F&M Notes Compiled

61
8/13/2019 F&M Notes Compiled http://slidepdf.com/reader/full/fm-notes-compiled 1/61  Game theory and strategic analysis: basic concepts Game theory: - Stylized portrayals of situations involving conflict and/or cooperation - Useful for analyzing small numbers situations - Applicable to oligopoly, customer-supplier relationships, and many other strategic situations in the  business world and outside the business world Interesting facets of game situations: - How many players? - How many choices of actions? - What are the payoffs? - Information -- What do the players know? What do they know about each other? About each other’s  payoffs? What do they know about what the other party knows? - Communication -- Can the parties communicate? Do they want to communicate? Do they want to receive communications? Do they know when communications have been transmitted? Received? Are the communications credible? --- Important communications: threat, promise, commitment - Structure of play -- Is it a single move play? Or multiple moves? -- Is the game played once (one-shot)? Or is it played multiple times, over and over again (repeat play)? --- Repeat play can build trust, encourage cooperation; but a finite end to a series of repeated games may cause unraveling (through backward induction – look ahead and reason back), Firms & Markets (Fall 2013) LJW Class Notes Oct 17

Transcript of F&M Notes Compiled

Page 1: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 1/61

 

Game theory and strategic analysis: basic concepts

Game theory:

- Stylized portrayals of situations involving conflict and/or cooperation

- Useful for analyzing small numbers situations

- Applicable to oligopoly, customer-supplier relationships, and many other strategic situations in the

 business world and outside the business world

Interesting facets of game situations:

- How many players?

- How many choices of actions?

- What are the payoffs?

- Information

-- What do the players know? What do they know about each other? About each other’s payoffs? What do they know about what the other party knows?

- Communication

-- Can the parties communicate? Do they want to communicate? Do they want to receive

communications? Do they know when communications have been transmitted? Received? Are thecommunications credible?

--- Important communications: threat, promise, commitment

- Structure of play

-- Is it a single move play? Or multiple moves?

-- Is the game played once (one-shot)? Or is it played multiple times, over and over again

(repeat play)?

--- Repeat play can build trust, encourage cooperation; but a finite end to a series ofrepeated games may cause unraveling (through backward induction – look ahead and reason back),

Firms & Markets (Fall 2013)

LJW Class Notes Oct 17

Page 2: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 2/61

Page 3: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 3/61

 

Firms & Markets Class Notes Oct 17 Page 3

- If one player has a dominant strategy, the game can also be solved, since the second player can

anticipate that the first will use its dominant strategy; again, order of play doesn’t matter

- If neither player has dominant or dominated strategies, there are no simple solutions; order of playwill matter; going first is usually a disadvantage (equivalent: knowing what the other player is going

to choose is usually advantageous)

-- Having special knowledge about the other party may help determine one's choice; butthere is the risk that the other party may “out-think” you

-- Simple randomization is a risk minimizing strategy for simple games (e.g., the game of

matching heads or tails of coins); someone who chooses to randomize doesn’t care if the other partylearns that randomization will occur

-- For more complex games without a dominant strategy for either side a “maximin” (or

“secure”) strategy may be useful, especially for someone who is risk averse:

--- Maximin strategy: first find the minimum possible outcome from each choice;then select the choice that yields the maximum among those minimums

--- Maximin is the appropriate strategy if a player must be the first to play

--- Maximin may be appropriate for someone who is risk averse, and the

game is being played simultaneously only once (or a few times); but

--- If the game is played repeatedly, randomization can usually improve theexpected outcome as compared to maximin; the goal of randomization is to find a set of

 probabilities (for playing each choice) such that you are indifferent to what strategy the other playerchooses; you are even indifferent to the other player’s discovering your randomization strategy or

even the randomization odds, so long as the other player does not discover the actual choice eachtime; see RANDOMIZED STRATEGY

Page 4: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 4/61

 

Firms & Markets Class Notes Oct 17 Page 4

ZERO-SUM GAMES

A. Look for a dominant strategy (I) (Nash equilibrium):

C

a b

a 3 1

R -3 -1

b -1 -3

1 3

B. Look for a dominant strategy (II) (Nash equilibrium):

C

a b

a 3 -3

R -3 3

1 -1

b -1 1

C.  Look for a dominated strategy (Nash equilibrium):

C

a b c

-3 -1 2a

3 1 -2

1 2 0

R b

-1 -2 0

-2 -3 -1

c

2 3 1

Page 5: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 5/61

 

Firms & Markets Class Notes Oct 17 Page 5

D. A simple game without a Nash equilibrium (randomized

strategy?):

C

a b

a 1 -1R -1 1

-1 1

b 1 -1

D. A more complex game without a Nash solution (maximin?

randomized strategy?):

C

a b

a 3 -3

R -3 3

-1 1

b 1 -1

Page 6: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 6/61

 

Firms & Markets Class Notes Oct 17 Page 6

RANDOMIZED STRATEGY

C

a ba 3 -3

R -3 3 b -1 1

1 -1

What is R's best strategy? If R plays b only (maximin), R could (if C figures it out) consistentlyreceive only -1. Suppose instead R randomizes:

R calculates: Regardless of what C does, I want to choose a probability (p) of playing a and

 probability (1-p) of playing b so that my outcome is the same:If C plays a, then my expected outcome is

 p * (-3) + (1-p) * 1 .

If C plays b, then my expected outcome is p * 3 + (1-p) * (-1) .So, if I am going to be indifferent to C's actions, the two expected outcomes must be

equivalent: p * (-3) + (1-p) * 1 = p * 3 + (1-p) * (-1)

8p = 2 p =! .

So, R should play a with a !  frequency and play b with a "  frequency. R's expectedoutcome is 0 (which is better than -1).

Similarly, what is C's best strategy? (Maximin is to play a only, bringing perhaps –1.) Suppose

instead C randomizes:

Regardless of what R does, I want to choose a probability q of playing a and a probability(1-q) of playing b so that my expected outcome is the same:

If R plays a, thenq * 3 + (1-q) * (-3) .

If R plays b, thenq * (-1) + (1-q) * 1 .

And equivalently:q * 3 + (1-q)*(-3) = q(-1) + (1-q) * 1

8q = 4q =# .

Expected value of the game to C is 0 (which is better than -1).

Page 7: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 7/61

 

Game theory and strategic analysis: extensions

 Non-zero sum games:

- The parties may care about only their own payoffs (absolute amounts); or they may care also about

what they received relative to what the other party received (relative amounts)

-- Most of the analysis will focus on maximizing one’s own payoffs (i.e., focusing onabsolute rather than relative amounts)

- Can start with simple games of pure cooperation (let’s agree on a technical standard, and it doesn’t

matter which one, so long as we agree); see NON-ZERO-SUM GAMES

-- The game is easy to solve if there is communication or if one party goes first

-- If the game is simultaneous, with no communication, then it becomes more “interesting”;the parties must try to figure out things about each other or about "the landscape" of the game to try

to come to a beneficial outcome

- A slightly more complicated game mixes cooperation and conflict (we’re better off agreeing on astandard than not agreeing, but I am better off if we choose my standard and you’re better off if we

choose your standard)

-- There’s a great advantage to going first

-- There is likely to be a great deal of early announcements, efforts to create a presumptionor a “bandwagon” effect; competition to go first

--- If the question of who goes first is unresolved, a credible commitment to have

already made an irrevocable choice serves as the effective first move; but the issue of credibility isalways crucial

- Even more complicated games can bring in elements of threat and promise (or both)

-- Threat: a communication concerning a contingent commitment: “If you take an action that

is adverse to me, I will take an action that is adverse to you and that appears otherwise to not be inmy interests to carry out.”

--- The threat is intended to deter behavior (and induce alternative behavior)

--- Since the threat involves taking an action that would otherwise not be in the first

 party’s interests to carry out, the credibility (and communication) of the threat is crucial

Firms & Markets (Fall 2013)

LJW Class Notes Oct 22

Page 8: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 8/61

 

Firms & Markets Class Notes Oct 22 Page 2

-- Promise: a communication concerning a contingent commitment: “If you take an actionthat is favorable to me, I will refrain from taking an action that is adverse to you (and instead will

take an alternative action that is favorable to you), even though it appears that it would otherwisenot be in my interests to refrain.”

--- The promise is intended to encourage behavior

--- Since the promise involves taking an action that would otherwise not be in the

first party’s interests to carry out, the credibility (and communication) of the promise is crucial

-- Issues such as who knows what and what is the order of “play” become quite important

The prisoner's dilemma:

- Game that contains a set of payoffs such that when each player follows his/her dominant strategy,

the stable (Nash) solution (confess, confess) that is reached causes the players jointly to be worse offthan if they could somehow reach a common understanding on a different set of strategies (don’tconfess, don’t confess); equivalently, when each party follows its dominant strategy, the parties are

 jointly worse off than if they could (somehow) escape from logic of following the dominantstrategy; see PRISONER’S DILEMMA

- The prisoner’s dilemma game (like most such games) can be portrayed through a game theory box

form or through a decision tree (extended form); see DECISION TREE (EXTENSIVE FORM)PRISONER’S DILEMMA

Parallels to a business pricing game:

- For example, two competing businesses face prospective choices as to whether they should

commit to a low price or a high price strategy; both are better off with (high, high); but low is adominant strategy for each, so both may end up with (low, low) where they are worse off

Applications to other business situations:

- The insights of the prisoner’s dilemma structure (each party, pursuing its own dominant

strategy, causes an unfavorable joint result) extend widely through the business world: e.g.,decisions to expand capacity; decisions to begin advertising campaigns; decisions to enter into a

 joint venture with other enterprises

-- In a joint venture to pursue R&D, for example, each member would like to “freeride” on the efforts of the other members (for example, send lesser quality employees to the joint

effort); but everyone tries to free ride, the joint venture will yield lower returns to everyone

- The insights also extend beyond the business world to explain other phenomena, such aslittering and pollution problems in general

Page 9: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 9/61

 

Firms & Markets Class Notes Oct 22 Page 3

NON-ZERO-SUM GAMES

A.  Pure cooperation:

C

a b

a 1 0

R 1 0

b 0 1

0 1

B. Cooperation and conflict; the role of credible commitment:

C

a b

a 1 0

R 2 0

0 2

b

0 1

C. Cooperation and conflict; the role of credible threat

(contingent commitment):

C

a b

a 2 3 Suppose C moves first -- but R

can communicate beforehand…

R 4 22 1

b

2 1

Page 10: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 10/61

 

Firms & Markets Class Notes Oct 22 Page 4

D. Cooperation and conflict; the role of credible promise

(contingent commitment):

C

a b

a 3 4 Suppose C moves first -- but R

can communicate beforehand…

R 3 1

1 2

b

4 2

E. Cooperation and conflict: the role of credible threat and

credible promise (contingent commitments):

C

a b

a 3 4 Suppose C moves first -- but R

can communicate beforehand…

R 3 2

1 2

b

4 1

F. Cooperation and conflict: when the principle of “looking ahead

and working back” overrides a dominant strategy:

C

a b

a 4 3 Suppose R moves first…

R 3 1

1 2

b

4 2

Page 11: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 11/61

 

Firms & Markets Class Notes Oct 22 Page 5

PRISONER’S DILEMMA

A.  The “pure” model:

C

don’tconfess confess

-5 -10

confess

-5 0

R

0 -2

don’t conf.

-10 -2

B.  Pricing problem:

C

low high

price price

5 0

low price

5 15

R 15 10

high price0 10

C.  A joint research venture:

C

a b

a 5 3 a: Send a low quality

employeeR 5 8

8 7 b: Send a high-quality

b employee

3 7

Page 12: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 12/61

 

Firms & Markets Class Notes Oct 22 Page 6

DECISION TREE (EXTENSIVE FORM) PRISONER’S DILEMMA

Outcomes: R CC’s decision nodes

Confess -5 -5

R’s decision node

Confess

Don’t confess0 -10

Confess -10 0Don’t confess

Don’t confess-2 -2

Page 13: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 13/61

 

Oligopoly basic concepts

Oligopoly:

- Small numbers of sellers (not monopoly; not competition)

-- The sellers are identifiable to each other

-- They are interdependent: What one seller does affects the others, and vice-versa

-- Mutual recognition: they all recognize this interdependence

-- They may be selling identical or differentiated products

- Strategic interactions are likely; what one firm should do depends on its anticipations of what theothers are likely to do; but they are making the same kinds of determinations

- Obtaining definitive results is difficult; there is a lack of predictability

- The prisoner's dilemma provides useful insights; but it involves discrete choices rather thancontinuous possible choices; and it often leads to the conclusion “it all depends...”

-- Consequently, many economists have tended to look elsewhere, looking forsimplification, mathematical tractability, definitive results; this has a long history

Cournot model

- Quantity as the strategic variable

- Each seller assumes (myopically) that the other(s) will hold their quantities fixed, and then the firstadjusts its quantity so as to maximize its profits

- All sellers behave similarly; a Nash equilibrium is reached

- Example of 2 sellers; linear demand curve; constant MC, identical for the 2 sellers

-- Equilibrium is reached; QM<QCOUR <QCOMP; PM>PCOUR >PCOMP 

-- Also as the number of sellers behaving in the Cournot fashion increases, quantityincreases and prices fall

Firms & Markets (Fall 2013)

LJW Class Notes Oct 24

Page 14: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 14/61

 

Firms & Markets Class Notes Oct 24 Page 2

-- These results have some appeal; and many economists love the tractability of the Cournotmodel; but

- The model is very formalistic, artificial; the firms are passive; they continue to act myopically;they never learn

Stackelberg model

- The Cournot model can be expanded by allowing one firm to “move” first with the knowledge thatthe second firm is going to follow; the first firm takes the second’s actions into account whenchoosing its quantity

-- There is a first mover advantage

-- There is a game theory flavor to this approach

-- But it still has the Cournot assumption and all that goes with it

Bertrand model

- Assumes that price is the strategic variable

- Each firm sets its price on the assumption that the other firm(s) price remains unchanged; again,myopic behavior

- Bertrand behavior and outcomes can be analyzed in the context of homogeneous goods(commodities) or differentiated products; see THE BERTRAND (PRICE AS STRATEGIC

VARIABLE) OLIGOPOLY MODEL

Homogeneous products

- Since each firm assumes that the other(s) will hold prices fixed, the first firm's best strategy is toset its price at a level that is just below the price of the others

-- But all firms are making the same assumptions and behaving in the same way

-- Because the product is a commodity, any time one firm charges a lower price, all the buyers flock to it

- The result: prices quickly ratchet down to the level of P=MC; this outcome is stable; it’s a Nashequilibrium; no firm wants to set its prices any lower than MC

-- This is the competitive outcome; it will occur even if there are only two firms

-- Even if there are fixed costs, if MC is constant (and the same for all firms in the industry),then the Bertrand assumption leads to P=MC (and the fixed costs won’t be recovered)

Page 15: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 15/61

 

Firms & Markets Class Notes Oct 24 Page 3

-- If MC is rising (difficulties of managing a larger enterprise, etc.), then Bertrand will stilllead to P=MC, but the rising MC means that fixed costs can be recovered (again, this is just anotherway of getting to the competitive outcome)

- The Bertrand outcome is another version of the (low,low) outcome for the prisoner’s dilemmaapproach to oligopoly

Differentiated products

- Each firm produces a somewhat similar but not identical product; each firm’s sales are affected byits own price and by the prices of its rivals

- Each firm sets its price myopically, assuming that the other firm(s) will keep their pricesunchanged

-- This maximization process leads to a “reaction curve” for each firm

-- Because the products are differentiated, one firm's lower price attracts some demand to it, but it doesn’t “steal” the entire market

- The outcome is a Nash equilibrium with P>MC; positive (above-competitive) profits

-- Product differentiation “softens” price competition

-- This result reinforces the result learned from the monopolistic competition model: if afirm has a niche that is not easily replicated, it can earn profits that are above the normal competitive

level

- The Bertrand differentiated-product model can be expanded to the case of a “first mover”, whosets his price first on the assumption that the second firm will continue to behave in the Bertrandfashion

-- The first mover is better off than if he moves simultaneously; but

-- The second mover is even better off, since he can choose his price in light of the price thatthe first firm chose (a relative second-mover advantage)

- Overall, the Bertrand result has some appeal; for the homogeneous product case, it shows how the(low, low) outcome for the prisoner’s dilemma could be reached; for the differentiated productversion, it shows the value of product differentiation, escaping from the harsh price competition ofcommodities; it does capture some of the flavor of the prisoner’s dilemma; the first-mover/second-mover outcome for differentiated products has intuitive appeal

-- But Bertrand, in its formal version, still has a myopic flavor; and it is limited as towhat other features of the market can be brought into the picture

Page 16: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 16/61

 

Firms & Markets Class Notes Oct 24 Page 4

- An expanded view of the prisoner’s dilemma can give us important insights for understandingoligopoly

Page 17: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 17/61

 

Firms & Markets Class Notes Oct 24 Page 5

THE BERTRAND (PRICE AS STRATEGIC VARIABLE) OLIGOPOLY MODEL

I. Homogeneous products.A. General assumptions:

- Only a “few” firms are in the industry; entry barriers are high

- Firms sell identical products (commodities)- Each firm considers price to be its strategic variable- Each firm sets its price myopically on the assumption that the other firm(s) will

keep its price unchanged

B. Specific example:- 2 identical firms produce widgets- MC1=MC2=$50- Linear (market) demand: Q = 10,000 - 20P; P = 500 - 0.05Q- Neither firm can be permanently driven out of the market; each firm can readily re-

enter the market

C. Specific outcome:- Since each firm assumes that the other firm’s price is fixed, each tries to set its

 price a little bit lower than the other's price (so as to capture the entire market)- Since both firms behave this way, there's only one stable (Nash) outcome:

P1=P2=$50; at that point, neither firm wants to reduce its price any further; Q=q1+q2=9,000;!1=!2=0; this is the competitive outcome 

(Note: If the two firms colluded and achieved the monopoly outcome, PM=$275;QM=4,500; !M=$1,012,500)

$500D

MR

275

50 MC

0 4.5K 9K 10K Q

D. General outcome:- Firms that behave according to the Bertrand assumption will reach equilibrium at

P=MC -- the competitive outcome

Page 18: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 18/61

 

Firms & Markets Class Notes Oct 24 Page 6

- Note: If the two firms above each also have fixed costs (say, FC1=FC2=$40,000),the Bertrand assumption still leads them to set P1=P2=$50; they will not be able to set prices thatcover their fixed costs; they will complain bitterly about "cut-throat competition"

II. Differentiated products

A. General assumptions- Only a few firms are in the industry; entry barriers are high- Each firm sells a somewhat similar but not identical product; each firm’s sales are

affected by its own price and by the prices of its somewhat similar rivals- Each firm considers price to be its strategic variable- Each firm behaves myopically with respect to the prices of its rivals

B. Specific example:- 2 firms each producing a somewhat different style of widget (e.g., blue widgets vs.

red widgets)- MC1=MC2=$50

- Firm 1: Demand: Q1 = 5000 - 20P1 + 10P2 - Firm 2: Demand: Q2 = 5000 - 20P2 + 10P1 !1 = P1Q1 - C1 = P1(5000-20P1+10P2) - 50(5000-20P1+10P2)d!1/dP1 = 5000-40P1+10P2+1000 = 040P1 = 6000 +10P2 P1 = 150+0.25P2

Similarly, P2= 150+0.25P1 These are "reaction curves" Note: Identical reaction curves are reached if demand is expressed “inversely” (i.e.,

P1=250-0.05Q1+0.5P2) and the firm sets MR=MC (i.e., d(Q1P1)/dQ1=50)

P1  Firm 2’s reaction curve275Firm 1’s reaction curve

200

150

0 150 200 275 P2

C. Specific result:P

1 = 150+0.25P

2P2= 150+0.25P1 P1 = 150+0.25*(150+0.25P1)

P1 = $200; P2 = $200; Q1 = 3000; Q2 = 3000; !1 = !2 = $450,000 Note: A purely “competitive” result (P=MC) would yield P1=$50; P2=$50 Note: If the two firms collude:P1 = $275; P2 = $275; Q1 = 2250; Q2 = 2250; !1 =!2 = $506,250

Page 19: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 19/61

 

Firms & Markets Class Notes Oct 24 Page 7

D. General result:-  Product differentiation “softens” the price competition that follows a Bertrand format

III. Differentiated products and one firm sets its prices first and the other firm follows

The firm that sets its prices first should “look ahead” and anticipate that the second firm will setits prices in accordance with its reaction function. In the example above, firm 1 would substitute

firm 2’s reaction function (P2 = 150+0.25P1) into its profitability condition (!1 = …). Firm 1 thentakes its first derivative and solves for its own maximizing price; and then firm 2 sets its price inaccordance with its reaction function. This will yield a result of

P1 = $210.75; P2 = $202.68; Q1 = 2812; Q2 = 3054; !1 = $453,476; !2 = $466,285Thus, though the first firm does better than if it doesn’t anticipate the second firm’s reactions, it isthe second firm that does even better by being the second firm. (There is a relative second-moveradvantage.) 

Page 20: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 20/61

 

Oligopoly extensions

The prisoner’s dilemma once again

- Whether the players reach the worse or better outcome will depend on a number ofstructural/behavioral factors surrounding the game:

-- Communication: can they communicate? If yes, can either or both marshal threats, punishments, and/or promises? Are the threats, promises credible? Why? Past experience? Futureexpectations?

-- Monitoring: how quickly do they learn about and can react to each other’s actions?

-- Sociology: do they trust each other? prior experiences?

-- Numbers of plays: is this a game that is played just one time only, or are there multiple plays of the game? in the latter case, the players may be able to learn to trust each other over time; but, does the game have a finite end? or no foreseeable end (or uncertainty and fuzzinesssurrounding the end)

-- Sequence of play

-- Size and relative magnitudes of payoffs: what are the gains to confessing? What are therisks to not confessing?

-- How many players are involved? (the confession of any one of whom can send the othersto jail)

-- How aggressive is the district attorney?

Application of the prisoner's dilemma to the business pricing game

- Though the prisoner’s dilemma is best at portraying discrete choices (the Bertrand model is betterat portraying continuous choices), the prisoner’s dilemma approach allows for more influences onthe outcome; it provides for a “richer”, more complete way of thinking about oligopoly (and theinsights carry over to a more continuous framing of the problem)

-- This approach is consistent with that portrayed by Michael Porter’s “Five Forces”

- Whether the sellers in an oligopoly end up at (high, high) or (low, low) depends on a number offactors (the same elements that indicate whether a cartel is likely to hold together):

Firms & Markets (Fall 2013)

LJW Class Notes Oct 29

Page 21: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 21/61

 

Firms & Markets Class Notes Oct 29 Page 2

-- Communication: Can the sellers communicate explicitly? Are there risks fromcommunication?

-- Are threats, punishments, promises possible?

-- How many sellers?

-- What are the relative sizes (market shares) of the sellers?

-- How easy is entry?

-- What do the industry participants know about each other's “payoffs”? What else do theyknow (or not know) about each other?

-- How easy is it for the smaller firms (fringe) to expand their output/sales? Do they facecost limitations? Brand name limitations?

-- How elastic or inelastic is demand? Lesser elasticity will mean greater rewards forcooperation that results in higher prices; it may be worthwhile to take greater risks to achieve acooperative outcome

-- Monitoring: how quickly do the firms learn about each other’s actions, and how quicklycan they react?

-- Do they have similar or disparate cost structures?

-- Is the industry characterized by low marginal costs relative to average costs (or relative to

"break-even" prices), which provides a substantial temptation to cut prices

-- Do they have any experience of trust? What is the culture of the industry?

-- Is the product uniform (a commodity) or differentiated; a commodity product makes iteasier for sellers to monitor each others’ actions; price cuts are more easily observed and matched(and therefore are deterred); but if price cutting should arise, a commodity market will likely resultin a “slide” to something close to a fully competitive outcome (in a differentiated market, the pricecutting would be “softened” by the differentiations)

-- What is the nature of the buying side of the market? Are there large, knowledgeable buyers who buy in large (tempting) orders? (this is like a single play of the prisoner’s dilemmagame) Or do most orders arrive as small “dribs and drabs”, which individually pose less temptationfor price cutting?

The strategic entrant

Page 22: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 22/61

 

Firms & Markets Class Notes Oct 29 Page 3

- How can/should an entrant into an industry with a dominant firm behave strategically so as toenhance its chances of success? How can the incumbent best fend off the entrant? See Mini-Case:THE INCUMBENT MONOPOLIST AND THE STRATEGIC ENTRANT

Page 23: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 23/61

 

Firms & Markets Class Notes Oct 29 Page 4

Mini-Case: THE INCUMBENT MONOPOLIST AND THE STRATEGIC ENTRANT

Suppose that the Onliwon Widget Corp. (OWC) is currently the only producer and seller ofwidgets. It has long-run marginal costs of $100 per widget, and it currently sells one widget peryear to each of 100 customers at $200 each (which it perceives to be its profit-maximizing price).

OWC has promised all of its widget customers that they will receive the lowest price that it chargesto any of them; this is often described as a “most favored customer” clause. In essence, OWC has promised price uniformity (at whatever price that it chooses).

The Weewantin Corp. (WC) is the only likely entrant into the widget business. If it decidesto enter, it has to decide how many of OWC’s customers to target, and what price to charge. WC

can only make a sale if it approaches a customer; equivalently, a customer can’t buy from WCunless it has been approached by WC.  WC knows the above information about OWC and knowsthat OWC might decide to “accommodate” its entry (by abandoning some customers) or might“fight” to retain its customers by lowering its price (to everyone). OWC knows that WC is a

 substantial company, which has entered many other industries and cannot be easily scared off by

low prices that are perceived to be temporary. 

You have been asked to be a consultant to WC. For each of the two scenarios described below, WC wants to know whether it makes sense for WC to enter. If not, why not? If so, thenhow many customers should WC target, and what price should it charge?

1. Suppose that WC can produce widgets that all of OWC’s customers believe to beequivalent to OWC’s widgets, although they recognize the different brand-name of WC. If WCapproaches a customer and offers to sell that customer a widget at a slightly lower price thanOWC’s price, it will gain that customer; at equal prices customers will stay with OWC.

WC has long-run marginal costs of $100 per widget. It will also have to spend $250 in

market development costs to enter the widget market, which will be sunk after they are spent.Should WC enter? How many customers should WC approach, and what price should itcharge? (Hint: WC wants OWC to “accommodate” rather than “fight”.)

2. Suppose instead that WC can only produce widgets that all of OWC’s customers believeto be inferior to OWC’s widget’s, and so they will not buy an WC widget unless its price is $40cheaper than the price of an OWC widget. If WC approaches a customer and offers to sell thatcustomer a widget at a price that is more than $40 lower than OWC’s price, it will gain thatcustomer; otherwise the customer will stay with OWC.

Also, WC now has long-run marginal costs of $120 per widget. And, again, it will also haveto expend an aggregate of $250 in market development costs to enter the widget market, which will be sunk after they are expended.

Should WC enter? How many customers should WC approach, and what price should itcharge?

3. Oooops. There's been a mix-up in assignment letters. You were really supposed toadvise OWC on its potential strategies for fending off entrants such as WC. Though your professional ethics code prevents you now from switching sides, WC is also interested in hearingyour views on how OWC might fend off entrants. So, what might OWC do?

Page 24: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 24/61

 

Uncertainty and risk; introduction to asymmetric information

Uncertainty

- Most of the discussion so far of demand, supply, monopoly, competition, etc., has been in the

context of full information; everyone knew everything that was important for making decisions

-- There have been a few instances of uncertainty: firms that want to practice pricediscrimination but don’t know which customers have which demand elasticities; some game theory

 problems, where the other party’s payoffs are not known and/or the other party’s likely actions aredifficult or impossible to predict

Uncertainty in general:

- Generalized uncertainty: What will the weather be tomorrow? next week? next year? What will

the price of Microsoft’s stock be tomorrow? next week? next year? What will the sales of the XYZcompany be next week? next month? next year? What will the XYZ company's input prices be next

week? etc. (Also, will a flipped coin land as heads or tails?)

- Why do individuals (owners, managers, etc.) care? It can be costly if individuals make wrongguesses, wrong choices

- Individuals often form estimates (e.g., the mean expected value; or the median); but there is

always a range around the estimate; and they care about that range; they care about variance

- Individuals care about variance because generally people are risk averse: we prefer a sure thingover an uncertain outcome that could yield higher or lower values but that has the same expected

(mean) value

-- Due to declining marginal utility of income: the utility of an additional dollar gained isless than the utility of a dollar lost; see “UTILITY” AND WEALTH OR INCOME

-- A risk-averse person should be willing to pay something to shed risk; the bigger the risks,

the more the person should be willing to pay; equivalently, if asked voluntarily to take on extra risk,the individual will require extra compensation (and if someone is increasingly risk averse, the extra

 payment will have to rise faster than the extra risk)

- These concepts of risk aversion underlie the modern theory and practice of finance; they underliethe risk-return indifference curves that are the staple of finance

Reducing the consequences of uncertainty

Firms & Markets (Fall 2013)

LJW Class Notes Oct 31

Page 25: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 25/61

 

Firms & Markets Class Notes Oct 31 Page 2

- Firms and individuals are not passive in the face of uncertainty; they take a panoply of risk-

reducing actions (often at the sacrifice of some gain), including

-- Buy insurance: pay something -- above the fair odds of the situation -- in order to reduceuncertainty and especially avoid a large downside outcome (even if it has a low probability of

occurring)

-- Diversify: portfolios; suppliers; channels of distribution; even lines of business

-- Hedge

-- Gather more information; hire experts, consultants (management consultants, investment bankers, financial advisors, etc.)

--- The advisors' value is partly in trying to improve expected values; but

also in reducing variance

-- Shed risk; refuse bets; walk away from risky situations; sell risky assets or risky lines of business

-- Build up assets; save for a rainy day; maintain inventories (of input supplies; outputs;

spare parts; machines; back-up workers; money in the bank; lines of credit); maintain redundancy

-- Vertically integrate (gain greater control over input processes)

Asymmetric information:

- In many business circumstances, one side of a transaction often knows things (usually about itselfand its behavior) that the other side doesn't know. This is a general problem of “asymmetric

information”

-- This problem has already appeared earlier in this course, in the circumstances where aseller with market power tries to determine which are the high willingness to pay customers

-- The problems of asymmetric information are much more general and will be addressed in

circumstances where market power is absent -- i.e., in markets where competition prevails

- The problems of asymmetric information can be classified into two general categories:

-- Problems of “hidden information”: one side knows about its own characteristics; in goodsmarkets the consequence of hidden information can be described as a problem of “lemons” (the

term is taken from the used-car business, where a poorly performing car is described as a “lemon”);in financial markets, especially insurance, the problem is often described as one of “adverse

selection”; this latter term is sometimes used more broadly to cover a wider range of hidden-information problems

Page 26: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 26/61

 

Firms & Markets Class Notes Oct 31 Page 3

-- Problems of “hidden action”: one side knows more about its own actions and whether

they are coincident with the interests of the other side; the consequence of “hidden action” may be problems of “agent-principal” or “moral hazard” (these phrases are often used interchangeably)

Page 27: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 27/61

 

Firms & Markets Class Notes Oct 31 Page 4

“UTILITY” AND WEALTH OR INCOME

Utility

U = f(W)

U’ = dU/dW > 0U” = d2U/dW

2 < 0

Wealth (or income)

Page 28: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 28/61

 

Asymmetric information applied

An example of hidden information:

- Suppose that there are two kinds of strawberries: “OK” strawberries, which will cost and sell for$2/quart in a competitive market; and “really wonderful” strawberries, which will cost and sell for$4/quart (in a competitive market)

-- With complete information, buyers who want and are willing to pay for high qualitystrawberries do so; buyers who want and are willing to pay for low quality strawberries do so; bothkinds of strawberries are sold

-- But what happens if buyers are unable to discern which are the high quality strawberries?Suppose that all strawberries look alike, and buyers can only figure out which is a high qualitystrawberry when they get home from the store (when it’s too late to do anything about it)

--- In this circumstance, in the absence of any corrective actions, the high qualitystrawberries will disappear from the market; only low-quality strawberries will be able to survive;competition will drive the price down to $2

- In general, if buyers cannot discern quality differences among sellers, then costly qualitydifferences cannot survive in the market; equivalently, if buyers have no appreciation for qualitydifferences, then they will view the items as commodities and only the low-cost low quality products can survive in a competitive process where unlimited numbers of low quality items can be produced (e.g., only low-quality cars will survive)

-- Though some buyers may be subject initially to fraud (i.e., low quality sellers whomasquerade as high quality sellers and charge high prices), the competitive process will drive downthe price of the items to the price that is appropriate to the low quality items, and buyers are likelyeventually to learn that only low quality products should be expected

-- The absence of high quality items, even though some buyers want them and are willing to pay for them (if only the buyers could discern which they really are), represents a marketinefficiency; this is the “lemons” problem (or, equivalently, a problem of adverse selection): The potential buyers of high-quality items are adversely selected against (and they must either buy lowquality items or not buy at all); it also represents a market opportunity/challenge to the potentialsellers of high quality items (and to the buyers who want them)

- A similar problem arises in insurance: Sellers (insurers) may not be able to discern differencesamong buyers; see Mini-Case: AN EXAMPLE OF THE ADVERSE SELECTION PROBLEM ININSURANCE

Firms & Markets (Fall 2013)

LJW Class Notes Nov 5

Page 29: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 29/61

 

Firms & Markets Class Notes Nov 5 Page 2

-- If there are individuals that represent different levels of risk to the insurer (e.g., differentfrequency of accidents), then insurers want to identify which individuals represent which risks andcharge the appropriate prices (insurance premiums) to each category; with complete informationthey will do so

-- But if insurers cannot discern who represents which levels of risk (or the insurers may noteven be able to discern that there are different levels), then they only observe average riskfrequencies and set their prices accordingly; but this average price is too high for the low-riskindividuals, who may well drop out of the insurance market and self-insure (and it is a bargain forthe high-risk individuals, who will want to buy as much as they can); the result is that only high-riskindividuals are left in the pool of insured individuals, and a high premium (appropriate to their risk)is charged to them

-- Again, this is the problem of adverse selection: The insurers have been adversely selectedagainst (since the low-risk individuals drop out of the market)

-- Since (from the perspective of the insurers) the low-risk individuals are high quality, theadverse selection problem is the same whether it is buyers or sellers suffer from asymmetricinformation: The high quality end of the market disappears

- Lenders face the same problem: trying to identify who (among loan applicants) are the ones whoare likely to repay the loan

-- A lender has to worry, “Are the loan applicants who are knocking on my door the onesthat other lenders have already rejected?”

- Employers face the same problem: trying to identify who is a “high productivity” employee and

who is a “low productivity” employee

-- An employer has to worry, “Are the job applicants who are applying for the position that Ihave advertised the ones who other employers have already rejected?”

-- If an employer cannot identify the “high productivity” employees, then the employer will be “adversely selected against” and will end up hiring only “low productivity” employees

Hidden action problems:

- Often in service markets, sellers have expertise that buyers do not have; it is difficult for buyers todiscern whether the actions that the sellers are taking, and the fees that they are charging, areappropriate to the task/problem at hand

-- This is the agent-principal (or moral-hazard) problem; at the limit, the service providerwill wholly exploit the ill-informed buyer: do very little and charge extremely high fees (moregenerally, moral hazard or agent-principal problems refer to the changed behavior by an individualor firm in response to the incentives of an arrangement that is clouded in asymmetric information)

Page 30: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 30/61

 

Firms & Markets Class Notes Nov 5 Page 3

-- As buyers learn about their exploitation, they will refrain from buying; buy much less; orsometimes “vertically integrate” and attempt to do the activity themselves

-- Again, the full-blown problem represents a market inefficiency, and a challenge to (butalso an opportunity for) market participants

- Insurers face a similar problem: once they have provided insurance, insured parties have lessreason to be careful, since the insurance will cover their losses from accidents

-- In insurance, this is described as the moral hazard problem: the frequency of accidentsincrease subsequent to the granting of insurance

-- The consequence is that insurance policies become more costly (insurers charge higher prices) than if the insureds took more care

- Lenders face a similar problem: once they have made a loan, how do they ensure that it will be

repaid? How do they prevent the borrower from “taking the money and running”?

-- This includes the problem of bond-holders vis-à-vis equity holders

- Managers face the same problem in trying to figure out which employees are working to their full potential and which employees are slacking off

-- This is a general problem for owners (shareholders) of a company vis-à-vis the CEO andsenior management; senior management vis-à-vis junior management, etc.

Solutions to the adverse selection (lemons) problem for goods:

- Sellers and buyers both are generally aware that they face an asymmetric information problem;they are usually not passive but try to use various means and mechanisms to pierce the “fog” ofasymmetric information: sellers of high-quality goods/services want to establish credibly that theirgoods/services are high quality (buyers of high quality goods/services want to assure themselves ofthe same); among the mechanisms

-- Establish a distinctive brand, so that buyers can identify your goods/services; use the brand as a means of establishing a reputation (buyers will seek out branded items)

-- Advertise; provide information (point-of-sale, etc.) (buyers will gather information)

-- Obtain certification, endorsements (buyers will seek certified items, etc.)

-- Offer samples, often at reduced prices (buyers will seek samples)

-- Offer warranties/guarantees; warranties may be too costly for low-quality sellers to offer(buyers will seek warranties)

Page 31: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 31/61

 

Firms & Markets Class Notes Nov 5 Page 4

--- In these circumstances, warranties constitute a “signal”: an action or mechanismthat is too costly for low-quality sellers to employ and therefore something that buyers can reliablyuse to identify high-quality items

--- “Signals” and “screens” are similar: Mechanisms that allow the market to

distinguish quality because it is too costly for the low quality party to send the signal (or passthrough the screen); signals are “sent” by the party that is trying to distinguish itself; screens arecreated by the other side of the market (but if buyers say, “I will only believe that a product is highquality if it carries a warranty,” and then sellers respond with warranties, is this a case of a screen ora signal?)

--- The offeror of a warranty has to make the offer credible to customers (“How do Iknow that the warranty will be honored?”)

--- The offeror must worry about abuse of the warranty by customers (which is aseparate problem of asymmetric information: in essence, the offeror of a warranty is providing an

insurance policy to its customers…)

--- A warranty is worthwhile for the seller to offer to buyers only if its benefits to theseller exceed its costs

-- Promote the firm's reputation (buyers will seek firms with good reputations)

Solutions to the adverse selection problem in insurance, lending, employment

- Insurance companies' efforts to deal with adverse selection:

-- Primarily, they try to gather extensive information about their insureds that will be linkedto the likelihood that they are high or low risk (i.e., the extent to which they are accident-prone)

--- The accident experience of the insured may well be important

-- Use deductibles and co-pay (loss sharing)

--- May use deductibles and co-pay in coordination with policy price and otherfeatures of the policy to induce high risk and low risk individuals to self-select to appropriate policies

-- Set limits on coverage

-- “Just say no”

-- Note that it is harder for low-risk (high quality) individuals to break through theasymmetric information “fog”, since they do not have branding and advertising and reputation andsampling to rely on (although their experience as an accident-free customer with an insurancecompany is usually a valuable piece of information that is used by the insurance company)

Page 32: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 32/61

 

Firms & Markets Class Notes Nov 5 Page 5

- Banks' efforts to deal with adverse selection (the basic problem: banks have to protect themselvesagainst lending to borrowers who won't repay):

-- Collect information

--- Personal loan: personal finances, tax returns, credit history

--- Business loan: business finances, balance sheet, P&L statement, tax returns, business plans, credit history

--- “Credit scoring” is the automated version of this collection of information

-- Ask for collateral as “security” for the loan

--- Collateral may be a “signal”: high-risk borrowers may be reluctant to “post”

collateral, since their default on the loan will cause them to lose their collateral

-- Ask for a co-signer

--- Again, this may be a signal (a cosigner may be reluctant to do so for a high-risk borrower)

-- Require a downpayment (e.g., 20% downpayment on a home); equivalently, have amaximum loan-to-value (LTV) ratio (e.g., maximum of 80%)

--- In the event of a loan default, the bank has a “cushion” to protect itself against a

reduction in the value of the collateral

--- This is similar in effect to a deductible in an insurance policy

--- Again, it may serve as a signal

-- Set limits on the size of loans

-- “Just say no”

-- Again, it is harder for low-risk (high quality) individuals to break through the asymmetricinformation “fog”, since they do not have branding and advertising and reputation and sampling torely on to demonstrate their high quality (high creditworthiness); but having a good credit historyhelps, etc.

- Employers’ efforts to deal with adverse selection:

-- Collect information

Page 33: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 33/61

 

Firms & Markets Class Notes Nov 5 Page 6

--- Hold interviews

--- Inspect resumes

--- Check references

-- Probationary employment (“sampling”)

-- Testing; certification

-- Require educational degree

--- Is this because the degree indicates the acquisition of knowledge? Or because thedegree is a “signal” (screen) for high productivity workers

--- If it is too costly (e.g., personal costs) for a low productivity person to persist

through an educational degree, then a degree may be (correctly) positively associated with high productivity, but the reason will be related to this signaling/screening rather than to any underlying“education” associated with the degree

- Overall, these mechanisms for dealing with adverse selection aren’t perfect; but they are used, because the parties on both sides of these kinds of transactions understand the adverseselection/lemons problem and actively try to deal with it

- Efforts to identify the better customers (lower cost, lower risk, more profitable, etc.) or betteremployees or better products, etc., that are embedded in a larger population is sometimes describedas “cherry picking”

Solutions to the agent-principal/moral hazard problem:

- Sellers and buyers are not passive in the face of these problems; sellers want to try to reassure buyers that the latter will not be exploited (buyers want to be reassured)

- Solutions in goods and services markets:

-- Branding, the establishment of reputation (buyers will seek out sellers with reputations fortrustworthiness)

-- Advertising (buyers will gather information about sellers)

-- Recommendations/referrals

-- Direct monitoring

-- Sampling

Page 34: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 34/61

 

Firms & Markets Class Notes Nov 5 Page 7

-- Certification or endorsements

-- Warranties and guaranties

-- Incentive pricing

--- Staged payments for work completed

--- Bonuses for early completion of projects; penalties for delays

--- Contingency fee arrangements

- Insurance companies address their problems of moral hazard behavior on the part of their insureds by:

-- Incentive pricing: deductibles and co-payment arrangements

-- Historical experience pricing: rewarding good risks (low accident claims) with lower prices (lower premiums); penalizing poor risks (high accident claims) with higher prices

-- Setting rules and cancelling policies (or not paying off) if an insured party has violated therules

-- Direct monitoring

-- Again, it is harder for individuals to convince insurers that they (the insureds) will not“exploit” the insurance company by taking less care; but personal experience may matter

- Banks address their problems of moral hazard behavior on the part of borrowers (banks have toassure themselves that borrowers, after receiving the loan, won't behave in ways that reduce the probability of repayment) by:

-- Direct monitoring of the borrower

-- Covenants/lending agreements

--- Restrictions on borrowers’ behavior; monitoring

-- Requiring periodic repayments (e.g., mortgage loan, car loan)

-- Threatening delinquent borrowers with loss of credit rating

- Managers address their principal-agent (moral hazard) problems by:

-- Direct monitoring

Page 35: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 35/61

 

Firms & Markets Class Notes Nov 5 Page 8

-- Incentive payments

--- Sales commissions

--- Piecework compensation

--- Bonuses

--- But incentive payments require monitoring and a belief by the relevant partiesthat the system is fair, transparent, etc.

-- Pay systems (e.g., a mixture of salary and commissions) that induce self-selection amongemployees into positions that are commensurate with the employees' abilities and proclivitiestoward risk-taking, etc.

-- Creating a company culture that discourages slacking and encourages on-the-job effort

- Overall, these mechanisms for addressing agent-principal/moral hazard problems are not perfect; but they're used because the various parties recognize that the agent-principal/moral hazard problems are present

- These asymmetric information problems generally can also be seen to be important as causes whylarge companies are harder to manage (diseconomies of scale, diseconomies of vertical integration) but also why some companies may chose to undertake some vertically related processes on an in-house basis rather than to outsource them

Page 36: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 36/61

 

Firms & Markets Class Notes Nov 5 Page 9

Mini-Case: AN EXAMPLE OF THE ADVERSE SELECTION PROBLEM IN INSURANCE

Assumptions:- Competitive insurance industry that sells “collision” automobile insurance -- i.e., insurance

that protects the driver’s own car in the event of a crash

- The cost of an auto crash is $1,000/crash; this includes repairs, administrative costs, anda normal profit for insurers- There are two types of drivers: “low-risk” drivers, who experience a crash (on average)

once every five years; and “high-risk” drivers who experience a crash (on average) once each year- Each driver knows whether he/she is a low-risk or a high-risk driver- Buying insurance is voluntary; drivers are generally risk-averse, and will buy insurance

that is “reasonably priced”: i.e., the annual premium (i.e., the price of insurance) is not too far awayfrom what the driver knows to be his/her expected annual costs from crashes; but if insurance is“too expensive”, drivers will forgo insurance and “self-insure”

- There are 5 million low-risk drivers and 5 million high-risk drivers in the general population

Outcomes:A. Complete information:

- Suppose that the insurance industry has complete information as to who is a low-risk andwho is a high-risk driver

- What will the (approximate) price of insurance be? Why?

B. Imperfect information:- Suppose that the insurance industry doesn’t know that there are different types of drivers;

the only thing that the insurance industry can observe is the annual rate of crashes among the

overall population of drivers (or the insurance industry may know that there are different types ofdrivers but can’t identify who they are; again, all that the industry can observe is the annual rate ofcrashes among the overall population of drivers)

- What will the (approximate) price of insurance be? Why?

- Is this last outcome stable? If not, what happens next? Why? What is the end result?How does this result compare with the complete information result?

Page 37: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 37/61

 

Asymmetric information extensions

Implications of asymmetric information for the structure of lending/financial markets:

- The informational-based problems and attempted remedies by banks and other lenders have an

important influence on the overall structure of financial markets: who gets finance, from whom, andunder what kinds of terms

-- Suppose that borrowers can be arrayed along a spectrum that represents the extent of their

informational opaqueness or transparency; see THE SPECTRUM OF INFORMATIONALOPAQUENESS/TRANSPARENCY

--- At one extreme, the most informationally opaque borrowers are likely to obtaintheir finance from friends, family, informal sources (who are likely to have special information orspecial means of obtaining repayment or may be willing to forgo repayment)

--- At the other extreme, the most informationally transparent borrowers are likely to

obtain their finance from the securities markets, where the lenders and investors have only limitedcapabilities of gathering information but can rely on the transparency of the borrower (and rely on

rating agencies and fixed income analysts and public accountants/auditors and on advisors tostructure covenants and on lawyers to enforce them)

--- Borrowers who are “in between” may be able to obtain finance from information-

specialist lenders, such as banks, insurance companies (as lenders), finance companies; the lendersare specialists in gathering information about potential borrowers and monitoring borrowers; also,

industrial suppliers (short-term “trade credit” or longer-term equipment financing) may offerfinance based on their observations as to the credit-worthiness of the customer

--- Specialist lenders will tend to offer loans that are short-term but renewable, attach

restrictions, insist on collateral, etc.

--- The boundaries between the categories of borrowers and where they obtain theirfinance are not rigid or fixed-for-all-time but are determined by the things that allow lenders to

gather information and to assure themselves of repayment, such as the technologies of datagathering and processing and of telecommunications, the standards of accounting, legal

arrangements such as bankruptcy codes

--- As technologies have improved, the boundaries have tended to move, such thatthe securities markets now encompass more types of lending (e.g., “junk bonds”, residential

mortgages, other asset-based lending or regular cash-flow-based lending) that previously werelargely done by specialist lenders (e.g., banks); but also the banks have been able to penetrate more

Firms & Markets (Fall 2013)

LJW Class Notes Nov 7

Page 38: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 38/61

 

Firms & Markets Class Notes Nov 7 Page 2

deeply into previously opaque areas, such as offering credit card loans to large numbers of

individuals and to small businesses

-- Two important determinants of the informational opaqueness or transparency or a borrower (if the borrower is an enterprise or a government) are the size and the age of the borrower;

see TWO DETERMINANTS OF INFORMATIONAL OPAQUENESS/TRANSPARENCY

--- Borrowers that are young have little or no “track record” and are likely to have torely on friends and family and other informal finance; as they grow older they can rely more on

information-specialist lenders, venture capital providers, etc.; as they develop even more of a trackrecord, they can rely on securities markets

--- Borrowers that are small are likely to have to rely on friends and family and on

informal finance; as they grow larger (and the size of the financing that they want grows larger),they can access the specialist lenders (since the size of the loan will justify the fixed-cost-

investments in gathering information and in monitoring), venture capital providers, etc.; with even

larger size they can access the securities markets

--- Again, the boundaries are not rigid but are determined by technologies, etc.

Page 39: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 39/61

 

Firms & Markets Class Notes Nov 7 Page 3

THE SPECTRUM OF INFORMATIONAL OPAQUENESS/TRANSPARENCY

Borrower: Opaque Transparent

Source of finance: Personal finance, Banks and other Securities marketsfriends, family financial intermediaries,

(credit cards) trade credit

Page 40: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 40/61

 

Firms & Markets Class Notes Nov 7 Page 4

TWO DETERMINANTS OF INFORMATIONAL OPAQUENESS/TRANSPARENCY

AGE

Securities markets

Banks and otherfinancial intermediaries,

venture capital,

trade credit

Personal finance,

friends, family(credit cards)

SIZE

Page 41: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 41/61

 

Auctions

Auctions

- Auctions represent a selling mechanism that may be useful when sellers are uncertain about thewillingness to pay on the buyers' side of the market (also auctions may be used by buyers who take bids to be supplied, but are unsure about the willingness to supply); the items are often limited innumber and the sales are time sensitive; see AUCTIONS

-- Generally, four types of auction mechanisms

--- Ascending-price, open-outcry

--- Sealed-bid, highest bidder at the highest price

--- Descending price, open-outcry

--- Sealed-bid, highest bidder at the second-highest-price

- Goal of the seller is to sell at the highest price; goal of the buyer is to buy at the lowest price (whilewinning the auction), and avoid the “winner's curse”; seller should “look ahead and reason back” inconsidering what type of auction mechanism to use

-- The general rule for the buyer is to determine the value of the item to himself and to bewilling to pay any price up to that value, but drop out after that value has been reached; but sometypes of auction structures may encourage more complicated strategizing

- Advantages/disadvantages of each type of auction

-- Open outcry, ascending price auction generally extracts most of the buyers’ willingness to pay (except for any remaining willingness-to-pay by the winning bidder, since the winning bidder pays a price that is only slightly higher than the runner-up bid); open-outcry does not encourageimplicit strategizing; but collusion among buyers may be a problem when there are a small numberof bidders

-- Sealed bid (highest bidder at the highest price) and descending price auctions encourageimplicit strategizing, since winning bidders will always worry as to whether they have paid morethan they needed to in order to win the auction

--- The bidder has to trade off the likelihood of winning against the possibility of paying more than is necessary; this can lead to strategies for bidding less than one's own maximum personal value for the item

Firms & Markets (Fall 2013)

LJW Class Notes Nov 12

Page 42: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 42/61

 

Firms & Markets Class Notes Nov 12 Page 2

--- Sealed bid makes collusion more difficult, but it can (and has) occurred,especially in situations where sealed bid auctions occur on a regular basis and the same group of bidders are generally present

-- Sealed bid (highest bidder at the second highest price) eliminates any incentives forstrategizing, encourages aggressive bidding; the outcome is similar to the open outcry outcome

--- This is equivalent, in the context of an open-outcry auction, to telling the auctionhouse (or E-bay) your maximum price that you are willing to pay and then letting them do your bidding for you

--- This mechanism and its advantages generalizes to multiple item auctions wherethe winning bidders will each take a fraction of the total number of items that is being auctioned(e.g., Treasury bills)

- All of these auction mechanisms could be used by buyers who are seeking to buy items at thelowest price (or are seeking a package of desirable attributes and a low price)

- Examples of multiple goods sealed-bid auctions: Treasury bills; buying electricity; seeUNIFORM-PRICE VERSUS DISCRIMINATORY SEALED-BID AUCTIONS FORMULTIPLE ITEMS

-- For Treasury bills (and other debt auctions), the Treasury arrays the bids of buyers,starting with the highest price (and whatever quantity the bidder wants to buy at that price)

--- The aggregated quantity by bidders that equals the amount to be auctioned

determines the “cutoff” or “marginal” bid

--- Until the mid 1990s the Treasury required each bidder to pay the price that he/she bid (“discriminatory price” auction)

--- Since then the Treasury has required that all bidders pay the price of the cutoff ormarginal bidder (“uniform price” auction)

--- The reason for the change: The Treasury was concerned that the discriminatory price auction was causing bidders to bid less aggressively because the winners were worried thatthey were paying “too much” (more than the marginal bid); the uniform price auction wouldencourage them to bid more aggressively (bid their maximum value), since, if they won, they wouldonly pay the marginal bid; Treasury expected that this more aggressive bidding would improve itsoverall revenues from the auctions

-- California has been buying electricity according to the same general principles and for thesame general reasons

Page 43: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 43/61

 

Firms & Markets Class Notes Nov 12 Page 3

--- But the number of suppliers has been few enough and the bid quantities by someof them has been large enough so that strategic withholding of quantities (so as to cause themarginal bid price to be higher) has been strongly suspected by some researchers (and by California political figures)

- Winner’s curse in auctions

-- The actual value of an item may be uncertain, but the bidders may form estimates of thatvalue, and knowledge of each others' estimates may affect their own estimates; see Mini-Case:BIDDERS’ INFORMATION AND THE “WINNER’S CURSE” IN AUCTIONS

--- This is often described as a situation of “affiliated” (or correlated) values; if thetrue (but unknown) value will be the same for any winner, then this is often described as a“common value” situation

--- Bids for oil-drilling rights on a plot of land are an example

-- In such situations, the winner of the bidding may be just the one who is the mostoptimistic; but such optimism may (on average) mean that the bidder has bid more than the truevalue of the item

--- Oil-drilling auctions, and auctions for the rights to use the electromagneticspectrum for cellular telephone service are examples where “winner's curse” problems are potentially or actually present

-- The general “solution” to the winner's curse problem is to be aware of it and to shadedownward one’s own estimate of the value of the item for bidding purposes

--- Knowledge of one’s own risk aversion tendencies, of the distribution (variance)of estimates around the true (but unknown) value, and of the bidding strategies of the other bidderswould help in determining the extent to which one’s own maximum bid should be shadeddownward

--- In a sealed bid auction, no further information is available

--- In an open-outcry auction the pattern of bidding by the other bidders may provideadditional information to help one’s own bidding strategy

Page 44: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 44/61

 

Firms & Markets Class Notes Nov 12 Page 4

AUCTIONS

Auctions are a device that sellers sometimes use to sell items, especially when the items arelimited in number, the sales are time sensitive, and the sellers are unsure (have incompleteinformation) about the state of market demand (willingness to pay). [Auctions are also sometimes

held by buyers, to solicit bids by suppliers, for similar reasons.]

There are, as a general description, four types of auctions:

1. Ascending-price, open-outcry. This is the familiar auction system that is used forantiques, paintings, etc., at fancy auction houses. The electronic version has become popular on theInternet. It is sometimes called the “English auction” system. The highest bidder (sole survivor)wins the auction, paying a price that is slightly above the second-highest bidder's price.

2. Sealed-bid, highest bidder at the highest-price. This also is a familiar auction system. Itis often used by governments. The highest bidder wins the auction, at the price that she bid.

3. Descending-price, open-outcry. This is a less well known system. It is used in the Dutchflower market. The price starts at an unrealistically high level and then descends gradually, until thefirst (highest) bidder enters a bid, winning the auction at that price.

4. Sealed-bid, highest bidder at the second-highest-price. This is a less familiar system (to be discussed below).

The goal of the seller is generally to receive the highest price for the item; the goal of the buyer generally is to pay the lowest possible price (while still winning the auction) and to avoid paying a price that is above the “true” value of the item. In considering auction strategies, the seller

should take into account the likely strategies of buyers.

The general rule for a buyer is a follows: Determine the maximum value (to yourself) of theitem. Be willing to pay any price up to that value. Drop out of the auction once that price has beenreached. (Also, be aware of the potential problem of “winner's curse” when there is generaluncertainty among bidders as to the item's true value.)

The ascending-price auction will generally extract most of the potential buyers’ willingnessto pay for an item. But the winning bidder is likely to remain with some unextracted willingness to pay. In essence, the winning bidder gets the item at a price that is only slightly higher than thesecond-highest bidder's maximum price. There is no implicit strategizing that will allow bidders (solong as they do not collude) to improve their outcomes.

By contrast, the sealed-bid highest price auction (and the descending price-auction)introduces the possibility of buyers' strategizing, especially if only a small number of bidders are present. In essence, in submitting a bid, a bidder must wonder if he/she will end up paying morethan he/she “needs to” -- i.e., more than just a small amount above the runner-up bid. He/she maythen be tempted to reduce the bid somewhat -- at the risk, of course, of losing the auction. There is

Page 45: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 45/61

 

Firms & Markets Class Notes Nov 12 Page 5

thus a tradeoff between bidding below the maximum value (and thereby saving some money) andlosing the auction. The following example shows the nature of the trade-off:

Suppose that there are two bidders in a sealed-bid auction for a single item. Let themaximum value that bidder A places on the item for herself be standardized to 1.0. Suppose also

that Bidder A is uncertain about what the maximum value of the item is to Bidder B, but thinks thatit could be anywhere between 0.0 and 1.0, with a uniform (flat) distribution.

Bidder A has to decide on the level of her sealed bid. She has to trade off the possibility ofgaining more surplus by bidding lower against the increased likelihood of not winning when she bids lower. Suppose that she bids a price of P, where 0.0 <_ P <_ 1.0. Her net gain (surplus), if shewins the auction, is 1 - P. Her probability of winning the auction is P (if her conjecture aboutBidder B is correct). Her expected gain (EG) from bidding P is therefore EG = P*(1 - P) = P - P2.To maximize her expected gain, take the first derivative, and set it equal to 0:

d(EG)/dP = 1 – 2P = 0; or P = 0.5.

The maximizing strategy for Bidder A is therefore to bid 0.5, even though her maximumvalue is 1.0; i.e., the best strategy is for Bidder A to shade her bid downward from her maximumvalue.

If Bidder B is actually similar to A (i.e., Bidder B values the item at 1.0 for himself, and hasthe same kind of uncertainty about A), then Bidder B's best bid is also 0.5. Therefore, the sealed-bidauction will yield only (about) 0.5 for the seller, even though both bidders actually place maximumvalues of 1.0 on the item. If the seller had used an ascending-price open-outcry auction, the winning price would have been very close to 1.0.

Here is where the paradoxical strength of auction system #4 can be seen. If the highest(winning) bidder in a sealed bid auction knows that he/she will only have to pay the runner-up bidder's price, then every bidder has the motivation to bid his/her maximum value for the item. (It isa dominant strategy.) It never pays to bid less (because that could mean that you might needlesslylose the item); it never pays to bid more (because that could mean that you might have to pay morethan the item is worth to you). In essence, auction system #4 recreates something very close to theoutcome of the ascending open-outcry system: The winning bidder wins the auction at a price thatis equal to the second-highest bidder’s maximum value. Equivalently, it encourages bidders to bidaggressively rather than strategically. If auction system #4 were applied to the problem describedabove, each bidder would be motivated to bid 1.0 (her dominant strategy).

In an open-outcry auction (including on-line auctions), a bidder who cannot attend theauction can leave a maximum bid with the auctioneer, with the assurance that if the bidder’s offer isthe highest bid, he/she will only pay a specified increment above the runner-up bid. Again, thisencourages the absentee bidder to bid his/her “full value”.

The contrast between the traditional sealed-bid auction and system #4 can also be seen whena large quantity of something is being auctioned, such that multiple bidders will share pieces ofwhatever is being auctioned (e.g., a large quantity of U.S. Treasury bills). The bids of each bidder

Page 46: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 46/61

 

Firms & Markets Class Notes Nov 12 Page 6

will state a quantity and a price. The traditional method would be to award quantities to the highest bidders, at the prices that they bid (a discriminatory or differential price auction). Thus, each bidderhas to worry, “Am I bidding higher than I need to?” By contrast, the generalization of system #4calls for awarding quantities to the highest bidders at the uniform price that is bid by the "marginal" bidder (i.e., the lowest bidder who still gets some of the quantity) (this is a nondiscriminatory or

uniform price auction). Now each bidder has the proper incentive to bid his/her maximum value forwhatever quantity is desired. Again, it encourages bidders to bid aggressively rather thanstrategically.

The uniform price auction system is used by the U.S. Treasury for all of its securitiesauctions, by W.R. Hambrecht for selling IPOs, and by the State of California for buyingelectricity; this method (it was called a “Dutch auction”) was used by Google to sell its IPO.

The formula for the optimal bid in a sealed-bid first-price auction can be generalized tomore than 2 bidders: With the same assumptions (i.e., Bidder A places a value of 1.0 on theitem, and she is uncertain about the value that others place on the item but believes that the range

is between 0 and 1, with a uniform distribution, for all the other bidders) and with n bidders(including A), then the optimal bid for Bidder A is P = (n – 1)/n = 1 – 1/n. Thus, the greater thenumber of bidders, the closer Bidder A’s bid should be to her personal maximum value (and the better will be the price received by the seller in the auction).

More generally, suppose that Bidder A’s maximum value of an item (for herself) is “V”and she faces one other bidder who (Bidder A believes) may have a range of possible values between H (the highest possibility) and L (the lowest possibility), with a uniform distribution. IfBidder A submits a bid Price of P, her gain if she wins is V – P. Her probability of winning is (P – L)/(H – L). Consequently, her expected gain (EG) is (V – P)*(P – L)/(H – L). To maximizeher expected gain:

d(EG)/dP = (V + L – 2P)/(H – L) = 0; or P = (V + L)/2.

Again, with n bidders, this generalizes to P = V – (V – L)/n.

Page 47: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 47/61

 

Firms & Markets Class Notes Nov 12 Page 7

UNIFORM-PRICE VERSUS DISCRIMINATORY SEALED-BID AUCTIONS FORMULTIPLE ITEMS

A. Seller Auctions:Treasury Bills:

$ A B Bidders (demand) Quantity to be sold atauction (supply)C D

E FPup

G

0 Qa  Q

B. Buyer Auctions: Electricity:

$

Quantity to be bought at auction(demand)

FBidders (supply) E

Pup

D

CB

A

0 Qa  Q

Page 48: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 48/61

 

Firms & Markets Class Notes Nov 12 Page 8

Mini Case: BIDDERS’ INFORMATION AND THE “WINNER’S CURSE” IN AUCTIONS

Bidders may have a range of information about an item that is to be auctioned. At oneextreme, bidders may have “perfect information” about the value of an item (e.g., if a $5 bill were being auctioned).

 Next, bidders may know their own valuation of an item, but not know (or only have guessesabout other bidders' valuations; but even knowledge of the others' valuations would not affect one'sown valuation. This would be described as “independent private values”.

Finally, suppose that the multiple bidders for a particular item that is up for auction areuncertain about the true value of that item. Some of the bidders may have better information thando some of the other bidders, but none can be completely certain about the item’s value. This isoften described as an “affiliated (or correlated) value” bidding situation. If the true (but unknown)value will be the same for any winner, then this is often described as a “common-value” biddingsituation.

An example of a common-value bidding situation would be as follows: the auction ofdrilling rights for underground oil on a plot of land, where all bidders have the same drilling andextraction capabilities and the value of whatever oil is present will be the same for any winner.

The danger for bidders in an affiliate+ (correlated) or common-value bidding situation isthat he/she ends up paying too much for the item. This is not the result of “excessive enthusiasm” inthe bidding process. Instead, it is the result of using an estimate of the item's value as the basis for bidding, where the estimate has inherent uncertainty (e.g., has “an error term” attached to it).

As a general example, suppose that all bidders for a plot of land have made geological tests

of the likely amount of oil that is present. But the estimates are just that. If the estimates are“unbiased”, but with “error”, then the winner of the auction will be the bidder whose estimate hadthe largest “positive error”. The winner will (on average) have paid too much.

The general strategy for bidding in such situations is to be aware of the “bidder's curse” problem and shade downward one’s own estimate of the value of the item. In a sealed bid (or Dutchflower) auction that is all one can do, since each bidder can’t learn about the other bidders’valuations. In an open-outcry auction, however, the process of bidding allows each bidder to learnmore about the other bidders’ valuation. If most of the other bidders remain in the bidding as the price approaches your estimate, this gives you valuable information about their valuations, and youshould revise your valuation upward. If most of the other bidders have dropped out, then youshould revise your estimate downward.

Consider the following specific (albeit artificial example): You, along with 4 other people,are bidding (at a standard sealed bid auction: highest bid wins at the highest price) for a suitcase thatis filled with money. You aren’t told how much is in the suitcase, but you and the other bidders areeach separately (privately) given an envelope with a piece of paper that lists an estimate of howmuch money is in the suitcase.

Page 49: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 49/61

 

Firms & Markets Class Notes Nov 12 Page 9

Further, you are told that the five estimates (one each to the five bidders) have the followingdistribution, where X is the true amount in the suitcase: X-$2000, X-$1,000, X, X+$1,000, andX+$2,000. But, of course, you have not been told who received which estimate, and you have not been told what X is.

When you open your envelope, the amount indicated on your piece of paper is $10,000.But, again, you don’t know which of the five estimates your $10,000 figure represents. You are not permitted to communicate with the other bidders.

In a standard sealed bid auction, how much should you bid?

Page 50: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 50/61

 

Externalities and public goods

Externalities: definition, examples, analysis:

- For “normal” goods, the costs and benefits of an activity (production or consumption) areinternalized; there are no "spillover" effects on others

- But some activities do have spillover effects or “externalities”: direct effects on others that do notoccur through a market transaction; externalities can be positive or negative; they can occur as aconsequence of production or consumption activity; see EXTERNALITIES

-- Negative externalities: costs that are imposed directly on others, outside of a market

context, as an incidental consequence of a production or consumption activity: e.g., industrial pollution; annoying the neighbors with loud music in your apartment; congestion; not painting theoutside of your house; a stockbroker cheating his/her customers (leading to a general publicimpression that all stockbrokers cheat); one branch of a franchise chain giving bad service

--- The social costs of the activity are higher than the private costs; in the absence ofany corrective mechanism, a profit-seeking firm or a gain-seeking individual ignores the spillovercosts; as a consequence, too much output (or consumption) occurs, with too little effort to correctthe negative spillover effects; this is an inefficiency (and there are distributional effects, since theactivity does impose uncompensated costs on others)

-- Positive externalities: benefits that are conferred directly on others, outside of a marketcontext, as an incidental consequence of a production or consumption activity; e.g., the downstreamflood control benefits that occur as a consequence of a forestry company’s tree planting andgrowing activity; planting flowers in the front garden of a home; painting the outside of your house;education; firm-specific advertising that also has the effect of increasing the demand for a widergoods or services category; exceptionally good service at one branch of a franchise chain

--- The social benefits of the activity are higher than the private benefits; in theabsence of any corrective mechanism, the firm or the individual ignores the spillover benefits; as aconsequence, too little output (or consumption) occurs; this is an inefficiency

-- The fundamental problem (root cause) is the absence of a property right (or the absence ofa clearly defined property right, or the difficulties of enforcing a property right, in the attribute thatis affected negatively or positively

--- If there is no property right in clean air, or it can’t be enforced effectively, thenthe firm will simply seek the lowest cost method of production and ignore the negativeconsequences of its actions; the same is true for individuals, who find it easier to throw trash on theground (i.e., littering) rather than to find a trash bin

Firms & Markets (Fall 2013)

LJW Class Notes Nov 14

Page 51: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 51/61

 

Firms & Markets Class Notes Nov 14 Page 2

--- If there is no property right in flood control, or it can’t be enforced effectively,then the firm will simply focus on selling its primary output and ignore the other beneficialconsequences of its actions; again, the same is true for individuals and their behavior in ignoring (ordownplaying) the beneficial aspects of their behavior for others

Externalities: resolution:

- Resolution of externalities may occur through private-sector actions or through public-sectoractions

-- Private-sector actions: where one firm is conveying an externality on another, a merger between the two firms will “internalize” the externality (positive or negative); the unified ownershipof the two activities will take it properly into account, reduce or eliminate the inefficiency

-- If a merger is not feasible or if the numbers of the parties are larger than two but still

manageable, then negotiated agreements among them may suffice

--- But negotiations bring into the picture bargaining problems, information problems, free-riding problems

--- Nevertheless, agreements do occur: e.g., joint advertising arrangements;cooperative arrangements in general among enterprises

--- Also, within a franchise chain, lots of joint advertising, and efforts to build acommon culture, enthusiasm, etc.

-- Where the numbers of parties are large and the externalities are substantial, the publicsector often plays a role

-- Positive externalities (e.g., education): The public sector provides free (to the user) primary and secondary education and plays a substantial role in undergraduate education

-- Negative externalities (e.g., pollution): Since 1970 the federal government (as well as thestates) have played a substantial role in requiring firms and industries to reduce their pollutingactivities

--- The typical method is detailed “command and control” regulation; this is oftenquite inefficient; among the problems, the regulators lack the detailed information to make thedetailed decisions as to pollution reduction (this is another problem of asymmetric information)

--- The efficient outcome would be achieved at levels of emissions where the MC ofthe emissions (e.g., health consequences) are equal to (the same as) the MC of abatement of theemissions; this yields a “price” (the opportunity cost of greater emissions or of tighter control) and aquantity

Page 52: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 52/61

 

Firms & Markets Class Notes Nov 14 Page 3

--- In principle, this outcome could be achieved through “effluent fees” (which havenot been widely used in the U.S.), or through the specification of maximum emissions quotas; thelatter can be made tradable “cap and trade”, and a market can be established; the market willestablish a price for the permits, and trading will induce efficient emissions control andencourage/induce technological improvements in emissions control

--- Tradable permits and consequent markets have been established in the U.S. since1990 in sulfur oxide emissions and more recently in nitrous oxide emissions

--- For dealing with greenhouse gases: a “carbon tax”; or a “cap and trade” system

Public goods:

- Most goods and services are “rival”: More for one individual means less for someone else; seePUBLIC GOODS

- Most goods and services are also “exclusive”: Individuals can be excluded from using/consumingthem

- Some goods are non-rival but exclusive: a toll highway or bridge at an off-peak time; the subwayat an off-peak time; an empty but fenced-off park; a scrambled radio wave; patented or copyright- protected information

- Some goods are rival but non-exclusive: fishing in the ocean for a limited stock of fish; whaling;traveling on a congested public street or highway; an unfenced but congested park

- Some goods and services are non-rival and non-exclusive: one person’s use or consumption does

not interfere with or reduce others’ use or consumption; and individuals cannot be excluded fromusing or consuming them; these are “public goods”; e.g.:

-- A radio wave

-- A light house

-- The police catching a criminal

-- National defense

-- A mosquito eradication program

-- Using an empty public street or highway or unfenced park

-- Using information/knowledge

- A few things to note about public goods:

Page 53: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 53/61

 

Firms & Markets Class Notes Nov 14 Page 4

-- A public good is not the same as “goods provided by the public sector”

-- A public good is an activity where the externalities are pervasive (and not just incidental)

- If individuals cannot be excluded, then it will generally be difficult for the private sector to provide

them:

-- Individuals cannot be excluded from the general benefits of police activity, nationaldefense, mosquito eradication; these tend to be provided publicly

-- An exception: over-the-air radio and television is provided privately, because advertisingcan be attached to programs (also, private security forces exist alongside public police forces)

- With government provision of public goods come a set of problems

-- How to keep the provision of the goods efficient; how to create appropriate incentives

-- Deciding on how much to provide, and at what quality

--- For normal goods, these are private, individual decisions; for public goods, sinceonly a single quantity and quality will be provided, there must be some mechanism for reaching asingle decision

-- Deciding who shall pay and how much and how

--- For normal goods, these are private decisions, there is a direct link between paying and receiving; for public goods, the payments have to be aggregated

- The private sector is involved with public goods in at least three ways:

-- As a supplier to the governmental provision of public goods (e.g., defense contractor;automobile supplier; gun supplier; uniform supplier; etc.)

-- As the direct contractor of a public service that is funded and supervised by government(e.g., private operation of a few prisons and of a few “public” schools)

-- As a direct provider when exclusiveness is possible

--- Exclusiveness is partly a technological phenomenon (e.g., electronic scramblingcan convert a radio wave into an exclusive phenomenon) and partly a legal phenomenon (e.g., someforms of new knowledge have been made exclusive through the legal system of patent andcopyright; a highway or a bridge can be “gated” and financed by tolls)

Page 54: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 54/61

 

Firms & Markets Class Notes Nov 14 Page 5

EXTERNALITIES

A. Negative externality (e.g., pollution by a competitive steel company):

$ MC priv+MCsoc

MC priv

Pst

MCsoc

0 qsoc  q priv  qst

B. Positive externality (e.g., flood control by a competitive forestry company):

$

MC

Pfp +MBsoc

MBsoc 

Pfp

0 q priv qsoc  qfp

Page 55: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 55/61

 

Firms & Markets Class Notes Nov 14 Page 6

PUBLIC GOODS

Rival Non-rival

Most goods and services (apples, Toll bridge (highway, subway,sweaters, etc.) etc.) at off-peak times

Fenced park at off-peakExclusive Scrambled radio wave

Patented or copyright information

PUBLIC GOODS:Congested public highway Radio wave (unscrambled)Fishing from a limited pool Light house beaconWhaling Police catching a criminal

 Non-exclusive Unfenced congested park National defenseMosquito eradicationEmpty unfenced parkInformation (with no property

rights)Empty public highway

Page 56: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 56/61

 

Externalities and public goods applied: intellectual property; Networks

Information as a public good

- Once information has been created, it’s a public good: it’s non-rival (more for you doesn't mean

less for me); it’s non-exclusive (transmission is very low cost, and excludability is often difficult orimpossible)

-- In the absence of any protective mechanisms, smaller amounts of new information would

 be created than if new information can be protected; if incentives matter -- if earning a return on thatinvestment matters -- then the public good nature of information (absent some protection) will

discourage the creative effort

--- Some painters would still paint; some writers would still write; some inventorswould still invent; some software writers would still write; etc.; businesses could protect some

 proprietary knowledge through secrecy and through unavoidable lags in copying by others; or theycan get returns in linked ways (e.g., concerts related to recorded songs; follow-up servicing related

to software)

--- But for many individuals, creating information involves the investment of timeand resources; if other individuals can “free ride” (information is non-exclusive) and quickly

compete with the original creator of the information, then the original creators will not be able torecover their investment and will be reluctant to engage in the information creation in the first place

Intellectual property

- Most societies choose to provide a property rights framework (“intellectual property”) to provide

an excludability aspect to newly created information: patent, copyright, trademark

-- The owner can prevent unauthorized replication or copying

-- With excludability, creators can charge a price for the fruits of their creation, earn a return

-- Trademark is primarily intended to allow/encourage distinctiveness of companies and products, support branding (help companies emerge from the fog of asymmetric information);

trademarks have unlimited lives but must be renewed every 10 years

-- Patents and copyrights have limited lives (unlike normal property, which persists in perpetuity); the limited life is a recognition of the tradeoff between the necessity for creating

incentives to create (hence establishing the property right) and the public good aspect of information(we would like to have it in the public domain, without exclusivity restrictions, since information is

non-rival)

Firms & Markets (Fall 2013)

LJW Class Notes Nov 19

Page 57: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 57/61

 

Firms & Markets Class Notes Nov 19 Page 2

- Patents currently have lives of 20 years from the time of filing; copyrights have lives of owner’slife + 70 years, or (if owned by a corporation) 95 years

-- An important feature of the patent system is that the essential features of the technology

must be revealed to the public at the time of the granting of the patent

- The legal tradition is to treat patent and copyright as inherently “monopolies”; but a betterapproach is to think of them as the application of property rights and excludability to a phenomenon

(information) that would otherwise be a public good and that would have inadequate incentives fornew creative effort

-- Does it make sense to describe as “monopoly” each of the new patents (276,788 in 2012!)

or each of the new copyrights (636,400 in 2010) granted in the U.S. -- plus the millions from pastyears that are still valid?

-- A few patents do qualify as “monopolies”: e.g., unique pharmaceuticals, some businessmodel patents, etc.; but the overwhelming majority are just another piece of property, a slightlydifferent way of doing things; it’s estimated that around 90% of patents that are issued are never

used (are these meaningful monopolies?)

- The analogy with real estate property is a better of thinking about intellectual property

-- Each piece of property is distinct, with unique attributes that affect its value (better view, better location, better minerals underground, etc.)

-- There may be a few real estate monopolies -- the ownership of the only water spring in a

50 mile radius; the top of the Empire State Building -- but most real estate is just another piece of(somewhat differentiated) property

- The limited life of patent and copyright represents the social recognition of the tradeoff between

the need to provide incentives for creation and the social gains from allowing the information toenter the public domain sooner

- When new ideas (new creations) are built on existing ideas, then transactions costs may prevent

the new ideas from coming to fruition

-- In a “frictionless” world, this would not be a problem

-- In a world with bargaining frictions, it can be a problem

- Similarly, if a production process must use multiple inventions, bargaining frictions may get inthe way

Page 58: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 58/61

 

Firms & Markets Class Notes Nov 19 Page 3

-- The problem is similar to the real estate developer who needs to assemble (buy)

multiple parcels of land in order to undertake a construction project; individual “holdouts” canstymie the process

 Networks

- Definition: Nodes connected by links; see NETWORKS

-- Example: a simple “star”: airline hub and spokes; local telephone exchange; local package

delivery; local electricity distribution; local radio broadcasting

--- Networks can also be configured as a ring or as all-points-directly-connected

-- Example: two stars with a trunk connection: long distance/local telephone; railroad;water/gas/oil pipeline; electricity distribution; Internet; cable TV; airlines; banks and ATMs; credit

card system (banks and merchants)

- Distinction between 2-way networks and 1-way networks

-- In 2-way networks, the transactions readily flow back and forth among nodes: a telephoneuser can be a caller or a receiver; a rail user can be a shipper or a receiver; an air traveler can go to

or from any point on the system; e-mail users can send or receive

-- In 1-way networks, the transactions naturally flow in only one direction: electricity flowsfrom generators to users; water flows from sources to users; in cable TV, the programs flow from

originators to viewers; on the web, information flows from web sites to users

- The importance of the distinction: In both kinds of networks there are often benefits from greatersize of network; but

-- In 2-way networks the positive network effects (network externalities) are direct: An

additional member of the network directly adds value (provides benefits) to other users (they cantransact with the new member)

--- There can also be benefits from economies of scale in the technology of the

network

--- There can also be diseconomies of scale: congestion problems

-- In 1-way networks the positive network effects (network externalities) occur indirectly:An additional member doesn't provide extra value directly to other members; but the members can

 benefit through the exploitation of economies of scale (lower unit costs) that occurs as moremembers join the network: lower electricity costs in a larger network; more locations of ATMs in a

larger networks; more issuing banks and more merchants in a larger credit card network (greatervariety of locations arise because of economies of scale

Page 59: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 59/61

 

Firms & Markets Class Notes Nov 19 Page 4

--- There can also be diseconomies of scale: congestion problems

- In networks compatibility is crucial: the nodes and links must be able to work effectively with

each other, “speak” to each other

-- The compatibility may be physical or technological; they may even be created by business practices (e.g., the airlines’ networks have become “incompatible” for travelers)

- Hardware-software systems constitute “virtual” networks

-- Compatibility is crucial

-- They can be portrayed as 1-way networks; the advantages of a larger “virtual network”

arise because of economies of scale (indirect externalities)

-- Computer operating systems and applications software constitute a virtual network

- How to achieve compatibility:

-- There may be a common agreement among industry members on an open standard, non- proprietary (e.g., railroad gauge; telephone standards)

-- There may be governmental forcing of compatibility; government-specified standard,

mandated interconnection

-- There may be competition to be the standard-setter, establish a proprietary standard:“standards wars”; tactics include

--- Building up a large installed base, so as to appear to be the “market consensus”;

this can be done through “anticipatory pricing” or “loss-leader pricing”; or through early productannouncements of promising future technological achievements (“vapor ware”?)

--- Also, network externalities provide an additional impetus for building up a large

initial installed base

-- If there are network externalities, and there are competing networks, then an initial lead(brought about through whatever means) may prove to be permanent: The largest network will have

the most economies of scale and/or the most variety in complementary components (e.g., locationsof merchants in a credit card network; extent of software in a computer system virtual network) and

thereby increase its attractiveness to customers; this is sometimes described as the phenomenon of“tipping”; it leads to “winner take most” market structures

--- Smaller networks (that are not compatible with the larger network) can continue

to survive only if they have some special characteristics that make them worthwhile to users despitetheir smaller size

Page 60: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 60/61

 

Firms & Markets Class Notes Nov 19 Page 5

--- And, also, leaders may stumble, make mistakes, and “smarter” entrants may be

able to overtake leaders.

-- With network externalities, market structures may show a considerable amount of inertia:With a large installed base and with compatibility essential, incremental change may be difficult (if

it involves incompatibilities)

--- The technology that survives may not necessarily be the best, if an early lead(perhaps due to chance phenomena) can dictate the eventual surviving technology; this is the notion

of “path dependence”: where we end up (technologically) depends on the path that was taken to getthere; and once we’re there, it’s difficult to change (because of the inertia of the network); e.g., the

U.S. now has a set of standards for our electrical system, or our rail system, or our telephonesystem, that are difficult to change

-- If incremental change is difficult, then entry may be difficult (for the same reasons)

-- The “tipping” phenomenon may also mean “stranded” technologies: technologies that areno longer supported, so that earlier investments are lost (e.g., Betamax hardware and tapes; early PCmodels and brands that no longer exist)

- Bottlenecks in a network (e.g., the local switch in a telephone system; the local marshalling yard in

a rail system; the local distribution facility in an electrical system; a proprietary technology) can become sources of market power, barriers to entry

Page 61: F&M Notes Compiled

8/13/2019 F&M Notes Compiled

http://slidepdf.com/reader/full/fm-notes-compiled 61/61