Transcript Woefully Imperfect Market Puzzle

Motivation and Objectives
 We have uncovered abundant evidence of woefully imperfect markets
Semiconductor Integrated Circuit (IC) market is one example
Used Textbook Market is another example
Conventional economic models were found inadequate in these markets
Our objectives are to bridge the gap between conventional models and
our model
A Familiar Conventional Model
 Hotelling’s Model is one of the earliest known models
It can be invoked to explain away trivial instances of market imperfection
Price of a soda can is known to be exorbitant at airports compared to
supermarkets
Customers are willing to pay for convenience and a cold can of soda!
Empirical Data
Semiconductor Market
Operational Amplifier (741)
Price /100 chips
Vendor A
$18
Vendor B
$22
Vendor C
$49
Vendor D
$95
Vendor C
$69
Vendor D
$30
Semiconductor Market
Transistor (TIP 31C)
Price /100
Vendor A
$29.30
Vendor B
$159
Empirical Data … continued!
Semiconductor Market
Memory (2114 RAM)
Price / chip
Vendor A
$13.75
Vendor B
$1.69
Vendor C
$1.29
Vendor D
$2.58
Vendor C
$55
Vendor D
$100
Textbook Market (used)
Electric Circuit Theory (Johnson)
Price /single copy
Vendor A
Vendor B
$10
$30
Empirical Data … characteristics
 These are undifferentiated products
 Same transportation costs
Synopsis of Market Models
Monopoly
Oligopoly Monopolistic Competition Competition
Maximize Profit
MR = MC
MR = MC
Price
price setter price setter
Market Power
p > MC
Entry
No Entry
MR = MC
p = MR = MC
price setter
price taker
p > MC
p > MC
Limited Entry
Free Entry
p = MC
Free Entry
Conventional Model Characteristics
 A monopoly does not care what the rival firm does … there are NO
RIVALS!
 A competitive firm does not care what the rivals do because it does not
matter!
An oligopolistic firm seriously considers how its actions affect its rivals
and how the actions of its rivals will affect it.
A monopolistically competitive firm seriously considers how its actions
affect its rivals and how the actions of its rivals will affect it.
These strategies (GAMES) lead to NASH EQUILIBRIA
Paradoxes of Imperfectly Competitive Markets
 Under smooth dynamics, outputs may follow chaotic trajectories
 Entry of a new firm in the market may actually decrease the total output
and increase the equilibrium price
Entry of a new firm in the market may actually increase the equilibrium
price
Entry of a new firm in the market may actually increase the profit of the
incumbent firm
A merger of two or more firms can decrease the profits of all merged
firms
The entry of a new firm in the market might decrease social welfare
Even if the entry of a firm would raise social welfare, this entry might not
be profitable
Justification for a New Model
 Hotelling’s model is not viable because there is no product
differentiation in our semiconductor and textbook markets
 Similarly, Chamberlin/Robinson monopolistic competition is not viable
because there is no product differentiation in our semiconductor and
textbook markets
Is Cournot’s model a viable candidate?
Cournot Model
In the Cournot Model of non-cooperative oligopoly, the
firms choose their output levels without colluding (no
cartels!) but they make conjectures about the reactions
of their rivals in response to their actions
Monopoly, Duopoly, and Oligopoly!
In order to set the stage for Cournot’s oligopoly, let us review the
structure of a monopoly
We posit a linear inverse demand function
p(q) = a – bq
The revenue is
R = pq
R = aq – bq2
Monopoly … continued!
The marginal revenue can be obtained as a partial derivative of R
with respect to the output q.
MR = R/q
MR = a – 2bq
In terms of elasticity ɛ
MR = p (1-1/ɛ)
Monopoly … pricing!
The cost curve C(q) = kq where k is a constant
The marginal cost MC = C/q = k
A monopoly sells where p = MC = MR so we have
k = a – 2bq
Hence the output and price for a monopoly are
qm = (a-k)/2b
pm = (a+k)/2
Cournot Oligopoly … pricing!
Without loss of generality, we posit a tractable linear demand curve
q = a- p
Total demand = q1 + q2 for two firms
In the Cournot model, each firm conjectures that the other firm will
act in a way to keep the quantity that it sells fixed. We will calculate
the reaction function of each firm to the quantity supplied by the
other.
Cournot Oligopoly … profit maximization!
/q1 [(a - q1 - q2 )( q1 ) - k q1] = 0
a - 2q1 - q2 - k = 0
So firm 1’s reaction function will be given as
q1* = (a - q2 - k) / 2
Following a symmetric procedure, firm 2’s reaction
function will be given as
q2* = (a – q1 - k) / 2
Cournot Equilibrium
A Cournot Equilibrium (C.E.), as in a Nash game (e.g.
prisoner’s dilemma) occurs when neither firm wants to
change and is content with its output and profit.
Imposing this criterion on q1* and q2* yields
C.E. = (a – k) / 3
q2
COURNOT REACTION FUNCTION
ak
Isoprofit curves
ak
q1
2
Straight line is REACTION FUNCTION for firm 1 reacting to firm 2
q2
COURNOT EQUILIBRIUM
ak
R12
ak
2
ak
C.E.
3
R21
ak
3
ak
2
ak
q1
R12 Reaction function of firm 1 reacting to firm 2
R21 Reaction function of firm 2 reacting to firm 1
Cournot Equilibrium … conclusion
At the Cournot equilibrium we have the following price / output equations:
q1 = q2 = (a – k) / 3
Q = 2 (a – k) / 3
p = (a + 2k) / 3
Conclusion: Cournot equilibrium price is only
marginally higher than the perfectly competitive price
and only marginally lower than the monopoly price
In general, for n firms in a Cournot oligopoly
qn = (a – k) / (n + 1)
Woefully Imperfect!
The plethora of market models cannot explain the
existence of glaring and woefully widespread price
differences that we have found in the semiconductor and
other markets.
This puzzling observation clearly challenges the
notions of efficient markets and rational and informed
buyers and sellers populating these markets.
Habit Persistence
We will attempt to explain this puzzle by invoking
behavioral and habit persistence hypotheses that appear
to override the efficient markets and the rational and
informed participant hypotheses
The equity premium puzzle, a term coined by Mehra
and Prescott (1985) is an analogous puzzle, and has been
reported previously for the stock market
Deep Habits
Ravn (2006) explores the concept of “Deep Habits”
which are the offshoots of Behavioral Science
Habit persistence is a preference specification that
yields a utility function that depends on the quasidifference of consumption
Utility Function for Habit Persistence
Ravn (2006) explores the concept of “Deep Habits”
which are the offshoots of Behavioral Science
Let us specify the utility function without habit
persistence as ∫[t]U(c[t]) where  is the subjective
discount factor, c is the consumption in a very small time
period t, U is the utility function and the integration is
performed from 0 to . Then the utility function with
habit persistence is hypothesized to be ∫[t]U(c[t] - 
c[t-1]) where the parameter  denotes the intensity of
habit persistence.
“Bounded Rationality” & Habit Persistence
It seems to us that that the resolution of our woefully
imperfect market puzzle lies in the domain of behavioral
science and habit persistence
 At the heart of the rational and efficient market
hypothesis is a fallacious assumption that market
participants will seek out the lowest price
Four Ways to Spend Money!
Milton Friedman made an astute observation about the four ways people
spend money
 1. You can spend your own money on yourself
2. You can spend your own money on somebody else. Then you
are more concerned about the cost and less concerned about the
content of the Birthday present!
3. You can spend somebody else’s money on yourself (Government
= 10%)
4. You can spend somebody else’s money on somebody
else(Government … not concerned about how much it is and what I
get = 40% of national income!!!)
Woefully Irresponsible Spending Habits
 The popularly held Weltanschauung that markets are generally
efficient and populated by rational individuals making informed
choices was found to be woefully inadequate
Individuals employed by state and federal institutions do not
exhibit a proclivity for seeking out the most competitive bids even
under the unlikely assumption of the non-existence of venal and
other ulterior motives
The size of this market is approximately 50% of the national
income!!!