Supplementary Lecture Notes

Advanced Microeconomic Theory
Lecture Notes
Sérgio O. Parreiras
Fall, 2016
Outline
Mathematical Toolbox
Decision Theory
Partial Equilibrium
Search
Intertemporal Consumption
General Equilibrium
Financial Markets
General Equilibrium with Production
Liquidity
How to use the toolbox:

1 Casually read it once so you can classify your
understanding of the topics in three categories: mastered,
familiar but not mastered, and never seen before.
2 Read it again but, skip the mastered topics.
3 In your second reading, make sure to have pen and paper
at hand and, Mathematica open and running.
4 Work the learning-by-doing exercises using pen and paper
and, verify—using Mathematica—if your answers are
correct.
5 When you have problems with Mathematica—as you will
for sure—refer to the crash tutorial and Mathematica’s
help documentation. As last resort, email me your .nb file.
Matrices
A matrix is just a convenient way of displaying information. A

n by m matrix A is composed of n × m entires. The entry Aij is
displayed in the ith row and jth column.
Example of a 3 by 3 matrix,
 
A11 A12 A13
 
A =  A21 A22 A23  .
A31 A32 A33
Matrices
A matrix is just a convenient way of displaying information. A

n by m matrix A is composed of n × m entires. The entry Aij is
displayed in the ith row and jth column.
Example of a 3 by 3 matrix,
 
A11 A12 A13
 
A =  A21 A22 A23  .
A31 A32 A33
Mathematical Toolbox: Matrix Multiplication
B : p rows q columns
 
 b11 b12 ... b1q 
 
 
 
 
 b21 b22 ... b2q 
 
 
 
2
b1
 .. .. .. 
×
 .. 
 . . . . 
1
+
2
a2
 
b2
 
×
 
2
bp1 bp2 ... bpq

a2
+
…+
b p2
×
p
a2
   
 a11 a12 ... a1p   c11 c12 ... c1q 
   
   
   
 a21 a22 ... a2p   c21 c22 ... c2q 
   
   
   
 .. .. ..   .. .. .. 
 ..   .. 
 . . . .   . . . . 
   
   
 an1 an2 ... anp   cn1 cn2 ... cnq 
A : n rows p columns C = A × B : n rows q columns

Matrix Multiplication
Examples
 
u(x1 )
 
 u(x2 ) 
A1×n = (p1 , p2 , . . . , pn ) and Bn×1 = 
 .. 

 . 
u(xn )
C1×1 = A × B = p1 u(x1 ) + p2 u(x2 ) + . . . + pn u(xn )

Examples
 
u(x1 )
 
 u(x2 ) 
A1×n = (p1 , p2 , . . . , pn ) and Bn×1 = 
 .. 

 . 
u(xn )
C1×1 = A × B = p1 u(x1 ) + p2 u(x2 ) + . . . + pn u(xn )

Mathematica
In Mathematica to enter the matrices,

 
( ) 1 2
1 0 3  
A= and B =  3 4 
5 4 7
7 0
we type:
A :={{1, 0, 3}, {5, 4, 7}}

B :={{1, 2}, {3, 4}, {7, 0}} shift+enter
To multiply the matrices, type: A.B shift+enter.

( )
1·1+0·3+3·7 1·2+0·4+3·0
A×B= .
5·1+4·3+7·7 5·2+4·4+7·0
Partial Derivatives
Often we wish to evaluate the marginal impact of ONE given

variable on some function of several variables.
∂ f(x, y + ∆, z) − f(x, y, z)
My f = f(x, y, z) = lim
∂y ∆→0 ∆
We refer to My as:
1 The marginal change f with respect to y.
2 The partial derivative of f wrt y.
3 The slope of f wrt y.
Partial Derivatives
How to Compute
∂ f(x, y + ∆, z) − f(x, y, z)
∂y ∆→0 ∆
To compute the partial derivative with respect a given
variable— y in the above example— we use the exact same
rules of derivation you learn in calculus with one variable.
What about the other variables? We treat all the other

variables that are not of interest (x and z in the example above)
as constants.
In Mathematica, we use the command D to compute partial

derivatives. For example, we use D[f[x, y, z], y] to compute My f.
See the crash tutorial for additional examples.
Partial Derivatives
How to Compute
∂ f(x, y + ∆, z) − f(x, y, z)
∂y ∆→0 ∆

as constants.

Partial Derivatives
How to Compute
∂ f(x, y + ∆, z) − f(x, y, z)
∂y ∆→0 ∆

as constants.

Partial Derivatives
How to Compute
∂ f(x, y + ∆, z) − f(x, y, z)
∂y ∆→0 ∆

as constants.

Partial Derivatives
Examples
∂ (√ √ ) 1
x+ y = √
∂x 2 x
∂ 2( )
x + 3xy + y2 = 2x + 3y
∂x
∂ y
(y log (x)) =
∂x x
Partial Derivatives
Learning-by-doing exercises
Compute the marginal utilities MUX and MUY for the following
utility functions:
1 u(x, y) = 14 x + 34 y
√ √
2 u(x, y) = 12 x + 12 y
1
3 u(x, y) = 3 ln(x) + 23 ln(y)
Mathemematical Tool Box,
The Chain Rule:
If h(x) = f(g(x)) then
h′(x) = f′(g(x)) · g′(x)

Chain Rule
1 For each of the composite functions below tell us, What are
the corresponding f and g and compute h′ .
√
a) h(x) = 2x
b) h(x) = − exp(−ρ · x)
1
c) h(x) = (4 + xσ ) σ
2 Use the Chain Rule to obtain the marginal utilities MX and
MY of the utility function,
2 1
u(x, y) = − exp(−x) − exp(−y).
3 3
3 If k(x) = f(g(h(x))) is a composition of three functions,
apply the chain rule twice to compute k′ (x).
4 Consider f(x, y) and g(x) compute the total derivative of
f(x, g(x)) with respect to x using the Chain Rule and partial
derivatives.
Taylor’s Approximation
Consider a function of one variable defined on the real line,

f : R → R. If f is differentiable, we write the first order Taylor
approximation:
f(x + h) − f(x) ≃ f′ (x) · h
The approximation works well only if |h| is ”small”.
For a function of two variables and h = (h1 , h2 ) we have a

similar expression:
∂ ∂
f(x + h1 , y + h2 ) − f(x, y) ≃ f(x, y) · h1 + f(x, y) · h2
∂x ∂y

approximation:
f(x + h) − f(x) ≃ f′ (x) · h

similar expression:
∂ ∂
∂x ∂y

approximation:
f(x + h) − f(x) ≃ f′ (x) · h

similar expression:
∂ ∂
∂x ∂y
Mathematical Toolbox,
Marginal Utility & Taylor’s Approximation:
∆U = U(x + ∆x, y + ∆y) − U(x, y)
≃ MUx · ∆x + MUy · ∆y
Using Taylor’s approximation, for each of the utility functions

below, compute the change in utility when the consumer moves
from consuming the basket (100, 100) to consuming the basket
(105, 99).
1 u(x, y) = 14 x + 34 y
√ √
2 u(x, y) = 12 x + 12 y
1
3 u(x, y) = 3 ln(x) + 23 ln(y)
4 u(x, y) = − 23 exp(−x) − 13 exp(−y)
Hint: What is the value of ∆X? What is the value of ∆Y?
What is the value of MUX and MUY when the basket (100, 100)
is consumed?
Implicit Functions
Let f(x, y) be a real-valued function of two variables and let g(x)

be a real-valued function of one-variable with the following
property:
If we set y = g(x), f remains constant as we change x.
that is, f(x, g(x)) = c for all x, where c is a constant.
We refer to the function g as an implicit function since g is

implicitly defined by the equation f(x, g(x)) = c.
Implicit Functions

property:

Implicit Functions

property:

Implicit Functions
For the utility curves below, find the equation of the

indifference curve that gives utility c:
1 u(x, y) = 14 x + 34 y
√ √
2 u(x, y) = 12 x + 12 y
1
3 u(x, y) = 3 ln(x) + 23 ln(y)
4 u(x, y) = − 23 exp(−x) − 13 exp(−y)
Hint: set u(x, y) = c and solve for y, this is your g function...
Implicit Function Theorem
If f(x, g(x)) = c for all x, where c is a constant.

∂
f(x, g(x))
′ ∂x
Then, g (x) = −
∂
f(x, g(x)).
∂y
Mathematical Tools
The Implicit Function Theorem
Proof: Taking the total derivative of f with respect to x,
d ∂ ∂ ∂ ∂
f(x, g(x)) = f(x, g(x)) · x + f(x, g(x)) · g(x)
dx ∂x ∂x ∂y ∂x
∂ ∂
= f(x, g(x)) + f(x, g(x)) · g′ (x)
∂x ∂y
∂
d f(x, g(x))
Since , f(x, g(x)) = 0 ⇒ g′ (x) = − ∂x .
dx ∂
f(x, g(x))
∂y
As f is constant along g(x), we also call g an iso-curve of f.
Indifference curves and iso-cost curves are examples of

iso-curves that you should be familiar.
The Implicit Function Theorem
For the utility curves below: a) find the marginal rate of

substitution; b) in the MRS, replace y by the implicit function g
you found in the previous learning-by-doing exercise and
simplify the expression; c) compute g′ (x) for the implicit
functions of the previous exercise; d) compare the results you
found in items (b) and (c).
1 u(x, y) = 14 x + 34 y
√ √
2 u(x, y) = 12 x + 12 y
1
3 u(x, y) = 3 ln(x) + 23 ln(y)
4 u(x, y) = − 23 exp(−x) − 13 exp(−y)
Interior Solutions
Maximizing a function of one variable defined on the real line,

f : R → R.
Maximization Problem max f(x) (P)

x∈R
First order condition f′ (x) = 0 (FOC)

′′
Second order condition f (x) ≤ 0 (SOC)
Any point x satisfying FOC and SOC is a candidate for an

interior solution.
Interior and Corner Solutions
Maximizing a function of one variable defined on an interval,

f : [a, b] → R. As before,
Maximization Problem max f(x) (P)

b≥x≥a
First order condition f′ (x) = 0 (FOC)

Second order condition f′′ (x) ≤ 0 (SOC)
Any point x satisfying FOC and SOC is a candidate for an

interior solution and now,
x = a is a candidate for a corner solution if f′ (a) ≤ 0.
x = b is a candidate for a corner solution if f′ (b) ≥ 0.
Concavity and Convexity
Consider any function f : Rk → R.
Definition: f is concave if and only if, for all α ∈ [0, 1], and any
two points x, y ∈ Rk , we have
f (α x + (1 − α) y) ≥ α f(x) + (1 − α) f(y).
Another definition: We say that f is convex if −f is concave.

Mathematical ToolBox
Global Maxima
Proposition. Assume f is concave and also assume that x

satisfy the FOC then x is a solution to the maximization
problem (i.e. x is a global maximum).
Choice Under Uncertainty
States of the World
Definition: State of the World

A state of the world ω is a complete description of reality.
The set of all possible sets of the world is denote by Ω.

The above definition is too ambitious to be of any use.
In practice, we assume Ω is small (finite).
Because we are forced to simplify reality:
– You should think of a state of the world ω as describing
all relevant information to some decision making problem.
States of the World


States of the World


States of the World


States of the World


States of the World
Example 1
Decision Problem In a sunny morning, choose whether to

carry an umbrella, a parasol or nothing to work.
Possible States of The World Ω = {ω1 , ω3 , ω3 }.

ω1 −→ The afternoon is also sunny.
ω2 −→ The afternoon is cloudy but it does not rain.
ω3 −→ The afternoon is rainy.
States of the World
Example 1
Decision Problem Buy, sell or take no action regarding

Alibaba shares in the NYSE.
Possible States of The World Ω = {ω1 , ω3 , ω3 , ω4 }.
ω1 −→ %∆ BABA > 20%.
ω2 −→ 20% > %∆ BABA > 5%.
ω3 −→ 5%> %∆ BABA > 0%.
ω4 −→ 0%> %∆ BABA>-10%.
where above, %∆ BABA is the rate of change of Alibaba’s

stock price, p1p−p
0
0
, where p1 is the future (six months ahead)
price and p0 is the current price.
Information Partitions
Individual i’s knowledge about the states of the world is

represented by a partition P of the set Ω.
Definition
A partition P of Ω is a collection of subsets of Ω such that:
1. If A ∈ P and B ∈ P and A ̸= B then A ∩ B = ∅.
2. ∪A∈P A = P.
If two states are in the same element of our knowledge

partition, it means we cannot distinguish between these states.
But if two states are if different elements of the partition, we
can distinguish them.

Definition
2. ∪A∈P A = P.

Information partition
Weather in the afternoon example

a) Pa = {{ω1 }, {ω2 , ω3 }}, we know if the afternoon is sunny or

not, but if it is not sunny, we cannot tell if it is rainy or
cloudy.
b) Pb = {{ω1 , ω2 , ω3 }}, we do not know anything about the
weather in the afternoon.
c) Pc = {{ω1 , ω2 }, {ω3 }}, we know if the afternoon is rainy or
not, but if it is not rainy, we cannot tell if it is sunny or
cloudy.

Definition
2. ∪A∈P A = P.


Definition
2. ∪A∈P A = P.

Information partition
Weather in the afternoon example

a) Pa = {{ω1 }, {ω2 , ω3 }}, we know if the afternoon is sunny or

not, but if it is not sunny, we cannot tell if it is rainy or
cloudy.
b) Pb = {{ω1 , ω2 , ω3 }}, we do not know anything about the
weather in the afternoon.
c) Pc = {{ω1 , ω2 }, {ω3 }}, we know if the afternoon is rainy or
not, but if it is not rainy, we cannot tell if it is sunny or
cloudy.
Cheryl’s birthday
Albert and Bernard just met Cheryl.

”When’s your birthday?” – Albert asked Cheryl. Cheryl thought a
second and said, ”I’m not going to tell you, but I’ll give you some
clues.” She wrote down a list of 10 dates:
May 15, May 16, May 19

June 17, June 18
July 14, July 16
August 14, August 15, August 17
”My birthday is one of these,” – she said.
Then Cheryl whispered in Albert’s ear the month and only the month
of her birthday. To Bernard, she whispered the day, and only the day.
”Can you figure it out now?” – she asked Albert.
Albert: I don’t know when your birthday is, but I know Bernard
doesn’t know, either.
MAY15 MAY16 MAY19
JUN 17 JUN18
JUL14 JUL16
AUG14 AUG15 AUG17

MAY15 MAY16 MAY19
JUN 17 JUN18
JUL14 JUL16
AUG14 AUG15 AUG17
”whispered in Albert’s ear the month”

MAY15 MAY16 MAY19
JUN 17 JUN18
JUL14 JUL16
AUG14 AUG15 AUG17
”To Bernard, she whispered the day”

MAY15 MAY16 MAY19
JUN 17 JUN18
JUL14 JUL16
AUG14 AUG15 AUG17

MAY15 MAY16 MAY19
JUN 17 JUN18
JUL14 JUL16
AUG14 AUG15 AUG17

Albert: I don’t know when your birthday is,
but I know Bernard doesn’t know, either.
MAY15 MAY16 MAY19
JUN 17 JUN18
JUL14 JUL16
AUG14 AUG15 AUG17

Bernard: I didn’t know originally, but now I do.
MAY15 MAY16 MAY19
JUN 17 JUN18
JUL14 JUL16
AUG14 AUG15 AUG17

Bernard: I didn’t know originally, but now I do.
Albert: Well, now I know, too!

Lotteries
Assume we have n states of the world.
Definition: A lottery is a list of prizes and probabilities,
ℓ = ((x1, x2, . . . , xn), (p1, p2, . . . , pn)) ,

where xk ∈ Rm is the prize the lottery gives when state k
occurs, and pk ≥ 0 is the probability that state k occurs.
Lotteries
Assume we have n states of the world.
Definition: A lottery is a list of prizes and probabilities,
ℓ = ((x1, x2, . . . , xn), (p1, p2, . . . , pn)) ,

where xk ∈ Rm is the prize the lottery gives when state k
occurs, and pk ≥ 0 is the probability that state k occurs.
Lotteries
In practice, it is very useful to depict lotteries as a graph.
x1 x2 ... xn
p1 p2 pn
ℓ
Lotteries
In practice, it is very useful to depict lotteries as a graph.
x1 x2 ... xn
p1 p2 pn
ℓ
The Certain Lottery
The lottery that gives prize x with probability one (with

certainty) is denoted by:
δx = ((x), (1)) .
δx
Expectation and Variance
The expected value of a lottery

ℓ1 = ((x1 , x2 , . . . , xn ), (p1 , p2 , . . . , pn )) is:
∑
n
E[ℓ1 ] = p1 · x1 + p2 · x2 + . . . pn · xn = pi · xi ;
i=1
and the variance of a lottery ℓ is:
Var[ℓ1 ] =p1 · (x1 − E[ℓ1 ])2 + p2 · (x2 − E[ℓ1 ])2 + . . . pn · (xn − E[ℓ1 ])2 =
∑
n
= pi · (xi − E[ℓ1 ])2 .
i=1
Composition of Lotteries
Given two lotteries, ℓ1 = ((x1 , x2 , . . . , xn ), (p1 , p2 , . . . , pn )) and

ℓ2 = (y1 , y2 , . . . , ym ), (q1 , q2 , . . . , qn )) and a number α in the
interval (0, 1), we can create a compound lottery ℓ that plays ℓ1
with probability α and ℓ2 with probability 1 − α.
ℓ = αℓ1 ⊕ (1 − α)ℓ2 =
= ((x1 , x2 , y1 , y2 ), (αp, α(1 − p), (1 − α)q, (1 − α)(1 − q)).
x1
x2
ℓ
y1
The compound lottery ℓ plays ℓ1 with probability α and ℓ2 with

probability ℓ2 :
ℓ = αℓ1 ⊕ (1 − α)ℓ2 =
= ((x1 , x2 , y1 , y2 ), (αp, α(1 − p), (1 − α)q, (1 − α)(1 − q)).
x1
p
ℓ1
α x2
1−p
ℓ
1−α q
ℓ2 y1
1−q y2
The compound lottery ℓ plays ℓ1 with probability α and ℓ2 with

probability ℓ2 :
ℓ = αℓ1 ⊕ (1 − α)ℓ2 =
= ((x1 , x2 , y1 , y2 ), (αp, α(1 − p), (1 − α)q, (1 − α)(1 − q)).
x1
p
α·
p) x2
α · (1 −
ℓ (1 − α)
·q
(1 − y1
α)
· (1
−q
)
y2
Preferences Over Lotteries
Given to lotteries ℓa and ℓb such that a decision maker (DM)

chooses ℓa over ℓb ,
the following statements are equivalent:
The DM judges ℓa no worst than ℓb (everyday language);
The DM prefers ℓa to ℓb (economics language);
ℓa ⪰DM ℓb (mathematics language).
For simplicity we write:
ℓa ≻ ℓb when ℓa ⪰ ℓb but ℓb ̸⪰ ℓa (strict preference)
ℓa ∼ ℓb when ℓa ⪰ ℓb and ℓb ⪰ ℓa (indifference).





A preference of the DM, ⪰DM , over the set of lotteries is

just the DM’s ranking of lotteries.
We wish (for convenience) a numerical score that reflects
the DM’s ranking.
von Neuman & Morgenstern’s Assumptions:
Completeness For any two lotteries ℓ1 and ℓ2 ,
ℓ1 ⪰ ℓ2 and/or ℓ2 ⪰ ℓ1 .
Transitivity For any lotteries ℓ1 , ℓ2 and ℓ3 ,
if ℓ1 ⪰ ℓ2 and ℓ2 ⪰ ℓ3 then ℓ1 ⪰ ℓ3 .
Continuity If ℓ1 ⪰ ℓ2 ⪰ ℓ3 then exists p ∈ [0, 1] such that
ℓ2 ∼ pℓ1 ⊕ (1 − p)ℓ3 .
Independence If ℓ1 ≻ ℓ2 then for any ℓ3 and any 0 < p < 1,
pℓ1 ⊕ (1 − p)ℓ3 ≻ pℓ2 ⊕ (1 − p)ℓ3 .
ℓ1 ⪰ ℓ2 and/or ℓ2 ⪰ ℓ1 .
ℓ2 ∼ pℓ1 ⊕ (1 − p)ℓ3 .
pℓ1 ⊕ (1 − p)ℓ3 ≻ pℓ2 ⊕ (1 − p)ℓ3 .
ℓ1 ⪰ ℓ2 and/or ℓ2 ⪰ ℓ1 .
ℓ2 ∼ pℓ1 ⊕ (1 − p)ℓ3 .
pℓ1 ⊕ (1 − p)ℓ3 ≻ pℓ2 ⊕ (1 − p)ℓ3 .
ℓ1 ⪰ ℓ2 and/or ℓ2 ⪰ ℓ1 .
ℓ2 ∼ pℓ1 ⊕ (1 − p)ℓ3 .
pℓ1 ⊕ (1 − p)ℓ3 ≻ pℓ2 ⊕ (1 − p)ℓ3 .
ℓ1 ⪰ ℓ2 and/or ℓ2 ⪰ ℓ1 .
ℓ2 ∼ pℓ1 ⊕ (1 − p)ℓ3 .
pℓ1 ⊕ (1 − p)ℓ3 ≻ pℓ2 ⊕ (1 − p)ℓ3 .
ℓ1 ⪰ ℓ2 and/or ℓ2 ⪰ ℓ1 .
ℓ2 ∼ pℓ1 ⊕ (1 − p)ℓ3 .
pℓ1 ⊕ (1 − p)ℓ3 ≻ pℓ2 ⊕ (1 − p)ℓ3 .
If ⪰ satisfy all of of the above, there exists u : R → R such that
((x1 , x2 , . . . , xn ), (p1 , p2 , . . . , pn )) ≻ (y1 , y2 , . . . , ym ), (q1 , q2 , . . . , qn ))
if and only if
∑
n ∑m
u(xk ) · pk > u(yk ) · qk .
k=1 k=1
Expected Utility
We write:
U(ℓ1 ) = u(x1 ) · p1 + . . . + u(xn ) · pn
and refer to U as the expected utility and to u as the:

1 utility for money
2 Bernoulli utility
3 von Neumann-Morgenstern utility (vNM)
Expected Utility
We write:
U(ℓ1 ) = u(x1 ) · p1 + . . . + u(xn ) · pn ,

1 utility for money
2 Bernoulli utility
Expected Utility
We write:
U(ℓ1 ) = u(x1 ) · p1 + . . . + u(xn ) · pn ,

1 utility for money
2 Bernoulli utility
How to use expected utility
1 List the states of the world: ω1 , ω2 , ...., ωn .

2 List states’ probabilities: p1 , p2 , ...,pn .
3 For each possible action:
a) List its possible outcomes (prizes): x1 , x2 , ..., xn .

b) Obtain outcomes’ utilities: u(x1 ), u(x2 ), ..., u(xn ).
c) Calculate the expectation of the utility:
p1 · u(x1 ) + p2 · u(x2 ) + . . . + pn · u(xn ).
4 Choose action that delivers the highest expected utility.



p1 · u(x1 ) + p2 · u(x2 ) + . . . + pn · u(xn ).



p1 · u(x1 ) + p2 · u(x2 ) + . . . + pn · u(xn ).



p1 · u(x1 ) + p2 · u(x2 ) + . . . + pn · u(xn ).

An Example
Suzan has to decide whether to insure (or not) her laptop which
is worth $ 1,500. The insurance premium is $ 20 and the
deductible is $ 200. The probability of theft or accident is 1/25
and the probability of nothing happening is 24/25. Her utility
for money is u(x) = 20000x − x2 . She has to choose between
insurance or no insurance.
1 What are the states?
2 What are the states’ probabilities?
3 If she does not insure, What are her prizes in each state?
4 If she does not insure, What is her utility in each state?
5 What is the expected utility of not insuring?
6 What is the expected utility of insuring?
An Example
An Example
An Example
An Example
An Example
ECON101 Review
Learning-by-doing exercise
Definition: Rate of Return

If an asset’s price today is p0 and its price tomorrow is p1 then
the asset’s rate of return is
p1
r= .
p0
With x dollars, how many units of the asset can you buy today?
What is the revenue you obtain by selling these units tomorrow?
Expected Utility
Learning-by-doing Exercises
For each of the following settings, describe: the states, the

probabilities, the prizes and, compute the expected utility.
√
Assume the utility for money is u(x) = x and the initial
wealth is w.
1 A monetary loss L, where w > L > 0 happens with
probability 0 < p < 1. The insurance premium and the
deductible are P > 0 and D > 0. The individual buys
insurance.
2 A fraction 0 < α < 1 of the initial wealth is invested in
bonds with a rate of return r ≥ 1, the remaining is invested
in a risky asset that has return RH with probability p and
RL with probability 1 − p, where RH > r > RL ≥ 0.
Extracting u from ⪰
≻
≻
≺
Risk Aversion
Let’s go back to expected utility theory, consider the two

lotteries:
1 1
ℓ1 = ((100, 102), ( , )
2 2
and
δ101 = ((101), (1))
We have
1 1
U(ℓ1 ) = u(100) · + u(102) · and
2 2
U(δ101 ) = u(101) · 1.
Risk Aversion
Let’s go back to expected utility theory, consider the two

lotteries:
1 1
ℓ1 = ((100, 102), ( , )
2 2
and
δ101 = ((101), (1))
We have
1 1
U(ℓ1 ) = u(100) · + u(102) · and
2 2
U(δ101 ) = u(101) · 1.
Risk Aversion
1
U(ℓ1 ) − U(δ101 ) = [u(102) − u(101) − (u(101) − u(100))] ·
2
1 1
U(ℓ1 ) = u(100) · + u(102) · and U(δ101 ) = u(101).
2 2
Risk Aversion
  
   1
U(ℓ1 ) − U(δ101 ) = u(102) − u(101) − u(101) − u(100) ·
| {z } | {z } 2
≃ Mu(101) ≃ Mu(100)
1 1
U(ℓ1 ) = u(100) · + u(102) · and U(δ101 ) = u(101).
2 2
Risk Aversion
  
   1
U(ℓ1 ) − U(δ101 ) = u(102) − u(101) − u(101) − u(100) ·
| {z } | {z } 2
≃ Mu(101) ≃ Mu(100)
U(ℓ1 ) > U(δ101 ) ⇔ Mu(101) > Mu(100).
1 1
U(ℓ1 ) = u(100) · + u(102) · and U(δ101 ) = u(101).
2 2
Risk Aversion
  
   1
U(ℓ1 ) − U(δ101 ) = u(102) − u(101) − u(101) − u(100) ·
| {z } | {z } 2
≃ Mu(101) ≃ Mu(100)
U(ℓ1 ) > U(δ101 ) ⇔ Mu(101) > Mu(100).
U(ℓ1 ) = U(δ101 ) ⇔ Mu(101) = Mu(100).
U(ℓ1 ) < U(δ101 ) ⇔ Mu(101) < Mu(100).
1 1
U(ℓ1 ) = u(100) · + u(102) · and U(δ101 ) = u(101).
2 2
Risk Aversion
u
u(102)
u(100)
x
100 101 102
U(ℓ1 ) E[ℓ1 ]
U(δ101 ) = u(101)
Risk Aversion
u
u(102)
1
2 u(100) + 12 u(102)
u(100)
x
100 101 102
U(ℓ1 ) E[ℓ1 ]
U(δ101 ) = u(101)
Risk Aversion
u
u(102)
1
2 u(100) + 12 u(102)
u(100)
x
100 101 102
U(ℓ1 ) E[ℓ1 ]
U(δ101 ) = u(101)
Risk Aversion
u
u(102)
1
2 u(100) + 12 u(102)
u(100)
x
100 101 102
U(ℓ1 ) E[ℓ1 ]
U(δ101 ) = u(101)
Risk Aversion
u
u(102)
1
2 u(100) + 12 u(102)
u(100)
x
100 101 102
U(ℓ1 ) E[ℓ1 ]
U(δ101 ) = u(101)
Expected Utility Theory
Attitudes Towards Risk
1 Diminishing marginal utility, u is concave, u′′ < 0 ⇒, the

consumer is risk-averse.
U(X) < u(E[X]) for all X
2 Increasing marginal utility, u is convex, u′′ > 0 ⇒, the

consumer is risk-loving.
U(X) > u(E[X]) for all X
3 Constant marginal utility, u is affine (linear plus a

constant),u′′ = 0 ⇒, the consumer is risk-neutral,
U(X) = u(E[X]) for all X













Measuring the Degree of Risk-Aversion
The Arrow-Pratt or Absolute Measure of Risk Aversion
Definition
The Arrow-Pratt absolute measure of risk-aversion of an agent
with VN-M utility u at wealth level w is:
−u′′ (w)
ρu (w) = .
u′ (w)
If for two individual with VN-M utilities u and e u we have that

ρu (w) > ρeu (w) for all wealth levels w then we say that the agent
with utility u is more risk-averse than the agent with utility eu.
The Arrow-Pratt or Absolute Measure of Risk Aversion
Definition
The Arrow-Pratt absolute measure of risk-aversion of an agent
with VN-M utility u at wealth level w is:
−u′′ (w)
ρu (w) = .
u′ (w)
If for two individual with VN-M utilities u and e u we have that

ρu (w) > ρeu (w) for all wealth levels w then we say that the agent
with utility u is more risk-averse than the agent with utility eu.
The Relative Measure of Risk-Aversion
We are not covering this material, please skip this slide...
Definition
The relative absolute measure of risk-aversion of an agent with
VN-M utility u at wealth level w is:
−u′′ (w) w
ru (w) = .
u′ (w)
The Coal Industry
A motivating example
Click to read the July of 2015, NY Times article,

Coal Miners Struggle to Survive in an Industry Battered by Layoffs
and Bankruptcy.
Since then, the trend described in the article has continued,
Arch Coal Files for Chapter 11 Bankruptcy Protection.
China has also experienced a similar trend (although for slightly
different reasons),
Mass Layoffs in China’s Coal Country Threaten Unrest.
Long-run (partial) equilibrium
Consider an industry (e.g. energy) where the cost function is

c(q) = q2 + 100 and demand function Q = 100 − 2p.
1 What is the long-run price?
2 What is the long-run aggregate quantity?
3 How many firms are in the market?
The marginal and average costs are MC(q) = 2q and
AC(q) = q + 100/q. So zero-profit, p = AC, and profit
maximization, p = MC, imply AC = MC so q = 10 and p = 20.
Aggregate demand is then Q = 100 − 40 = 60, so there are
N = Q/q = 60/10 = 6 firms.

N = Q/q = 60/10 = 6 firms.

N = Q/q = 60/10 = 6 firms.
Suppose now there is technological progress (e.g. fracking) that

allows some firms to produce (energy) cheaper (e.g. natural gas
instead of coal).
The new cost function is b d = q.
c(q) = q2 /2 + 80 and so MC
Assume there is only one firm using the new technology and 6
using the old one.
1 What is the equilibrium price?
2 What are the firms’ profits?
If the price is p, an old firm supplies q = p/2 and a new firm
supplies b
q = p. Total supply is then Q = 6p/2 + p = 4p.
Equating it with total demand, Q = 100 − 2p, give us
p = 100/6 < 20. Profits are π = 100/6 · (50/6) − (50/2)2 − 100
and πb = (100/6)2 − (100/6)2 /2 − 80.

instead of coal).
c(q) = q2 /2 + 80 and so MC
using the old one.
supplies b
p = 100/6 < 20. Profits are π = 100/6 · (50/6) − (50/2)2 − 100
and πb = (100/6)2 − (100/6)2 /2 − 80.

instead of coal).
c(q) = q2 /2 + 80 and so MC
using the old one.
supplies b
p = 100/6 < 20. Profits are π = 100/6 · (50/6) − (50/2)2 − 100
and πb = (100/6)2 − (100/6)2 /2 − 80.

instead of coal).
c(q) = q2 /2 + 80 and so MC
using the old one.
supplies b
p = 100/6 < 20. Profits are π = 100/6 · (50/6) − (50/2)2 − 100
and πb = (100/6)2 − (100/6)2 /2 − 80.

instead of coal).
c(q) = q2 /2 + 80 and so MC
using the old one.
supplies b
p = 100/6 < 20. Profits are π = 100/6 · (50/6) − (50/2)2 − 100
and πb = (100/6)2 − (100/6)2 /2 − 80.

instead of coal).
c(q) = q2 /2 + 80 and so MC
using the old one.
supplies b
p = 100/6 < 20. Profits are π = 100/6 · (50/6) − (50/2)2 − 100
and πb = (100/6)2 − (100/6)2 /2 − 80.
Beyond partial equilibrium
Labor Market
After the old technology firms are displayed by the new ones,
What happens with the labor force? To answer this, we have to
model both technologies and so, we assume:
1 Interest rate, r = 1.
2 Old firms need 100 units of capital to produce k = 100.
3 New firms need 80 units of capital to produce b k = 80.
4 The wage rate is w.
√
5 Old production function is q = f(k, l) = l if k ≥ 100 and
zero otherwise.
√
6 New production function is bq = f(k, l) = 2l if k ≥ 80 and
zero otherwise.
Cost functions are:
w 2
c(q) = w · q2 + 100 and b
c(q) = · q + 80.
2
Labor Market
√
zero otherwise.
√
zero otherwise.
Cost functions are:
w 2
c(q) = w · q2 + 100 and b
c(q) = · q + 80.
2
Labor Market
√
zero otherwise.
√
zero otherwise.
Cost functions are:
w 2
c(q) = w · q2 + 100 and b
c(q) = · q + 80.
2
Labor Market
√
zero otherwise.
√
zero otherwise.
Cost functions are:
w 2
c(q) = w · q2 + 100 and b
c(q) = · q + 80.
2
Labor Market
√
zero otherwise.
√
zero otherwise.
Cost functions are:
w 2
c(q) = w · q2 + 100 and b
c(q) = · q + 80.
2
Labor Market
√
zero otherwise.
√
zero otherwise.
Cost functions are:
w 2
c(q) = w · q2 + 100 and b
c(q) = · q + 80.
2
Labor Market
√
zero otherwise.
√
zero otherwise.
Cost functions are:
w 2
c(q) = w · q2 + 100 and b
c(q) = · q + 80.
2
Labor Market
√
zero otherwise.
√
zero otherwise.
Cost functions are:
w 2
c(q) = w · q2 + 100 and b
c(q) = · q + 80.
2
Labor Market
√
zero otherwise.
√
zero otherwise.
Cost functions are:
w 2
c(q) = w · q2 + 100 and b
c(q) = · q + 80.
2
Labor Market
The demand for labor, for the old and new technology firms are:
( p )2 ( p )2
d(w) = and b
d(w) = /2.
2w w
What about the labor supply?
Labor Market
The demand for labor, for the old and new technology firms are:
( p )2 ( p )2
d(w) = and b
d(w) = /2.
2w w
What about the labor supply?
Labor Supply
a first look
Assume individuals have utility for two goods: leisure and a

consumption good, ℓ and x:
u(ℓ, x) = ln(ℓ) + ln(x).
Also assume that: individuals are endowed with h hours that

can either be sold at the labor market or used for leisure and;
individuals only source of income is labor. Let px be the price of
the consumption good and w be the wage rate. The individuals
budget constraint is:
px · x ≤ w · (h − ℓ).
Labor Supply
a first look
Solving for the optimal basket,

Muℓ w
MRS ≡ = ⇒ w · ℓ = px · x
Mux px
px · x = w · (h − ℓ)
Thus,
w
ℓ(w, px ) = h/2 and x(w, px ) = · h.
2px
So individual labor supply is inelastic,
l ≡ h − ℓ(w, px ) = h/2.
Labor Supply
a first look

Muℓ w
MRS ≡ = ⇒ w · ℓ = px · x
Mux px
px · x = w · (h − ℓ)
Thus,
w
ℓ(w, px ) = h/2 and x(w, px ) = · h.
2px
l ≡ h − ℓ(w, px ) = h/2.
Labor Supply
a first look

Muℓ w
MRS ≡ = ⇒ w · ℓ = px · x
Mux px
px · x = w · (h − ℓ)
Thus,
w
ℓ(w, px ) = h/2 and x(w, px ) = · h.
2px
l ≡ h − ℓ(w, px ) = h/2.
Labor market equilibrium
a first look
If the population size is M then the labor supply is
LS (w, px ) = M · h/2.
If there are N1 firms with the old technology and N2 with the
new technology, the labor demand is:
( p )2 N ( p )2
2
LD (w, p) = N1 · + .
2w 2 w
√
p
If N2 = 0, the equilibrium wage is w = 2 2N 1
.
√ Mh
N2
If N1 = 0, the equilibrium wage is w = p Mh .
But p also depends on w !
a first look
If the population size is M then the labor supply is
LS (w, px ) = M · h/2.
If there are N1 firms with the old technology and N2 with the
new technology, the labor demand is:
( p )2 N ( p )2
2
LD (w, p) = N1 · + .
2w 2 w
√
p
If N2 = 0, the equilibrium wage is w = 2 2N 1
.
√ Mh
N2
If N1 = 0, the equilibrium wage is w = p Mh .
But p also depends on w !
Long-run equilibrium
beyond partial equilibrium
Learning-by-doing exercise (PS 03)
c(q) = w2 · q2 + 80.
Given the cost curve b
1 What are the MCd and AC?d
2 What is the long-run equilibrium quantity of an individual

firm q?
3 What is the long-run equilibrium price p?
4 If the demand is Q = 100 − 2p, how many firms operate?
5 Re-do the above questions for the cost curve
c(q) = w · q2 + 100.
a second look
The previous model has the following shortcomings:

Labor may not be infinitely divisible.
There may be frictions when moving labor across firms.
Fault Lines, pp. 21–24
1 What is Jane’s story?

2 Does technological progress destroy or create jobs?
3 What has happing with the fraction of high-school
graduates in every cohort since the 1970’s?
4 Read Bessen’s article in The Atlantic.
5 PS 04: Obtain census data (1970–last census) for the
fraction high-school graduates and the fraction of college
graduates ( for each age cohort). Do these data support
Rajan’s conclusions?




Labor Market
Search Frictions
We consider a search model, which incorporates search frictions,

which were ignored in the previous supply-demand analysis. Though,
we still make several simplifying assumptions:
1 The probability of not receiving any wage offer is 1 − p and, if a
wage offer is made, the probability it is wk is qk .
2 If the worker accepts a wage offer wk , he or she receives wk in
every period (no job termination is zero and, no
search-while-one-the-job).
3 A worker receives zero income while unemployed (no
unemployment benefits).
4 If the worker refuses a wage offer, he or she remains unemployed.
5 The utility of a stream of income (x0 , x1 , x2 , . . .) is
U(x0 , x1 , x2 , . . .) = u(x0 ) + δ · u(x1 ) + δ 2 · u(x2 ) + . . ., where
u(0) = 0, u′ > 0, u′′ < 0 and, 0 < δ < 1.
Job Search
An Example with three wage levels
A
w1 s1
p · q1
p · q2 A
1−p U w2 s2
R
p · (1 − q1 − q2 )
A
w3 s3
R
Job Search
An Example (cont.)
In the previous diagram, the possible “states” are: U

(unemployed), wk (wage offer wk received), sk (wage offer wk
accepted), where k = 1, 2, 3.
The edge label is either the probability the edge is followed or

the action, wage offer is accepted (A) or rejected (R), that
makes the edge to be followed with certainty.
There are seven states and we want to find the value of being at
each state given the strategy the worker follows. We consider
three possible strategies: σ1 (accepts all offers), σ2 (accepts only
w2 and w3 ) and, σ3 (accepts only w3 ).
Job Search
An Example (cont.)
We have that:
V(sk ) =u(wk ) + δu(wk ) + δ 2 u(wk ) + δ 3 u(wk ) + . . . =

( )
=u(wk ) + δ · u(wk ) + δu(wk ) + δ 2 u(wk ) + δ 3 u(wk ) + . . .
And so,
u(wk )
V(sk ) = u(wk ) + δ · V(sK ) ⇒ V(sk ) = for k = 1, 2, 3.
1−δ
Job Search
An Example (cont.)
We have that:

( )
| {z }
V(sk )
And so,
u(wk )
1−δ
Job Search
An Example (cont.)
We have that:

( )
And so,
u(wk )
1−δ
Job Search
An Example (cont.)
The value of being unemployed is not zero as it includes the

future, expected payoff:
V(U) =u(0) + δ · ((1 − p)V(U) + p · (q1 V(w1 ) + q2 V(w2 )+

+(1 − q1 − q2 )V(w3 ))) .
Job Search
An Example (cont.)
We need to find the values of the states wk for k = 1, 2, 3 but, as

the values depend on the worker’s actions, we need to specify
the worker strategy.
If the worker follows strategy σ1 , that is, accepts any wage, we

have:
V(w1 ) =V(s1 )
V(w2 ) =V(s2 )
V(w3 ) =V(s3 ).
Job Search
An Example (cont.)
If the worker follows strategy σ2 , that is, accepts only the wages
w2 and w3 we have:
V(w1 ) =V(U)
V(w2 ) =V(s2 )
V(w3 ) =V(s3 ).
If, instead, the worker follows strategy σ3 , we have:
V(w1 ) =V(U)
V(w2 ) =V(U)
V(w3 ) =V(s3 ).
Job Search
An Example (cont.)
We have seven (Bellman) equations and seven unknowns (the

values of the state). The equations are linear in the unknowns
so, in principle, we can solve (by elimination) using paper and
pencil. But, as it it is tedious, we use Mathematica. Please see
the file Search.nb.
After we solve the equations, for each strategy, we must

compare the value of unemployment (the initial “state”) under
each of the strategies: V(U)σ1 , V(U)σ2 and, V(U)σ3 . We select
the one strategy with the highest value.
Consumer Search
1 The value of the good is v.

2 At each period the consumer has to decide to buy the good
at the current price or decline.
3 If the consumer does not buy, her/his utility is u(0) = 0.
4 Future utility is discounted at rate, 0 < δ < 1.
5 If the consumer buys, the utility is u(v − p), where p is the
price paid.
6 There are three price levels: pL < pM < pH (low, medium
and high), which change accordingly to the probabilities
bellow.
1/2
1/2 1/2
1/2 pL pM pH
1/2
1/2
Consumer Search

price paid.
bellow.
1/2
1/2 1/2
1/2 pL pM pH
1/2
1/2
Consumer Search

price paid.
bellow.
1/2
1/2 1/2
1/2 pL pM pH
1/2
1/2
Consumer Search

price paid.
bellow.
1/2
1/2 1/2
1/2 pL pM pH
1/2
1/2
Consumer Search

price paid.
bellow.
1/2
1/2 1/2
1/2 pL pM pH
1/2
1/2
Consumer Search

price paid.
bellow.
1/2
1/2 1/2
1/2 pL pM pH
1/2
1/2
PS 05
Let σ1 be the strategy where the consumer only buys at pL

and waits otherwise.
Assuming the consumer follows σ1 write the recursive
equations fro V(pL ), V(pM ) and V(pH ).
Solve these equations for V(pL ), V(pM ) and V(pH ).
Re-do 2 and 3 for the strategy σ1 be the strategy where the
consumer only buys at pL and pM and waits at pH .
Assume v = 20, δ = 18 , pL = 5, pM = 10 and pH = 15.
Which strategy σ1 and σ2 does the consumer prefers?
Market For Lemons
George Akerlof
Amanda has an used car to sell. She knows its value.

On average her car model (regardless of being on sale or not)
has the following values with frequencies (i.e., probabilities):
2 3 2 1
(2K, 5K, 10K, 15K) ( , , , ).
8 8 8 8
The buyer makes an offer.
Amanda says yes or no.
Fault Lines, Chapter 2: Exporting to Grow
Who were the late developers?

Why does more money not flow from rich countries to poor
countries so as to enable the poor countries?
Which were two (distinct) strategies to grow used by late
developers?
What were the consequences of these strategies for the
households?
What is one way to discipline inefficient firms?
What was the role of the banking system in Japan and
Germany?
What happens with wages in Germany and Japan?
What did Germany and Japan to stay competitive?
What happen with Japan exchange rate? What was
Japan’s response and its unintended consequences?
What is the problem faced by banks of export-led
economies?

developers?
households?
Germany?
economies?

developers?
households?
Germany?
economies?

developers?
households?
Germany?
economies?

developers?
households?
Germany?
economies?

developers?
households?
Germany?
economies?

developers?
households?
Germany?
economies?

developers?
households?
Germany?
economies?

developers?
households?
Germany?
economies?

developers?
households?
Germany?
economies?
Pareto Efficiency
Definition: Pareto Efficiency

An allocation (of good, services, inputs, etc...) is Pareto
efficient if there is no way of making an agent strictly better off
without making another agent worse off.
Trade and Efficiency
An Example
1 Two countries: A and B.

2 Each country with two agents.
3 One consumption good.
4 Inputs: capital and labor.
√
5 One firm in each country with technology: F(K, L) = K L.
6 Endowments: ωi = (ki , li , sA B
i , si ) in which
1 ki is the amount of capital,
2 li is the amount of labor,
3 s1i is the share of firm in country A, and
4 sBi is the share of firm in country B that
agent i owns.
Intertemporal Model (no uncertainty)
t = 0, 1, . . . , T periods.
one good at each period, ct consumption at period t
πt = 1 is the spot price for all t (pay at the ”spot”)
pt is the forward price (contingent price) (pay today)
t=0
p 0 = π0
p1 t=1 t
t=2 t=T
p2 π2 πt πT
λ
..
.
pT
Definition: A forward contract is a non-standardized contract
between two parties to buy or to sell an asset at a specified
future time at a price agreed upon today.
t = 0, 1, . . . , T periods.
t=0
p 0 = π0
p1 t=1 t
t=2 t=T
p2 π2 πt πT
λ
..
.
pT
t = 0, 1, . . . , T periods.
t=0
p 0 = π0
p1 t=1 t
t=2 t=T
p2 π2 πt πT
λ
..
.
pT
t = 0, 1, . . . , T periods.
t=0
p 0 = π0
p1 t=1 t
t=2 t=T
p2 π2 πt πT
λ
..
.
pT
t = 0, 1, . . . , T periods.
t=0
p 0 = π0
p1 t=1 t
t=2 t=T
p2 π2 πt πT
λ
..
.
pT
t = 0, 1, . . . , T periods.
t=0
p 0 = π0
p1 t=1 t
t=2 t=T
p2 π2 πt πT
λ
..
.
pT
OTC = over the counter

t = 0, 1, . . . , T periods.
t=0
p 0 = π0
p1 t=1 t
t=2 t=T
p2 π2 πt πT
λ
..
.
pT
The Relationship between Forward and Spot Prices
In practice, forward contracts are over-the-counter (OTC) -

bilateral contracts between two parties that are customized as
opposed to standard contracts that are traded in markets.
Here, however, we assume forward contracts are traded in a
competitive market. As a result, by arbitrage, we must have:
πt
pt = .
(1 + r)t
Can you explain why?

Present Value
It cash-flow in period t
r interest rate period t to t + 1 (constant)
Present value formula:

I1 I2 IT
PV(I0 , I1 , I2 , . . . , IT ) =I0 + + + ... +
1 + r (1 + r) 2 (1 + r)T
∑
T
It
= (PV)
(1 + r)t
t=0
Present Value
It cash-flow in period t
r interest rate period t to t + 1 (constant)
Present value formula:

I1 I2 IT
PV(I0 , I1 , I2 , . . . , IT ) =I0 + + + ... +
1 + r (1 + r) 2 (1 + r)T
∑
T
It
= (PV)
(1 + r)t
t=0
Inter-temporal Consumption
2-Period (T = 2) Consumer Problem
c1 max U(c0 , c1 )
c1 , c2
subject to
c0 + 1
≤ Y0 + (1+r)
(1+r) c1
1
Y1
c0 ≥ 0 and c1 ≥ 0
(1 + r)Y0 + Y1
Y1
c0
Y0 Y1
0 Y0 +
1+r
c1 max U(c0 , c1 )
c1 , c2
subject to
c0 + 1
≤ Y0 + (1+r)
(1+r) c1
1
Y1
(1 + br)Y0 + Y1
c0 ≥ 0 and c1 ≥ 0
(1 + r)Y0 + Y1 r ↗ br
Y1
c0
Y0 Y1
0 Y1 Y0 +
Y0 + 1+r
1 + br
c1 max U(c0 , c1 )
c1 , c2
subject to
c0 + 1
≤ Y0 + (1+r)
(1+r) c1
1
Y1
c0 ≥ 0 and c1 ≥ 0
(1 + r)Y0 + Y1 r ↘ er
Y1
(1 + er)Y0 + Y1
c0
Y0 Y1
0 Y0 +
1+r
Y1
Y0 +
1 + er
Understanding Present Value
Arbitrage
Assume you can borrow or lend at the interest rate 1 + r:

1 What is thevalue today of Y1 dollars tomorrow?
2 What should we do if someone else thinks that x dollars
x
tomorrow are worth less than today?
1+r
3 Or alternatively, believes x dollars tomorrow are worth

x
more than dollars today?
1+r
Understanding The Solution to the Consumer Problem
MU0
MRIS ≡ =1+r
MU1
1
c0 + c1 = PV(Y0 , Y1 )
(1 + r)
To understand the first equation: MRIS is how many units of

consumption tomorrow are equal to one unit of consumption
today for the consumer and 1 + r is how many units of
consumption tomorrow are equal to one unit of consumption
today for the market. In equilibrium they ought to be equal.
The second equation just says the consumer expends her

income during her lifetime.
General Equilibrium
Contigent Goods
Definition: A continent good is defined by:

1 Its physical properties.
2 The location where it is available.
3 Its delivery time.
4 The state of the world in which it is available.
If there are K distinct physical attributes, L locations, T time

periods and S states, then the total number of contingent goods
is n = K · L · T · S.
General Equilibrium
Contigent Goods


is n = K · L · T · S.
General Equilibrium
Contigent Goods


is n = K · L · T · S.
General Equilibrium
Contigent Goods


is n = K · L · T · S.
General Equilibrium
Contigent Goods


is n = K · L · T · S.
Contingent Goods
An Example
Assume that:
K ∈ {umbrella,parasol},
K ∈ {Hillsborough,Chicago},
T ∈ {today,tomorrow}, and
S ∈ {sun,rain}.
Then we have 16 goods!

The first good is an umbrella in Hillsborough available today
provided today is sunny, the second good is an umbrella in
Hillsborough available today if it is rainy, the third good is is an
umbrella in Hillsborough available tomorrow if it is sunny—and
so on—the last good is a parasol available tomorrow in Chicago
if it rains.
General Equilibrium
A Pure Exchange Economy
There are n individuals indexed by i = 1, 2, . . . , n and k

goods indexed by j = 1, 2, . . . , k.
A pure exchange economy is a list of individuals’ utilities
and endowments, (ui , ω i )ni=1 .
An equilibrium is a price vector p = (p1 , . . . , pk ) and an
allocation of goods, where xi = (xi1 , . . . , xik ) is i’s basket of
goods.
1 The bundle xi solves i’s utility maximization problem,
max ui (x).
x
s.t.
p·x ≤ p·ω i
2 Markets clear, for each good k, if pk > 0 then

∑n i
∑n i ∑n i
∑n i
i=1 ωk = i=1 xk and, if pk = 0 then i=1 ωk ≥ i=1 xk .
General Equilibrium

goods.
max ui (x).
x
s.t.
p·x ≤ p·ω i

∑n i
∑n i ∑n i
∑n i
General Equilibrium

goods.
max ui (x).
x
s.t.
p·x ≤ p·ω i

∑n i
∑n i ∑n i
∑n i
General Equilibrium

goods.
max ui (x).
x
s.t.
p·x ≤ p·ω i

∑n i
∑n i ∑n i
∑n i
General Equilibrium

goods.
max ui (x).
x
s.t.
p·x ≤ p·ω i

∑n i
∑n i ∑n i
∑n i
An Example
Individuals: Anna and Bob

Goods: present and future consumption, c0 and c1 .
Utilities: uA (c0 , c1 ) = ln(c0 ) + δA ln(c1 ) and
uB (c0 , c1 ) = ln(c0 ) + δB ln(c1 ).
Endowments: ω A = (wA A B B B
0 , w1 ) and ω = (w0 , w1 ).
An Example

uB (c0 , c1 ) = ln(c0 ) + δB ln(c1 ).
0 , w1 ) and ω = (w0 , w1 ).
An Example

uB (c0 , c1 ) = ln(c0 ) + δB ln(c1 ).
0 , w1 ) and ω = (w0 , w1 ).
An Example

uB (c0 , c1 ) = ln(c0 ) + δB ln(c1 ).
0 , w1 ) and ω = (w0 , w1 ).
An Example (cont)
cA
MRSA ≡ MUA A
0 /MU1 =
1
δA cA
= pp01 and
0
p 0 · cA A A A
0 + p1 · c1 = p0 · w0 + p1 · w1 implies
1 p0 · wA0 + p1 · w1
A δA p0 · wA A
0 + p1 · w1
cA
0 = and cA 1 = .
1 + δA p0 1 + δA p1
Likewise for Bob,
1 p0 · wB B
0 + p1 · w1 δB p0 · wB B
0 + p1 · w1
cB
0 = and cB
1 = .
1 + δB p0 1 + δB p1
Aggregate demand for present consumption,
1 p0 · wA A
0 + p1 · w1 1 p0 · wB B
0 + p1 · w1
C0 (p0 , p1 ) = + .
1 + δA p0 1 + δB p0
The equilibrium price solves,
C0 (p0 , p1 ) = wA B
0 + w0 .
An Example (cont)
cA
MRSA ≡ MUA A
0 /MU1 =
1
δA cA
= pp01 and
0
p 0 · cA A A A
0 + p1 · c1 = p0 · w0 + p1 · w1 implies
1 p0 · wA0 + p1 · w1
A δA p0 · wA A
0 + p1 · w1
cA
0 = and cA 1 = .
1 + δA p0 1 + δA p1
Likewise for Bob,
1 p0 · wB B
0 + p1 · w1 δB p0 · wB B
0 + p1 · w1
cB
0 = and cB
1 = .
1 + δB p0 1 + δB p1
1 p0 · wA A
0 + p1 · w1 1 p0 · wB B
0 + p1 · w1
C0 (p0 , p1 ) = + .
1 + δA p0 1 + δB p0
C0 (p0 , p1 ) = wA B
0 + w0 .
An Example (cont)
cA
MRSA ≡ MUA A
0 /MU1 =
1
δA cA
= pp01 and
0
p 0 · cA A A A
0 + p1 · c1 = p0 · w0 + p1 · w1 implies
1 p0 · wA0 + p1 · w1
A δA p0 · wA A
0 + p1 · w1
cA
0 = and cA 1 = .
1 + δA p0 1 + δA p1
Likewise for Bob,
1 p0 · wB B
0 + p1 · w1 δB p0 · wB B
0 + p1 · w1
cB
0 = and cB
1 = .
1 + δB p0 1 + δB p1
1 p0 · wA A
0 + p1 · w1 1 p0 · wB B
0 + p1 · w1
C0 (p0 , p1 ) = + .
1 + δA p0 1 + δB p0
C0 (p0 , p1 ) = wA B
0 + w0 .
An Example (cont)
cA
MRSA ≡ MUA A
0 /MU1 =
1
δA cA
= pp01 and
0
p 0 · cA A A A
0 + p1 · c1 = p0 · w0 + p1 · w1 implies
1 p0 · wA0 + p1 · w1
A δA p0 · wA A
0 + p1 · w1
cA
0 = and cA 1 = .
1 + δA p0 1 + δA p1
Likewise for Bob,
1 p0 · wB B
0 + p1 · w1 δB p0 · wB B
0 + p1 · w1
cB
0 = and cB
1 = .
1 + δB p0 1 + δB p1
1 p0 · wA A
0 + p1 · w1 1 p0 · wB B
0 + p1 · w1
C0 (p0 , p1 ) = + .
1 + δA p0 1 + δB p0
C0 (p0 , p1 ) = wA B
0 + w0 .
An Example (cont)
1 p0 · wA A
0 + p1 · w1 1 p0 · wB B
0 + p1 · w1
C0 (p0 , p1 ) = + .
1 + δA p0 1 + δB p0
C0 (p0 , p1 ) = wA B
0 + w0
p1 (1 + δB )δA wA B
0 + (1 + δA )δB w0
=
p0 (1 + δB )wA B
1 + (1 + δA )w1
An Example (cont)
1 p0 · wA A
0 + p1 · w1 1 p0 · wB B
0 + p1 · w1
C0 (p0 , p1 ) = + .
1 + δA p0 1 + δB p0
C0 (p0 , p1 ) = wA B
0 + w0
0 + (1 + δA )δB w0
=
p0 (1 + δB )wA B
1 + (1 + δA )w1
An Example (cont)
1 p0 · wA A
0 + p1 · w1 1 p0 · wB B
0 + p1 · w1
C0 (p0 , p1 ) = + .
1 + δA p0 1 + δB p0
C0 (p0 , p1 ) = wA B
0 + w0
0 + (1 + δA )δB w0
=
p0 (1 + δB )wA
Total + (1 +
1supply δA )wB
1
(inelastic)
An Example (cont)
1 p0 · wA A
0 + p1 · w1 1 p0 · wB B
0 + p1 · w1
C0 (p0 , p1 ) = + .
1 + δA p0 1 + δB p0
C0 (p0 , p1 ) = wA B
0 + w0
0 + (1 + δA )δB w0
=
p0 (1 + δB )wA B
1 + (1 + δA )w1
An Example (cont)
1 p0 · wA A
0 + p1 · w1 1 p0 · wB B
0 + p1 · w1
C0 (p0 , p1 ) = + .
1 + δA p0 1 + δB p0
C0 (p0 , p1 ) = wA B
0 + w0
p1
= 1/(1 + r)
p0
0 + (1 + δA )δB w0
=
p0 (1 + δB )wA B
1 + (1 + δA )w1
An Example (cont)
Note: We can only solve for the price ratio p0 /p1 . In the
general case (with k goods) we can solve for k − 1 price ratios.
That is, we can pin-down real prices but not nominal prices.
Also notice, we only used one market (present consumption

good) to find the equilibrium. In general, if k − 1 markets are in
equilibrium then all markets are in equilibrium (Walras’ Law).
A numerical example
To make the example more concrete, let’s further assume that:

Time preferences (discount factor) are the same for all,
δA = δB .
Anna’s endowment is wA = (7, 0).
Bob’s endowment is wB = (0, 5).
So Anna and Bob’s demand for present consumption are:
7 5p1
cA
0 = and cB
0 =
1+δ 1+δ
so that the equilibrium price is the p1 that equates supply and
demand (clears the market):
7 + 5p1 7
7= ⇒ p1 = δ.
1+δ 5
and so the equilibrium consumption levels are:

7 7δ
cA
0 = , cB0 = ,
1+δ 1+δ
5 5δ
cA
1 = and cB
1 = .
1+δ 1+δ
Financial Markets
Assume that there are no markets at t = 0 to buy consumption

for future delivery c1 . There are only spot markets. To buy c1 ,
one needs to wait for time t = 1. Also assume the spot prices
for consumption are one, π0 = π1 = 1.
In the previous example, Anna did not consume her entire

endowment of present consumption, she sold a fraction of it in
order to buy future consumption (since her endowment of
future consumption was zero). Now, she can still sell c0 at the
spot price but how she can transfer this income to the future?
The only way is if she has access to financial markets. In

particular we assume there is a bond, which returns 1 + r in
period 1 to each dollar invested in period 0.
Financial Markets
The bond
Let zA and zB how much money Anna and Bob invested in a

bond that has rate of return 1 + r (from period 0 to 1).
We allow zA and zB to be positive or negative. If zA is positive,

Anna buys bonds. If zA is negative, she sells bonds. The same
applies to zB .
In equilibrium, if Anna is selling then Bob must be buying, and

vice-versa. Therefore, in equilibrium, we must have:
zA + zB = 0 .
Financial Markets
The bond


applies to zB .

zA + zB = 0 .
Financial Markets
The bond


applies to zB .

zA + zB = 0 .
Financial Markets
Anna’s maximization problem
Anna’s problem is to maximize her utility ln(c0 ) + δ ln(c1 ) so

that she can satisfy her ...two budget constraints:
(t = 0) c0 + zA = 7
(t = 1) c1 = 0 + (1 + r)zA
The problem seems more complicated because she has three
variables to choose: c0 , c1 and zA but, this is mere appearances!
Using the budgets, we can write Anna’s utility as:
UA (zA ) = ln(7 − zA ) + δ ln((1 + r)zA ),
which depends only on one variable: zA !

Financial Markets

that she can satisfy her two budget constraints:
(t = 0) c0 + zA = 7
(t = 1) c1 = 0 + (1 + r)zA
UA (zA ) = ln(7 − zA ) + δ ln((1 + r)zA ),

Financial Markets

(t = 0) c0 + zA = 7
(t = 1) c1 = 0 + (1 + r)zA
UA (zA ) = ln(7 − zA ) + δ ln((1 + r)zA ),

Financial Markets

(t = 0) c0 + zA = 7
(t = 1) c1 = 0 + (1 + r)zA
UA (zA ) = ln(7 − zA ) + δ ln((1 + r)zA ),

Financial Markets

(t = 0) c0 + zA = 7
(t = 1) c1 = 0 + (1 + r)zA
UA (zA ) = ln(7 − zA ) + δ ln((1 + r)zA ),

Financial Markets
Anna’s maximization problem (cont).
UA (zA ) = ln(7 − zA ) + δ ln((1 + r)zA ), and so

−1 (1 + r)
U′A (zA ) ≡ +δ =0⇒
7 − zA (1 + r)zA
7δ
zA =
1+δ
For Bob, his budget constraints are: c0 + zB = 0 and
c1 = 5 + (1 + r)zB and so
UB (zB ) = ln(−zB ) + δ ln(5 + (1 + r)zB ), and so

−1 (1 + r) −5
U′B (zB ) ≡ +δ = 0 ⇒ zB = .
−zB 5 + (1 + r)zB (1 + δ)(1 + r)
Finding the equilibrium interest rate
Continuing the example
In equilibrium zA + zB = 0, so
7δ −5 51
+ =0⇒ 1+r=
1 + δ (1 + δ)(1 + r) 7δ
And so:
−7δ 7δ
zB = and zA =
1+δ 1+δ
. By Bob’s budegt, c0 = −zB and c1 = 5 + (1 + r)zB so
7δ −7δ 5 1 5 δ
cB
0 = and cB
1 =5+ =5− = 5.
1+δ 1+δ7δ 1+δ 1+δ
Beyond the example
Consider an economy exactly as in the previous example but

with:
N individuals index by i = 1, 2, . . . , N
Individual i’s utility is
cα0 cα
Ui (c0 , c1 ) = +δ· 1,
α α
where 0 < α, δ < 1.
Individual i’s endowment is ei = (ei0 , ei1 ).
cα
Remark: from calculus, we have that lim = ln(c).
α↘0 α
For simplicity, we write u(c) = cα /α.
Beyond the example
Individual i’s budget constraints are:

c0 + zi = ei0 and c1 = ei1 + (1 + r)zi ,
and so writing the utility as a function of zi (using the budgets),
we have:
Ui (zi ) = u(ei0 − zi ) + δu(ei1 + (1 + r)zi
so
U′i (zi ) = −u′ (ei0 − zi ) + δ · (1 + r)u′ (ei1 + (1 + r)zi )
u′ (ei0 −zi )
and so U′ (zi ) = 0 ⇔ ′
δu (ei1 +(1+r)zi )
= 1 + r, that is:
u′ (c0 )
MRSc0 ,c1 ≡ MU0 /MU1 = =1+r
δu′ (c1 )
Since u(c) = cα /α, u′ (c) = cα−1 so
1
c0 = (δ · (1 + r)) α−1 · c1
Beyond the example
Even if consumers have identical utilities, their consumption c0

and c1 may differ as endowments (income) are not identical.
From the previous work, we have that for every consumer i:
1
ci0 = (δ · (1 + r)) α−1 · ci1
Summing over i we have
∑
N
1 ∑
N
ci0 = (δ · (1 + r)) α−1 · ci ,
i=1 i=1
∑N i
As i=1 c0is the total demand of c0 , which in equilibrium has
∑
to be equal the total supply N i
i=1 e0 , so:
(∑ )α−1
N i
e 1
1 + r = ∑i=1N
0
i
· .
i=1 e1
δ
Important remark: notice that α − 1 is a negative number!!!

Problem Set 6
Refer to the previous example.

1 Explain each of Bob’s two budget constraints. What is the
economic meaning of the right-hand side? the left?
2 Explain in economic terms why we know that zB ≤ 0?
3 Briefly explain what is a spot price?
4 Re-write Bob’s budgets, assuming that spot prices are
π0 = 3 and π1 = 2.
5 Find the value of 1 + r such that the bond market is in
equilibrium.
6 For this value of 1 + r, what are Anna and Bob’s
consumption levels?
7 How these consumption levels compare with the economy
with future delivery and no financial markets we saw
earlier?
Fault lines: Flighty Foreign Financing
1 What is the sign of the correlation between growth and

domestic savings?
2 What happens with the financial sector in a
producer-biased economy?
3 What is the current account? savings – investment
4 What was the sign of the current account in developing
countries in the 1980’s and 1990’s?
5 What two types of developing countries Rajan’s addresses?
6 How is foreign financing different in developing countries
from foreign financing in developed countries?

domestic savings?

domestic savings?

domestic savings?

domestic savings?

domestic savings?
General Equilibrium vs Financial Markets
Another Example, Modeling Uncertainty
We consider an economy with:

Two periods, t = 0, 1.
Two states, s = 1, 2.
Probability of state 1 is p.
We learn/discover which state happened, only in period 1.
In period 0, there is uncertainty regarding the state of the
economy in period 1.





One consumption good, c.





Modeling Uncertainty
Continuing the example...

We have 3 contingent goods:
1 Consumption today c0 .
2 Consumption tomorrow if state 1 occurs, c1 .
3 Consumption tomorrow if state 2 occurs, c2 .
Prices are:
The period zero (t = 0, today) prices for the contingent
goods are p0 , p1 and p2 .
The respective spot prices are π0 , π1 and π2 .
For a financial asset a, we write:
za for the dollar amount invested in the asset (positive or
negative).
1
ra for the rate of return of the asset if state 1 happens.
r2a for the rate of return if state 2 happens.
Example
cont.
We assume two financial assets, a and b and also that the

consumer utility is given by,
U(c0 , c1 , c2 ) = u(c0 ) + δ · (p · u(c1 ) + (1 − p) · u(c2 )) and,
her endowment is e = (e0 , e1 , e2 ).
General Equilibrium: Financial Markets:

One budget (at t = 0): Three spot budgets:
π0 · c0 + za + zb ≤ π0 · e0
p0 · c0 + p1 · c1 + p2 · c2 ≤
π1 ·c1 ≤ π1 ·e1 +za ·r1a +zb ·r1b
≤ p0 · e0 + p1 · e1 + p2 · e2 π2 ·c2 ≤ π2 ·e2 +za ·r2a +zb ·r2b
Choice variables are: Choice variables are:

c0 , c1 and c2 c0 , c1 , c2 , za and zb .
No Arbitrage prices (returns)
Example cont.
If we maximize U subject to the budget constraint (general

equilibrium case) we will always find an optimal c0 , c1 and c2 .
However, in the case of financial markets, we may not have an

optimal c0 , c1 , c2 , za and zb .
It depends on the prices of the assets (or their rates of return).

For example if r1a > r1b and r2a > r2b then if we take za = x and
zb = −x for some x > 0, you can see the spot budget at t = 0
does not depend on x but the income of the budgets in states 1
and 2 is an increasing function of x! Increasing x, allows the
consumer to increase the consumption in both states!
No Arbitrage
We saw that if the returns of the asset a are always larger than
the returns of asset b then selling b and using the proceeds of
the sale to buy asset a, will increase the consumer income in the
future at zero cost! We call this an arbitrage, we shall rule out
returns that allow arbitrate, thus:
If r1a > r1b then it must be that r2a < r2b .
If r1a < r1b then it must be that r2a > r2b .
Arbitrage
Basic Concepts
There are K financial assets and z1 , z2 , . . . , zK are the amounts

invested in each asset by some individual— remember that if zk < 0,
the individual sells the asset k (and is liable to pay its returns, no
default allowed); conversely, when zk > 0, the individual buys the
asset k (and is entitled to receive its returns).
Definitions
A (trading/investing) strategy is a z = (z1 , . . . , zK ).
A strategy z is self-financing iff z1 + z2 + . . . + zK = 0.

Arbitrage
Basic Concepts
There are K financial assets and z1 , z2 , . . . , zK are the amounts

invested in each asset by some individual— remember that if zk < 0,
the individual sells the asset k (and is liable to pay its returns, no
default allowed); conversely, when zk > 0, the individual buys the
asset k (and is entitled to receive its returns).
Definitions
A (trading/investing) strategy is a z = (z1 , . . . , zK ).
A strategy z is self-financing iff z1 + z2 + . . . + zK = 0.

Arbitrage: classroom practice PS 08
Consider a economy with two periods (0 and 1) and two states

in period 1.
The rate of returns of an asset are written as r = (r1 , r2 ) (return
in state 1 and state 2).
Consider only three assets: a, b and c with rate of returns
ra = (4, 1), rb = (1, 3) and rc = (1, 1)
1 For any (self-financing or not) strategy z = (za , zb , zc ) what
are its respective returns/payoffs ?
2 For a self-financing strategy, if zc = 0 what is the value of
zb in terms of za ?
3 For a self-financing strategy that does not invests on c—
with zc = 0 —what are the returns/payoffs of the strategy
as a function only of za ?
4 Find one (there are many), not necessarily self-financing
strategy with zc = 0 such that its returns/payoffs are higher
than the payoffs of the strategy (0, 0, 1) for every state.
5 Find one (there are many), self-financing strategy with
zc ̸= 0 such that its returns/payoffs are positive in every
state.

in period 1.
ra = (4, 1), rb = (1, 3) and rc = (1, 1)
zb in terms of za ?
state.

in period 1.
ra = (4, 1), rb = (1, 3) and rc = (1, 1)
zb in terms of za ?
state.

in period 1.
ra = (4, 1), rb = (1, 3) and rc = (1, 1)
zb in terms of za ?
state.

in period 1.
ra = (4, 1), rb = (1, 3) and rc = (1, 1)
zb in terms of za ?
state.
The idea of arbitrage
The A’s front office realized right away, of course, that they
couldn’t replace Jason Giambi with another first baseman just
like him. There wasn’t another first baseman just like him and
if there were they couldn’t have afforded him and in any case
that’s not how they thought about the holes they had to fill.
“The important thing is not to recreate the individual,” Billy
Beane would later say. “The important thing is to recreate the
aggregate.” He couldn’t and wouldn’t find another Jason
Giambi; but he could find the pieces of Giambi he could least
afford to be without, and buy them for a tiny fraction of the
cost of Giambi himself. – Moneyball by Micheal Lewis, p. 103
continuation
The A’s front office had broken down Giambi into his obvious
offensive statistics: walks, singles, doubles, home runs along
with his less obvious ones: pitches seen per plate appearance,
walk to strikeout ratio and asked: which can we afford to
replace? And they realized that they could afford, in a
roundabout way, to replace his most critical offensive trait, his
on-base percentage, along with several less obvious ones. The
previous season Giambi’s on-base percentage had been .477, the
highest in the American League by 50 points. (Seattle’s Edgar
Martinez had been second at .423; the average American
League on-base percentage was .334.) There was no one player
who got on base half the time he came to bat that the A’s could
afford; – Moneyball by Micheal Lewis, p. 103
continuation
on the other hand, Jason Giambi wasn’t the only player in the
Oakland A’s lineup who needed replacing. Johnny Damon
(onbase percentage .324) was gone from center field, and the
designated hitter Olmedo Saenz (.291) was headed for the
bench. The average on-base percentage of those three players
(.364) was What Billy and Paul had set out to replace. They
went looking for three players who could play, between them,
first base, outfield, and DH, and who shared an ability to get on
base at a rate thirty points higher than the average big league
player. – Moneyball by Micheal Lewis, p. 103
Assets’ returns and no-arbitrage
Let rks be the return of asset k in state s.

No-Arbitrage
Assets’ returns r = (r1 , . . . , rK ) satisfy the no-arbitrage
condition if there is no self-financing strategy z such that
r1s · z1 + . . . rK
s · zK ≥ 0 for every state s and the inequality is
strict for at least some state.
Consumption Based Asset Pricing
Consider an economy with:

1 One consumption good (for simplicity).
2 Time/state-contingent markets for the consumption good,
contingent prices are pts
3 Spot markets for the consumption good, spot prices are πs .
4 Spot financial markets, where ras is the vector of returns (in
future states) of asset a, which is traded at period/state s.
Remark: In our previous examples, assets where traded only at
period/state t = 0
Consumption Based Asset Pricing (cont.)
Assume there are no arbitrage opportunities in this economy

and consider a financial asset that is traded in state s, with
return r in state bs and return zero in all the other states.
Then,
πbs ps
r = ·
pbs πs
Assume there are no arbitrage opportunities in this economy

and consider a financial asset that is traded in state s, with
return r in state bs and return zero in all the other states.
Then,
πbs ps
r = ·
pbs πs
Explaining the pricing rule... Notice that:

πs
1 One dollar in period/state zero is worth ps in state s. Why?
2 So by (1): one dollar in period/state zero is worth r · πs
ps in
state bs. Why?
πbs
3 One dollar in period/state zero is work pbs in state bs. Why?
π
4 So by (2) and (3): r · p s = pbs .
π
s b
s
Why?
π p
5 We conclude that r = pbs · πs .
b
s s
PS 09
s3 Spot prices of financial

37 asset a.
s1
s4
35
s0 33
33 s5
s2
33
31
s6
29
t0 t1 t2
PS 09
1 Explain why the rate of return of the asset in state s1 is the

ratio of spot prices 35/33.
2 Re-draw the previous tree, but instead of prices write the
rate of return of asset a in each state (with the exception of
s0 ).
3 Consider an option that allows you to buy one unit of asset
a at price 31 in each of the third period states (s3 , s4 , s5
and s6 ). What is the value (spot price) of the option in
each of the final states (s3 , s4 , s5 and s6 )?
4 Assume we have a bond at each period with rate of return
1.03 between two periods. Find a portfolio consisting of the
bond (zb dollars invested in the bond in state s1 ) and the
asset (za dollars invested in asset a in state s1 ) such that
the payoff of the portfolio is equal to the payoff of the
option in states s3 and s4 .
5 What is the cost of the portfolio you found, za + zb ? What
is the value of the option at s1 ?
PS 09
s3
Option values.
6
s1
s4
?
s0 2
? s5
s2
2
?
s6
0
t0 t1 t2
Value of the option at s1
Consider the investment strategy that at s1 invests za dollars in

the asset and zb in the bond.
For the strategy to replicate the value of the option in the
future (states s3 and s4 ), we must have:
37 103
(1) 35 za + 100 zb =6
33 103
(2) 35 za + 100 zb =2
Solving this system give us:
3100
za = 35 and zb = −
103
This strategy is not self-financing; it costs:
505
za + zb = ≈ 4.9
103
Since the strategy has the same payoff as the option in the
future (s3 or s4 ), then, in the present at s1 , it has to have the
same price/cost as the option. We conclude the option is worth
4.9 at s1 .
If the price of the option were different, you can find a self
financing-strategy (trading the option, the asset and, the bond) that gives
you gains with positive probability and losses with zero prob..
Production in The General Equilibrium Model
How does production changes our model?

1 In the consumer budget constraint, the consumer i’s
income is the value of the endowment (goods and inputs)
but now consumer i is also endowed with shares θij of firm j.
2 As before (no production case), consumers choose basket of
goods to maximize utility subject to their budget
constraint.
3 Firms maximize profits.
4 If there is entry, profits are zero.
5 Markets for inputs must clear (supply equals demand).
6 Markets for goods clear as before (supply equals demand).
Production in The General Equilibrium Model
A simple Example
1 One consumer who owns the firm and has k units of capital
and l units of labor.
2 One consumption good.
3 One firm that produces the consumption good using the
1
technology (production function), Q = (K · L) 4 .
4 The firm’s demand for capital and labor are rrespectively
K and L.
5 Price of consumption good is p, price of capital is r and
price of labor is w.
Equilibrium conditions
p · c = r · k + w · l + 1 · π consumer budget,
1
π = p · (K · L) 4 − r · K − w · L, firm’s profits
∂π 1 1
= 0 ⇔ p · (L/K3 ) 4 − r = 0, max. profit cond.
∂K 4
∂π 1 1
= 0 ⇔ p · (K/L3 ) 4 − w = 0, max. profit cond.
∂L 4
K = k and L = l inputs markets clear,
1
c = (K · L) 4 output markets clear.
We have seven equations and also seven variables: c, K, L, p, r,

w, π. Notice that k and l (the capital and labor endowment are
not variables since their values are given, k and l are
constants!). But note we can re-scale the prices p, r, w and the
equilibrium will not change, so we can set p = 1.
Fault Lines: A weak safety net
1 What’s the story of Badri?

2 In a recession: what banks and venture capitalists do?
what happens with (some) existing firms?
3 What used to be the case with economic recoveries before
1990? What has happened post 1990 with output growth
and job growth in the recoveries?
4 Why in the US there is a strong political will do engage in
stimulus (monetary or fiscal) during recessions relatively to
Europe?
5 What is the first of Rajan’s explanation for the jobless
recoveries?
6 What is the relationship of search technology and the
jobless recoveries (i.e., Rajan’s second explanation).

Europe?
recoveries?

Europe?
recoveries?

Europe?
recoveries?

Europe?
recoveries?

Europe?
recoveries?
Liquidity
Let’s assume:
Three dates (0, 1 and 2) and one consumer.
Investment occurs at dates 0 and 1.
Consumption occurs at dates 1 or 2.
With prob. λ consumption takes place only at date 1.
With prob. 1 − λ consumption takes place at date 2.
Safe (short asset/cash/zero interest bond) investment of x
yields x at next date.
“Risky” (long asset) investment of x at date 0 yields R x at
date 2 where R > 1. The long asset is illiquid at date 1.
Initial wealth: W0 = 1.
The fraction of wealth in short asset is x.
Individual goal: max
x
λu (x) + (1 − λ)u (x + (1 − x)R)
st.
0≤x≤1
Liquidity Shocks, the Autarchy case
t=0
x in short asset
1 − x in long asset
λ no shock
shock 1−λ
t=1 t=1
c1 = x c2 = x + (1 − x)R
Without shock, consumer waits to con-

sume at t=2 and receives the return of
long-asset in addition to short-asset.
With shock, consumer must

liquidate assets and consume
everything at t=1 but long-
asset is worthless at t=1!
Liquidity Shocks
An Example
Dates: = t = 0, 1, 2.
Consumer with utility u(c) = log(c).
Safe (short asset): investment of x yields x at next date.
Risky (long asset): invest. x at t = 0 yields 4 x at t = 2.
Long asset is illiquid at t = 1.
Initial wealth: W0 = 10.
Investment at t = 0, 1.
Consumption at t = 1, 2 but not both.
1
With prob. 2 consumption takes place only at date 1.
The fraction of wealth in short asset is x.
1 1
max
x
log (x 10) + log (x 10 + (1 − x)40)
st.
2 2
0≤x≤1
Liquidity Shocks
An Example
1 1
U(x) = log (x 10) + log (x 10 + (1 − x)40)
2 2
max
x
U(x)
st.
0≤x≤1
10 10 − 40
U′ (x) = + =0 (FOC)
2 (x 10) 2 (x 10 + (1 − x)40)
2
x=
3
Liquidity Shocks, the Market case
with a spot market for the long-asset, pp 60–64 in Allen & Gale book
The model is the same as before in the risk-pooling case with

the exception that consumers investment decisions are again
individual and there is a market at t = 1 for the long-asset.
0 ≤ 1 − x fraction of wealth invested in the long-asset,
0 ≤ x fraction of wealth invested in the short-asset.
P is the spot price of the long-asset at date t = 1 (you can
think of it also as the return of liquidating at date t = 0
one dollar invested in the long-asset at date t = 0).
Liquidity Shocks
with a spot market for the long-asset
t=0
λ t=1
x in short asset
shock c1 = x + (1 − x)P
t=2
c2 = ( Px+ (1 − x))R
1 − λ no shock et
ss
g-a
lon
buy
stay put t=2
t=1
c2 = x + (1 − x)R
sell
long
-ass
et
t=2
c2 = x + (1 − x)P
Liquidity Shocks
with a spot market for the long asset
It must be that P ≤ R in eq.

λ prob. investor wants to consume only at t = 1.
( x )
max λu(x + P(1 − x)) + (1 − λ)u ((1 − x) + )R
0≤x≤1 P
In eq. we have that
P = 1 ⇒ c1 = 1 & c2 = R ⇒ u∗ = λu(1) + (1 − λ)u (R)
Law of Large Numbers
eliminating uncertainty thru averages, side comment
Here after we shall assume that if there is a large number of

consumers and if liquidity shocks are independent, then:
The fraction of consumers hit by a liquidity shock is λ,
which is also the individual probability an individual
consumer is hit by a liquidity shock.
The fraction of consumers who are not hit by a liquidity
shock is 1 − λ.
Liquidity, Risk Pooling
Let’s assume:
Three dates (0, 1 and 2).
Infinitely many consumers i ∈ [0, 1], each one with Wi = 1
and same preferences ui = u.
Investment opportunities and consumption are as before.
Probabilities of liquidity shocks are independent.
λ prob. of a ‘bad’ liquidity shock or fraction of consumers
who suffer a ‘bad’ shock.
Company decides on investment decision for the pool of
consumers, it promises c1 to early consumers and c2 to late
consumers.
Company faces no risk (Law of Large Numbers), its plans
are feasible if λc1 = x and (1 − λ)c2 = (1 − x)R.
max
x,c1 ,c2
λu (c1 ) + (1 − λ)u (c2 )
st.
0≤λ≤1
λc1 =x
(1−λ)c2 =(1−x)R
Banks’ goals
Obviously, we see the ”company” that collects the deposits of

consumers and invests them as a bank. So, why a bank’s goal
would be to maximize the expected utility of its costumers?
Utility max. is an indirect consequence of competition, if bank

1 promises less to depositors, bank 2 can take away the clients
of bank 1 by offering better terms to the depositors. In the
limit, even if banks aim to maximize profits, they ear zero
(economic) profits and depositors utility is maximized.
Aggregate Shocks
1 Diversifiable vs Non-Diversifiable Risk

2 Idiosyncratic vs Aggregate Shocks
3 Un-systematic vs Systematic Risk
Reading assignment: excerpt of Nate Silver’s “The Ratio and

the Noise”.
Liquidity Model with Aggregate Shocks
The set-up:
A deposit contract d is the promise that gives depositors the right (not obligation) to
withdraw d at t = 1.
Depositors who do not withdraw at t = 1 receive the value of the bank’s residual assets
at t = 2 (provided the bank is solvent).
As before, there is a liquid asset, the short-asset with return 1 (cash) and, a long-asset,
which is iliquid in period 1 (no secondary/spot market) item The bank chooses d
(deposit contract) and x fraction of deposits invested in the short-asset at period t = 0.
A “bad” aggregate shock happens with probability π, B state; and no shock happens
with probability 1 − π, G state.
Long-asset rate of return (period t = 0 to period t = 2) are RL = 0 in the B state and
RH = R in the G state, where (1 − π)R > 1.
Each depositor receives a liquidity shock at t = 1 with probability λ.
Depositors learns the state of the economy B or G, and whether they received a shock
at t = 1.
Shocks are independent across depositors.
Let fB and fG be the fraction of depositors who with withdraw in period t = 1 in each
respective state.
The bank is solvent at t = 2 if the B state happens, if and only if d · fB ≤ x. Similarly
the bank is solvent in the G state when, f d · fG ≤ x.
Liquidity Model with Aggregate Shocks
t=1
c1 = min(d, x/fB )
λ o ck
sh
t=0 di
ty
ui
d deposit contract liq
π
1−λ
x in short asset B no li
quidit
y sho ck t=2
B ·d
t=1 ĉ2 = max(0, x−f
1−fB )
c1 = min(d, x/fG )
1−π G
λ
1−λ
t=2
ĉ2 = (x − fG · d + (1 − x)R) /(1 − fG ) if x ≥ fG · d
ĉ2 = 0 if x < fG · d
Bank Runs
Notice that in the previous diagram, ĉ2 , is the actual period 2

consumption, c2 , if only if the depositor decides to not
withdraw in period 1. In case the depositor withdraws, c2 will
be equal to the corresponding c1 .
When should the consumer withdraw? When c1 > ĉ2 . Thus,
c2 = max(ĉ2 , c1 )
Bank Runs
Definition 1: A bank run occurs when depositors who did not

receive a liquidity shock withdraw in period 1.
Definition 2: A bank run is essential (not expectation driven) if

a bank run occurs even if depositors expect that f = λ, that is:
even if the depositor expects that only the depositors who
receive a shock will withdraw, she still wants to withdraw even
when she does not receive a shock.
Hereafter, we only consider essential bank runs. There are three
possible cases, a bank run:
1 happens in both states B and G.
2 only happens in state B.
3 never happens.
The bank’s decision of d and x will determine in which case we
will fall. That said, it is easy to see that it is never optimal of
the bank to allow the bank run always happen. So we only will
look at the last two cases.
Case 2
Bank Run Only Happens After the B shock
In this case the problem of the bank is:

( )
x−dλ+R(1−x)
(1 − π)λu(d) + (1 − π)(1 − λ)u 1−λ +
max
x,d +πu(x)
subject to
x−λ·d x−λ·d+R(1−x)
1−λ
<d≤ 1−λ
Case 3
Bank Run Never Happens
In this case the problem of the bank is:
π(1 − λ)u((x(− λ d)/(1 − λ))+

)
max x−λ d+R(1−x)
x,d +(1 − π)(1 − λ)u 1−λ + λu(d)
subject to
d≤ x−λ·d
1−λ
Fault Lines: chapter 5
1 What is the dual mandate of the FED ?

2 What were the two mistakes made by the FED in the
2000’s ?
3 Explain the sentence: “it (the FED) would not lean against
a potential unsustainable rise in asset prices ”.
4 How does the FED usually conducts monetary policy?
5 What are the effects of low interest rates?
6 How “rising asset prices” can be “self-sustained” to a
certain degree?
7 What message “the willingness to flood the market with
liquidity in the event of a severe downturn” sends to
bankers?
Final Exam Coverage
Although the final exam is focusing on the liquidity models, the following topics might be included:
1 Expected Utility
2 Contingent Markets
3 Financial Markets
4 Arbitrage
With topics 2 and 3, focus on how to write the budget constraint(s) of a consumer and solve for the
consumer utility maximization problem, find prices such that aggregate supply equates aggregate
demand. With topic 1, focus on how to set-up an expected utility and evaluate whether the agent
is risk averse or not. With topic 4, focus on the formal definition we saw in class. Regarding the
main topics, we saw three liquidity models:
(M–1) Model without banks but with a secondary market for the long-asset in period 1.
(M–2) Model with a bank but no aggregate shocks.
(M–3) Model with bank and aggregate shocks.
For each model, focus on explaining its set-up: How many periods? Which assets? What con-
sumers/depositors do in each period? What the bank does in each period? What is the expected
utility of a consumer? Why in model (M-1), the equilibrium price in the secondary market must
be P = 1? Describe the meaning of each variable: λ, π, R, etc... Notice that not all models have
the same variables. In model (M-3), if agents have optimistic beliefs, when a bunk run still happens
regardless of their beliefs?
Fault Lines: chapter 6
1 Describe the incentives that agents in the financial sector face.

2 What is the expectation of banks regarding financial crisis?
3 Rajan wrote the book before TARP. Explain the idea behind
TARP and how it fits with Rajan’s account.

Supplementary Lecture Notes

Uploaded by

Copyright:

Available Formats

Supplementary Lecture Notes

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Supplementary Lecture Notes

Uploaded by

Copyright:

Available Formats

Advanced Microeconomic Theory

How to use the toolbox:

A matrix is just a convenient way of displaying information. A

A matrix is just a convenient way of displaying information. A

bp1 bp2 ... bpq

A : n rows p columns C = A × B : n rows q columns

C1×1 = A × B = p1 u(x1 ) + p2 u(x2 ) + . . . + pn u(xn )

C1×1 = A × B = p1 u(x1 ) + p2 u(x2 ) + . . . + pn u(xn )

In Mathematica to enter the matrices,

A :={{1, 0, 3}, {5, 4, 7}}

To multiply the matrices, type: A.B shift+enter.

Often we wish to evaluate the marginal impact of ONE given

What about the other variables? We treat all the other

In Mathematica, we use the command D to compute partial

What about the other variables? We treat all the other

In Mathematica, we use the command D to compute partial

What about the other variables? We treat all the other

In Mathematica, we use the command D to compute partial

What about the other variables? We treat all the other

In Mathematica, we use the command D to compute partial

If h(x) = f(g(x)) then

h′(x) = f′(g(x)) · g′(x)

Consider a function of one variable defined on the real line,

The approximation works well only if |h| is ”small”.

For a function of two variables and h = (h1 , h2 ) we have a

Consider a function of one variable defined on the real line,

The approximation works well only if |h| is ”small”.

For a function of two variables and h = (h1 , h2 ) we have a

Consider a function of one variable defined on the real line,

The approximation works well only if |h| is ”small”.

For a function of two variables and h = (h1 , h2 ) we have a

∆U = U(x + ∆x, y + ∆y) − U(x, y)

Using Taylor’s approximation, for each of the utility functions

Let f(x, y) be a real-valued function of two variables and let g(x)

that is, f(x, g(x)) = c for all x, where c is a constant.

We refer to the function g as an implicit function since g is

Let f(x, y) be a real-valued function of two variables and let g(x)

that is, f(x, g(x)) = c for all x, where c is a constant.

We refer to the function g as an implicit function since g is

Let f(x, y) be a real-valued function of two variables and let g(x)

that is, f(x, g(x)) = c for all x, where c is a constant.

We refer to the function g as an implicit function since g is

For the utility curves below, find the equation of the

If f(x, g(x)) = c for all x, where c is a constant.

Proof: Taking the total derivative of f with respect to x,

As f is constant along g(x), we also call g an iso-curve of f.

Indifference curves and iso-cost curves are examples of

For the utility curves below: a) find the marginal rate of

Maximizing a function of one variable defined on the real line,

Maximization Problem max f(x) (P)

First order condition f′ (x) = 0 (FOC)

Any point x satisfying FOC and SOC is a candidate for an

Maximizing a function of one variable defined on an interval,

Maximization Problem max f(x) (P)

First order condition f′ (x) = 0 (FOC)

Any point x satisfying FOC and SOC is a candidate for an

Consider any function f : Rk → R.

Another definition: We say that f is convex if −f is concave.

Proposition. Assume f is concave and also assume that x