econom-
ics of collective bargaining between firms and labor unions are discussed. The second
considers alternative objectives that might influence labor union behavior.
Labor Unions and Collective Bargaining

Labor unions represent workers in negotiations with management concerning wage
rates, fringe benefits, and working conditions. Where the workers of a firm are members
of a union, wages are not set unilaterally by management in response to market condi-
tions. Rather, they are determined by negotiations between management and union
representatives. Of course, general labor market conditions, including wage rates for
comparable workers (both union and nonunion) in other firms and industries, influence
these negotiations.
In some i ndustries, collective bargaining is done on an industrywide basis. Repre-
sentatives from the major companies comprise an industry bargaining team and the la-
bor union has a bargaining team that represents the workers. The result is similar to a
Wage rate
(&'unit) Margipal expendituce
Supply
W,
W,
Demand
MN
Rate of
tabor input
L
bilateral monopoly situation, where a single seller faces a single buyer in a market. The
industry negotiating team acts as a mM1opsonist or the sole buyer.of labor in the in-
dustry, and the union acts as a monopolist or the sole seller of labor in the industry.
The relevant labor demand and supply functions for such a situation are shown in
Figure 13.6 Consider the union as a seller of labor. Industry demand for this labor is
is the
shown by the demand curve DD, the marginal revenue product of labor, and MR
marginal revenue curve associated with the labor demand curve. The supply function
for labor is shown by line SS, and the marginal expenditure for labor is shown by ME.
If it could act without restraint, the industry bargaining team would maximize profit
by equating the labor demand curve and the marginal expenditure curve at point a and
would hire Lm units of labor. At this input rate, the marginal revenue product is equal
L
o the marginal expenditure on labor. The price or wage rate is determined by the point
on the supply curve corresponding to the quantity L,,1 , that is, Wrn. Note that this wage
is less than the marginal revenue product, meaning that workers are not paid their en-
tire marginal contribution to output.
In contrast, assume that a labor union acting as a monopoly seller of labor wants to
maximize the total surplus paid to industry workers. In Figure 13.6, the area below the wage
rate but above the supply curve can be thought of as a surplus paid to workers over and
above the amount necessary to have them offer their labor for sale. In a sense, it is a form
of economic rent. For example, workers would offer L units of labor for sale at a wage of
L 1, units of Ia-
W"' . 1-hat is, W* corresponds to that point on the laboi supply curve where
CHAPTER 13 Pricing and Emplo y ment of Inputs 453
hor are offered for sale. if they are paid a wage above W* ,
they are earning a surplus. The
total surplus accruing to all workers is maximized by equating the marginal revenue func-
tion and the supply curve at point b to determine the quantity hired of
of W. The wage rate W is that point L, and a wage rate
on the industr y labor demand curve corresponding
to quantity L. Not surprisingly, the union approach results in a higher wage rate and more
workers employed than occurs if the management objective is followed.
The exact result of the bargaining process is i ndeterminate. If management has
much more bargaining power than the union, the wage rate may end up being close to
Wm. Con versely, if the relative power lies with the labor union, the negotiations may re-
suit in a wage rate closer to W.
One factor determining the outcome is the negotiating skill of those at the bar-
gaining table. However, other forces will also affect the outcome of the negotiations. For
example, if the economy is in a recession and the firms have built up large inventories,
management may be in a position to weather a prolonged strike. In such a period, la-
bor's position may he weak because of limited availability of alternative jobs for work-
ers should there be a strike. In contrast, in a period of strong demand for industry out-
put, a strike may mean the loss of otherwise large profits. This potential for large profits
may also make it difficult for management to plead poverty at the bargaining table.
Example Wage Rate Determination—The Bilateral Monopoly Case

Teams representing the United Steelworkers of America and management teams rep-
resenting all major steel producers in the United States are negotiating a new labor cOn-
tract. Assume that all workers and all firms are represented by these teams. The industry
demand for labor is w 200 - 2L and the associated marginal revenue function is
MR 200 - 4L
The supply curve for labor and marginal expenditure or labor function are
w 8+4L and ME=8+8L
where w is the daily wage rate and L
is the number of workers employed in thousands.
Determine the wage rate and level of employment that will result from these negotiations.
Solution Because virtually all firms and workers are represented by the two teams,
this is a bilateral monopoly problem. As such, the solution will depend on the relative
bargaining power of the two groups and their negotiating skills, and a unique solution
does not exist. However, two extreme SO!Utiofls can be determined.
First, assume that management is ab16 to dominate the negotiations and dictate the
terms of the contract. Ihe amount of labor hired is determined by equating the labor
demand and marginal expenditure functions
200 - 2L= 8 + 8L
M-1
d solving for L:
L = 19.2
The wage rate is found by substituting L 19.2 into the labor supply function
w 8 + 4(L) = 8 + 4(19.2)
or
w = 84.80
If the union is able to dictate terms of the negotiations, the quantity of labor hired
would be found by equating the marginal revenue and labor supply functions and solv-
ing for L, that is,
200 - 41. = 8 + 4L
or
=
The wage rate would he determined by substituting L =24 into-the labor demand function:
w=200-2L=200-2(24)
or
w 152
The union solution, of course, has both a higher wage rate and a greater employ-
ment than the management solution. In general, neither side would be able to dictate
the tentis of the aveement, and therefore the actual solution would He somewhere be-
tween these two extremes. - -
Union Objectives
In the bilateral monopoly example, it is assumed that the goal of the union is to maxi-
mize the net surplus to its members. This surplus is maximized where the marginal rev-
enue curve intersects the supply curve., Note that this approach is analogous to the mo-
nopolist maximizing profit by equating marginal revenue and marginal cost. In a sense,
the labor union is a monopolist in this labor market and the supply function serves as
the marginal cost curve.
Obviously, the union may have other objectives. For example, it may want to have
all union members fully employed or it may want to maximize the aggregate income of
the membership. To analyze the implications of these alternative goals, consider the de-
mand for labor function and the associated marginal revenue function shown in Figure
13.7. Suppose that there are L2 workers in the union To keep all these workers em-
ployed requires that the firm bargain for a wage rate of W2, the wage rate associated
with L2 on the demand curve.
Conversely, if the objective is maximizing aggregate wages, the union would want
to move to L 1 , the point where the MR curve intersects the horizontal axis. The wage
rate associated with this point is W1 , again determined by the demand curve. In this case,
however, there is an unemployment problem, as L2 - L 1 members of the union will not
be employed. Which members are to be unemployed? Will it he those with the least se-
niority or those with the fewest skills? This is a difficult problem for union leaders to re-
solve, but it is clear that some union leadets have bargained for and obtained wage
agreements that have led to unemployment for a significant number of the union mem-
bers. For example, shortly after organizing as the United Mine Workers, coal miners bar-
gained for and won much higher wage rates, but at the cost of a significantly higher un-
employment rate for mine workers.
CHAPTER 13 Pricing and Emplo y ment of Inputs
4,515
Wage rate
(S,u nit)
WI
W)
Demand for
labor
Rate of labor input
There are still other objectives the union leaders may have, such as maintaining
their positions of leadership and power in the union. Usually, objectives of union lead-
ers are stated in vague terms such as "higher wages," "better working conditions." or
more job security," and it is very difficult to identify the true objectives of the union.
The final result of collective bargaining may reveal more about the true objectives than
would any statement by union leaders.
Key Concepts * ' ... . .......... . . . .. ....

Collective b. irgainiiigcanb thought of asabiiateral . uionopk ituaticin where
the firm arts :is anonnpsonistor the sole buyer of labor and the union acts as
a monqVOW or the sole seller of labor.
• The outcome of collective hartining will depend not only nn the relevant de-
maiid and upp fuaeUou but ;iAai oi the r&atie bargaining skill and strerLgth
Of manageme and union ieders.
• Alternative labor union objectives Include full employment of all members of
the=* maximizatj(n of the net Surplus to workers,, and maximization of ag-
greate wages paid 10 members.
MMMUM WAGE LAWS
Managers in the United States and in many other countries work in an environment
where the gov ernment has set a minimum hourly wage rate for most workers. Such laws
generally require that workers be paid not less than a ecified rate per hour. Although
Marginal revenue product.

wage rate
(S/unit)
Marginal revenue product
Rate of labor
input (L)
LM
most economists argue that such laws reduce total employment, especially for young
workers with limited job skills, these laws have persisted for decades. The appropriate
managerial response to a minimum wage law is the focus of this section.
Consider a firm that operates in product and input markets that are both perfectly
competitive. The firm's labor demand (MRPL) function and the supply curve for labor
arc shown in Figure 13.8. initially, assume that the wage rate (w 0 ) has been set in the
market, and the firm's only decision is the amount of labor to hire. This amount is de-
termined by the intersection of the demand function with the horizontal supply curve
and results in a labor input rate of L0 units per period.
If a government-mandated minimum wage rate is imposed at any wage rate less
than w0, it has no effect on the amount of labor employed or the market wage rate.
However, if a minimum wage rate higher than w0, say w 0, is legislated, that rate effec-
" tively becomes the supply function of labor facing the firm. The firm uses the same
/ profit-maximizing rule (i.e., employ labor until MRPL = w3, but now employs only Lm
units. Employment by the firm has been reduced by L0 - units per period. While
\ those workers who have kept their jobs are better off because their wage rate has in-
\ creased, other workers are worse off because they are no longer employed.
Labor unions have generally supported increases in the minimum wage Thic is
somewhat curious because the vast majority of union workers earn substantially more
than the minimum wage. One explanation is that skilled workers are a substitute for un-
skilled workers. Therefore, the cross-price elasticity between the quantity of skilled
union labor employed and the wage rate for unskilled labor is positive If the price of
unskilled work increases because of a higher minimum wage. the demand for skilled
workers should increase, resulting in higher wages for them. Thus one possible result of
a higher minimum wage is more unemployed low-wage workers and more employed
skilled workers who will be paid a higher wage than before.
CHAPTER 13 Pricing and Employnint of inputs 457
Xc*pts V
* Mlmimi wage tw iirjLy r u*dLt fit ru
dwtg rebQ
4 rirm at e-r1
tne(tb Uie
ioy
SUMMARY
Input prices are determined by the interaction of supply and demand. The demand for
inputs is derived from the demand for the output that those inputs are used to produce.
The marginal revenue product function, computed by multiplying marginal revenue
and marginal product, is the firm's input demand function.
Determining input prices is important because these prices are used in deciding on
the optimal mix of labor, capital, and natural resources to be employed by the firm. The
approach to in put price determination dcpcnds on the structure of both the input and
output markets. A perfect competitor in an input market faces a constant input price.
Thus, the only decision made by the firm is the quantit y of the input to be employed. A
monopsonist has market power'in the input market, and thus faces an upward-sloping
input supply function. The input price and quantity employed by the monopsonist are
determined by the intersection of the input demand function, with the firm's marginal
expenditure on input function.
There is a unique correspondence between output and input decisions. A decision to
produce a given rate of output implies a decision to hire certain rates of capital and labor
inputs and vice versa. The condition for a profit-maximizing output rate (i.e., marginal
cost equals marginal revenue) is equivalent to the condition for a profit-maximizing
input rate (i.e., marginal revenue product equals input price).
Economic rent is a return to a factor of production in excess of the minimum pay-
ment necessary to keep that factor in its present employment, Wage differentials among
workers reflect both demand-side forces, such as differences in productivity, and sup-
ply-side factors, such as training requirements, health risks, and working conditions.
Collective bargaining refers / to the process of determining wage rates and other
working conditions through negotiation between labor union representatives and
management from one or several firms. In some cases, collective bargaining can be
thought of aq a bilateral monopoly situation where the firm acts as a monopsonist, the
sole buyer of labor, and the union acts as a monopolist, the sole seller of labor. The out-
come of collective bargaining will depend not only on the relevant demand-and-
supply functions but also on the relative bargaining skill and strength of management
and union leaders, Objectives of labor unions include full employment of all members
of the union, maximization of aggregate wages paid to members, and maximization of
the net surplus to workers.
Minimum wage laws require that firms pay labor at least a specified wage per hour.
If this wage is above -that prevailing in a perfectly competitive labor market, the firm
will rcducc the amount of labor hired, and unemployment will result.
13-1. What determines the mix of labor and capital used by the firm to produce out-
put? Explain.
13-2. Explain why the term derived demand is used to describe the demand for factors
of production.
13.3. Provide an intuitive explanation for the profit-maximizing rule that the firm
should hire an input until the marginal revenue product of the input is equal to
the additional cost of the input.
13-4. Why is the input supply function horizcntnl for a perfcct competitor in the input
market and upward sloping for a monopsonist?
13-5. Why is there no unique input price and quantity of input hired in the bilateral mo-
nopoly case? What factors will play a role in determining that price and quantity?
13-6. Given that a firm always should be producing efficiently, explain why it is impos-
sible for management to separate the decisions about the rate of output to be pro-
duced and the rates of capital and labor to be employed?
13. 7. The annual salary of many players in the National Basketball Association is more
than $1,000,000. Is it "right" for these athletes to make many times more than the
average American worker? If that average salary were reduced by one-half, how
many players do you think would leave the NBA for other occupations? Explain.
13-8. For many years the United States and other countries used a draft system to
maintain required personnel levels in their military forces. During most of the pe-
riod when the draft existed, earnings in military service, including such fringe ben-
efits as housing and meals, were substantially less than in comparable civilian oc-
cupations. Supporters of the draft system argued that military service was
sufficiently dangerous that regardless of the wage rate, there would not be ade-
quate volunteers. Do you agree or disagree with this position? Explain.
13-9. The leader of a major labor union claims that the union's objective is full em-
ployment for all members. Despite an unemployment rate of 15 percent for work-
ers in this union, labor negotiators are demanding a 6 percent increase in wage
rates for each of the next three years. Workers in other industries have been re-
ceiving wage increases of 3 or 4 percent per year. Are the actions of the union ne-
gotiators consistent with the stated objective of the union? Explain.
Problems
13-1. Suppose that the demand and supply functions for unskilled labor in an economy are
DEMAND: L d 10-0.5w
SUPPLY, L, = 6 + 0.9w
where L is millions of workers and w is the hourly wage rate.
a. Determine the equilibrium wage rate and number of workers employed.
b. If the U.S. Congress sets a minimum wage of $5.15 per hour, what will be the
wage rate and the quantity of labor employed? Compared to your answer for
part (a), how many workers will lose their jobs?
c. If the minimum wage rate is sett $6.00, what will be the wage rate, the num-
ber of workers employed, and the number unemployed? Compare these an-
swers to those in part (a).
CHAPT,R 13 Pricing and Emplo y ment of Inputs 459
13-2. Suppose the demand for centers on NBA basketball teams is given by the equation
Ld=400,000-2W
while the supply of centers is given by the function
Ls
[32 if i-$40,000
I. Oifw <$40.000
where L is number of players and w
is the annual salary.
a. Graph the demand and sup p ly functii-rnc.
b. Determine the equilibrium wage rate for a center in the NBA.
c. Suppose a sudden drop in attendance and number of television viewers shifted
the demand function to
L = 300,000 - 3w
How would the annual salary change? Would the number of workers playing
professional basketball be any different than before the demand curve shifted?
Based on the supply function, what is your best estimate of the opportunity
cost of playing center in the NBA?
1w/11)e labor demand and supply functions for unskilled labor are
Demand: L = 20 0.6w
upp1y: L = 4 + 1.4w
where L is millions of workers and w
is the hourly wage rate.
a. In the absence of any minimum wage legislation, what is the equilibrium wage
rate and level of employment?
b. ¶
minimum wage rate is set at $7.00 per hour, how
much labor will be employ d? What wage rate will workbpaid?
ç. If a governnient .
/ much m ate minimum wage rate is set at $10.00 per hour how
will be embed ? What wage rate will workers be paid?
13-4. Consider a firm in a perfectly competitive output market where the price of the
product is $10 per unit. Given the following information on the total product of
labor function, determine the firm's demand schedule for labor (i.e.,
ompleting the table. byc
Quantity Total
of Labor Product Marginal
Marginal Product Revenue Product
0 0
1 21
2 30
3 38
4 42
5 43
6 40
Suppose this firm is a monopsonist that faces the following labor supply sched-
ule. Use these data to determine the firm's total and marginal expenditure on la-
bor schedules.
Wage Quantity of T)tal Expenditure Marginal Erpenditure

Rate Labor Supplied on Labor on Labor
20 0 -
22
25 2
30 3
35 4
40
50 6
Use these data to determine how much labor the firm should employ. What will
the wage rate be? Explain.
13-5 Demonstrate that, for a monopolist s the profit-maximiing condition that mar-
ginal revenue equals marginal cost (i.e., MR = MC) implies that the marginal
revenue product of capital must equal the price of capital (i.e., MRPK = r).
13-6. A team representing management of all firms in the automatic widget industry
is currently negotiating a new three-year contract with the leaders of the United
Widget Workers of America (UWW) labor union.The industry demand function
for labor (i.e., the marginal revenue product of labor) is
MRPL =20 - 2L
the marginal revenue function associated with the demand curve is
MR =20— 4L
and the labor supply and marginal expenditure on input functions facing the in-
dustry are
w = 5 + 2L
ME = 5 + 4L
where L is the number of workers in thousands and is the hourly wage rate.
a. If the management team can dominate the negotiations and dictate the terms
of the agreement, what wage rate and level of employment will be determined?
b. If the labor union team can dominate the negotiations and dictate the terms of
the agreement, what wage rate and level of employment will be determined?
13-7 The supply of apartments in a city is fixed at 2,000 units; no new units can be built
because of shortage of water. The demand for these units is given by Qd = 5,000
- SP where P is the monthly price of a unit and Q is number of units. The reser-
vation price for each unit is $300 per month.
a. Determine the equilibrium monthly price of an apartment.
b. What is the economic rent per unit and the total economic rent?
c. A government rent-control ordinance is passed, limiting the monthly price to
$400 per month. Determine the number of units demanded at this price. How
might apartment owners respond to a situation when the quantity demanded
exceeds the quantity supplied, but when rent cannot be increased?
13-8. Smith owns the only water-producing well in the county. The well produces 130
units of water per period at zero cost. Demand for water in the county is given
CHAPTER 13 Pricing and Eniplownent of Inputs
461
by Q = 200 - 2P where P is the price per unit and Q is the number of Units de-
manded. Smith's reservation price is $25 per unit.
'a. What is the equilibrium price of water?
b. What is the economic rent per unit and the total economic rent?
c. The county commission passes a law limiting the price per unit of water to
$40. What is the economic rent per unit and the total economic rent now?
d. The county commission passes a law limiting the price per unit of water to
$20. How much water will be sold?
13-9. Assume that all jobs are identicM erept for the risk of being killed at
work.
job with zero risk of death pays $20,000 per y ear, determine the annual wageIfina
each of the jobs listed here. (Use the information provided on page 451 regard-
ing the additional amount necessary to compensate for the risk of being killed.)
Annual Death Rate
Occupation Per Thousand Workers
----------
Police officer 0.78
Crane operator 1.47
Mine worker 1.76
Guard 2.67
Problems Requiring Calculus -

13-10. The market dem and-and-supply functions for Rhubarb Field Dolls are
QD = 2,000 - 250P
and
Qs 600+ lOOP
where QD is quantity demanded. Qs is quantity supplied, and
P is the price.
Adamco Inc. manufactures these dolls using the following production function:
Q=
where Q is the rate of output, K is capital, and L is labor. Adanico is the largest
employer in a small town and faces the following labor supply function:
Hi
where w is the wage rate.
If the capital input is fixed at 225, determine the profit-maximizing labor input,
rate of output., and wage rate.
intQE
Northern Lumber Products, Inc.
PART!
Northern Lumber operates a large lumber-processing 'mill in a small town in
Washington state. It is one of the larger lumber producers in the region and has some
market power in the sale of that product. A recent consulting study has indicated that
the price elasticity of demand for the firm's product is about —3.0. Also, Northern is the
dominant employer in the local labor market, and effectively can be considered as a
monopsonist in the purchase of labor. The firm's labor demand (i.e., marginal revenue
product) function is:
MRP = 1,000 - 2L
where L is the number of workers. Because of its size relative to the labor supply in the
area, Northern faces an upward sloping labor supply function,
= 50 + 0.025L
where w is the daily wage rate.

Once the firm determines the optimal rate of labor input and the wage rate,
The rate of output is determined. The firm uses a cost-plus pricing formula that in-
cludes the price elasticity of demand as a determinant in setting product price. The
same study indicated that average cost is about $300 per unit (1,000 board feet) of
lumber.
Requirements
1. Determine the amount of labor that the firm should employ in order to maxi-
mize profit.
2. Determine the wage rate the firm will have to pay.
3. What price will the firm charge per unit of output?
PART II
Another lumber producer may locate a plant in the same area. If it does, there will
be more competition for labor and the labor supply function facing Northern will shift to
w 50 + 0.04L
Because of increased competition in the market for lumber, Northern's demand func-
tion will become more elastic with the price elasticity of demand decreasing to —4.0.
Integrating Case Study V 463
Requirements
I. Determine the amount of labor that the firm should employ in order to maxi-
mize profit under this new labor market condition.
2. Determine the new wage rate that the firm will have to pay.
PART
Ir
Risk and Jta _tudgeting
CHAPTER 14
Risk and Decision Making
CHAPTER 15
Capital Bni)qet/ng
INT EGRATED CASE

Study VI: Bentley EnterprLes, Inc.
CHAPTER
Risk and1ec4 Making
N Preview
I The Concept of Risk
N Risk and Decision Making
Risk—Return Evaluation Statistics
Risk Preference
• Risk Management
Insurance
Gambling and Insuring: A Contradiction?
Adjusting the Discount Rate
Diversification
Hedging I
• Decision Tree Analysis

• Summary
U Discussion Questions
I Problems
467
468 PART VI Risk and Capital Budgeting
PREVIEW
Before introducing new products, managers undertake consumer surveys, assess prod-
uct lines of their competitors, and closely check estimated production costs. These ac-
tions are taken to increase the information base on which the decision will be made.
Still, there are few sure things in the world of management. Many spectacular failures
have occurred despite management having taken all possible steps to ensure success. It
is simply not possible to predict consumer behavior or changes in production technol-
ogy with complete accuracy. Every decision carries with it the prospect that something
wilt gu wiung iiu iiiii lIjsLeau 01 eitimg iigc plullis dUU 111cluawligmict 1It01UCi va,u
as expected, losses and decline in shareholder value will be incurred. Despite all of the
modern techniques, learning to make decisions where there are risk and uncertainty is
the essence of modern management.
In marketing goods and services, firms often try to reduce the risk faced by poten-
tial buyers. For example, the Ford Motor Company offers a three-year or 36,000-mile
warranty on most important parts of its cars and trucks. The Mauha Kea Beach Hotel
in Hawaii offers its guests one free night of lodging if it rains for 30 minutes on any day
that the guest is at the hotel. These actions reduce the risk to the customer, but they in-
crease risk for the firm.
Managerial decisions are made in different risk environments. In the case of deci-
ion making under certainty, all relevant dimensions of the decision are known. For ex-
ample, in chapter 8 a linear programming problem was formulated in terms of deciding
how much of each of two products should be produced given a specified profit rate for
each and specified processing time required for each product on each of these ma-
chines. All of the relevant information was assumed to be known with certainty. The re-
sult was a unique solution that maximized the firm's profit. That is, there was only one
relevant outcome of this decision.
In decision making under uncertaithy, there are several, perhaps many, outcomes of
a decision, but the probability of each of those outcomes occurring is unknown. For ex-
ample, a firm may be considering the production of an altogether new product. Because
consumers have had no experience with the product, there is no way to estimate the po-
tential demand for it. The marketing department may undertake various market sur-
veys in an attempt to estimate how many people might buy the product, but it is doubt-
ful that much confidence can be placed in such efforts. There may be no way to provide
anything but some rough guesses about demand.
Finally, decision making under risk applies when all significant outcomes of a deci-
sion are known as is the probability of each outcome occurring. For example, consider
the drilling of a well in an already proven oil field. Suppose that 100 wells have been
completed, of which 10 are producing 2,000 to 3,000 barrels per day (BPD), 40 are pro-
ducing 5,000 to 10,000 BPD, and the remaining 50 are producing 10.000 to 15.000 BPD,
This-kind of historical experience allows the decision maker some basis for assessing the
probability of success of the new well.
The focus of this chapter is on decision making under risk. The objective will be to
develop guidelines for making rational decisions given the decision maker's attitude
about (i.e., preference for) risk. First, the concept of risk is formally defined, and then
the principles of probability developed in chapter 2 are used to quantitatively measure
risk. Next, the role of risk in making dccisions is analyzed; basic to this is an under-
CHAPTER 14 Risk and Decision Making 469
standing of the concept of the manager's preference for risk. Then methods for manag-
ing risk are developed and applied. Finally, the use of decision trees is explained,
THE CONCEPT OF RISK

The analysis of risk is based largely on the concepts ofprobability and the probability dis-
tributions that were developed in chapter 2. First, terms such as strategy, state of nature,
and outcome need to be defined. A strategy refers to one of several alternative plans or
courses of action that could be implemented in order to achieve a managerial goal. For
example, a manager might be considering three strategies designed to increase profits:
(1) build a new, more efficient plant that will allow production at lower cost; (2) develop
a new marketing program designed to increase sales volume; or (3) redesign the product
in order to allow lower-cost production and increased sales due togreater consumer ac-
ceptance.A state of nature refers to some condition that may exist in the future that will
have a significant effect on the success of any strategy. In making the decision about build-
ing a new plant, an important state of nature is the economic climate expected to prevail
for the next few years. In this context, the possible states of nature might be (1) recession,
(2) normal business conditions, or (3) economic boom. An outcome specifies the gain or
loss (usually measured in dollars) a ssociated with a particular combination of strategy and
state of nature. For instance, the outcomes associated with the decision to build a new
plant might be the present value of all future net profits. Finally, a payoff matrix shows the
outcome associated with each cdmhination of strategy and state of nature. An example of
a payoff matrix is shown in Table 14.1. For example, if a new plant is built and a recession
occurs, the loss is estimated to be $40 million. Conversely, the combination of a new mar-
keting program and normal business conditions would yield an estimated profit of $35
million. If the probabilities of each state of nature are known, then the combination of
those probabilities and the payoffs for each strategy constitute a probability distribution.
Risk refers to the amount of variability among the outcomes associated with a par-
ticular strategy. Where there is Only one probable outcome of a decision, there is said to
be little risk; where there are many possible outcomes with substantially different dol-
lar returns, there is said to be substantial risk. For example, a manager with $1 million
to invest may be faced with two alternatives. She could buy a one-year Treasury bill
yielding 6 percent interest. At the end of a year, the $1 million investment will return
$1,060,000. The only risk associated with this investment is that the federal government
might be unable to pay its debts. This is so unlikely that there is essentially no risk as-
sociated with this investment (i.e., there is onl y one outcome). Indeed, Treasury bills are
often referred to as a riskiess or risk-free investment.
State of Nature (economic condition s)

Strategy
Recession Normal Boom
New plant $-40 $25 $40
New marketing program —20 35 70
New product design —15 30 60
The second alternative is the drilling of an oil well in an unproven field. If oil is
found, the well will immediately be worth $50 million, the present value of all net prof-
its from selling this oil. If oil is not found, the well will be worth nothing. This is a risky
investment because there is great variability in the range of possible outcomes. In gen-
eral, the greater the variation in outcomes, the greater is the risk.Thus, the definition of
risk is based on the variability in the outcomes of a particular decision.
Key Concepts
Riskexists when there is a range . posiblc outcomes associated with a dcci-
h -prba*ihties of those outcomes occurring tire known, if those
5kM) and the
probabilities are nknown, the decision maker is said idJace uncertainty.
A strategy iz,.A plan or co.rse of action desig n c to .ch i eve m;.mnagennt goal.
A state of nature refers to a condition that ma y exist in the future that will have
a significant effect on the success of a stratest
An outcome is the gain or loss associated with a particular combination of strat-
egi and state of nature.
A payoff matrix lists the outcomes associated with each strategy-state of nature
nmhi n1 inn -
RISK AND DECISION MAKING

Having defined risk and reviewed some of the related terminology, the task now is to
develop quantitative measures of return and risk and to show how they are applied in
decision making. Because individuals have different preferences concerning risk taking,
it also is important that such preferends be identified and their effect on decisions eval-
uated. Obviously, individuals respond to the same risky choices in different ways. For
example, a few people are willing to risk death or serious injury in their recreational
pursuits (e.g.. hang gliding), while others prefer a game of Monopoly. One reason for
these preferences is their attitude about risk. Rational decision making requires that the
expected return be determined and the risk be measured, and that there be information
about the manager's preference for risk.
Risk—Return Evaluation Statistics

Recall from chapter 2 that three statistics were developed to describe a probability dis-
tribution. Given a set of outcomes. X, and the probability of each occurring, P, these
statistics are
Expected value or mean:
p.. = ' 'P1 (X) (14-1)
Standard deviation:
(7 \.PI(X(_.L)2 (142)
I oefficient of variation:
Cr
(14-3)
p.
These statistics have a direct application in measuring the expected return and risk as-
sociated with any business decision for which a set of outcomes and their probabilities
have been determined. The expected value, the standard deviation, and the coefficient
of variation will be referred to as risk - return evaluation statistics.
Suppose that two investments, I and II, are being considered. Both investments re-
quire an initial cash outlay of $100 and have a life of five years. The dollar return on each
depends on the rate of inflation over the five-year period. Of course, the inflation rate
is not known with certainty, but suppose that the collective judgment of economists is
that the probability of no inflation is 0.20, the probability of moderate jnfiaton is 0.50,
and the probability of rapid inflation is 0.30. The outcomes are defined as the present
value of net profits for the next five years. These outcomes for each state of nature (i.e.,
rate of inflation) for each investment are shown in Table 14.2.
Analysis of these investments can be made by calculating and comparing the three
evaluation statistics for each alternative. The expected value (p.) is an estimate of the
expected dollar retuin for the investment. Because risk has been defined in terms of the
variability in outcomes, the standard deviation (o •) is a measure of risk associated with
the investment. The larger o • is, the greater the risk. Risk per dollar of expected return
is measured by the coefficient of variation (v).
The evaluation statistics for each investment alternative are computed as follows:
INVESTMENT I
0.2(100) + 0.5(200) + 0.3(400) = 240
- p.)2 =V .2(100 - 240)2 + 0.5(200 - 240)2 + 0.3(400 - 240) 111.36

111.36
V1 =--1 = = 0.46
240
State of Nature Probability (P 1) Outcome (X,)

Investment I
No inflation 0.20 100
Moderate inflation 0.50 200
Rapid inflation 0.30 400
Investment II
No inflation 0.20 150
Moderate inflation 0.50 200
Rapid inflation 0,30 250
INVESTMENT II
= 0.2(150) + 0.5(200) + 0.3(250) = 205

= V'(150 - 205) 2 + 0.5(200 _205)2 +0.3(250_205)2 = 35.00
35.00
V11 = --- = 0.17
The expected return for investment I of $240 is higher than for 11 ($205). but I is a
riskier investment because a, = 111.36 is greater than a ll 35. Also, risk per dollar of
expected returns for I ( v 11 0.46) is higher than for II (v11 0.17). Which is the better
investment? The choice is not clear. It depends on the investor's attitude about taking
risks. A young entrepreneur may well prefer I, whereas an oldr worker investing a few
dollars in a retirement account where risk ought to be minimiZed might prefer II. The
entrepreneur is in a better position to absorb a loss if it occurs and is probably accus-
tomed to investing in risky ventures.
Key Concept
•e-Given a probabilit y distribution for the outcomes of ahusinessdecision,the sta-
tistics of that distribution can be nv , , I i o evaluate -retn rnd risk:
• -.---The expected value or rue-an is a measure of expected return.
.-The standard deviation a a measure of risk. - -
.—.-The coefficient of variation is measure of risk per ddlar of return.
Risk Preference
Generally, higher returns are associated with higher risk. Indeed, the essence of eco-
nomics is making choices where there is a trade-off. Virtually all investment choices re-
quire giving up expected returns for higher risk or taking more risk in the expectation
of a greater expected return. Thus, any decision will reflect the manager's attitude or
preference for risk, and these preferences differ substantially among individuals. For ex-
ample, some investors are so averse to taking risk that they keep all their assets in bank
deposits insured by the federal government. Others are willing to risk all they own and
can borrow to finance extremely risky ventures such as drilling oil wells or buying con-
tracts in the fast-moving commodities market. There must be a reward for taking risk,
and this comes in the form of higher returns. The following example will help to explain
the concepts of risk and risk preference
Consider the following gamble: "1 will toss a coin 10 times- If a head appears each
time, you must pay me $1,000. If tails appear one or more times, I will pay you a dollar."
Most people would reject this offer even though it is a fair game. That is, if the game
were played many times, the amount lost would equal the amount won. However, the
nature of the game is such that on any one trial the player could lose a large amount
and stands to win only a small amount. The probability distribution for the return or
payoff for each outcome of the game is shown in Table 14.3. The probability of a head
appearing on 10 consecutive tosses of a coin is given by ()10 or 0.001. The probability
that there is at least one tail in 10 tosses is I minus the probability that there would be
10 heads or (1.0 - 0.001) 0.999. These probabilities multiplied by their associated
payoffs yield an expected value of zero; therefore, it is a fair game.
L
Ctn Tq in Eainp1
Probab11i' x
Outcome Probability Payoff Payoff
10 heads 0.001 $-1,000 $—i.Oo
At [east one tail 0.999 ± 1 ±1.00
Expected payoff = 0
To understand why most people would reject this gamble, one must understand the
concept of preference for risk. It is assumed that individuals can associate satisfaction
with money. More money usually means more satisfaction, although for man y peopJ
each additional dollar brings less satisfaction than the previous dollar did. Economists
use the term utthty as a measure of satisfaction. Although an individual's satisfaction is
v irtually impossible to measure, it will be assumed that this can be done for a h
y po-
hetical individual. This assumption will make it easier to develop several important
concepts. One relationship between utility and money is shown in Figure 14.1a. The
curve shows that relationship just suggested; that is. utility increases as money inc
reases
but at a continually decreasing rate. Thus, the function is concave to the horizontal axis.
A person having a concave utilit y function is said to be risk averse-
Suppose that this person ha $10,000 and is offered the opportunity to bet $5,000
on the toss of a coin. If a head is tossed, she wins $5,000; if a tail is tossed, she loses
$5,000. She now faces the choice of having $10,000 with certainty if she chooses not to
bet (this option is called the certain prospect) or of having $15,000 with probability 0.5
or $5,000 with probability 0.5. The latter option is called the uncertain prospect or the
gamble. The expected monetar y value of the uncertain prospect is $10,000, that is,
= 0.5($5,000) + 0.5($15,000) = $10,000
which is the same as the dollar value of the certain prospect.
Now, consider the analysis in terms of the utility associated with each dollar payoff.
For a risk-averse person, the dollar payoffs are translated into utility by using a func-
tion like that shown in Figure 14.1 a. The certain prospect (i.e., having $10.000) is asso-
ciated with a utility of about 34. The expected value in utility of the uncertain prospect
is determined as follows. A $5,000 payoff corresponds to a utility value of 25, and a
$15,000 outcome is associated with a utility of 37. Thus, by substituting these utility val-
ues for the payoffs, the expected value in1terms of utility is computed as
= 0.5(25) + 0.5(37) = 31
This is less than the utility (34) associated with certain prospect. Thus, this utility-
maximizing person prefers the certain prospect and chooses not to accept the wager.
Note that an amount of $8,500 with certainty would also yield a utility level of 31.
This amount is said to be the certainly equivalent of the gamble. The individual would
P ay up to $1,500 (i.e., the difference between the expected value of the gamble and the
certainty e quivalent) in order to avoid the gamble. This suggests a formal definition of
risk aversion. If the certainty equivalent of the gamble is less than the expected dollar
value of the gamble the person is said to be risk averse.
Utility
371
Utility
34 -------
25 H -
12 L
0 Money ($) Money ($)

5,000 8.500 10,000 5,000 10,000 12,000 15
(a) Risk-averse (b) Risk-seeking
Utility
36k-
241— -------
5,000 10,000 15,000 Money($)

(c) Risk-neutral
J— - --
Next, consider a risk-seeking person whose utility function has a continually in-
creasing slope as shown in Figure 14.1b, Here, each additional dollar increases utility
more than did the previous dollar. The utility level associated with the certain prospect
of $10.000 is 12. The expected utility for the uncertain prospect is 16; that is,
lt = 0.5(5) + 0.5(27) = 16
where the utility levels associated with $5,000 and $10,000 are 5 and 27, respectively.
This expected value of utility for the uncertain prospect is greater than the utility (12)
associated with the certain prospect. Thus, this person would prefer the uncertain
prospect even though the expected dollar returns are the same. The certainty equiva-
lent of the gamble is S12.000, determined by the amount of money associated with a util-
ity level of 16 on the utility function. In this case, the individual would pay up to $2.000
to take the gamble.
In general, if the certainty equivalent of the gamble is greater than the expected
dollar value of the gamble, the individual is said to he risk seeking.
Finally, in the risk-neutral case depicted in Figure 14.1c, the expected value of
the utility associated with an uncertain prospect is equal to that of a certain prospect
where both have equal expected dollar values. The expected utility associited with
the uncertain prospect is 24; that is, 0,5(12) + 0.5(36) 24.The utility associated with
the certain prospect of $10,000 also is 24. Thus, the certaint y equivalent of the gam-
ble ($10,000) equals the expected value of the gamble; this person is said to be risk
neutral.
The concept of the preference for risk can be formalized in the following way. At
any time, an individual has a given level of-wealth, W which is known with certainty.The
utility of that wealth is also known with certaint y if the utility function is specified. As-
sume the initial level of wealth is $100, and that the utility function for this person is
given by the natural logarithm of wealth, i.e.,
U=inW
This is simply a mathematical function used to describe this individual's transformation
of wealth into utility or satisfacti'on. Note that the utility of the initial wealth of $100 is
4.61 (i.e., ln(100) = 4.61). Now assume the individual is faced with a gamble that has the
following probability distribution:
Probability (P,) ..t!aioff (X1)
0.5 —80
0.5 +80
This is a fair game because the expected outcome is zero.

After the game is played, the individual's wealth is either $20 or $180, depending
on which outcome occurs. That is, either $80 is won or $80 is lost from the initial wealth
of $100. The expected wealth if the game is played is $100, i.e.,
E(W) = 0.5(20) + 0.5(1 80) = 100
which has a utility of 4.61, that is, the utility of expected wealth is
(4E(W)] = In[E(W)] = In(100) 4.61
The expected utility of wealth is determined as
E{U(W)] = 0.5[U(20)] -- 0.5[U(180)]
= 0.5[ln(20)] + 0.5[ln(180)]
4.10
Given these probabilities and the outcomes, the expected wealth is $100. The util-
ity of expected wealth is simply the utility generated if this person had $100. In contrast,
after the game is played the person actually will have either $20 or $180 in wealth. Each
of these amounts of wealth has a level of utility (i.e., ln(20) = 3.00 and ln(180) 5.19).
By multiplying each of these utility levels by the appropriate probability, the expected
utility of wealth (4.10) is determined.
Now, the preference for risk can be formally defined. An individual is said to he risk
averse if, when faced with a gamble, the utility of expected wealth is greater than the ex-
pected utility of wealth, that is if
U[E(W)] > E[U(W)j Risk Aversion
Clearly, the individual in the preceding example is risk averse by this definition.
An individual is said to be risk seeking if the utility of expected wealth is less than
UI UL111L Ui WAIL, LiIL L
L4E(W)] <.E[U(W)] Risk Seeking

Finally, if the utility of expected wealth is equal to the expected utility of wealth, the in-
dividual is said to be risk neutral, that is,
UIE( W] = E[U(W)] ' Risk Neutrality
These principles can be demonstrated with reference to the utility function U = In
W shown in Figure 14.2. Note that the utility is shown at the initial wealth level ($100)
and at $20 and $180, the levels of wealth after the game is played. The notion that
U[E(W)] > E[U(W)] is another way of saying that this person would prefer wealth of
$100 with certainty to a gamble that had an expected wealth of $100.The risk associated
with the gamble has reduced utility.
Notice that the expected utility of wealth of 4.10 could also be achieved by having
a certain wealth level of $60.04, the certainty equivalent of the gamble. This risk-averse
individual would be willing to pay up to $39.96 (i.e., the difference between expectd
wealth given the gamble and the certainty equivalent of the gamble) to avoid taking the
gamhle.'[his difference is called the risk premium. The fact that risk-averse persons are
willing to pay up to the level of the risk premium to avoid gambles (such as the risk of
loss from automobile accidents, home fires, etc.) creates a market for insurance. As we
will see, because some people are risk seekers in some circumstances (i.e., they have
negative risk premiums), a market for gambling also exists. That is, firms will respond to
risk-averse persons' willingness to pay positive premiums to avoid risk by providing in-
surance on automobiles, homes, and lives. Other firms respond to the willingness of risk-
seeking individuals to pay negative premiums for taking risk by providing various gam-
bling games.
These gambles can be either explicit or implicit. An explicit gamble is one where
the individual seeks out a risk decision such as a bet on a horse race or the purchase of
common stock, An implicit gamble occurs when the risk is incidental to the primary ac-
tivity. For example, a home owner faces the risk that her house might burn down. This
gamble is not sought out but rather is implicit in the decision to purchase the home.
While the risks arise in different ways, they are dealt with in exactly the same manner
by the rational decision maker.
Consider another example, involving a person who has initial wealth of $200 and a
utility function given by
U=w+
10.000
CHAPTER 14 Risk and Decision Making
4.77
Utility
U(w)
U(180) =5.19
1J[E001 = 4.61
E [U( w)] =4.10
U(w) = 3.00
Wealth (w)
100 180 2.
E(w)
Assume that this person faces a gamble where he will win $100 with probability 0.5 or
lose $100 with probabilit y O.S. The probabilit y distribution is shown here:
Probability Outcome (X,) I Wealth (W,) U(W,)
05 +100 • 300 3,000
0.5 —100 . 100 200
Note that this is a fair game because the expected value (in dollars) is zero. Thus, the ex-
pected wealth is $200, the original amount. The utility of expected wealth is 1,000, that is,
U[E( w)1 = U(200) = 200 .. = 1,000

But, the expected utility of wealth is
F[U(W)] = 0.5[U(300)] + 0.5[U(100)]
U(w)
U(300) = 3,000
El U(w)] = 1,600
E[L(w)] = 1,000
U(100)= 20€
Wealth (w)
:.'--
= 0.5(3,000) + 0.5(200) = 1,600

Because the expected utility of wealth (1,600) is greater than the utility of expected wealth
(1,000), this person is clearly a risk seeker and can be expected to take the gamble.
This example is depicted graphically in Figure 14.3. Note that expected utility of
wealth of 1,6100 also cotild he achieved by n certain wealth level of $239. Thus, the
certainty equivalent of the gamble is $239. Therefore, the risk premium is negative
(i.e., the expected wealth,.given the gamble of $200, is less than the certainty equiva-
lent wealth of 239). This means that this risk-seeking person would pay up to $39 to
take the gamble.
There are many people who are risk seekers, especially when small amounts of
money are involved; they are willing to pay a premium for a gambling opportunity. The
success of casinos in Las Vegas, Atlantic City, and other places is clear evidence that a
market for gambling exists.
CHAPTER 14 Risk and D ecision Making
479
C€zse Study
The State Lottery
In recent years, a number of states have developed lotteries in order to generate

more revenue for government programs, especially education. In a
typical state lot-
tery, of each dollar wagered, about $0.05 is used for administrative ex
p ense, $0.45 is
used for state government programs, and $0.50 goes into the pool for prizes fu the
winnerS.
Consider a lottery program that generates $1 million a week by selling I million
tickets at $1 each. A typical distribution of the prize pool of $0.5
million is shown here.
Prize
Value Total
iVumber Probability
(X,) Prize
of Prizes of Winning(P1) Value
$150,000 1 0.000001
75,000 $150,000
2 0.000002 150,000
25,000 4 0.000004
5,000 100,000
10 0.000010
2,000 50.000
25 0.000025 50,000
$500,000
This is
not a fair game (i.e.. one where the expected value of the prize equals the amount
wagered) because the expected value is not zero. The expected return on a $1 ticket is
only $0.50, that is,
= 1pXi = 0.00oOO1 0 50000 ) +0.000002(75,000) +

p.
0.000004(25,000)'+ 0. 000010(5,000) + 0.000025(2,000)

+ 0.999958(0) $0.50
Note that the certainty equivalent (the $1 cost of a ticket) is greater than the expected
value of the gamble. Because millions of people are willing to play these games regu-
larly implies that their behavior should be characterized as risk seeking, at least when
the wagers are for a small amount of money.
Another interesting dimension of these games is that the large prizes often are
paid in the form of an annuity. That is, the $150,000 first prize in the preceding example
may reflect the present value of the awrd, but more often it is advertised as a prize
worth more than $300,000 Consisting of 20 annual payments of $15,000 each. At a dis-
count rate of 8 percent, this annuity has a present value of about $150,000.
Supporters of the state lottery concept argue that it is a relatively painless wa y to
generate revenue for important state programs. Unlike a tax that one is forced to pay.
purchasers of lottery tickets
voluntarily contribute to the government. Further, it is ar-
gued that if people understand that the expected return on a $1 ticket is only $0.50, one
must conclude that the fun and the excitement associated with playing are greater than
the expected loss.
480 PART VI Risk and capital Budgeting
Some critics of state lotteries claim that people do not really understand the prob-
abilities of winning and that the present value of the large prizes is really less than the
amount advertised. Further, there is good evidence that the largest share of the revenue
from these programs comes from low-income people; for example. it is estimated that
about 75 percent is paid by those earning less than $15,000 per year. I
Key Concepts -
• In generatinveStflleflts that offer higher expected returns also involve greater risk.
• Behavior is iU to be
--Risk averse if the certainty equivalent of a gamble is less than the expected
value of the amhie.
—Risk neutral if the certainty equivalent of a gamble is equal to the expected
value of the gamble.
- - Risk seeking if the certainty equivalent of a gamble is ereater than ih '-
pecied dhiC ot the arnb1e.
Equi''alentlY, behavior is
-Risk averse if th e utility of expected wealth is greater than the expected util-
ity of wealth. -
•--Risk neutral if the utility of expected wealth is equal to the expected utility
of wealth. than the expected utility
-Risk seeking if the uiiity of cected wealth is less
f wealth.
RISK MANAGEMENT
Decision makers manage risk in four fundamental ways. First, insurance can be pur-
chased to protect against losses associated with fire; natural disasters such as flooding;
theft; and accidents. Second, hedging by buying various contracts in the financial mar-
kets provides another form of insurance against financial risks such as a crop failure for
an agricultural enterprise, stock market decline, and currency fluctuations. Third, the
firm can reduce risk by diversifying its activity-SO that a decline in one market may be
offset by better conditions in another. Finally, the-differential risk associated with vari-
ous actions being contemplated by the firm can he adjusted for by adding a risk pre-
mium to the interest rate used to discount projected cash flows and'or using certainty
equivalents as the basis for comparison.
Insurance
Firms and individuals buy insurance to protect against the financial loss associated with a
variety of risks, including fire, theft. floods, earthquakes, and death. The market for insur-
ance exists hecause pcopk arc risk averse. Consider a manager who knows that his tire-
also faces a prob-
\L rks manufacturing firm will eani :i net profit of $ 100.0(K) each year but
VW*
-,-
Event
PoaIjl
Outcome
I
Prohabilif v Outcome < Pro/,abi1ir'
Fire S 20,000 0.5 $10,000

No fire 100000 0.5 50.000
IL = $ 60,000
ability of 0.5 that the plant will burn down.' Assume that if the plant burns down, it will cost
$80,000 to replace. Now there are two possibié outcomes: (1) the plant burns down. so net
returns are $100,000 - $80,000 - $20,000; or (2) the plant does not burn down, so the net
return is $100,000. Each outcome has a 0.5 probability of occurrence. The expected mone-
tary value is $60,000, as shown in Table 14.4. Essentially, the manager is facing a gamble.
Suppose that manager's utility function is similar to that shown in Figure 14.4. Point
.4 corresponds to the outcome associatedwith the plant burning down (i.e.. a dollar pay-
off of $20,000 and utility = 100) and point B corresponds to the "no fire" outcome, when
the dollar payoff is $100.000 and the utility is 200. Thus, the expected utility for this un-
certain prospect is 150. That is,
p. = 0.5(100) ± 0.5(200) = 150
Now the certainty equivalent is $40,000 (i.e., $40,000 with certainty generates a util-
ity level of 150, as shown by poitit C in Figure 14.4). So the decision maker would he in-
different between a certain prospect of an income of $40,000 and the gamble just de-
scribed, having an expected dollar value of $60,000.
Utility
triction
200
162
ISO
100
Money
20 40 52 100 (Sl.000s)
Obviousi, this is an unrealistic probabilit y. It is used simply for case of calculation.

482 PART VT Risk and Capital Budgeting
This implies that the profit-maximizing (but risk-averse) manager would pay up to
$60,000 for a fire insurance policypn the factor y, because having this policy would guar-
antee a certain outcome of $40,000 whether or not the plant burned down. That is, the
manager would know with certainty that the firm would net $40,000 (i.e., $100,000 in
profitless the $60,000 insurance premium). If there is a fire, the insurance company will
rebuild the factory at no cost to the firm.
Now consider the situation from the insurance company's perspective. The ex-
pected payout by the insurance company is determined by multiplying the probability
of a fire by the dollar loss, that is, 0.5 X $80.000 = $40,000. This claim plus, say, 20 per-
, ç,. opera t;,-,, '-'"... .1-.. .-. €:..-. :......... 1:.,,..
(flat a LLL Ii1ULcU1'.c f.J1iLy It.Jl L1&1
plant could be offered for about $48,000. The result is that the manager of the fire-
works company could actually have a certain outcome of $52,000 (i.e., the $100,000
profit less the $48,000 insurance premium). This outcome, showi as point D in Figure
14.4, has a utility level of 162 and is preferred to the utility level 150 that corresponds
to the uncertain prospect. Thus, in this example, the manager can choose between the
certain prospect of having a utility level of 162 or an uncertain prospect with an ex-
pected utility of 150. Clearly, the rational manager would opt for the certain prospect
by purchasing insurance.
Although utility is impossible to measure, the best evidence that most people are
risk a'eie and rational is the existence of a large number of insurance companies and
the variety of risks that they insure against. For example, private firms have issued poli-
cies that insure pianists' hands and dancers' legs. Ski resorts can buy insurance to pro-
tect against a lack of snow, and insurance companies paid more than $150 million in
claims when rocket malfunctions sent two communications satellites into the oblivion
of outer space.
Cate Study
Preference for Risk Among Physicians
Jury awards and out-of-court settlements of $1 million or more are not uncommon in
malpractice suits against medical doctors and hospitals. Because one mistake could
mean financial ruin, virtuall y all doctors carry insurance against the risk of malpractice
claims. Some in the so-called high-risk specialties, such as obstetrics and orthopedic
surgery, face premiums in excess of $100,000 per year.
It is estimated that medical doctors pay about $8 billion each year for insurance
protection against malpractice claims. Of this amount, only about 35 percent, or $2.8 bil-
lion, is paid out to injured people. The remaining 65 percent ($5.2 billion) goes for in-
surance company operating expenses and profits. A large part of the expenses are legal
fees to hire lawyers to defend doctors who are sued and payments to the man y expert
witnesses (including economists) who testify in malpractice suits. Thus, the average
physician pays an insurance premium of about $3 for every dollar of loss expected. This
suggests risk-averse behavior on the part of physicians.
483
Consider a hypothetical physician whose annual income is $100,000 after all ex-
penses except malpractice insurance. Assume that the award for any
s
practice suit would be $100,000 and that the probability of that occurringuccessful ma!-
is 0.10. Thus
the expected loss in any one year is $10,000 (i.e.. 0.10 X ,
$100,000). Supp ose that the cost
of an insurance policy against this risk is $30,000.
If the doctor buys an insurance policy, ris k
has been eliminated, and net income
would be $70,000, that is, $100,000 less the $30,000 insurance premium. If
not purchased, income will be (1) $100,000 with p r i nsurance is
where an injured patient wins a lawsuit with p r obability 0.9; or (2) zero, in the case
of the risky aItcrnatj- is $90,000, that is, obability ojo. Thus the expected vahi
P, = 0.9($100,000) ± 0.1(0) = $90,000

If the decision were h.ased solely on the expected value of dollar retur
n. . the physi-
clan would not buy the insurance. But assume that the doctor is risk averse, as indicated
by the utility function
U = 401 - 0.212
where I
is income in thousands, Again, this is simply a mathematical relationship
that
allows income to be translated into utility for this person. The utility associated with the
certain income of $70,000 is 1,820, that is,
U ,40(70) - 0.2(70)2 1,820
which is higher than the expected utility (p.) associated with not having insurance.
The
latter is determined in the following way. First, the probability distribution is outlined
and the two income possibilities ($ 1
00,000 and zero) are converted into utility.
Event Probability Income Utllht
No suit 09 $100,000 2,000
Suit 0.1 0 0
Then the expected value (in terms of utility) is computed:
= 0.9(2,000) + 0.1(0) = 1,800
T
hus, this doctor is willing to spend $30,000 each year for insurance against
Possi ble claims having an expected value of only $10,000. Clearl
behavior. y, this is risk-averse
That v irtually
all doctors have malpractice insurance suggests that they are risk
averse, However, there may be another exp
lanation. Hospitals can be held responsible
for a doctor's malpractice As a result, almost all hospitals require that doctors have
malpractice insurance before they are allowed to use
Of the hospital. This reduces the risk
loss for the Owners of the hospital. Because having hospital privileges is essential for
imost medical practices, one would find that even risk-seeking physicians would have
nsurance. Thus, the high proportion of doctors who have malpractice insurance could
be the resultI of risk-averse hospital administrators forcing their risk preferences on
physicians,
484 PART VI Risk and capiral Budgeting
Utility
itiCn
Money (5)
q
Fall
Gambling and Insuring- A Contradiction?

Individual behavior is generally defined as either risk averse, risk neutral, or risk seek-
ing. Thus, one might expect to find risk-averse individuals spending thcir time review-
ing their insurance coverage and to find risk-seeking persons in casinos in Las Vegas or
Atlantic City. But if a survey were made, it probably would show that many people both
buy insurance and engage in gambling games of one sort or another. For example, it is
probably true that most people who engage in gambling in Las Vegas drove there in in-
sured cars and live in insured homes. This appears to be a contradiction, as it suggests
that people are risk seeking and risk averse at the same time. Actually, there is no con-
tradiction, but a complete explanation is beyond the scope of this book. Suffice it to say
that this behavior depends on the type of gambling games available and the nature and
cost of insurance that can be purchased.
One rather simple explanation for this phenomenon is offered by Friedman and
Savage,2 who argue that the utility function may look as shown in Figure 14.5. It de-
scribes the person who is risk seeking when small amounts of money are involved and
risk averse when larger amounts may be at risk. A person having such a utility function
would engage in gambling games for relatively small amounts of money, but would in-
sure against large losses, such as those associated with a house fire or automobile acci-
dent. A utility function of this type is a sufficient condition for a iatiouai person to gam-
ble and have insurance at the same time.
2 M. Friedman and L. Stvae. "The Utility Analysis of Choices Invo l ving Risk," Journal of Po!ucul
Econom y 56 (1948):279-304.
CHAPTER 14 Risk and Decisin Making 485
Example Risk Preference and Decision Making
Consider.the problem facing a manager who must choose between two investments,
A and B, each having an initial cost of $10. The probabilit
y distributions for the pa y
-ofsreachinvtmsowheflingtab.Ecpyofresnth
present value of all future net profits.
A B
Probability (J') Payoff (X,) Probabi(ay(P1) Payoff (X1)
0.10 $-20 0.20 $10

0.50 20 0.40 20
0.40 50 0.40 30
The decision maker's utility function is U = IOOX - X2 , where K is the dollar payoff.
I. Would you characterize this decision maker as risk seeking, risk neutral, or risk
averse? Explain.
2. if the objective is to maximize expected net present value, which investment is the
better choice? (For the moment, disregard risk.)
3. Evaluate the risk associated with the dollar returns for each investment.
4. If the objective is utility maximization, which investment should be chosen?
Solution
I. Based on the utility function, it is seen that the decision maker is risk averse. This can
be shown in the following way. Select any two dollar payoffs, say, $20 and $40, and
arbitrarily assign them probabilities of occurrence, say. 0.4 and 0.6, respectively. The
expected value of these payoffs is
i.(dollars) = 0.4(20) 4- 0.6(40) = $32
From the utility function, the utilities associated with $20 and $40 are 1,600 and 2,400
respectively. Thus the expected utility value is
p.(utility) = 0.4(1,600) + 0.6(2,400) 2,080
But if this person were offered $32 with certaint y (i.e., a certain prospect having the
same value as the expected dollar value of the investments), the utility would be
2,176. Because the utility is higher for the certain prospect, the person is risk averse.
2. The expected returns for each investment are
P-A 0.10(-20) + 0.50(20) + 0.40(50) = 28
= 0.20(10) ± 0.40(20) + 0.40(30) = 22
Thus,A is the preferred choice because it has the highest expected value.
3. The evaluation statistics for risk and risk per dollar of expected return for both in-
vestments are
Investment A:
= V"0.10(-20 — 28) + 20)2 + 040(50.28) 2 = 21.4

- - ---
- 016
it, - 28 -
Investment B:
if8 = V0(10_22)2 +0.40(2O - 22) 2 + 0.40(20 - 22) 2 75
= = = 034
.t,3 22
The risk is greater for A (ffA = 21.4 compared to o = 7.5), and the risk per dollar
of expected return is also greater for A ( V A = 0.76 compared to v = 0.34).
4. To make the decision, assuming the goal is utility maximization, the payoffs must be
transformed from dollars into utility using the utility function 11 = IOOX - X.2 For
example, the utility associated with a net payoff of —20 is
U = 100(-20) - 20 2 = — 2,400
Repeating this for each outcome will yield two new probability distributions where
the payoffs are in terms of utility.
Probability (P,) Payoff (U,) Probability (I',) Payoff (tJ)
0.10 —2,400 0.20 1,900
0.50 1,600 0.40 1,600
0.40 2500 0.40 2,100
The expected utility returns are
P-A =0.10(-2,400) + 0.50(1,600) + 0.40(2,500) 1,560
= 0.20(900) + 0.40(1,600) + 0.40(2,100) = 1,660
Hence, in terms of utility, investment B is the preferred alternative.
Key Concepts
• The market for insuranceexists because niany peupk arc risk averse.The mar-
ket for garnbling exists because some peuple are risk seekers.
• Some people exhibit risk-seeking behaitr and risk-averse behavior at the
,
same time. One ex planation for thLs sth" t.ey are risk sekecswhen relatively
snmH amounts of money are wolvcd but are risk averse when large amnunts
are at stake.
Adjusting the Discount Rate

Another way to adjust for risk is to use a risk-adjusted discount rate in determining the
present value of the future profits associated with an investment. Given the stream of
future profits, ire, the basic evaluation framework
n -
-" (14-4)
PV -
-F- r)t
is modified by using an appropriate risk-adjusted interest or discount rate for r. Most

investors are willing to accept greater risk only if there is the promise of greater returns
when compared to an investment with iCSS risk. For example, suppose that the typical
Required rate of return

(percent)
30
Is
20
10
Risk (a
LU 20 3.0
return on an insured bank certificate of deposit is 10 percent per year. Clearly, no ra-
tional investor will invest in very risky ventures such as drilling for oil or mining for gold
unless the expected return is considerably higher than 10 percent per year.
Suppose that the line R in Figure 14.6 shows all combinations of risk and return for
which a hypothetical investor is indifferent. That is, this function shows the williuguess
of this investor to trade off risk against return. Clearly, the shape of this function will
vary for different individuals, depending on their preference for risk. A very risk-averse
- person might have a trade-off function similar to the dashed line R'—any increase in
risk must carry with it a significant increase in retu rn. Conversely, another individual's
trade-off function might be described by line R". Here only a small increase in the rate
of return is required to compensate for a rather large increase in risk.
Assume that the rate of return associated with a riskiess investment is 10 percent. Re-
call that a riskiess investment would have a standard deviation of zero. For the investor
with trade-oft function I?, if the risk increases to, say. cr = 1.0, a 15 percent rate of return is
required. The difference between this 15 percent return and the riskiess rate of 10 percent
is referred to as the risk premium. If a 2.0, the trade-off function R indicates that a 22
percent return is required for this person. Thus, the risk premium is 12 percentage points.
In evaluating investments, these differential discount rates would be used to evaluate the
Present value of future profits, That is, net cash flows for a high-risk investment would be
discounted using a higher discount rate than would be used for a low-risk alternative.
It should be emphasized that there is no equation or table that relates risk and the
discount rate. Clearly, there is a positive relationship between these two factors, but the
relationship between them is strictly judgmental and must be made by individual deci-
scn nia1ccr.
Example Using Risk-Adjusted Discount Rates

Suppose that management at Showrnax Theaters must decide whether to expand by
building a few large theaters in large cities or building a number of minitheaters in small
cities and towns. Each of the alternatives would require an initial investment of $2 mil-
lion. Although the large theaters have a greater expected return, this option has greater
risk because there is more competition in the larger cities. In contrast, there is less po-
tential for profit in the small markets, but in many of them there is little or no competi-
tion. The expected value of the net profits in each of the next 10 years is $600,000 per
year in the large markets and $500,000 per year in the small markets.
The value of a is estimated to be 1,5 in the large markets and 0.5 in the small cities.
Management has a risk - return trade-off function similar to that shown by curve R in
Figure 14.6. This means that a rate of about 17 percent would be used to discount cash
flows in the large-city alternative and a rate of about 12 percent would be used to dis-
count cash flows in the small-city alternative.
Solution The net present value of an investment is determined by subtracting the
initial cost of the investment from the present value of future profits. Note that the ap-
propriate risk-adjusted discount rate is used in each case.
Large-City Alternative:
10 6W,000
NPv = —$2,000,000
$2,795,162 - $2,000,000 - $795,162

Small-City Alternative:
to 500,000
NPV i(i±0.12)t - $2,000,000
= $2,825,112 - $ 2,000,000 = $825,112
Although the nondiscounted cash flows each year are greater for the large-city alter-
native, when adjusted for risk, the present value of the cash flows is greater for the small-
city alternative. Thus, building minitheaters in smaller cities is the preferred investment.
Diversification
The firm can reduce risk by diversification of its assets. That is, it can invest in a variety
of activities so that a decline in one area of. the business might be offset by an expan-
489
sion of another. For example, farmers generally plant several different crops as protec-
tion against a single crop failure
or an unusually low price on one crop. A firm provid-
ing hotel services might decide to enter the movie business, reasoning that if for some
reason travel declined and people spent more time in their home cities, they might go
to more movies. The products are also complementary; the firm could show its own
movies in its hotel rooms.
The principle of risk reduction through diversification is most easily understood by
thinking in terms of buying two securities, A
and B, which have the same expected rate of
return of = 0, 10 or percent, and the same
lO level of risk, ie = cr8 = 0.20. As-
sume that 50 percent oftheortfolio
p is es
inv
ted in sset
a A (i.e., w = 0.5) and 50 percent
is invested in asset B.(i.e., w =
O.50).Thus, the weighted average return on the portfolio is
TP = w 4 r 4 + wBrB=
(145)
0.5(0.10) + 0.5(0.10) = 0.10
or 10 percent.
By investing in both assets, the expected return is no different than had onl y
one of
the assets been selected. The advantage of diversification is that risk can be reduced
with no reduction in expected rate of return. Recall that the risk on A and B
or 20 percent. The risk on a portfolio of two assets Is given by was 0.20.
= (14-6)
where CrAB
is the correlation between the returns on the two investments (i.e., the de-
gree to which the returns move together over time). If the returns are perfectly corre-
lated, where 0A,6 = + 1.0, then the risk on the portfolio is
0.20
Note there has been no reduction in the level of risk that would have been faced by
investing in only one of the assets. But if the correlation is less than perfect, then risk
will be reduced. To reduce risk, the returns on the two assets have to be less than per-
fectly correlated. For any given value of crAB, the
risk for this portfolio is
V0.02 + 0.02 crAe
Now, the maximum reduction in risk is achieved when there is perfect negative corre-
lation. That is, when a AB = —1.0.
Now, the risk on the portfolio is zero, i.e,
cr Vöö^o,o2 T l) = 0
Here risk has
been completely eliminated, The portfolio risk for various other values of
aAB is given below:
Q4R +0.50 = 0.17

O'J)
+0,25
o' = 0.16
Grp0.l4
AB = —0.25
o' = 0.12
QB = 0.50 0.10
o'
This is an important result and was an instrumental part of the work in finance that led
to the Nobel Prize being awarded to Harry Markowitz, Merton Miller, and William Sharpe
in 1992. The essential point is that by developing an efficient portfolio of assets, the risk as-
sociated with a given level of return can be reduced or, alternatively, the return on a given
level of risk can be increased compared to a nonefficient (i.e., a nondiversifled portfolio).
Thus, one way for the firm or individual to reduce risk is through diversification of
assets. Individuals can invest in a variety of companies; firms can invest in several differ-
ent product lines or have stores in various parts of the country or world. If the portfolio
of assets is structured correctly, risk can be reduced with no reduction in expected return.
Hedging
It is possible for firms to insure or hedge against adverse price movements by using fu-
tures contracts. These contracts involve entering into a contractto buy or sell a speci-
fied quantity of a good of a given quality at a specified price in the future. For example,
an agricultural producer in eastern Kansas plants a wheat crop many months prior to
its being ready to harvest. At the time of planting, there are a variety of possible risks,
including having a poor crop due to bad weather conditions and having the price de-
cline. It is possible to purchase insurance against crop failure, thus eliminating that risk.
Further, the producer can eliminate most price risk by entering into what is called a for-
ward contract (i.e., essentially, a contract to sell the. wheat crop at a specified price at the
time of harvest regardless of what the market price is at that time), or a futures contract.
The latter is a contract traded in such places as the Chicago Board of Trade and the
Chicago Mercantile Exchange, where the producer buys a contract that specifies the
right to sell a given amount of wheat at a specified price in the future.
Contracts in these markets trade continuously, and the value of the contract will go
up or down in inverse relationship to the price of wheat. If the cash price of wheat in-
creases, the right to sell at a specified price becomes worth less, and if the cash price goes
down, that right to sell becomes more valuable. The producer ultimately will sell wheat
at the current cash price, but if that price has gone down, the producer will make a profit
on the futures contract that will offset the loss associated with the decline in the cash
price. If the cash price goes up, the value of the futures contract declines so the additional
profit from selling at the higher cash price is offset by the decline in the value of the fu-
tures contract. If set up correctly, all price risk to the producer is virtually eliminated.
For example, assume the cash price is $4.00 per bushel and the producer plants 500
acres in September and expects a yield of 100 bushels per acre for a total expected crop
of 50,000 bushels- The farmer is confident about production costs and has decided that
an acceptable profit can be earned if the crop could be sold at about the current cash
price of $4.00 per bushel. Of course, output will not be available until the following Au-
gust. In the futures market, assume the price of wheat for the August contract is $4.00
per bushel, and the producer buys a contract to sell 500,000 bushel of wheat at that
price. This brokerage cost of these contracts would be about $1,000 or about $0.02 per
bushel. Essentially, the producer has paid $1,000 to guarantee a sale of the crop at $4.00
per bushel for a total of $200,000. Should the price of wheat fall to $3.00 per bushel the
"mere are some additiocal considerations that complicate these transactions, including the need to
maintain a cash reserve with the broker in the event that the value of the contract moves in the wrong
direction.
CHAPTER 14 Risk and Deci,sion Making
491
farmer will sell the crop at the cash price for $150,000. The value of the contract will
have increased by $50,000 and the farmer will sell the contract and capture that profit.
Thus, to.al revenue will be $200,000 on the two transactions. Should the cash price rise
to $5.00 by August, the crop will sell at that price for revenue of $250,000, but the fu-
tures contract will be sold at a loss of $50,000, again providing net revenue of $200,000.
By using the futures market, the producer has paid someone else to assume this price
risk. That risk has been transferred to a speculator, who is betting on being able to predict
future movement in the price of wheat and make a profit. The wheat producer in our ex-
ample can sell wheat at a net price of $3.98 per bushel (the $4.00 contract price less the
$0.02 per bushel commission) with certainty or opt for the risky alternative where the ac-
tual selling price is unknown until the time of sale. Some firms use hedging to reduce risk,
while others do not. Part of this is explained by their different preferences for risk. Another
part may be ex p lained by ignorance of the ways to reduce risk through this mechanism.
There are futures markets in a wide variety of commodities (e.g., so y beans, cattle,
sugar, oil, gold, lumber, etc.) as well as for a variety of foreign currencies and common
stock indices such as the S&P 500 and the Nasdaq indexes. A fairly recent development
is a futures market in electricity (e.g.. the California Oregon Border Electricity contract).
I,
Ke'
• Risk can be managed by V
• --Purchasing insurance to c&er losses in case of fire, thefts flood, or loss of a

key anaer. V
—Adjusting the rate used to discount futu cash ftow to their pre.sent value.
V - Divers ific atj g of assets.
—
V
V
V -
'—Purchase of forward and/or futures contracts that guarantee aspecified price.
DECISION TREE ANALYSIS

The previous discussion of risk began with a set of outcomes for each investment and
the probability of each outcome occurring. However, some strategic decisions involve
a sequence of decisions, states of nature, and possibly even subsequent decisions. Given
this complexity, how are the alternative strategies to be evaluated? One approach is to
use a decision tree that traces the sequence of events and/or decisions that lead to each
outcome. Such a diagram shows two or more branches at each point where a decision
or event (i.e., a state of nature) leads to the various outcomes, These branches are sim-
ilar to those on a tree. Understanding the concept of a decision tree will help in under-
standing how the data in a probability distribution are determined.
To illustrate the concept, consider the following gambling game. An initial bet of
$10 is made. A card is drawn from an ordinary deck of playing cards and then a coin is
tossed. If the player draws a heart and then tosses a head, he receives $30, for a net pay-
off of S20. If a heart is not drawn, but a head is tossed, the player is paid $15, for a net
payoff of $5. For all other outcomes, the player loses his initial $10 bet.
When drawing the card, the relevant probabilities are 0.25 that a heart is drawn
(there are 13 hearts in a deck of 52 cards, so 13/52 = 0.25), and the probability that a
Spade, club, or diamond is drawn is 0.75. When tossing a coin, the probability of a head
and the probability of a tail are both equal to 0.5.
Net
payoff Probability
0 1-lead 10 0.125
Flip
coin
Tail -10 0.125
Dra-
card
c
C sJHead -I-s 0.375

Flip
coin
050
Tail -10 0,375
•••• -• .....,.
The game can be analyzed using the decision tree shown in Figure 14.7. The prob-
ability of the outcome at each step in the process is shown on the branches of the tree.
The probability of each final outcome is equal to the product of the probabilities along
the branches leading to that outcome. For example, the probability of drawing a heart
and then tossing'a head for a net payoff of $20 is 0,25 X 0.50 0.125. By repeating this
for each path along the decision tree, the probability of each outcome is determined.
By associating the net payoff for each outcome with its respective probability, the
probability distribution is determined. Then the usual evaluation statistics that de-
scribe expected return, risk, and risk per dollar of expected return can be computed.
The decision tree approach can be directly applied to managerial decision making.
Suppose a firm is considering entering a new market. This entry would require building
either a large or small plant. This decision is shown in part I of Figure 14.8. A square is
used at each branch to show decisions. Note that there are no probabilities associated
with such decisions. A diamond is shown at those branches associated with the various
states of nature that may occur.A probability must be assigned to each of these branches.
In this example, there are two stochastic elements associated with each decision
(the term stochastic refers to an outcome that is determined by chance): (1) the reac-
tion of a major competitor in the business and (2) the economic conditions that will pre-
vail. Suppose it is learned that the competitor may respond to the new plant with a new
national or regional advertising prOgram or with no new advertising program. Assume
that the probability of each occurring will depend on the size of plant built, as shown in
Table 14.5. The probabilities for each alternative competitive reaction are entered on
the appropriate branches of part II of Figure 14.8. Note that the probabilities depend
On the size of plant built. If a large plant is built, the probability is high that the com-
petitor will respond with a major advertising program. Conversely, building a small
plant would result in a regional program or no program at all.
CHAPTER 14. Risk and Decision Making
4.93
I Outcome
I Probahljljtjes
IBOOM
• ( 1) 80
Normal
/o0E2: 0
0. (4) 100 0.06

Normal (0.60) (5)
0t2
N-20
J2Q.
Normal (0 60
- _Q__
--- ------------------------
2A\ —"-"
No (n
cS) (10)
(11) 0
(I2)
,F] Dec; sion point (0. I (13) as
-0-
Normal (0.6())
0 Stochastic point
0.24
-5
(16) _3_ 0.18
(17)
Managerial decision Competitor reacti (18)

reaction: condition 5
size of advertising 0.06
program
II H
III I
IV
Pro babilitv That Competitor Responds wlth.

Plant Size National Program Regional Program
Large plant No New Program
0.70 0.20 0.10
Suppose that the possible economic conditions and their probabilities are recession
(0.10), normal business conditions (0.60), and boom conditions (0.30). These states of
nature and their respective probabilities are shown in part Ill of the decision tree. Fi-
nally, the outcome for each branch is the present value of profits under the states of na-
ture (i.e., competitor's reaction and economic conditions) along that branch.
Note that there is a payoff (i.e., profit) associated with each of the 18 possible com-
binations of decisions and stochastic events. Each combination consists of a plant size, the
competitor's reaction, and an economic condition.1bese payoffs have been listed in part
IV of Figure 14.8 together with the probability of that sequence occurring. The probabil-
ai e found by multiplying the probabilitY along each of the branches leading to the
ities
outcome. For example, consider the top branch of the decision tree. The probability that
the competitor responds with a national advertising program is 0.70, and the probability
of an economic boom is 0.30. The product of these probabilities is 0.21, and this is shown
d ha side of the flgure
n
as the first entry in the probability distribution on the right
There are two distinct probability distributions outlined in Figure 14.8. The first,
consisting of the payoff_prObabthtY pairs (1) through (9), corresponds to the decision
to build the large plant. Pairs (10) through (18) correspond to the decision to build the
small plant. Note that the probabilities sum to one for each distribution. y, and
, based on these probability distributions, the evaluation statistics,
Finally
for each investment alictflatiVe can be computed. For example, consider the compu-
v
tation of these statistics for the large-plant alternative:
+ 0.06(100) ± 0.12(40)
0.21(80) + 0.42(0) + 0.07(-40)
+ 0.06(60) + 0.01(20) = $31.8 million
+ 0.02(-20) + 0.03(120)
31.8
These statistics for both plant sizes are summarized in Table 14.6. As usual, the decision
expected
The return or expected net present
will not be an easy one for management .
value of all future profits increases with the scale of the plant, but so does risk, as mea-
For example, the expected return on the large plant, $31.8 million, is
sured by a' and v.
higher than the expected return for the small plant, but the risk per dollar of return (1.4)
is also higher-The manager's preference for risk will be a fundamental factor in making
the decision. --
Alternative Plant Size

LargeSmall,
(dollar figurer In millions)
-:
$31.8 $16.3
Expected return 60 $8.9
Risk (a)$43.6 0.55
Risk pet dollar of 1.4
expected return (v)
-----------
CHAPTER 14 Risk Decision Making 495
Clearly, decision makers must keep the problem in proper perspective. One way is
to include only those alternatives and outcomes that are relevant and significant in
terms of both the probability of occurrence and the net payoff. Consider including the
possibility of terrorist actions as a possible state of nature when evaluating a decision.
For military planners, this state of nature is both relevant and significant. However, for
most business decisions such a possibility probably is not relevant, even though the pay-
off associated with this outcome could be a large negative number. -Mat is, the firm may
incur huge losses in the event of such an attack, but this outcome is sufficientl y im-
probable that it is not relevant for most business decisions.
Even with careful attention to including only relevant decisions and stochastic fac-
tors, decision trees can easily become very large and complex. The total number of out-
comes is equal to the product of the number of decisions or stochastic alternatives along
each branch on the tree. For example. a decision to build a new pl ant might involve six
different plant-size alternatives, five different ways a competitor might react, and three
different economic environments. Hence the number of outcomes in this case would be
90 (i.e., 6 x 5 x 3). One could easily imagine decision trees with thousands of possible
outcomes. Whereas a computer would have no problem computing the probabilities
and evaluation statistics for each of a large number of intricate decision trees, the man-
ager's capacity to evaluate the options is limited.
key Concepts................
' A decision true traces sequences of strategie. and stats of nature to arrive it a
stof outcomes. Th probability or an y outcome is found by j.miultipjying the
pi obahi!iies on each branch leading tn that nztconie.
is a probability distribution fo r each . hoce i ncluded in a decision tree.
SUMMARY
A decision maker faces risk when there are several outcomes associated with a decision
and the probabilities of those outcomes are known. The greater the variation in those
outcomes, the greater the risk. The set of outcomes and their associated probabilities
comprise a probability distribution. If the probabilities of the outcomes are not known,
the decision maker faces uncertainty.
The evaluation of a decision where risk is present is made by determining the ex-
pected value (), the standard deviation ( o'), and the coefficient of variation (v) of the
probability distribution for the investment. The expected value estimates the expected
return, the standard deviation measures risk, and the coefficient of variation measures
risk per dollar of expected return. In general, decisions that promise higher expected
returns also carry greater risk.
Behavior is said to be risk averse if the certaint y equivalent of a gamble is greater
than the expected dollar value of the gamble. In contrast, behavior is risk seeking if the
certainty equivalent of a gamble is less than the expected dollar value of the gamble. A
market for insurance exists because many people are risk averse. A market for gam-
bling exists because there are risk-seeking individuals. Some people buy insurance and
also engage in gambling games because they are risk seekers when small amounts of
money are involved and risk averse when larger amounts are at stake.
A utility function indicates the amount of utility or satisfaction a person receives
from wealth or income. It can be used to determine the individual's preference for risk.
Risk aversion is indicated by a utility function that increases but at a decreasing rate. A
risk-neutral person has a utility function that is linear, and the utility function for a risk-
seeking person increases at an increasing rate.
Risk can be managed by using insurance, diversifying, and hedging with various fi-
nancial instruments. Risk can also be managed by using a risk adjusted interest rate in
discounting cash flows. This luvuives adding a risk premium to the rate used to discount
future profits. The risk premium increases as the risk associated with the decision in-
creases. An alternative method for adjusting for risk is to evauate the certainty equiv-
alent of a risky decision.
A decision tree traces the sequence of strategies and states of nature to determine
the set of outcomes associated with a decision. The probability of any outcome is found
by multiplying the probabilities on each branch leading to that outcome. There is a
probability distribution for each choice included in a decision tree.
14-1. Explain how the principle of risk aversion can be used to explain why the owner
of a business would buy insurance on the life of a key employee.
14-2. Assume the risk-free interest rate (e.g., the rate on 91 -day 'Theasury bills) is 5 per-
cent. What interest rate would you use to discount the projected profits for each
of the following investments? Explain. (Hint: What risk prernium .would you add
to the risk-free rate in each case?)
a. A new motel iii your city.
b. An exploratory oil well in an unproven oil field..
c. A videotape rental store in your home town.
14-3. One manager has said: "Because an individual's utility cannot be measured, there is
no reason to consider it when making decisions." Do you agree or disagree? Explain.
14-4. Provide two examples of decisions made by administrators at your college or uni-
versity where a decision tree could have been used.
14-5. In what areas can a firm use insurance to reduce risk?
14-6. Firm A sells one product to a single group of consumers, whereas firm B sells a
variety of products to different groups of consumers. Which firm is taking more
risk? Explain.
14-7. What are two examples of decision making under risk and two examples of deci-
sion making under uncertainty?
14-8. Consider a gamble that involved winning or losing an amount equal to your en-
tire initial wealth. Would you take this gamble if the probability of each outcome
was 0.5? What does this imply about your preference for risk? Explain.
`ob1ems
14-1. Do each of the following distributions meet the requirements for a probability
distribution? Why or why not?
(a) F, X, (b) F,. X, (c) P1 X,
—0.10 10 0.30 —40 0.20 4
—0.20 15 0.30 —50 0.40 8
0.30 20 0.40 —200 0,30 0
0.50 40 0i5 12
0.50 50
14-2. For each of the following probability distributions, calculate the expected value
(i.'.), standard deviation (a), and coefficient of variation (v).
(a) r, x, (b) P,
0.8 20 0.-1, 0
0.2 —5 0.2 20
0.4 15
0.3 30
14-1 For each of the following probability distributions, where F, is probability and
X1 and Y1 are outcomes, determine the expected values
(u and u), the standard
deviations (a and a,), and the coefficients of variation (v and v).
(a) F, I F, (b) i' x.
0.20 —5 15 0.10 0 50
0.40 0 29 0.20 10 40
0.10 10 30 0.60 20 30
0.30 20 50 0.10 30 20
14-4. Given the following probability distribution for returns on an investment in the
common stock of.two firms:
Percentage Returns
Probability Amalgamated Gold First National Bank
Compute the expected value, standard deviation, and coefficient of variation for
each investment.
14-5. Given the following probability distribution for the percentage returns on the
common stock of General Motod and S&P 500 Index, determine the expected
return and standard deviation for each.
Percentage Returns
Probability General Motors S&P Index
0.2 —10 I.)
0.3 10 5
0.5 20 10
14-6. A gamble consists of a choice between two games. The first game requires rolling
an ordinary die and receiving $10 times the number shown. The second game in-
498 PART V1 Risk and Capital Budgeting
volves flipping a coin and receiving $70 if it comes up heads and nothing if it comes
up tails. Which game would be selected by a person with utility function U = in W?
What choice would be made if the utility function is U = W27
14-7. Consider a person with initial wealth of $1,000 who has a utility function
U=InW
where W is wealth. Suppose this person is offered a gamble that involves win-
ning $100 with probability 0.50 and losing $100 with probability 0.50.
a. Compute the level of utility at the initial wealth.
b. Given that the gamble is accepted, compute the expected wealth, the utility
of expected wealth, and the expected utility of wealth.
c. Will this person take the gamble? Explain.
d. What is the certainty equivalent of the gamble.
14-8. A sales representative for the Rapid Vacuum Cleaner Company knows from
past experience that she will sell a vacuum cleaner in two of every five homes
in which she gives a demonstration In a typical day she will give demonstrations
in 10 homes. If it costs $15 to give each demonstration (this includes all implicit
and explicit costs) and if her commission is $60 for each cleaner sold, determine
expected net profit per da
14-9. The Lac DuFtambeau Corporation manufactures a broad line of fishing tackle
and accessories. The firm has surplus cash and is considering the acquisition of a
firm that manufactures fly rods. A broker identifies two firms, Flyrite and Perfect
Rod, that could be purchased for the same amount. The probability distributions
for the present value of all future profits for each of these acquisition candidates
are as follows:
Flyrite Perfect Rod

Probability Outcome Probability Outcome
0.40 0 0.20 100,000
0.20 200,000 0.50 200,000
0.40 500,000 0.30 250,000
a. Make a complete investment analysis of each investment (i.e., compute the

expectedpresent value of all future profits, a measure of risk, and a measure
of risk per dollar of expected returns).
b. Which of the two firms should be acquired? Why?
14-10. For each of the following functions relating utility (U) and money (M), deter-
mine if the function implies risk-averse, risk-seeking, or risk-neutral behavior.
Explain.
a. U=M+0.25M2
b.U=1OM
c. U = 500M - 2M2 (relevant only for M250)
14-11. A gambler has initial wealth of $200, a utility function U U2 , and is offered a
gamble involving a bet of $100. The payoffs are $10,$50, and $200, with proba-
bility 0.4,0.4, and 0.2, respectively.
a. What is the investor's utility if the gamble is not taken?
CHAPTER 14 Risk and Decisirn Making 499
b. If the gamble is taken, what is the expected wealth? What is the utilit
y of ex-
pected wealth and the expected utility of wealth?
c. What is the certainty equivalent of the gamble?
d. Will this person take the gamble if he is a wealth maximizer? Explain.
e. Will this person take the gamble if he is a utility maximizer? Explain.
f. What is the most this person would bet in order to take the gamble?
1442. United Steel, Inc., has experienced losses for several years. The production de-
partment has determined the Only hope for reversing the negative trend in prof-
its is to build a new plant outside the United States. Two alternative locations
have been identified. Production in country A would be very efficient and low-
cost, but the government is unstable. In the event of a revolution, United's assets
might be appropriated. Country B has a much more stable government, but pro-
duction costs are considerably higher than in A.
The following data show the probability distribution for the present value of
all future profits for each alternative.
Country Country B
0,40 0 0.40
0.20 20 0.40 20
0.40 60 I 0.20 30
a. Make a complete analysis of both alternatives. Which alternative should be
chosen? Why?
b. If the collective utility function of the board of directors is
U= 200X— X2
where Xis measured in millions of dollars of profit, which alternative should be
selected? Explain.
14-13. Management at Unique Publishing has been paying an annual premium of
$3,000 for fire insurance on their $100,000 plant. Net profit for Unique consis-
tently is $100,000 per year after deducting the insurance premium. Any unin-
sured loss due to fire would be an expense in computing net income. A study has
shown that the probability of fire during a year is only 0.004. (Assume that any
fire would destroy the plant.) A consultant suggests that Unique cancel its fire
insurance. Evaluate this recommendation for each of the following assumptions:
a. The sole objective of management is to maximize profit (i.e., no considera-
tion is given to risk).
b. The sole management objective is maximizing utility and the relevant utility
function is U = 100ir - 0.51r2, where 'rr is net profit in thousands of dollars.
c. Given the utility function from part (b), compute and compare the expected
utility of profit and the utility of expected profit. What does this tell you about
the risk preferences of management at Unique Publishing?
14-14. An individual has a total wealth of $1,000 including a home valued at $250. The
probability that the home will be totally destroyed by fire in any given year is
0.2. An insurance policy that would cover 100 percent of any fire loss is available
for $60, Assume the individual's objective is utility maximization.
a. If the individual's utility function is U = 2 W, should the insurance policy be

purchased? Explain.
b. If the individual's utility function is U = \/i should the insurance policy be
purchased? Explain.
c. If the insurance company's expenses are 20 percent of claims paid out, what
is the company's expected profit on this policy?
14-15 State University is considering an evening MBA program. Dean Stephens de-
termines that a minimum enrollment of 100 in a graduate program is necessary
for a breakeven operation. As the only MBA program in the state, enrollment
probably would be 125 students. Southwestern, a private university, is the only
other school in the area that offers graduate programs. The dean thinks the
probability is about 0.67 that Southwestern would respond to the State program
with their own MBA program. If it did. enrollment probably would only be 70
students in the State program. However, if State offered both MBA and MPA
(Master of Public Administration) programs, Dean Stephens thinks that enroll-
ment would be 175 with no competition and 90 if Southwestern offers a pro-
gram. Because of similarities in course requirements, the dean thinks that the
MBA and MPA programs could be run as one graduate program. The dean be-
lieves that offering both degrees would reduce the probabihly that Southwesi-
era would enter the MBA degree market to about 0.2.
a. Construct a decision tree that shows each possible outcome and the proba-
bility of its occurring.
b. Determine the expected number of students in State's MBA program if of-
fered by itself and as a combined MBA—MPA program, where both degrees
would be offered.
14-16. The manager of the student union building has to decide between introducing a
low-cost pizza outlet or a gournet sandwich shop in the basement of the build-
ing. The university's enrollment may increase (with probability = 0.8) or de-
crease, and there may be a tuition increase (probability = 0.6) or decrease. An
enrollment increase would be good for sales of either type of food unit. A tuition
increase would have a negative effect on sales but would affect the sandwich
shop more than the pizza stand.Annual profits for each alternative and state of
nature are shown below.
Pizza Stand Profit:

Tuition
Up Down
Up $100 $200
Enrollment
Down 50 90J
Gourmet Sandwich Shop:

Tuition
UP Down
Up $60 $300
Enrollment
Down 40 160
a. Set up the decision tree for this problem.
b. Make a complete investment evaluation of the two alternatives. 'Which
should be selected? Explain.
14-17. Acme Manufacturing is considering three alternatives, A, B, and C, for a new
plant. Cost data for each of these plants are shown here. The cost of shipping one
unit of output to the market will vary, because each plant would be at a differ-
ent location.
A B C
Annual fixed cost $100,000 $200,000 $300,000
Production cost per unit 20 18 15
Shipping cost per unit 5 6 6
The quantity sold will depend on economic conditions for the next year, as indi-
cated here:
Economic Condition Probability Quantity Sold
Normal 0.8 100,000
Recession 0.2 80,000
The per-unit price of the product will depend entirely on the price set by Zenith
Steel, the primary competitor in the market. It is thought that the probability of
various prices being set will depend on the plant built by Acme, as shown here:
/ Probability That Zenith
Responds with a Price oft
Acme Builds Plant $20 $25 $30
A 010 0.40 0.50
B 0.30 0.30 040
C 0.60 0.20 0.20
Because of Zenith's dominance in the market, whatever price it sets will be fol-
lowed by Acme.
a. Construct a decision tree showing the first-year net profit for Acme for each
combination of plant size, economic condition, and pricing reaction.
b. Determine the expected first-year profit for each plant alternative. Ignoring
risk, which plant should be built?
14-18. Sharp Products, a major paper recycling firm, must replace its processing equip-
ment.The only alternatives are the Century Processor, a very efficient but somewhat
unreliable machine, and the Suresliot Processor, a less efficient but almost repair-
free piece of equipxnenLme two machines have equal initial costs and have a three-
year life. Due to rapid technological change, the machines will have no salvage value
at the end of that period. Sharp's engineering department has estimated the
expected net cash flows associated with each machine over their three-year lives.
Expected Net
Cash Flows for Year
Machine 1 2 3
Century 300 400 400
Sureshot 250 350 450
502 PART Vi Risk and Capital Budgeting
The management at Sharp uses a 14 percent discount rate for most equipment
purchases. However, because of the unreliability of the Century Processor, a dis-
count rate of 18 percent is appropriate for that machine. Which machine should
be purchased? Explain. (Assume that all cash flows occur at the end of each year.)
14-19. A firm has two investment options, with the expected return and risk for each
shown in the table below.
Dynotronics Macrop hone
Expected return (r) 0.12 0.09
Risk (T) 0.11 0.07
If the firm puts 50 percent of its investment funds into each, what is the portfo-
lio rate of return and the portfolio risk if:
-a. The correlation coefficient for the returns on the two investments is 0.50?
b. The correlation coefficient for the returns on the two investments is -0.50?
14-20. An investor is offered the opportunity to buy shares of stock in one of two com-
panies, both of which have a 20-year life. Company A pays an annual dividend of
$6.70 per share and has a r = 1.0. Company B pays an annual dividend of $10 per
share and has a a 3. The investor uses the following formula to determine the
decimal equivalent of the risk-adjusted rate to use in discounting future dividends:
r OAO + 0.02r
The investor can buy shares in either company at $55 per share.
a. Which shares if any should be purchased? Explain.
b. What is the maximum price that this investor would pay for a share of each
stock?
Computer Problems
The following problems can be solved by using the TOOLS program (download-
able from www.prenhail.com/petersen) or by using other computer software.
14-21. High-Risk Strategies, an international commodity trading company, is faced with
four alternative strategies for trading wheat. The profit for each strategy will de-
pend critically7 on theweather conditions (i.e., the state of nature) prevailing dur-
ing the next 12 months. The set of profit outcomes for each strategy - state of na-
ture combination is outlined here:
Weather Strategy
Conditions
(probability) A B C D
Hot-dry (0.10) $ 2,800,000 $1,700,000 $-2,410,000 $-3,625,000

Warm-wet (0.20) 7,400,000 1,920,000 -3,440,000 -5,100,000
Warm-dry (0.30) 4,520,000 1,800,000 2,140,000 1,460,000
Cool-wet (020) -2,900,000 1,400,000 6,240,000 7,800,000
Cool-dry (0.20) -2,100,000 1,100,000 5,110,000 6,125,000
Evaluate each alternative and determine which strategy should be implemented.

14-22. United Fabricated Products is faced with the decision to build a new plant near
the site of a proposed new steel mill. Three plant sizes are under consideration-
small, medium, and large. Unfortuilately, if United is to build the plant, it must
make the commitment now before it is certain that the steel mill actually will be
built. If United delays the decision, its option on the only available site will ex-
pire, and a major competitor, Walters Steel Products, will exercise its subodi-
nated option on the ]and. The bët estimate available sug g ests that the proba-
bility the mill will be built is &.'Z5. The costs are: small plant. $12.1 million:
medium plant, $18.0 million; and large plant, $25.7 million.
The profit outcomes for each possible outcome and the initial cost of each
plant site are shown here:
Present Value of
Profits (millions) for Each
Plant Size
Steel Mill .S'n"li Medium Lame
Built $14.6 $29.4 $45.7
Not built -3.4 -11,5 -26.5
Should United build a plant and, if so, which size? Explain.
14-23. Secure Money Managers limits its investments to five securities. The annual es-
timated rate of return on each of these for each state of the economy is shown
here. The probabilities of the latter are: recession, 0.12; moderate economic
growth. 0.65; and rapid economic growth, 0.23. Compute the expected return.
risk, and risk per Unit of return for each security.
Security___________________
Economic International United Outland
Condition Southern HA
Meats Trucking Steel Telephone Inc
Recession 11.4 -14.0 -9.5 11.1 5.0
Moderate growth 11.9 9.7 11.0 12.0 13.0
Rapid growth 12.2 21.6 16.2 12,4 14,2
14-24. Chi-Town Promotions must select a rock group for its next concert tour. The fi-
nancial and marketing vice-presidents have estimated the following probability
distributions for profits for each of four groups. Which should be selected?
The Grandpas and
Electric Banana Zinc Dirigible the Grandmas Hug
Prob. Profit Prob. Profit Prob.Profit Pro!,. -Prof"
0.73 $2,450,000 0.42 $-5,400,000 0.22 $1705,000 Q'tW\
0.12
0.17 -1,400,000 0.21 1,650,000 0.23 2,124,000 0.23 -1,024,000
0.10 200,000 0.13 4,800,000 0.55 1,925,000 0.29 2,950,200
0.24 6,450,000 0.36 8,100,200
CHAPTER
capil
I Preview
• Mumndzatlon of Shareholder Value and Capital Budgeting
I The Capital budgeting Process
Projecting the Cash Flows
Evaluating the Capital Project
Capital Rationing and the Profitability Ratio
Linear Programming and Capital Rationing
• The Cost of Capita!
Cost of Debt Capital
Cost of Equity Capital
The Composite Cost of Capital
• Mergers and Acquisitions
Types of Mergers
Merger Incentives
Merger Procedures
N Summary
N Discussion Questions
• Problems
CHAPTER 15 Capital Budgeting .505
PREVIEW
Most ofthe principles of managerial economics covered thus far have focused on short-run
decisions. For example, determining the profit-maximizrng price and output rate are im-
portant short-run decisions. Of equal importance are decisions about the nature of the fin-n
in the long run. What new items should be added to or eliminated from the product line?
Should old capital equipment be replaced? Should a new plant be built or should a com-
petitor be acquired to broaden the product line or increase the number of retail outlets?
Decisions such as these are absolutely crucial to the lon g-runprofitability of the firm
Recall that the objective is to make those decisions that will maximize shareholder value,
which has been defined as the present value of all future profits or net cash flows. To meet
this objective, assets must be continually deployed and redeployed to capture new profit op-
portunities. For example, General Motors' entry into the credit card market and the dcvcl-
opment of cable television services by several of the regional telephone companies (e.g..
NYNEX and Bell Atlantic) represent significant new resource commitments by these firms.
In this chapter, the firm's long-run decisions are considered. First, the nature of cap-
ital budgeting decisions and the process of making long-run investments are consid-
ered. Next, alternative evaluation techniques for evaluating capital projects are devel -
oped and applied. Several sp
ecial topics are considered, tnc!uding capital budgeting in
an inflationary environment and allocating limited capital funds under capital ra-
tioning. Methods for estimating the cost of capital funds are reviewed, including the
Capital Asset Pricing Model (CPM), which has proven to be one of the fundamental
developments in the modern theory of finance. Finally, mergers and acquisitions are dis-
cussed. This topic is a special case of capital budgeting. For example, a firm may decide
to enter the computer business by building a new plant, staffing it with new production
personnel, and developing a marketing force. Alternatively, the firm may seek to ac-
quire (or merge with) an existing firm that already is in the computer market. Consid-
eration of either alternative requires the application of capital budgeting principles.
MAXIMIZATION OF SHAREHOLDER VALUE

AND CAPITAL BUDGETING
Capital projects are those that are expected to generate returns for more than one year.
Capital budgeting refers to the process of planning capital projects, raising funds, and
efficiently allocating resources to those capital projects. Examples of capital projects in-
clude new factories, machines, automobiles and trucks. and computers. Outlays for re-
search and development and advertising Programs are also capital expenditures if the
returns on those projects will flow for more than one year.
It is useful to categorize capital projects in the following way:
1. COST REDUCTION. Investments in training, machinery, or other capital assets
that reduce the cost of producing output.
2. OUTPUT EXPANSION: Investments that accommodate increased output in re-
sponse to actual or expected increases in demand.'
'The same investments may result in both output expansion and cost reduction sinnuitaneously. This
certainly would be true it there are increasing returns to scale in p roduction. In that case an increase in
output due to a capital expenditure necessarily would reduce
the average cost of output.
3. EXPANSION BY DEVELOPING NEW PRODUCTS AND/OR MARKETS:

Expenditures for the development and production of new products and/or the de-
velopment of new markets by adding sales staff or by opening new outlets.
4. GOVERNMENT REG ULATION: Expenditures made to meet government
safety, environmental, and other rules.
Capital projects typically are very costly. Indeed, many firms seek external financ-
ing to implement a capital spending program. Furthermore, most capital spending pro-
jects are not easily reversed. For example, once a new manufacturing plant is built, it
may have no other use than the one intended. Thus, it would be difficult to e1l if a
change in market conditions rendered it unnecessary. For these reasons, capital plan-
ning decisions may determine the course for the firm for many years to come.The prob-
lems associated with a poor pricing decision or incorrect etimate for one production
run may be rectified quickly. Generall y, this is not true for a major capital project. if cap-
ital spending decisions are poorly made, the firm's existence may be threatened.
Capital budgeting requires information on sales, production costs, advertising, and
availability of funds, and therefore generally involves all areas of management. Fur-
thermore, because of its critical long-run importance, the capital budgeting process usu-
ally is reviewed on a continuous basis by the top management of the firm.
The principle underlying capital budgeting decisions is that expenditures are made
until the marginal return on the last dollar invested equals the marginal cost of capital.
The cost of capital is the rate that must be paid on money raised externally by the firm
(e.g., by borrowing or selling stock) or the opportunity cost (i.e., the foregone return)
on funds the business has that it would have invested elsewhere or that the owners
could have spent on consumption.
Recall that the value of the firm is equal to the sum of all future net profits reduced
to their present value using an appropriate discount rate, that is,
R
Value = (15.1)
where R1 is revenue in the tth period, C1 is cost, and r is the discount rate, which is the firm's
opportunity cost of the funds used to make the investment. The basic capital budgeting
question is: Will the capital expenditure, which itself will increase the firm's costs, increase
revenue and/or reduce other costs sufficiently to increase the value of the firm? If the rate
of return on the project is greater than the marginal cost of capital, the value of the firm
will increase, and therefore the proposed capital project should be implemented.
Suppose that management has identified and evaluated each of the capital projects
shown in Table 15.1. The projects are listed in decreasin g order of rate of return. The
schedule can be used to determine the quantity of capital demanded by the firm de-
pending on the cost of capital. Thus, these data represent the firm's demand for capital
function, which is shown in Figure 15.1.
Using the marginal revenue/marginal cost principle, the firm will invest as long as
the return on an investment is equal to or greater than the cost of capital. For example,
if the cost of capital is 20 percent, none of the investment projects would be imple-
mented because that cost exceeds the highest return (18 percent) available on a capital
project Thus if the firm borrowed $4.5 million at 20 percent interest to implement the
first capital project, the annual interest of $900,000 would exceed the average, annual
CHAPTER 15 Capital Budgeting 507
Type of Capital Cost Rate

Project-Description Expenditure (millions) Return (0/a)
I. Replacement of obsolete Cost reduction
equipment on an assembly line $4.5 18
2. Opening of three new stores on Market expansion 6.2
the East Coast 17
3. Formal training programs for all Cost reduction 1.5
production employees 14
4. Replacement of outdated heating Cost reduction
and air-conditioning system 3.9 12
5. New production facility Output expansion 5.1 10
return ($810,000) on the project, and both profit and the value of the firm would be re-
duced. However, if the cost of capital was 15 percent. both projects I and 2 would he im-
plemented because their rates of return are 18 and 17 percent, respectively In this case,
the firm would spend $10.7 million on these two capital items. If the cost of capital was
lower, the firm would make even more capital outlays.
The marginal cost of capita! function (MCC in Figure 15.1) shows the cost to the
firm of obtaining various amounts of capital. Determining the firm's Cost of capital is
discussed later in the chapter. The cost is shown as a gradually rising function because
most firms are required to pay a higher cost to obtain increasing amounts of capital. For
example, if the firm is borrowing those funds, the more that is borrowed, the greater is
the risk that the firm will he unable to repay the lender. To be compensated for taking
that additional risk, the lender must earn a higher return, which is accomplished by in-
creasing the interest rate charged to the firm.
Figure 15.1 indicates that the firm would make investments 1,2, and 3 because the
rate of return exceeds the cost of capital for these projects. Note that the rate of return
on investment. 3 is 14 percent, and the rate of return on investment 4 is 12 percent.
Rate of return,
of capital (%)
cost
Dr ni and for
capital
Capital
(millions)
However, because the firm's cost of capital is now more than 12 percent, investment 3
is made but 4 is not (i.e., in the figure, the MCC curve lies above the demand for capi-
tal schedule to the right of project 3).
}ey Concepts
Capial budgetin g refers tl process of planning capital projects, raising
funds,.and i.fflcientiv allocatmg those funds to capital projects.
I •. Capital expenditures are made in order to reduce cost, increase output, expan..
into new products or markets, and/or meet government regiJations.
• In general, capital expenditure .are tnade until the rate of return on the last do!-
lar invested equals the marginal cost of capital
• The demand for canital function shows ihe amount of carnithl snrndmg that wil!
e made at each cost of capital
THE CAPITAL BUDGETING PROCESS

in most firms, capital budgeting is a continuous process, with proposals being made reg-
ularly in all areas of the organization. Typically, each level of the organization has au-
thority to make capital expenditures up to some dollar limit. Proposals in excess of that
limit must be screened and approved by higher levels of management to ensure that the
projects are consistent with the overall plan for the firm. For example, the production
line foreman may have authority to make capital expenditure decisions up to a level of
$25,000. Proposals above that amount must be approved by the plant manager, who
may be able to authorize expenditures up to $250,000 but must have approval from the
division vice-president for projects that cost more than that amount.Thus, the larger the
dollar cost of the proposal, the more screening steps it must go through.
The first step in this evaluation is to determine the cost of the project. Next, the net
cash flows are estimated. Finally, evaluation techniques are used to compare cash inflows
to the cost of the project. In the following, it is assumed the cost of the project has already
been determined so that the focus is on projecting and evaluating the cash flows.
Projecting the Cash Flows

Suppose that a firm is considering the addition of a new product line. It is estimated that
the cost of new production machinery, reorganization of the production line, and addi-
tional working capita] for inventory and accounts receivable will require an initial in-
vestment of $20 million. The breakdown of these costs is as follows:
Machinery $15,000,000
Reorganization 2,000,000
- Working Capital 3,000,000
Total $20,000,000
Sales revenues from this product are projected to be $12 million the first year, but
because the new product is a substitute for one of the firm's existing products, the
incremental sales revenue will be only $10 million. That is, the new product will add
$12 million to sales, but sales of an existing product will fall by $2 million, yielding a net
or incremental sales increase of $10 million.The marketing staff estimates that sales rev-
enue will increase 10 percent each y ear. The research and development department has
indicated that in 5 years, it can develop an entirely new product to replace the one un-
der consideration. Therefore. management has decided to assume a 5-yea, life for this
product and that any evaluation of the proposed project be limited to a 5-year period.
The production and ergineering departments of the firm have developed the fol-
lowing incremental production cost estimates. Variable costs will be 40 percent of rev-
enue and additional fixed costs will he $100,000 per year. The F;,, dcpartmat i--
ports that for depreciation purposes, the new machinery has a 5-year life and that the
straight-line depreciation method will be used. That is, the annual depreciation charge
against income is one-fifth of the initial cost of the machinery, or $3 million per year:
annual depreciation charge = - cost = $ 15.000,000= $3.000,000

years 5
The financial vice-president also indicates that the combined federal and state marginal
income tax rate for the firm is expected to be 40 percent. Note that depreciation is a
noncash expense that is included as an expense in comouting income taxes but is then
added back in to determine the net cash flow, The only reason it is considered is because
of its effect on the firm's income tax liabilit y, The salvage value of the equipment at the
end of the 5 years is estimated tp be $4 million. Also, at the end of the 5-Year life of this
project, the $3 million that was needed for additional working capital is recovered. To
simplify the analysis, it is assumea that all revenues and expenses occur at the end of
each year.
The detailed cash flow projection for each y ear is shown in Table 15.2. Note that
there has been no consideration given to the costs of financing this project. The defini-
tion of cash flow when making capital budgeting decisions is the after-tax cash flow, as-
suming that the firm has no debt. The reason for this is that when using an appropriate
evaluation technique, future cash flows are discounted at a rate equal to the firm's cost
Year
Sales $10,000,000 $11,000,000 $12,100,000 $13,310,000 $14,641,000

Less: Variable costs 4,000.000 4.400,000 4,840.000 5,324,000 5,856,400
Fixed costs 100,000 11J,Q00 100,000 100,000 100.000
Depreciation 3,000.000 3,000,000 3,000,000 3,000,000 3,000,000
Profit before tax $ 2,900,000 T3,500.000 $4,160,000 S 4,886,000 $ 5,684.600
Less: Income tax 1,160.000 1.400.000 1,664.000 1,954,400 2,273,840
Profit after tax $ 11740,000 $ 2,100,000 5 2,496,000 $ 2,931,600 $ 3,410,760
Plus: Depreciation 3,000,000 3.000.000 3,000.U00 3,000,000 3,000.000
Net cash flow $ 4,740,000 $ 5.100.000 $ 5.496.000 S 5.931,600 S 6.410.760
Plus: Salvage value of rnachiner 5 4.000,000
Recapture of working capital 3.000.000
Net cash flow ( y ear 5) ¶1 $.410.76()
of capital. This automatically accounts for the financing costs. To include such costs as a -
subtraction from the net cash flows would effectively count them twice, and thus would
be incorrect.
Evaluating the Capital Project

The remaining problem in this example is to rationally compare the projected net cash
flows to the initial $20 million investment that is required so that the proposed project
can be accepted or rejected. In order to maximize the value of the firm, the evaluation
ek,..I.l
1. Consider all relevant cash flows.

2. Discount cash flows using the firm's opportunity cost of capital.
3. Select the one project from a set of mutuall y exclusive projects that maximizes
the value of the firm.2
4. Allow each project to be evaluated independently of all others being considered.
Three evaluation techniques will be considered: payback period, internal rate of return,
and net present value. As will be shown, the payback period method, although still used
in some circles, fails to meet any of the four criteria just outlined and will be dismissed
as a viable method. Under most circumstances, the internal iate of ietuin method yields
value-maximizing results, but it fails to meet criterion (2) and sometimes (3) and (4).
Thus, it is inferior to the net present value method, which meets all four criteria and will
always guarantee that the project or projects selected for implementation will maximize
the value of the firm.
Payback Period Under this method, the number of years that it takes for the net
cash flows (undiscotmted) to equal the cost of project is defined as the payback period.
The decision rule is to select the project tl,at has the shortest payback period. Consider
the projected net cash flows for two proposed capital priects, A and B:
Initial Payback
Project Cost 1 2 3 4 5 Period
A 1,000 500 500 1,000 1.000 1,000 2
B 1,000 250 250 500 5,000 10,000 3
Note that for project A, the sum of the cash flows equals the cost of the project in 2
years, whereas it takes 3 years in the case of project B. Thus, using the payback period
rule. project A would be selected. However, even the most casual analysis would indi-
cate that project B should be preferred because it has much larger cash flows in periods
four and five. The net present value will be higher for B than for A for any normal dis-
...ount rate. The payback period method fails because it does not consider all cash flows
and does not discount those flows to their present value. Obviously, this method is a
poor evaluation technique that warrants no further consideration.
2 Mcuuallv exclusive means that the selection of one project from a set of alternative projects precludes the
others from bein g implemented. For example. a firm ma y he considering alternative design proposals for a
new plant where only one will be selected.
Internal Rate of Return The implicit rate of return on a capital expenditure can be
measured using the internal rate of return (IRR) evaluation method. The internal rate
of returi is the discount rate that equates the present value of the cash flows with the
initial investment cost. Define A1 as net cash flow in year t, C as the initial cost of the
project, and n as the life of project. Now the IRR is determined by setting
.1 A -1
=C (15-2)
and solving this equation for r*, the internal rate ef re. turn.The decision TUiC is that if 1*
is greater than the cost of capital, the investment should be made, That is, profits, and
therefore the value of the firm, will be increased by making this investment because the
firm is using capital that costs, say, 12 percent to earn a return greater than 12 percent.
Clearly, solving equation (15-2) for r* by hand is difficult when ii is greater than 1.
Fortunately, firancial calculators and easily used computer programs are available that
can solve IRR problems quickly. In their absence, trial-and-error methods can also be
used. To illustrate the trial-and-error approach, arbitrarily select a discount rate and
evaluate the present value of the cash flows. If the present value is higher than the cost.
increase the discount rate and repeat the process. If the present value is lower than the
cost, reduce the discount rate. By gradually adjusting the discouitt rate, this iterative
process will ultimately lead to the IRR.3
Again using the data from t1e example in Table 15.2, the IRR method would re-
quire setting up the equation
4,740,000 5, 100,000
± + 5,496,000
- + 5 ,931,600 13,410,760
- + ------ + = $20,000,M)
1 + r* (1 + r*)2 (1 + r*) 3 (1 ± r* )4 (1 +
and solving for r*, which yields r* = 0.179, or 17.9 percent. Because this rate exceeds
the cost of capital (12 percent), the investment should be made.
Net Present Value The net present value (NPV) method of evaluation consists of
comparing the present value of all net cash flows (appropriately discounted using the
firm's cost of capital as the discount rate) to the initial investment cost. If the present
value of the cash flows exceeds the cost, the proposal meets the evaluation criterion:
that is, the value of the firm will be increased by making the investment. Equivalently,
if !VPV>0 the implicit rate of return on the capital expenditure exceeds the firm's cost
of capital, and thus future profits will be higher if the investment is made. Using the
NPV method, the rate of return on the investment is not determined explicitly.
Therefore, it is referred to as an implicit rase of return. If the net present value is nega-
tive, the implicit return is less than the cos t of capital.
Formally, the NPV rule is that if
NPV = - C >0 (15-3)
"There are spectal cases where it is possible that two or more discount rates will equate the present value of
the cash flows to the cost.That is, multiple values of the internal rate of return are obtained. This can occut
in unusual situations where cash flows vary from positive to negative from one period to another. Such
cases are so special as not to be of concern in this text. Interested readers are refrrod toT. Copebnd and 1
Westoi, Financial Theo, and Corporate Policy, 3rd ed. (Reading. MA: Addison- Wesle
y, 1988).
the capital expenditure should be made. For the example shown in Table 15.2. suppose
that the firm's cost of capital is 12 percent.The net cash flows at the end of each year
are given in the table. Note that at the end of the project's life (i.e., 5 years), the cash in-
flow from the salvage value of the machinery and the recapture of dollars tied up in in-
ventory and accounts receivable are included. Based on a cost of $20000,000. the NPV
calculation would be
14,740,000 5,100,000 5,496.000 5,931 600 13,410,7601 L0,000,000
NPV + (1.12)2 + +
+
23,589,040 - 20,000,000 = 3,589,004
As the NPV> 0, the proposal passes the test. This means That the return on the new
product line exceeds the firm's cost of capital, and the investment should be made be-
cause it will increase the value of the firm.
Comparison of NPV and IRR Evaluation Methods Generally, when evaluating a
single project, the JRR and NPV methods will yield consistent results. That is, if the net
present value is positive, the internal rate of return will be greater than the firm's cost
of capital, and vice versa. This is because for net present value to be positive, the implicit
rate of return on the pi'ojeci roust be greater than the discount rate. Therefore, for any
single project, the two approaches usually give the same accept/reject signal.
However, in the case of two or more mutually exclusive projects (i.e., where only
one of the investments will be made), the two evaluation techniques can result in con-
tradictory signals about which investment will add more to the value of the firm. Con-
sider the cash flow and evaluation criteria values for investments A and B in Table 15.3.
Project A has a lower net present value but a higher internal Tate of return. The reason
for this inconsistency is that there is a difference in the implied reinvestment rate for
the annual cash flows each year. That is, the cash returns that flow from the capital pro-
ject each year are reinvested, but the rate earned on those reinvested dollars is not
known. Therefore, an assumed rate must be used. In the NPV approach, it is implicitly
assumed that the cash flows are reinvested at an interest rate equal to the firm's cost of
IABLE 35.3 , Cph

Data lir'ca
- -
'
ProjectA Prjeci B
Initial cost $1.000 $1,000

Net cash flows (year)
1 450 —300
2 450 0
450 300
A. 450 500
5 450 2,000
Evaluation criteria
NPV(12% discount rate) 622 675'
IRR 34,9 24.2
capital. Under the IRR method, the implicit reinvestment rate is the computed rate of
return on the capital investment being considered. This is an optimistic assumption be-
cause the rate of return on the firm's investment opportunities in future years may not
be as high as for the project being considered. Thus, the NPV method offers the more
conservative approach.
The. NPV method meets all of the four criteria presented previously and will always
select that project or projects that maximize the value of the firm, whereas under some
circumstances, the IRR technique will riot. For this reason, the NPV approach is preferred.
Essentiall y, the value of the firm will be maximized by aggresivciy dcvclopimg cap-
ital project proposals, carefully estimating their initial cost and future cash flows, dis-
counting those flows using the firm's opportunity cost of funds, and then implementing
all proposals having a positive net present value.
Case Study
Capital Budgeting in the Real World
Surveys have been made concerning the procedures used by managers to evaluate pro-
posed capital projects. A summary of the results from one survey on the primary and
secondary methods used is shown here:
Percentage of Firms
Using Method
1959 1985
Primary evaluation method
Net present value 7 21
Internal rate of return - 49
Payback period 42 19
Other 51 11
Secondary evaluation method
Net present value 1 24
Internal rate of return 1 15
Payback period 15 35
Other 82 26
Although most firms now use the net present value and internal rate of return
techniques, a significant number still rely on the payback and other methods that do not
use discounted cash flow analysis. However, when compared to the results obtained
some 26 years earlier, there has been a marked increase in the number of firms using
the better evaluation approaches. In 1959, less than 10 percent of the firms surveyed
were using discounted cash flow analysis (i.e., either NPV or JRR). By 1985, about 70
percent were using one or the other as their primary method and 39 percent were using
one of them as their secondary method.
Competitive pressures will ensure that managers will adopt state-of-the-art tech-
niques in all phases of their business operations. To do otherwise will result in lower
profits and perhaps even losses.The firm that is not efficient in all areas, including cap-
ital budgeting, is not likely to be successful in a competitive business environment.
SOURCE: Suk H. Kim, Trevor Crick, and Seung H. Kim. "Do Executives Practice What Academics Preach?"
Management Accounting (November 1986):42-52.
kLyi'uncept
he internal raiL uf rëturn is that disc'untrte that eqithtes the preeñt value of
all future net cash flows to the'cost of an investment. It the iiitrnJ rate tfrelurn
on an in exceeds the cost capital, the investment will inereaprkit.
Using the net pre:enf value evalUation technique, if the present value ofail fu.
tur cash flows (diseounteJ b y the firm": cr c capti) cds tLe nftid ct
nf ht. nri LPrt
the use of thetile—
A'1V techiiique will :iJwavs lead to i vestin.ni teisiore; thet
mxim!,.c the. value of the firm.
• & erieraUv. the ?iI1' iuid IRT methods"ield '.nsistent ::eepl!rejct signai.
I L'.'wcvei. he . n cornpnne two mutual! exclusive pmojccs. the :datveink-
• in l the projects can be different using the two methods hecau iJt TRR ieh
• niquL ;'um • s that-future cash flows i.an he reinvested at the tnter;ial rate of re-,,
Aurn for thc projec.tin L1.
Capital Rationing and the Profitability Ratio

Up to this point, it has been assumed that the firm would make all capital expendi-
tures that meet the criteria that NPV> 0 or that JRR> cost of capital. However, in
some cases, the total amount of money the firm has and that can be obtained by bor-
rowing or by selling stock is less than the total that would be spent if all projects were
undertaken. In this case, choices must be made among those projects that meet the
evaluation criteria. That is, available capital funds must be rationed among these com-
peting projects.
There are several reasons why a business may be subject to capital rationing. First,
the sheer number of capital proposals, if they all were implemented, may exceed man-
agement's ability to develop-and manage them. Clearly, there is a limit to the scope and
number of projects that can be effectively monitored at any given time. Second, inter-
nal funds are limited, and management may prefer not to take the risk associated with
additional borrowing and/or the possible reduction in control associated with the sale
of additional shares of common stock. Finally, an operating unit of a larger firm may ar-
bitrarily be assigned a maximum capital expenditure budget for the year.
Strict adherence to the NPV evaluation approach (i.e., selecting the project with
the greatest NPV) can lead to nonoptimal decisions if capital is rationed. Suppose that
CHAPTER 15 Capital Budgeting ri'-
01)
Project A Project B
Cost Project C
$20,000 $10,000 $101000
?et cash flo ws per year
Period I
W,UO(J $4,700
Period 2 $4,700
91000 4,700
Period 3 4,700
9,000 4,700 4.71w
NPV (12% discount rate) 1,616 1,289 1,289
the firm's capital budget was limited to $20,000, and it was

co nsidering the three in-
jects byshown
vestm in Table 15.4. Further, assume that the firm's policy is to rank pro-
their NPV
continuing down until andthethen implement projects beginning at the top of the list and
available funds are exhausted. This approach would be sub-
optimal because only project A
the firm. But if projects B and C would be selecttj. thus adding $1,616 to the value of
values would be $2,578. were selected instead of A the sum of the net present
That is, the combination of inv estments B and
more to profit than would investment A. C would add
The problem here is that the basic NPV ap-
proach
tial costdoes notproject.
of the compare the relative magnitudes of the net present value and the ini-
Consider another set of capital projects where each has a positive

capital rat NPV but where
these COfldioning
jtjn 5,
restricts choice to some subset of the entire array
of projects. Under
the problem
the capital constraint summingreduces to finding all combinations of projects that meet
the NPV5
for each conibblation, and selecting the sub-
set that has the highest sum of the individual
vidual projects A through F NPVS. IxWüJjy, Wnsid&r *e set of indi-
ited to $1 million. in Table 15.5, and assume that total capital spending is lim-
PV of
Project Profltablifty
Cash Flows Initial Cost NPV
Ratio (R4,)
A - 300,000 250,000 50,000
B 1.20
510,000 500,000
C 10,000 1.02
790,000 750,000
D 40,000 1.05
600,000 500,000
B 100,000 1.20
280,000 250,000 30,000 1.12
G 1,020,000 800,0[)0
220,000 1.28
516 PART VI Risk and Capita! Budgeting
The subsets of projects that meet the $1 million capital budget, their cost, and total
NPV for each set are listed here:
Combination Cost I NFV
A,B.E $1,000,000 $ 90,000

•A, B 1,000,000 120,000
A,C 1,000,000 90,000
A,D,F 1,000,000 210,000
&D 1,000,000 110,000
13, F., F 1,000.000 100,000
C, E 1,000,000 70,000
C, F 1,000,000 100,000
0, E, F 1.000,000 19000
of $210,000, is the
Clearly, the combination A, D, I which yields a combined NPV
choice that maximizes the value of the firm.
What if another project (G) were added to the list that had an initial cost of
$800,000 and a net present value of $220,000? Note that implementation of this project
would preclude investing in any of the oihers. As its NPV of $220,000 is higher for than
any other feasible combination of projects, G should be implemented and the remain-
ing $200,000 in the capital budget left in the firm's bank account.
An equivalent but less cumbersome approach uses the concept of the profitability
ratio (Rn) to rank projects. The profitability ratio is computed as one plus the present
value of all future cash flows divided by the initial cost of the project; that is,
NPV (15-4)
RP
COST
If the firm ranked the projects A through F in Table 15.5 by the profitability ratio, pro-
F has
jects F,.4, and D would be identified directly as those to be implemented. That is,
the highest ratio (1.24); A and D are next, both having a ratio of 1.20. The total cost of
these three projects is $1 million, which exhausts the capital budget. When project G is
included, it has the highest profitability ratio (i.e., 1.28) and has an NFV greater than
any other feasible combination of the other projects. Thus, it should be selected even
though it does not exhaust the available budget of $1 million. The profitability ratio ap-
proach is more efficient than the alternative of determining the cumulative net present
value of each feasible combination of projects.
Linear Programming and Capital Rationing

The linear programming tool developed in chapter 8 can be used to solve more com-
plex capital budgeting problems. Consider a firm that has developed five capital pro-
jects, each of which requires an investment outlay in each of the first 2 years and has a
positive net present value as shown in Table 15.6.
To implement all five projects would require $112 in year 1 and $42 in year 2. But
the firm's capital budget constraints for years 1 and 2 are $55 and $30, respectively. That
is, capital rationing applies—the firm does not have the resources to undertake all the
projects. Because of the separate budget limitation in each of 2 initial years, the straight-
Outlay in Year Net Present Value

roJecr 2 (c& 1 V1)
1 24 5 30
2 45 8 40
3 10 5 35
4 8 4 25
5 25 20 42
forward method used earlier will not work. However, the problem can be setup as a lin-
ear program and solved.
Define X1 as the proportion of proje,ct j undertaken. The objective function is
Maximize V = 30X1 + 40X, + 35X + 25X4 + 42X5,
(That is, maximize the sum of the net present values of all projects undertaken.) Sub-
iectto:
(1)24X1 -s-45X2 + 10X3+8X4±25x555
(2) 5X1 +, 8X2 + 5X3 + 4X4 + 20X5 30
(i.e., these constraints require that the amount spent on all projects in each year not ex-
ceed the available budget),
(3)-(7) X3lforj=1,...,J
(i.e., these constraints require that no more than 100 percent of a project's cost be un-
dertaken) and the nonnegativity constraints
(8)—(12) X1:50,j=1,,..,5
The solution is
0.6056,X2 = 0.0,X3 1.0,X4 = 1.0,X5 = 0.8986
The interpretation is that the firm would fully fund projects 3 and 4 and would fund
60.56 percent of project 1 and 89.56 percent of project 2. Partially funding a capital pro-
ject may be interpreted as essentially extending the original investment period to more
than 2 years. Alternatively, the firm might adopt some sort of decision rule that would
allow the capital budget limitation to be raised for projects where ,,%. wi less than 1•4
Solving the dual problem yields a shadow price for cqwh eonstraint. The shadow
price for constraints 1 and 2 are 1.0986 and 0.72m,pectively. This means that a $1 in-
crease in the budget for the first' would.increase net present value by $1.10, and a
A ldget ierese in y ear 2 would increase net present value by $0.72. Clearl y, the value
of this firm could be increased if it could shift part of its budget from year 2 to year 1.
413y using the iuieger programming variant, all values of N. wu.i1d be forced to zero or One.
518 'ART.y! Rs/aid Capital Budgeting
This might be ac plisbedy cbagg the timing of the firm's financing plans (i.e.,
borrowing, sale of stock, etc.). For èxample,.a on-dollar shift would increase the value
of the objective function by $0.38 (i.e., $1.10 - $0.72).
-
tue tuftth athrt
t tle f n must select that combtnatiøn çf pro
t1ie-iie r'resent valnes..
one Ibe net present value- divided by
ti çn be trnat'ive ue.tmcnt wiien theficx i. subject
I-.
• . . - ....,, i,... ...- S
.............................5,. ........
'S
1 2 2L
, 0
THE COST OF CAPITAL . •
The use of the net present value and internal rate of return methods requires that fu-
ture cash flows he discounted by 'the firm's cost of the funds used to pay for the costs of
: the project. The cost of such funds is referred to as the cost of capital. In general, the
cost of capital is the return required by investors in the debt and equity securities of the
firm. Basically, there are three sources of funds to the firm for capital spending: retained
earnings, debt, and equity; In the following, the cost of debt financing and the cost of eq-
uity financing through the sale of 'common stock are considered. The cost of using re-
tained earnings for capital spending is approximately the same as the cost of common
• stock, so the cost of retained earnings is not discussed explicitly.5
Cost of Debt Capital . ..

There is little controversy about the cost of capital raised by borrowing from banks or
by selling bonds. That cost is the net or after-tax interest rate paid on that debt. Because
interest, unlike dividends on stock, is deductible from income when computing income
taxes, it is the after-tax cost that 'is important. For a given interest rate (i) and marginal
tax rate((t) the after tax cost of debt (id) is given by
r4 = .i(1 t) (15-5)
For example, if the firmbotrovis at'a l0perceni t interest rate and faces a 40 percent mar-
ginal tax rate, its after tax cost 61 debt capital is
ref — 0.10(1 —0.40)=0.06
or 6 percent.6 ' '!
uve manager .ultghL vloW the, cost of reamed earnings as bc rig zcro Th .. clear', is not the case if for
no other reason than that these s a sigalfican..óppQrtunity cost to the owners of the business, who could
use- those funds for personal consumption or for investment in some other business.
6Note that in the equations, the decimal equivalent of the interest rate is used. That is, a 10 percent rate of
• interest appears as 010 in equation.
Two important considerations should be mentioned. First, if the firm is not earning
profits, the pretax and after-tax interest rates will be the same because the marginal tax
rate is zero. Second, the concern is with the marginal cost of capital and not the average
cost of capital for the firm. For example, the fact that the average cost of debt in the
firm's balance sheet is 9 percent is irrelevant. The only important consideration is the
cost of raising new capital. This is because the investment decision requires comparing
the rate of return on new projects (i.e., the marginal return) with the cost of acquiring
additional capital (i.e., the marginal cost of capital).
Cost or, Equity Capital

In general, determining the cost of equity capital from retained earnings or sale of
common stock is more complicated and more controversial than determining the cost
of debt. In the following discussion, three approaches to estimating this cost are pre-
sented. Note that dividends are not deductible from income when computing income
taxes, so the firm's tax rate does not play a direct role in determining the cost of eq-
uity capital.
Method I: The Risk-Free Rate Plus Risk Premium Generally, investment in
comnipn stock is considered to be riskier than investment in bonds. The firm has a
contractual obligation to make interest and principal payments to bondholders, and
such payments must be made before dividends can be paid to stockholders. If prof-
its rise and fall, it is likely that dividends also will fluctuate. However, except in the
case of severe financial problems, bondholders will be paid. Thus, it is thought that
investors will demand a return on equity (re) composed of a risk-free return (rf), usu-
ally considered to be the rate of return on long-term government bonds, plus a pre-
mium for accepting additional risk. This premium reflects the two sources of risk.
First, there is the risk associated with investing in the securities of a private company
as opposed to buying federal government securities. Second. there is the additional
risk associated with buying stock rather than bonds of a business. The premiums
associated with the two types of risk are labeled e1 and e2 . Thus, the cost of equity
capital is
Te = T+ e1 + e2 (15-6)
One way to measure the first type of risk (e1) is to use the difference between the
rate of interest on the firm's bonds (rd) and the rate of return on government bonds (Tf).
This difference should increase as risk of default increases. For e 2 a rule of thumb is used
to approximate the risk of buying the common stock of a firm rather than bonds. Nec-
essarily, this is based on judgment rather than any formula or equation. A tvnical ap-
proach used by financial analysts is to assume that the return on a firm's common stock
should be about 3 to 5 percentage points greater than on its debt. Using the midpoint
of this range (i.e., 4 percent) as an estimate of e2, the total risk premium (e) would be
calculated as
C e1 ± e,
or
e = ( Td r1) + 0.04
For example, suppose that the risk-free rate is 8 percent and the firm's bonds are
yielding 10 percent. Therefore, the total risk premium would be 6 percentage points,
that is,
e = (0.10 - 0.08) + 0.04 0.06
and the firm's cost of equity capital would be
e = 0.08 + 0.06= 0.14
or 14 percent.
Method El: Discounted Cash Now Method 1 includes risk but fails to consider the
possibility of growth in dividends and the value of common shares over time. An alter-
native approach is the discounted cash flow method. Just as the, value of the firm is the
present value of all future profits, the value of a share of common stock is the present
value of all future dividends using the investor's required rate of return (re) as the dis-
count rate. That is, one share of stock entities the owner to receive a series of payments
(i.e., dividends) roughly equivalent to an annuity. Thus, the value of that share should
equal the present value of the annuity. if the current dividend per share (D0) is expected
to remain constant, the value or price (P) of a share will be
pD0
+ re)
or
11
P= (15-7)
+ r)'j
It can be shown that the bracketed term reduces to
Ta
so that the value of a share of stock is simply the dividend rate divided by the required
rate of return, that is,
Ta
However, if the dividend is expected to increase over time at an annual rate of g, it can
be shown that the price per share will be given by
=Do
(15-8)
Ta - g
Solving equation (15-7) for r yields an equation for the cost of equity capital:
D0
r=+g (15-9)
That is, the return required by the investor is equal to the current dividend yield on the
common stock (DWP) plus an expected growth rate for dividend payments. The esti-
mate of that growth rate might be the historic rate of growth for the firm or, perhaps,
the consensus growth rate being used by financial analysts who study the firm.7
For example, suppose that a firm is paying a dividend of $8 per share on common
stock that sells for $100 per share and that there is agreement among financial analysts
that the growth rate of dividends of this firm will be 6 percent per year. Using this ap-
proach, the firm's cost of equit y capital is estimated to be 14 percent. That is,
+ 11 .06 = 0.14 or 14°i
Obviously, variations in the price of the stock win change the firms cost of capital. 1-br
example, if investors bid up the price of this stock from $100 to $120 per share, the cost
of capital will fall to 12.7 percent. That is,
8
r + 0.06 = 0.127 or 12.7%
Method ifi: Capital Asset Pricing Model (CAPM)

The capital asset pricing model,
widely used in the world of quantitative finance, emphasizes not only the risk differen-
tial between common stock and government bonds but also the risk differential among
stocks!'The risk differential between stocks and government bonds is estituated by
(r0,
- rf), where rm is the return on the average common stock and r
f is the rate on risk-free
U.S. government securities. /
Relative risk among stocks is measured using the beta coefficient, 1 3 , as a risk index.
The beta coefficient is the ratio of variability in return on a given stock to variability in
return for all stocks. Return on a stock for a period of time, typically one year, is the div-
idend plus the change in the value of the stock during the period. Using k1i as the total
return in the ith period on an investment in one share of common stock of company S
and k7' as the total return on all common stocks (or a representative sample) in the ith
period, the beta coefficient is determined by estimating the parameters of the following
regression equation:
= a ± 13k7'
The estimated value of 13 is the beta coefficient.
A stock with average risk will have a beta value of l.O,meaning that the returns on that
stock vary in proportion to returns on all stocks. A higher-risk stock might have a beta of
2.0, meaning that the variation in return on that stock is twice that of the average stock.
For example, if 13 2, when the return on an average stock increase 10 percent, the return
on this risky stock increase by 20 percent. Conversely, a beta of 0.5 would be associated with
a low-risk stock where return varied only one-half as much as for the average stock.
Most firms that have issued common stock that is actively traded in organized markets are followed by
one or more analysts in banks, brokerage companies, mutual funds, and other financial institutions. Their
forecasts are readily available, and rriauagers usually make it a point to keep track of what these analysts
are predicting for their firm. For some large firms, there ma y he hundreds of analysts who follow each
company.
877he theory underlying the development of the capital asset pricing model was an integral part of the
revolution that transformed the world's financial markets in the 1970s and 19805. William Sharpe, Harry
Markowitz, and Merton Miller shared the 1992 Nobel Prize for their pioneering work in this area.
Using the capital asset pricing model, the cost of equity capital is the risk-free rate
(rf) plus a weighted risk component, that is,
re Tf+ (rm - rf)t3 (1540)
In this model, the overall risk premium for common stock is measured by (rm - r) and
the beta weight (f3) then adjusts for the risk associated with the specific firm in question.
For example, suppose that the risk-free rate (Tf) is 8 percent and the average return
on common stock is 11 percent. For a firm having a beta of 1.0 (i.e., the risk for the firm
is the same as that for the market average), the cost of equity capital is
-= 0.08 -r i.uu.ii - 0.08) = 0.1i or ii percent
Conversely, the cost of capital for a firm having a beta of 2.0 would be significantly
higher. That is,
r = 0.08 + 2.0(0.11 - 0.08) 0.14 or 14 percent
Estimates of beta for many publicly traded companies are available from various
financial service companies. For example, the Value Line Investment Survey reports es-
timated betas for several thousand firms. 9 Recent estimates for some major U.S. corpo-
rations are listed here.
American Telephone & telegraph (0.85) BankAmerica Corp. (1.25)
Consolidated Edison (0.80) Delta Air Lines (1.10)
Ford Motor Corp. (1.05) General Mills (1.00)
Nike (0.95) Pepsico (0.95)
Sears, Roebuck (1.05) Zenith Electronics (0.85)
Note that American Telephone and Consolidated Edison offer below-average risk,
while Nike, BankAmerica, and Zenith carry significantly above-average risk.
The Composite Cost of Capital

Many firms attempt to maintain a constant or target capital structure. For example, the
management of a manufacturing business may prefer a capital structure that is 30 per-
cent debt and 70 percent equity. In contrast, managers of an electric utility company
may prefer a 60 percent debt and 40 percent equity structure. In either case, capital
would he raised periodically both by incurring debt and by selling stock.The differences
in capital structure reflect the preference for risk on the part of owners and managers
and the nature of the business. A capital structure heavily weighted toward debt implies
greater risk because of the greater interest and principal payments that are required.
Public utilities tend to have a higher percentage of debt than most other firms because
they have a monopoly position, and usually the product they sell is a necessity. This
means that the firm Will have a reasonably stable and dependable flow of revenue and
profit, and this offsets part of the high risk associated with a large proportion of debt in
their capital structure. -
9Valuc Line PublLshing Company, New York.

A firm with a target capital structure may maintain two separate ledgers—a list of
capital projects and a list of financing plans (borrowing, sale of stock, etc.). A particular
financing option, say selling bonds having a cost of 10 percent, is not tied to one capital
project. Rather, the firm uses an overall or composite cost of capital as the evaluation
criterion for each capital project.
This composite cost of Capital (re) is a weighted average of the after-tax cost of debt
(rd)and equity (re) capital.The weights are the proportions of debt
( Wd) and equity ()
in the firm's capital structure. That is,
= "d'd + were 15-11)
For example, suppose that a firm's target capitalization structureis 40 percent debt and
60 percent equity. Over the next twelve months it plans to raise $100 million by selling
$40 million in bonds at a cost of 8 percent and issuing $60 million of stock at $60 per
share at a cost of 12.5 percent. Thus, the firm's weighted or composite cost of capital
would be 10.7 percent. That is,
r (0.40)(8) + (0.60)(125)
10.7
The eoiposie rate of 10.7 percent would be used as the cost of capital to evaluaie all
proposed capital expenditures.
I- - .- -
Key Coeepts - -
The cost f taitalis the re'turn equirIby ivestrs In the debt
and equity se-
curities of the
• The c of debt capital h the after-tax interest tate on the flrth's bonds or
borrowing. ---- : -
Three mothoct of determinrng the cost of equity pitaI are
L The risk4ree rate plus a risk prethiurn.
2.. Discvx ted cash flow.
3. The capital asset pricing model.
• The ompoii(e cost of capital is a wei gh ted average of the cost
of debt and eq-
uity where the weights are the proporticjn of debt and equity in the firm's tar-
get capItal
Case Study
Measuring the Cost of Equity Capital
Having several ways to determine the cost of equity capital can lead to different cost
estimates, d
epending on the method used. The differences can be significant and may
make the difference in a decision to make or not make a capital expenditure. Also, a
public utility such as a gas or electric company generally has to offer evidence on the
cost of capital in hearings about the rates the companies may charge its customers. Ob-
viously, in such cases, the company may want to document a high cost of capital to jus-
tify higher rates. In contrast, the staff of the regulatory body, who represent the inter-
ests of the consumer, may use lower estimates of the cost to keep the rates down.
Although many cases could be cited, a good example is an application for a rate
increase by South Central Bell Telephone Company before the Tennessee Public Ser-
vice Commission. A financial expert hired by the telephone company used the dis-
counted cash flow method to estimate the cost of equity capital at 14 percent. A mem-
ber of the commission staff used the same technique to estimate the cost at about 11
Je1tcilt. 11111U --Apia,iiaiiou foi the discepuiiey wa ihat the two experis differed in theit
opinions about the growth rate of dividends.
Clearly, the choice of a growth rate is important. The diffrence between an 11 per-
cent and a 14 percent cost of equity capital translated into $20 million in annual profits
for the firm. In its decision, the Commission opted for a rate of just over 11 percent.
SOURCE: Public Utilities Reports, 4th Series, vol. 22. pp. 257-220.
MERGERS AND ACQUISITIONS

The growth of the firm can occur either internally or externally. When the firm builds a
new plant, it is growing internally. For example, General Motors' construction of a new
plant is an example of internal growth. Alternatively,the firm may grow by purchasing
the assets of another firm (i.e., growth by acquisition) or by agreeing to join with that
other firm under single ownership (i.e., growth by merger). Mergers and acquisitions
are special cases of capital budgeting.
Types of Mergers
Mergers are typka1ly divided into three categories: horizontal, vertical, and conglom-
erate. Horizontal mergers involve firms that directly compete for sales of similar prod-
ucts or services. A merger of General Motors and Ford would be a horizontal merger.
Vertical mergers occur when firms that had been operating at different stages in the pro-
duction and distribution of a product combine to form a single firm. Vertical mergers
may consist of a firm acquiring a seller of its output, such as IBM purchasing a chain of
computer stores that sell the firm's computers. Sometimes this is referred to as down-
stream integration. Another type of vertical merger is the acquisition by a firm of a sup-
plier of resources or components needed to make its product. An example would be
IBM purchasing a manufacturer of computer chips or disk drives. Mergers with suppli-
ers often are referred to as upstream integration. Finally, conglomerate mergers involve
the joining of firms producing unrelated products. An example would be a merger be-
tween Delta Airlines and the Levi Strauss Corporation.
Figure 15.2 depicts the three basic categories of mergers. It shows two industries,
construction equipment and bakery products. A horizontal merger is represented tv
the joining of two manufacturers of parts. The combination of a bakery and bed
CHAPTER 15 Capital Budgeting
525
CONSTRUCTION EQUIPMENT
BAKERY PRODUCTS
t
0 (1 :1 ):
Manufacture
of par s
(DH, rizo n tro
ai^j Baking
Assembly of
equipment
0a 0 0 Baked
goods
retailing
goods retailer would be a vertical merger. The merging of an equipment assembl y

and a bakery would constitute a con g lomerate merger. firm
Case Study
The Poker Game for Gulf Oil
In March 1984, Standard Oil of California (Socal) and Gulf Corporation signed a pre-
liminary agreement to merge. The terms were that Socal would pay $80 per share—a to-
tal of over $13 billion—for Gulfs stock. The result was a merger between the nation's
9th and 10th largest industrial corporatibns. Together, the two firms had 1983 assets of
$45 billion and sales of nearly $60 billion.
The Socal–Gulf merger emerged as the product of a concerted effort by Gulf to
avoid a hostile takeover by an investment group headed by T Boone Pickens, Jr., the
chairman of Mesa Petroleum, in late 1983, Pickens and his colleagues purchased 13 per-
cent of Gulfs stock with the avowed purpose of restiucturing the firm in order to in-
crease profits. In response, Gulf's management convened a special shareholders' meet-
ing to change corporate rules in order to make a takeover more difficult. About the
same time, Gulf announced that it was considering the acquisition of Superior Oil Co.
Although never completed, this merger would have increased Gulfs size and made a
takeover more difficult because of the higher cost of the combined
firm.
526 PART VI Risk and Capisal Budgeting
In February 1984, the Pickens group announced that it would pay Gulfs stock-
holders $65 per share for 13.5 million shares of stock. If completed, this purchase would
have raised the group's stock ownership to over 20 percent, and might have enabled it
to replace some of Gulfs directors with individuals loyal to Pickens at the firm's annual
stockholder meeting in May.
Gulfs public reaction to the offer by the Pickens group was to label it as "unfair
and inadequate." Privately, the company's management began to contact other firms
that might be interested in acquiring Gulf. By March, the list of prospective suitors had
been pared to three: Socal, Atlantic—Richfield, and an investment firm, Kohlberg,
Kravis, Roberts. and Co. Representatives of each of these firms were invited to come to
Gulf's headquarters in Pittsburgh and make a careful examination of the firm's records.
Following this investigation, each firm submitted a sealed offer for Gulf Oil.
Socal's bid of $80 per share was higher than Atlantic—Richfield's $72 but less than
the $87.50 offered by Kohlberg, Kravis, Roberts, and Co. However, the Socal bid was ac-
cepted because the firm had access to the $13 billion necessary to complete the pur-
chase, whereas the Kohlberg firm would have required several months to assemble the
necessary funds.
Lest one feel sorry for the unsuccessful effort of T. Boone Pickens and his associ-
aLes, it should be noted thai Socal's offer enabled them to earn a capital gain of 760
million on the Gulf stock that they owned. I
Merger Incentives
Some acquisitions take place because m'nagers embark on the creation of a vast em-
pire over which they can preside. The experience of James Ling, former president of
Ling—Ternco—VaUght, a conglomerate involved in the defense industry, is a case in point.
In 1960, LTV ranked as the 335th largest firm in the United States on the basis of as-
sets. After a flurry of acquisitions, the corporation had climbed to 22nd place by 1968.
Mr. Ling's merger efforts were well rewarded in the stock market. In 1963, the price of
LTV common stock was $9 per share. By 1967, the price reached a high of $169.50 per
share. Unfortunately, the bubble finally burst, as cutbacks in defense spending after the
Vietnam War and the prospect of legal actions against the firm caused the share price
to plummet to about $7 in 1970.
Although ego and empire building sometimes play a role in merger activity, usually
mergers take place because managers believe that the value of the combined firm is
greater than that of its constituent parts. Among the possible sources of this synergy are
the following.
Increased Market Power A merger between two competing firms can result in in-
creased market power for the combined firm. In some cases, this increased market
power stems from the elimination of an aggressive competitor. Consider an industry
dominated by a few large firms. If one firm aggressively reduces prices, the other firms
may be forced to respond with price reductions of their own. The result of this price
competition will generally be reduced profit levels. However, one firm may acquire the
aggressive rival by merger. By eliminating the source of active competition, the result
may be higher prices, and hence increased profit, for all firms in the industry, albeit at
the cost of reduced consumer surplus to buyers of the product.
In other situations, a horizontal merger may increase market power by eliminating
excess capacity in an industry. Consider the tobacco industry at the turn of the century.
In 1890, five cigarette manufacturers formed a cartel to stabilize prices. In 1898, the car-
tel expanded its activities into chewing tobacco by purchasing five other manufactur-
ers. One year later, production facilities of 30 other chewing tobacco manufacturers
were purchased and immediately closed. The result was a classic example of monopoly
in action. B y restricting output, the cartel was able to increase prices and profits.
Technical Economies of Scale If technological conditions in an industry result in in-
creasing returns to scale, firms with large production facilities will be able to produce at
lower average cost per unit than will smaller firms. Thus, small firms that merge may be
able to reduce per-unit costs b y expanding the scale of their operation. Such technical
economies of scale may result from increased specialization of labor, greater use of au-
tomated production techniques, and/or sharing of overhead and management expenses.
Pecuniary Economies of Scale Pccuniary economies of scale are cost savings that
result from increased monopsony power of large firms. That is, large buyers may be able
to obtain lower prices for inputs that they purchase. Consider small farming operations
seeking to ship fruits and vegetables. If the firms merge to form a single corporation,
the resulting firm may require a significant proportion of total trucking capacity in its
region. Because the single corporation has greater market power, it is in a better posi-
tion to negotiate favorable rates with truckers than would many small firms acting in-
dependently. The same principle applies in capital and advertising markets. As a major
borrower, a firm such as AT&T may be able to obtain lower interest rates than those
offered to smaller borrowers. Similarly, with aggregate expenditures in excess of $1 bil-
lion, Procter & Gamble is able to obtain volume discounts for advertising time and
space. Firms acquired by AT&T or Procter & Gamble also would be able to buy at lower
prices. These cost savings provide an incentive to merge.
Reduced Transaction Costs Vertical mergers may reduce transaction costs for the
acquiring firm. In the absence of vertical integration, purchasing inputs from suppliers
or selling to distributors can involve substantial transaction costs. First, suitable trading
partners must be located. Changing prices or needs may dictate periodic reassessment
of the suitability of these partners. Next, the terms of transactions with these partners
must be established. Because the interests of the trading partners may differ, skill and
time may be required to reach agreement. When all parties are satisfied, there is still the
problem of monitoring performance. Dissatisfaction by one party may result in a ter-
mination of the agreement OF costly litigation.
The difficulty of negotiating and maintaining agreements with trading partners is
compounded by the incomplete flow of information. Each firm will use caution to avoid
disclosing facts that might allow competitors an advantage or permit the trading part-
ner to better its relative position. Even the attempt to be completely open may be par-
tially thwarted by differences in training and procedures of each firm.
Vertical integration reduces transaction costs by establishing permanent linkae.s
among the different stages of production. Thus, there is less need to search for suppliers
and distributors and the details of interaction between divisions can be dictated by the
management of the firm. Similarly, a common set of procedures can be specified for use
throughout the firm. Being verticall y integrated also reduces the risks associated with
dependence on only one or two firms for the supply of important inputs.
Risk Spreading Conglomerate mergers may be a means of spreading risk. By diver-
sifying its product line, managers can reduce fluctuations in profit rates. Consider a hy-
pothetical merger between Holiday Inn and the Twentieth-Century Fox Corporation.
The hotel and resort business is adversely affected by gasoline shortages and increasing
tra-el cnsts. In Contrast, motion pictuie producers tend to benefit from conditions that
keep people close to home, where they are more likely to attend movies. Thus, a merger
between these two firms could produce a conglomerate with sin1ler fluctuations (i.e., a
lower variance) in profits than for either of the firms operating independently. This re-
duced variability in profit should imply a lower beta value and, hence, a reduced cost of
capital for the combined firm.
In this example, there may also be complementarity in the two product lines. Holi-
day Inn could show Twentieth-Century Fox movies in its motel rooms and Twentieth-
Century might include references to Holiday Inns in its movies.
Valuation Discre pancies There ma y
be a substantial discrepancy between the ac-
quisition price of a firm and its value to a potential merger partner. In some cases, an
acquiring firm may be able to purchase production capacity at a much lower cost than
it could build comparable capacity. In other circumstances, resources may be available
at a bargain price. The merger between Gulf and Socal discussed in the case study on
pages 525-26 is an excellent example. In early 1984, the exploration cost of crude oil was
about $12 per barrel. But by acquiring Gulf, Socal increased its oil reserves by over 1.5
billion barrels at a cost of about $5 per barrel.
Key Concepts . -
• Horizontal mergers involve the c&tmbinai.ion t.if t_wia or more finns that produce
and sell siinilar Products or services.
•• A vertical merger occurs when a finn acquires- a seVer of
its product or a sup-
plier of niateri•ils used to manufacture the product.
• Conglomerate mergers combine two firms pmJicing unrelated products or
services.
HorizonI mergers may result min
market power by eliminating an ag-
gressive competitor or by taking excess capaciI from the market.
• Mergers may reduce costs as a result of
- - —Technical CcOnOflhj es of scale --
- - Pecumar CconOmies of scale - -- .-: - - -- -
—Reduced transaction COStS .
—Valuation discrepancies
-. A merger ma y be a wayof reducing fluctuations in profits, thus reducing risk.
CHAPTER 15 Capital Budgerin, 529
Merger Procedures
Typicaijy, three basic steps are involved in a merger. First, a suitable acquisition candi-
date must be identified. Second, the value of the target firm to the acquiring company
must be determined. Finally, the managers and/or stockholders of the target firm must
consider and approve the offer.
Identification of a Merger Partner Several factors must be considered in selecting

an acquisition target. One is the size of the prospective merger partner. Normally, a
merger with a small firm is easier than with a larger tirm because fewer dollars are in -
volved. For example, the acquisition of Gulf cost Socal over $13 billion. 1 -o consummate
the merger. Socal had to establish a multibillion-dollar line of cr' e'dit with banks around
the country. A number of other firms were interested in Gulf, but were unable to
arrange the necessary financing. Possible antitrust problems also favor the acquisition
of small firms. If the participants in a horizontal merger control a substantial share of
the market, the combination may he prohibited by the courts. Antitrust law and its ap-
plication to mergers are considered in chapter 19.
The expected reaction of the owners and managers of the target firm is another fac-
tor to be considered in selecting a merger partner. Although hostile takeovers have be-
come more frequent in recent years, this type of acquisition is more difficult than if both
parties favor the action. Again, the Socal—Gulf merger is a good example. Gulf's man-
agement strongly opposed a taktover by Mesa Petroleum. The firm issued public state-
ments advising shareholders not to tender shares to Mesa. At one point. Gulfs man-
agers went so far as to change the corporation's rules to make a takeover more difficult.
In contrast, when Socal expressed interest in a merger, Gulf cooperated in every way.
Thus, the time and dollar costs for an acquisition by Socal were much less than they
would have been for Mesa Petroleum.
The extent that resources and facilities of a target firm complement those of the ac-
quiring firm is a third factor in selecting a merger partner.A firm needing a stable cash flow
to finance expansion or modernization of its facilities may seek a merger partner with sub-
stantial cash reserves.A firm that perceives that it has an important gap in its product line
will be attracted to a producer of a product that fills that gap. Similarly, firms requiring spe-
cific resources are likely to select the owners of those resources as targets for acquisition.
Finally, acquiring firms tend to select merger partners that appear to be good bar-
gains. It may be possible to purchase a firm at a price substantially below its potential
value if its owner wants to retire or if the firm has encountered financial difficulties be-
cause of poor management or unfavorable economic conditions.
Determination of the Value of the Target Firm Once the merger partner has been
identified, the value of the firm to the acquiring company must he determined.
Conventional analysis suggests that the value of an enterprise is equal to the present
value of future profits. That is, annual profits over some planning horizon are estimated
and then discounted back to their present value. This amount would represent the max-
imum price that should be paid for the firm.
Mergers take place because of an anticipated synergism between the two firms.
Thus the' value of the target firm to the acquiring firm may be considerably greater
than the present value of its profits as an independent entity. Hence, an acquiring firm
may be willing to pay a premium price. Fundamentally, the maximum price should still
be based on an analysis of the present value of profits. However, in this circumstance.
it is incremental profit to the acquiring firm that is the relevant consideration. That is,
management should estimate its profit over the planning horizon with and without
the merger. The discounted value of the difference is the maximum price that should
be paid.
For example, assume that the present value of all future profits is estimated to be
$400 million for Acme Inc. and $200 million for Peabody and Co. if the two firms con-
tinue to operate independently. Also assume that if Acme acquires Peabody. the com-
bined business could take advantage of economies of scale and that the present value
of Acme's profit (including the acquisition of Peabody) would e $800 million. Thus, the
present value o f the incremental profit for Acme would increàe by $400 million as a
result of the merger. If Peabody had 5 million shares of stock outstanding prior to the
merger, Acme could bid as much as $80 for each share (i.e., $400 million ^ 5 million
shares) of Peabody's stock.
Theoretically, the value of the firm as an independent entity should be approxi-
mated by the market value of its outstanding stock. For example, with the present value
of profit equal to $200 million, the market value of Peabody's stock should be about $40
per share (i.e., $200 million ± 5 million shares). If this is true, the likelihood of a merger
will depend on the amount of synergism associated with the merger. If the merged firm
is no more valuable than its constituent parts, there is no incentive for the acquiring firm
to offer a price greater than the current market price for the target. In that case, own-
ers and managers of the prospective merger partner will not benefit, and the combina-
tion probably will not take place. In contrast, if there are substantial s ynergistic bene-
fits to the merger, a premium over the market value of the target firm can be offered.
Because the benefit to managers and owners of the target firm can be significant, the
merger offer may be viewed favorably. In;he example, owners of Peabody's stock could
receive as much as $80 per share for stock that was selling for $40.
Presentation of the Offer Once the maximum value of the target firm has been de-
termined, the next step is to formulate a plan to facilitate the acquisition. If the man-
agement of the target firm is favorably disposed to the merger, the procedure should be
relatively straightforward. Managers from both firms will meet to determine the details
of the arrangement. Once an agreement has been reached, they will make a joint state-
ment indicating the terms of the merger and urging acceptance by the acquired firm's
stockholders.
The offer for the shares of stockholders may be in cash or it may be structured as
an exchange of stock. A cash offer usually includes a premium over the current market
price of the acquired firm's stock. Offers to exchange stock specify that shareholders of
the target firm will receive a specified number of shares of the acquiring firm's stock for
each share that they hold. The actual number of shares depends on the relative stock
price of the two firms and the extent of synergistic benefits of the merger. Consider the
hypothetical acquisition of Peabody by Acme. Suppose that Acme's stock is currently
selling for $80 per share, while that of Peabody is at $40 per share. If these market prices
accurately reflect the values of the firms as independent corporations, one share of
Peabody stock would be worth as much as two Acme shares. But if combining the two
firms increases their joint value, Acme will he willing to offer a more favorable ex
CHAPTER 15 Capital Budgeting s31
change rate. For example, Acme maybe willing to offer as much as one share of its stock
for every Peabody share. The greater the synergistic benefits, the more attractive the
rate of exchange can be made to Peabody's stockholders.
If the management of the target firm opposes the merger, the endeavor is more
complicated. In this case, the acquiring firm must make its appeal directly to the target
firm's stockholders through a tender offer. Such an offer typically indicates that a cer-
tain price will be paid for a specified number of shares of the acquiring firm's stock ex-
changed for that of the target firm. Frequently, the offer has a fixed termination date
and is valid cmlv if a minimum number of shares are submitted, or tendered, to the ac-
quiring firm. For example, suppose that Peabod y 's management opposed the Acme
takeover. Acme responds with a tender offer to purchase Peabody's stock at $60 per
share. However, the offer is to he valid for only 60 days and is conditional on at least 51
percent of Peabody's stock being tendered to Acme. If Acme obtains the necessary Si
percent, the voting rights associated with this majority interest can be used to elect a
new board of directors who will favor the'merger. If sufficient shares are not tendered,
the offer may be withdrawn and the shares returned to their original owners.
Key Concepts
• Prospective n1er!er partners are evaluated in ternis of siw ,^. apticipated reaction to
a merger proposaL price relative to value, and extent of comp l ernewary rcsiurcea
• lie maximum value ut a lret firm is the present value of the addi:ional prof-
its that would be earned b y the acquiring firm after , tb e merr.
• An acquiring firm may offer cash or an exchange of stock for the sharer, of a tar-
get firm.
Case Study
The White Knight and Other Antimerger Strategies
Hostile takeovers can be extremely bitter and costly. As a means of thwarting such ac-
tions, several defenses can be used by managers of potential target firms. One option is
to identify another firm, referred to as a white knight, that agrees to purchase the tar-
get firm to prevent its acquisition by a less favorable merger partner. A good example
is Gulf which used Socal and Atlantic Richfield as white knights to ward off unwelcome
advances by Mesa Petroleum.
Another strategy is the sale of previoulv authorized but unissued shares to a sym-
pathetic firm, financial institution, or group of investors. By having more shares out-
standing, it is more difficult to obtain the number of shares required for a takeover. A
related tactic is to initiate a merger with another firm. The advantage of this approach
is that the resulting larger firm may be more difficult to acquire and perhaps less at-
tractive with the new addition.
Then there is the Pac-Man strategy. It requires that the firm that is the target of a
hostile take-over retaliate by using a tender offer of its own for the shares of the ag-
gressor. That is, the target firm becomes the hunter instead of the hunted, as with the
Pac-Man video game. Using the earlier example, if Acme had offered to pay $60 for
Peabody shares, Peabody might respond with an offer to buy Acme shares for $90.
A firm subject to an unwanted takeover may swallow what is called a poison pill
to reduce its attraction to the other firm. One form of poison pill is to borrow money to
buy its own stock. As the firm increases its debt and reduces its equity, it becomes a
more risky business.That is, the likelihood of insolvency increases, and it becomes a less
desirable merger candidate.
Finally, managers may protect themselves against the: consequences of a hostile
takeover by the inclusion of a golden parachute provision in their contracts. Such pro-
visions are simply short-term employment contracts that become thective when there
is a change of corporate control. They provide displaced managers with compensation
in the event that they are replaced as a result of a merger.0
STiM1RY
Capital projects are those investments that are expected to generate returns for
more than one year. Capital expenditures are made to reduce costs, increase output,
expand into new products or markets, and/or meet government regulations. Capital
budgeting is the process of planning capital projects, raising funds, and efficiently al-
locating those funds to capital projects. In general, the firm continues to make capi-
tal expenditures until the rate of return on the last dollar invested equals the mar-
ginal cost of capital.
The capital budgeting process consists of two phases. First, the net cash flows are
estimated, and second, evaluation techniques are used to compare those cash flows
to the initial cost of the projectThe payback method determines the number of years
necessary for the cumulative net cash flows to equal the cost of the project. Because
it fails to account for all cash flows and does not discount these cash flows at the
firm's opportunity cost of capital, the payback method is an inferior evaluation tech-
nique. The net present value (NPV) and internal rate of return (IRR) concepts are
the two most commonly used evaluation methods. Using the JVPV method, a capital
project is determined to be profitable (i.e., would increase the value of the firm) if
the present value of all future cash flows exceeds the initial cost of the project. In
contrast, the internal rate of return is the discount rate that equates all future cash
flows and the project's cost. An investment should be made if the IRR exceeds the
firm's cost of capital.
Generally, the NPV and IRR methods yield consistent accept or reject signals for
proposed capital projects. However, the !RR method has several problems. For exam-
ple, it sometimes results in multiple solutions for the internal rate of return. Also. it dis-
counts future cash flows at the rate of return for the project being evaluated and not at
the firnis opportunity cost of capital. Because of this, if the firm is subject to capital ra-
tioning (i.e., where the firm has moie feasible capital projects than it can finance from
both internal and external sources), the JRR method may not select that project that
maximizes the value of the firm.
The profitability ratio is defined as one plus the ratio of net present value to initial
project cost. It is used to rank the relative profitability of projects where the firm is sub-
ject to capital rationing. Linear programming can be used to determine which capital
projects to implement when capital rationing applies.
The cost of capital is the return required by investors in the debt and equity secu-
rities of the firm. The cost of debt is the net (after-tax) interest rate paid on that debt.
Alternative methods for estimating the cost of equity capital include (1) the risk-frc
rate plus a risk premium, (2) discounted cash flow, and (3) the capital asset pricing
model.Thc last method uses the beta coefficient (a measure of the variabilit y in return
on a given stock relative to variability in return on the average stock) to measure rela-
tive risk among the common stock of different companies.
As firms often use both debt and equity financing sources, a weighted cost of capi-
tal measure is used. The composite cosf of capital is a weighted average of the cost of
debt and the cost of equity capital. The weights are the proportions of debt and equity
in the firm's target capital structure.
A merger is defined as the joining of two or more firms under single ownership. Usu-
ally, mergers take place be management believes that tile value of the combined tirm
is greater than the sum of the individual values. Horizontal mergers involve firms that com-
pete directly for sales of simi1argoods or services. Vertical mergers occur when firms that
had been operating at different stages in the production and distribution process combine
to form it firm. A conglomerate merger combines essentially unrelated firms.
From the standpoint of the acquiring firm, a merger may be a means of increasing
its market power by eliminating competitors. Other incentives to merge include cost re-
ductions resulting from economies of scale, reduced transaction costs, and valuation dis-
crepancies. A merger also may be a way of smoothing fluctuations in profits.
Several factors affect the choice of a target for acquisition. A merger with a very
large firm may present financial problems and raise antitrust issues. If management of
the target firm is expected to oppose the merger, the acquisition is likely to be more
costly and difficult than if the proposal is accepted by those managers. The best merger
partners are those that have complementary resources and can be purchased at a cost
that is low relative to their value to the acquiring firm.
The maximum price that one firm would he willing to pay for another is the present
value of the additional profits that are estimated to result from the merger. An acquir-
ing firm may offer cash or an exchange of its stock for that of the target firm.
15-1. What is the basic principle underlying the approach used by a prcfit-maximizing
manager in making capital budgeting decisions?
15-2. Why is the marginal cost of cap i tal schedule or function upward sloping for most
firms?
15-3. Contrast the net present value and internal rate of return approaches in the eval-
uation of capital projects. In general, if the NPV is greater than zero, what does
this imply about the internal rate of return?
15-4. Why is it that the net present value and internal rate of return methods can yield
contradictory results when evaluating two mutually exclusive investments? De-
velop an example of two investments in which this would happen.
15-5. What is the relationship between a firm's niarginal income tax rate and the net
cost of debt capital to the firm?
15-6. Some analysts claim that the cost of debt capital is lower than the cost of equity
capital for most firms. If this is true, why don't firms rely exclusively on debt fi-
nancing and not sell any additional common stock? Explain.
15-7. One firm in an industry may have 50 percent debt while another may have no
debt. Why do some firms in the same industry have substantially different capi-
tal structures? Why do most firms in some industries (e.g., public utilities) have
a large debt component in-their capital structures while most firms in some other
industries have relatively little debt?
15-8. What is syuefgy as it applies to mergers? How does it affect the amount that one
firm would be willing to pay for another?
15-9. Give at least one example (other than those used in the chapter) of each of the
following types of mergers:
a. Horizontal
b. Conglomerate
c. Vertical
15-10. How could .a conglomerate merger be used to reduce risk?
15-11. Explain how a vertical merger between an electric utility and a coal company
could reduce transaction costs for the combined firm.
15-12. What are three strategies a firm might use to avoid being acquired by another
company?
15-13. Suppose that the capital market has correctly valued the price of a firm's stock.
Why would an acquiring firm be willing to pay more than the market price?
What would determine the prenum over the market price that the firm would
be willing to pay?
Problems
15-1. Staff members of the financial analysis department of Global Electronics have
determined the required investment and the rate of return on each of the follow-
ing capital projects.
Required
Capital Investment Internal Rate
Projects (millions) of Return (%)
A 5.2 12.9
B 8.6 15.2
C 34 10.0
D 5.1 14.8
E 11.2 19.0
F 6.5 7.9
The firm's marginal cost of capital is given by the function

r = 9 ± abC
where r is rate of return (in percent) and C is millions of dollars of capital raised
for investment.
a. Graph the firm's marginal cost of capital function and the firm's capital de-
mand function.
b. Which capital projects should be implemented? What should be the firms to-
tal capital investment?
c. If a general tightening in the financial markets shifts the firm's marginal cost
of capital function to
r = 8 ± 0.35C
determine which projects should he implemented and the total amount spent on
capita] items.
15-2. Tarnutzer Construction Company must replace its front-end loader. The initial cost
and annual net cash flows for the two models under consideration are shown in the
table, Given the heavy use of the machines, the y will be completely worn out at the
end of 5 years and have no salvage value. The firm's cost of capital is 12 percent.
Model
Heavvdujv Sure Shovel
initial cost $10,000 $10,000
Net cash flows (),ear)
1 / 5,000 —4.000
2 5,000 0
3 5.000 6,600
4 5.000 6.600
5 5,000 20,000
Use both the IRR and NPV methods to evaluate these capital proposals. Which
should be purchased? Explain. (Except for the initial cost, assume that all cash
fl ows are received or paid at the end of each year.)
15-3. Top management at Transworld, Inc., a large multinational conglomerate, uses
the net present value method for evaluating capital expenditure proposals. The
firm's cost of capital is 16 percent. Currently, the following eight proposals are
under review:
Initial Net Present
Proposal Cost (millions) Value (millions)
S11.6 . $3.6
2 94 —13
3 8.7 2.2
4 13.2 2,4
5 14.0 2.0
6 7.5 1.4
7 6.5 1.7
8 9.5 —2.4
a. If the corporation has a capital spending budget of $iOO million br the com-
ing year, which capital projects should be implemented? Explain.
b. If the capital spending budget is limited to $47.5 million, which projects should
be undertaken? Explain.
15-4. The board of directors of Alder Enterprises wants to maintain a capital structure
that is 30 percent debt and 70 percent equity (i.e., retained earnings and common
stock). For the current year, the firm expects to earn $2 million after taxes (the
firm's marginal tax rate is 40 percent) on sales of $15 million. Company policy is
to pay Out 50 percent of net after-tax income in dividends to the holders of the
firm's one million shares of common .tuck. Management and outside analysts
project a growth rate of 8 percent for sales, profits, and dividends. The market
value of the firm's common stock is S20 per share.
The average dividend yield on all common stocks is 14 percent, and the in-
terest rate on U.S. government securities is 12 percent. Roth management and the
firm's investment advisers, Saunders & Wennergren, agree that Alder could sell
bonds at an interest rate of about 14 percent. The estimated beta coefficient for
Alder is about 1.5.
a. Determine the cost of debt capital for Alder Enterprises.
b. Del ermine the cost of equity capital using each of the three methods described
in the chapter.
c. Determine the composite cost of capital for the firm using the capital asset
pr icing mode! to compute the cost of equity capital.
15-5. Quicksior, Inc., builds and manages storage units that are rented to individuals
and firms. Typically, the firm builds a number (usually 200 to 500) of these garage-
like units at sites on the periphery of growing urban areas. Most of the time the
storage units are demolished and the site cleared for housing and/or commercial
development within 5 to 10 years.
Management is deciding whether to build a 300-unit storage center between
Oklahoma City and Norman, Oklahoma. Currently, similar units in the area are
being rented for $85 per month. Fch unit would coutain 400 square feet and the
construction cost per square foot, including sitedevelopment costs, would be $8.
The operating costs of the project (i.e., insurance, management, property taxes,
electricity) are estimated at $50 per unit per month. Both the rental rate and op-
erating costs are expected to remain constant for the foreseeable future. Quick-
stor's cost of capital is 10 percent and its marginal income tax rate is 30 percent.
Because of the growth of the University of Oklahoma in Norman, it is likely
that the storage units will he demolished after 10 years to make room for audent
apartments. Demolition costs will be insignificant, and there would be no salvage
value for any of the materials. The entire cost of the project will be depreciated
in 10 years, using the straight-line method.There is no alternative use for the land
for the next 10 years.
a. Given that Quickstor's objective is profit maximization, should this project be
implemented? Use both the NPV and JRR evaluation criteria.
b. If your answer to part (a) is that the project should not be implemented, de-
tcrminc thc maximum initial investment that could be made so that the pro-
ject would be profitable-
1 .5-6. The head of the marketing department (who has had antg in capital bud-
geting). has proposed anew direct telephone advertising program that would re-
quire an initial investriient of S100,000 in equipment and S50,00() per year in ad-
ditionai labor costs, It is estimated that the program would be effective for 5
years and would increase revenue in each of those years by $75,000 over the cur-
rent level. After 5 years, the equipment would be obsolete with zero salvage
value and the advertising program ineffective because of the entrance of new
competition. She argues that the 4-year payback period makes the proposal a
"can't miss" proposition. Do you have any comments on this evaluation proce-
dure? Use the NPV method to evaluate this proposal. (Note: The firm's cost of
capital is 20 percent, it uses straight-line depreciation, and the marginal tax rate
is 50 percent.)
15-7. The engineering department of McDougal Steel has developed the following cap-
ital project evaluations:
Present Value of All
Future Aret Cash Flo W3 lniriai Cost
A $190.000 $200,000
B 210.000 150.000
C 600.000 500,000
1) 490,000 400,000
E 350,000 300,000
a. Assuming an unlimited capital budget, which projects should be imple-

mented? Why?
b. Due to a change in top management philosophy that precludes additional ex-
ternal financing of any kind, the department is faced with a capital budget con-
straint of $500,000. Which projects should be implemented? Explain.
c. Five minutes before the department head is to take the results from part (b)
to the corporate capital budgeting meeting, Smith comes in with a revolu-
tionary new capital item that has a cost of $450,000 and an NPV of $120,000.
Assuming Smith's analysis is accurate, what should the department head
do? Why?
15-8. Given the following information on three major corporations, determine the cost
of debt and the cost of equity capital for each firm using the three methods de-
scribed in the chapter (i.e., the risk-free rate plus risk premium, the discounted
cash flow, and the capital asset pricing methods).
Raze on Price of Dividend
Long-Term Common Annual Growth
Firm Tax Rate Bonds. Stock Beta Dividend Rate
Chrysler 39% Q('OJ-,
U., O 1.20 1.60 14%
Johnson & Johnson 36 7.9 59 1.15 0.88 15
Ohio Edison 38 8.5 22 0.80 1.50 2
Assume that the risk-free interest rate is 5 percent and that the average return in
the stock market is 12 percent.
15-9. Determine the cost of equity capital (using each of the three methods de-
scribed in the chapter) and the cost of debt capital for each of the following
firms. The risk-free rate is 5 percent and the average return on common stock
is 12 percent.
Rate on Price of Dividend

Long-Term Common Annual Growth
Thin Tax Rae Bonds Stock Beta Dividend Rate
Minnesota P&L 34% 7.9% $31 0.70 $2.04 4%
Exxon 38 8.6 63 0,70 3.00
Oracle 34 9.2 47 1.5 20
15-10. World Airways plans to expand its fleet of planes and maintenance facilities in
a number of international locations. Management has identified the following
six projects, each of which requires investment in each of 2 years. The firm faces
a limited capital budget of $200 million in year 1 and $300 million iny ear 2.
Present Value of Outlay Net Present Value
In Year (millions) ,(millions)
Project I(C) 2(C (V
1 40 50 70
2 80 70 100
3 50 60 50
4 40 50 40
5 50 80 90
6 30 50
Set up the linear programming problem to determine which capital projects to

implement.
15-11 The common stock of Throckmorton Machinery is currently selling for $40 per
share, while the price of McKnight Equipment stock is $10 per share. Top man-
agers at McKnight are planning an offer for the shares of Throckmorton. They
believe that by applying their considerable managerial talents, the value of
Throckmorton could be increased 25 percent by a merger with McKnight.
a. If a cash offer is to be made to owners of Throckrnorton common stock, what
is the maximum price that McKriight's managers should offer?
b. If the acquisition is to be made by an exchange of stock, what are the most fa-
vorable terms that McKnight should offer?
c. What will probably happen to the share price of Throckmorton's stock when
the offer is announced? Explain.
15-12. International Steel consistently earns a profit of $100 million per year, and it has
10 million shares of common stock outstanding that have a market value of $150
per share. East Asia Steel earns a profit of $20 million per year, and it has 5 mil-
lion shares of stock outstanding with a market value of $40 per share. Interna-
tional's management sees the acquisition of East Asia Steel as a way to expand
in the Asian market, thus increasing its profits.
a. What is the maximum cash price that International should pay if there is no
synergy in the acquisition (i.e., if the total profit of the combined firm simply
would be equal to the sum of their current profits)?
b. What is the maximum cash price if there was synergy that increased the com-
bined profit by 50 percent?
c. -Assuming no synergy, what is the most faorable stock exchange that Inter-
national could offer?
CHAFFER 15 Capital Budgeting 539
d. Assuming the synergy described in part (h), what is the most favorable stock
exchange that International could offer?
Computer Problems
The following problems can be solved by usin g the TOOLS program (downloadablc
fr om www. prenh alLcom/petersen) or by using other computer software.
15-13. international Communications offers discount long-distance telephone service
to businesses throughout the world.To meet the improved service offered by its
compctitors, 'Naboaall must upgrade its entire switching system. Three alterna-
tive s y stems are available. The initial cost and estimated net year-end cash flows
over the 10-year lives of the systems are shown here. Use both the net present
value (NPV) and internal rate of return (IRR) criteria to determine which sys-
tem should be selected. The firm's cost of capital is 13.4 percent.
System Interstate Scrambler Regent
Initial Cost $300,000 $135,000 $130,000
Net profits (year)
1 45,620 —40,000 5.200
2 51,900 —22.000 8,100
3 55,800 —6,900 10.500
4 !
60,100 5,500 8,100
5 61,000 30,200 10,500
6 60.000 40.800 8,100
7 55,900 60,400 10,500
8 52,000 80,200 88,100
9 50,600 90,000 105,000
10 40200 75,400 115,200
15-14. Margaret Vangilder manages the annual bazaar at the Pacific United Church.
The most recent bazaar resulted in profits of $10,000, but shoppers had to be
turned away because the church recreation room is too small. Given population
and income growth in the area, profits could be expected to increase by 10 per-
cent each year if the church could be expanded. A local contractor estimates that
such an expansion would cost $35,000. The market interest rate is 10.125 percent,
and the church will be torn down at the end of 10 years to make room for a foot-
ball stadium. Should the investmçnt be made?
1515. The Sloan Corporation is considering two projects with the following cash flows:
End of Year
1 2 3 4 5
X $--100,000 $125000 0 0 0 0
V —100,000 0 0 0 0 $228,000
a. Compute the net present value for each project using discount rates of 3,7,9,
11. 13, and 15 percent. Graph the net present value of each project as a func-
tion of the discount rate.
b. Determine the internal rate of return for each project. For what range of dis-
count rates is the ranking based on the internal rate of return criterion con-
sistent with the ranking based on net present value?
15-16. A temporary dam being planned by the federal government has an initial cost
of $60 million and will have net cash benefits extending over a 12-year period as
shown,
Net Cash Flow Net Cash Flow
Year (millions) Year (millions)
1 0 7 15
2 0 8 15
3 5 9 20
4 5 10 21)
5 10 11 25
6 10 12 25
a. Assume a discount rate of 10 percent. Compute the net present value and the
internal rate of return for the project.
b. Repeat part (a) using discount rates of 5 and 15 percent. Does the choice of
the discount rate significantly affect the evaluation of the project? Explain.
15-17. A new technique has been proposed for surfacing roads to prevent potholes. It
can be used for a maximum of 10 years, but its effectiveness decreases somewhat
over time. The initial cost to resurface the road is $90 million, and the future net
cash benefits are shown below.
Net Cash Benefits Net Cash Benefits
Year (millions) Year
1 0 6 16
2 20 7 16
3 20 8 16
4 20 9 16
5 20 10 16
a. Assume a discount rate of 10 percent and that all benefits accrue at the end of
the year. Compute the net present-value and the internal rate of return for the
proposal.
b. Repeat part (a) using discount rates of 5 and 15 percent.
Bentle y Enterprises, Inc.
Bentley Enterprises is a manufacturer and nationwide distributor of specialty food

p roducts, such as ice cream , candy, and bakery goods.
The firm's strategic planning
group has developed a proposal for locating a number of "shaved ice" retail outlets in
the parking lots of shopping centers. The product consists of crushed ice in a paper cone
that is mixed with one or more flavors. Management decides that the l5roduct can be
sold for $0.80 each. Two alternatives have been proposed. In the first, called the "na-
tional" proposal, 500 stands would he located at shopping centers of various sizes across
the Uii.ed States. The second proposal, referred to as the "limited" plan, would place
fewer stands (200) in th large malls of major urban areas.
The initial cost of the "igkolike" structures, including the necessary equipment, is
projected to be $55,000 each. Because of their specialized nature, the buildings will have
no value at the end of the life of the project. The average payment of a 4-year lease for
each parking lot space would be $25,000. The entire lease payment would be paid at the
beginning of the project.Thus, it is a one-time outlay that would pay for space for 4 years
in advance.
At this time, the only competition in this field comes from a few "mom-and-pop"
type operations scattered around the country. However, management at Bentley is con-
cerned that Unique Foods, Inc., its major competitor, might respond to the Bentley plan
with a similar program. Over the past 20 years, Unique has had a tendency to watch the
actions of Bentley and then attempt to duplicate its new programs. In fact, for new prod-
ucts introduced on a national basis, Unique has responded with a similar program about
60 percent of the time. For programs developed on a more limited basis by Bentley,
Unique has come up with an almost identical program in 30 percent of the cases.
Average daily volume for each unit for the first year of operation will depend on
(1) whether the national or limited program is undertaken and (2) whether Unique re-
sponds with a similar program. Estimaths of the volume under each combination of
events are outlined below. The average volume is higher under the limited program be-
cause of the large number of shoppers in the large malls. The stands would be open 365
days a year.
First-Year Average Daily Volume un Its sold)

Program
Unique National Limited
Responds 250 300
Doesn't respond 400 500
Because of the novelty nature of the product, management is assuming a "worst-

case" scenario that there will be-no demand for the product after the 4th year. During
this period, both price and units sold are expected to remain constant.
Although the profit associated with any one unit cannot be estimated with accu-
racy, management thinks that annual profit for the average unit can be determined
based on the following operating profit data for eight experimental units that were in
operation for 1 year at random locations throughout the country. The data does not in-
clude the cost of the structure or the lease payments for parking lot space.
Annual Revenue Annual Operating Profit
Unit (thousands) (thousands)
1 $80,000 0000
2 120,000 28,000
3 60,000 11,000
4 90,000 16,000
5 110,000 20,000
6 100,000 25,000
7 7000 12,000
8 601000 14,000
The management decision rule at Bentley for evaluating projects such as this is
that a project should be implemented only if shareholder value is increased (i.e., if the
present value of all future operating profits exceeds the initial project cost). The only
alternative for the firm would be to invest in a 4-year bank certificate of deposit that
would yield 10 percent interest. Management thinks that the risky nature of the "shaved
ice" stands implies a risk premium of four percentage points over the rate available on
the baijk deposit. Management also insists on having a quantitative measure of the risk
associated with each alternative. (Note:'A general rule at Bentley is to assume that all
annual revenues are received and operating costs incurred on the last day of each year.)
Requirement
What action should the firm take? Should either of the "shaved ice" stand pro-
posals be implemented or should the firm simply invest its money in a certificate of de-
posit? Be sure to evaluate risk. U

Chapter11 15

Uploaded by

Copyright:

Available Formats

Chapter11 15

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Chapter11 15

Uploaded by

Copyright:

Available Formats

CHAPTER

• !ntroducfion to Game Theory

INTRODUCTION TO GAME THEORY

• A p ure sIrit4'l inyolvas ilsay, inakrnj tb mt. 'inui c'vhir cI:n J

Noncooperative Gaines: The Prisoner's Dilemma -

Cooperative Games: Enforcing a Cartel

can -be en1'rc,.'ecL5 ucli 'njj ay bei y

Repeated Games: Dealing with Cheaters

York: Basic Books. 1984).

Price, cost per unit

Present versus Future Profits: Limit Pricing

Profit Stream I Profit Stream It

The Organization of Petroleum Exporting Countries (OPEC) is a cartel consisting of

The Value of a Bad Reputation: Price Retaliation

Example Evaluating Price Retaliation

• Limit pricing involves setting lower prices, which

Establishing Commitment: Capacity Expansion

Preemptive Action: Market Saturation

The analysis of geographic saturation can also be applied to product characteristics.

Pacific Cüpper, a family-owned business, produces copper that is purchased by other

PRICING OF MULTIPLE PRODUCTS

Example Calculating the Profit-Maximizing Prices for Joint Products

Fu.11y Distributed versus Incremeittal Clost Pricing

If an enterprise is providing several goods, Ramsey pricing suggests guidelines for

Yxainple Using Ramsey Pricing

Intermediate Products (Transfer Pricing)

With a perfectly competitive external market, there really is no price decision to be

Because Q. is the profit-maximizing output of writing tablets for the combined

Necessary Conditions for Price Discrimination

Although airline pricing practices have the appearance of price discrimination,

pes of Price Discrimination

Price; cost Price per Unit

() First degree price (b) Second degree

In addition to electricity, second-degree price discrimination is often used in setting

Price, cost Price, cost Price cost

iSi by .CU bflC4..

Q1 =7— and Q11=10—P

Equating marginal revenue to marginal cost gives

Midnight Time of day

COST-PLUS OR MARKUP PRICING

Mechanics of Cost-Plus Pricing

AVC— and AFC=

P='AVC+ AFC ±- (12-4)

Example Cost-Plus Pricing of Automobiles

AFC = !on = $10,000

Evaluation of Cost-Plus Pricing

Cost-Plus Pricing and Economic Theory

But P + Q can be rewritten as -1- Note that (dP/dQ)(Q/P) is 1/EAr,,

Profit maximizatithi requires that MR = MC. As a simplifying assumption, let MC =

Solving for P yields

and the demand for rabbit's feet is given by

- Problems Requiring Calculus

INPUT PRICING AND EMPLOYMENT

Key Concep ': ' . -........

Market Structure I: Monopolist (Product

Market Structure II: Monopolist (Product

Rate Marginal Labor MarginaL

3 18.50 11.50 18.50

Maigioat revenue produc: