Lecture 5

Preliminaries: Building Dynamic Models

Econ 602 Spring 2020

Ibn Haldun University

April 13, 2020

Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 1 / 37
 A Simple Dynamic Economy General Principles

General Principles for Specifying a Model

A model typically has at most 3 types of entities that take decisions.

1 Households: Preferences over commodities,endowment of commodities.
2 Firms: Production technology available
3 Government: Policy instruments controlled.
Information set of decision makers
Equilibrium concept: How agents perceive their power to affect
market prices.
Competitive equilibrium: All agents in the model are price takers.

Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 2 / 37
 A Simple Dynamic Economy An Example Economy

An Example Economy
Consider a discrete time, pure exchange economy where t=0,1,2...
2 individuals that live forever.
No firms or government.
There are countably infinite number of commodities or consumption
in period t=0,1,2..
Definition
∞
An allocation is a sequence (c 1 , c 2 ) = (ct1 , ct2 )t =0 of consumption in each
period for each indiviudal.

The preferences of individuals over consumption allocations can be

represented by the utility function
∞
U (c i ) = ∑ βt ln(cti )
t =0

where β ∈ (0, 1)
Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 3 / 37
 A Simple Dynamic Economy An Example Economy

Assumptions about the utility function

The above utility function satisfies the following assumptions.

1 Time separability: Total utility from a consumption allocation c i equals
the discounted sum of period utilities.
Period utility at time t only depends on the consumption in that
period, not on consumption in other periods.
2 Time discounting: β < 1 implies that agents are impatient. β is the
subjective time discount factor.
The subjective discount rate ρ is defined as β = 1+ 1 . Often intimately
ρ
related to the interest rate in the economy.
3 Other standard properties of the utility function is that it is continuous,
twice continuously differentiable, strictly increasing (U 0 (c ) > 0),
strictly concave (U 00 (c ) < 0) and satisfies the Inada conditions.
limc →0 U 0 (c ) = +∞
limc →+∞ U 0 (c ) = 0

Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 4 / 37
 A Simple Dynamic Economy An Example Economy

Endowment Streams

Agents have deterministic endowment streams ei = (eti )t∞=0 of the

consumption goods given by
( (
1 2 if t is even 2 0 if t is even
et = et = (1)
0 if t is odd 2 if t is odd
There is no risk in this model and both agents know their endowment
pattern perfectly in advance.
All information is public i.e. all agents know everything.
At period 0, two agents meet at a central market place and trade all
commodities i.e. trade consumption of all future dates.
Let pt denote the price in period 0 of one unit of consumption good
to be delivered in period t.
Both agents behave competitively, they take the sequence of prices
(pt )t∞=0 as given and beyond their controls and they make decisions.
Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 5 / 37
 A Simple Dynamic Economy An Example Economy

Equilibrium under Arrow Debreu Markets Structure

We will define the equilibrium of this economy first under

Arrow-Debreu markets structure
In these markets trade takes place at period 0. (Before any
uncertainty has been realized in cases of uncertainy.)
Given a sequence of prices (pt )t∞=0 , households solve the following
optimization problem

max{c i }∞ ∑t∞=0 βt ln (cti ) (2)

t t =0

s.t. (3)
∑t∞=0 pt cti = ∑t∞=0 pt eti (4)
cti ≥ 0∀t (5)

Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 6 / 37
 A Simple Dynamic Economy An Example Economy

Arrow-Debreu Equilibrium

Note that the budget constraint can be written as

∞
∑ pt (eti − cti ) ≥ 0 (6)
t =0

The quantity eti − cti is the net trade of consumption of agent i for
period t which may be negative or positive.
In equilibrium, prices are right in the sense that they induce agents to
choose consumption so that total consumption equals total
endowment in each period.

Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 7 / 37
 A Simple Dynamic Economy An Example Economy

Definition of Equilibrium

Definition
∞
A competitive (Arrow-Debreu) equilibrium, are prices {p̂t }t =0 and
∞
allocations ({cti }t =0 )i =1,2 such that
∞
1 Given {p̂t }t∞=0 , for i=1,2 {ĉti }t =0 solves the optimization problem
expressed in 2.
2 Goods market clearing condition holds. i.e. there is no free disposal of
goods
et1 + et2 = ĉt1 + ĉt2 (7)

Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 8 / 37
 A Simple Dynamic Economy An Example Economy

Solving for the Equilibrium

First set up the Lagrangian associated with the problem in 2.
There are two individuals in the economy labeled i = 1, 2. Hence we
are solving two optimization problems.

∞ ∞
" #
L = max
∞ ∑
{cti }t =0 t =0
βt ln (cti ) + λi ∑ pt {eti − cti } (8)
t =0
The first order necessary conditions are then
βt
= λ i pt (9)
cti
β t +1
= λ i pt + 1 (10)
cti +1
and hence (11)
pt +1 cti +1 = βpt cti ∀t (12)
for i=1,2.
Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 9 / 37
 A Simple Dynamic Economy An Example Economy

Solving for the Equilibrium

Using the optimality condition expressed in Equation (12) together

with the budget constraint we can solve for the optimal sequence of
consumption bundles of household i as a function of the infinite
sequence of prices and endowments in the economy. i.e.

cti = cti ({pt }t∞=0 )

Plugging this into the market clearing condition
cti ({pt }t∞=0 ) + cti ({pt }t∞=0 ) = et1 + et2 ∀t

This is a system of infinite equations in an infinite number of

unknowns ({pt }t∞=0 ).
Hard to solve. Remedy: Reduce the number of equations and
unknowns to a smaller number.

Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 10 / 37
 A Simple Dynamic Economy An Example Economy

Solving for the Equilibrium

For our simple economy it is easy to achieve a solution.
Sum Equation (12) across agents to obtain
pt +1 (ct1+1 + ct2+1 ) = βpt (ct1 + ct2 )
Using the goods market clearing condition, we find that

pt +1 et1+1 + et2+1 = βpt et1 + et2

and hence

pt +1 = βpt
therefore equilibrium prices are of the form

pt = βt po
Without loss of generality, we can set po = 1, i.e. make consumption
in period 0 the numeraire.
Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 11 / 37
 A Simple Dynamic Economy An Example Economy

Solving for the Equilibrium

Then equilibrium prices have to satisfy p̂t = βt .

Since β < 1, the price of ct is lower than that of co .
This reflects the impatience of both agents.
Using Equation(12), we have that cti +1 = cti = coi ∀t i.e. consumption
is constant across time for both agents.
This reflects the desire of the agents to smooth consumption over
time.
Observe that the budget constraint of both agents will hold with
equality since period utility is strictly increasing.
The left hand side of the budget constraint for individual i becomes
∞ ∞
coi
∑ p̂t cti = coi ∑ βt =
1−β
t =0 t =0

for i=1,2.
Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 12 / 37
 A Simple Dynamic Economy An Example Economy

Solving for the Equilibrium

The two agents differ only along one dimension: agent 1 is rich first.
This is an advantage given that prices are falling over time.
The right hand side of the budget equation becomes
∞ ∞
2
∑ p̂t et1 = 2 ∑ β2t = 1 − β2
t =0 t =0
For agent 2 it becomes
∞ ∞
2β
∑ p̂t et2 = 2β ∑ β2t = 1 − β2
t =0 t =0
The equilibrium allocation is then given by
2 2
ĉt1 = ĉo1 = (1 − β) 2
= >1
1−β 1+β
2β 2β
ĉt2 = ĉo2 = (1 − β) = <1
1 − β2 1+β
Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 13 / 37
 A Simple Dynamic Economy An Example Economy

Solving for the Equilibrium

These allocations obviously satisfy

ĉt1 + ĉt2 = 2 = êt1 + êt2 ∀t
The mere fact that agent 1 is rich at first makes him consume more
in each period.
Note that there is substantial trade going on
2
In each even period, the first agent delivers 2 − 1+ β to the second
agent.
2β
In all odd periods, the second agent delivers 2 − 1+ β to the first
agent.
Note that this trade is mutually beneficial for both agents

Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 14 / 37
 A Simple Dynamic Economy An Example Economy

Solving for the Equilibrium

Without the trade they would obtain a lifetime utility of

u (eit ) = −∞ whereas with trade they obtain
ln 1+2 β
u (ĉ 1 ) = ∑t∞=0 βt ln ( 1+2 β ) = 1− β >0
2β
ln 1+ β
u (ĉ 2 ) = ∑t∞=0 βt ln ( 1+ β ) =
2β
<0 1− β
Hence both agents are better off in the competitive equilibrium than
by just eating their endowments.
Next we will show that the equilibrium consumption allocation is also
socially optimal.

Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 15 / 37
 A Simple Dynamic Economy An Example Economy

Pareto Optimality Welfare Theorems

For this economy the competitive equilibrium is socially optimal.
First let’s define what socially optimal means.
Notion of optimality: Pareto efficiency
Loosely speaking an allocation is Pareto efficient if it is feasible and if
there is no other feasible allocation that makes no household worse
off and at least one household better off.
The precise definition of a feasible allocation is
Definition
∞
An allocation (ct1 , ct2 )t =0 is feasible if and only if
1 cti ≥ 0 ∀t, for i=1,2.
2 ct1 + ct2 = et1 + et2 ∀t
Feasibility requires that consumption is nonnegative and satisfies the
resource constraint for all period t=0,1...
Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 16 / 37
 A Simple Dynamic Economy An Example Economy

Pareto Optimality and Welfare Theorems

A more proper definition of Pareto efficiency would then be

Definition
∞
An allocation (ct1 , ct2 )t =0 is Pareto efficient if it is feasible and if there is
∞
no other feasible allocation (c̃t1 , c̃t2 )t =0 such that
u (c̃ i ) ≥ u (c i )for both i=1,2.
u (c̃ i ) > u (c i ) for at least one i=1,2.

Important note: Pareto efficiency has nothing to do with fairness in

any sense.
An allocation in which agent 1 consumes everything in every period
and agent 2 starves is Pareto efficient.
We can only make agent 2 better off by making agent 1 worse off.

Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 17 / 37
 A Simple Dynamic Economy An Example Economy

Pareto Optimality and Welfare Theorems

Definition
First Welfare Theorem
∞
Let (ct1 , ct2 )t =0 be a competitive equilibrium allocation. Then this
allocation is at the same time Pareto optimal.

Definition
Second Welfare Theorem
∞
Let (ct1 , ct2 )t =0 be a Pareto efficient allocation. Then there is a
competitive equilibrium with prices {p̂t }t∞=0 that decentralizes this
allocation.

Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 18 / 37
 A Simple Dynamic Economy An Example Economy

Negishi’s Method to Compute Equilibria

In the example above, it was straightforward to compute the
competitive equilibrium by hand.
This is usually not the case for the dynamic general equilibrium
models.
Now we describe a method to compute equilibria for economies in
which the welfare theorem holds.
The main idea is to compute the Pareto optimal allocations by solving
an appropriate social planner’s problem.
The social planner problem is a simple optimization problem and does
not involve any prices.
If the first welfare theorem holds, then all competitive equilibrium
allocations are Pareto optimal.
Hence by solving for the Pareto optimal allocation we will then have
solved for the competitive equilibrium.
Equilibrium prices that support these economies will be derived from
the appropriate Lagrange multipliers.
Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 19 / 37
 A Simple Dynamic Economy An Example Economy

Negishi’s Method to Compute Equilibria

Benevolent social planner
Objective: To maximize the overall welfare in the economy. The
weighted sum of utilities of all agents
Constraint: Does not take prices into account. Only feasibility of the
allocations matter.
Consider the following social planner’s problem.

max∞ αu (c 1 ) + (1 − α)u (c 2 ) (13)

{ct1 ,ct2 }t =0
∞
= max∞ ∑ β [αln(ct1 ) + (1 − α)ln(ct2 )]
{ct1 ,ct2 }t =0 t =0
t

s.t.
cti ≥ 0 ∀t, for i=1,2.
ct1 + ct2 = et1 + et2 = 2 ∀t
for a Pareto weight of α ∈ [0, 1].
Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 20 / 37
 A Simple Dynamic Economy An Example Economy

Negishi’s Method to Compute Equilibria

The weight α indicates how important agent 1’s utility is to the

planner relative to agent 2’s utility.
The solution to this problem depends on the Pareto weights. i.e. the
optimal consumption choices are functions of α.
We have the following proposition
Proposition
∞
An allocation {ct1 , ct2 }t =0 is Pareto efficient if and only if it solves the
social planner’s problem in Equation (13) for some α ∈ [0, 1].

This proposition states that we can characterize the set of all Pareto
efficient allocations by varying α between 0 and 1 and solving the
social planner’s problem for all α’s.
For a particular value of α, the associated efficient allocation turns
out to be the competitive allocation.
Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 21 / 37
 A Simple Dynamic Economy An Example Economy

Negishi’s Method to Compute Equilibria

Attach the Lagrange multiplier µt to the resource constraint
Ignore the non negativity constraints on cti since they never bind due
to Inada conditions.
The first order necessary conditions are
αβt
= µt (14)
ct1
(1 − α ) βt
= µt (15)
ct2
Combining these yields

ct1 α
= (16)
ct2 1−α
α
ct1 = c2 (17)
1−α t
Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 22 / 37
 A Simple Dynamic Economy An Example Economy

Negishi’s Method to Compute Equilibria

The ratio of consumption between the two agents equals the ratio of
the Pareto weights in each period.
A higher Pareto weight for individual 1 results in this agent receiving
more consumption in every period.
Using the resource constraint in conjunction with Equation (17) yields

ct1 + ct2 = 2
α
c 2 + ct2 = 2
1−α t
ct2 = 2(1 − α); ct1 = 2α (18)

Note that the division is the same in each period, independent of the
agents’ endowments in that particular period.

Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 23 / 37
 A Simple Dynamic Economy An Example Economy

Negishi’s Method to Compute Equilibria

Hence the set of Pareto optimal allocations is given by
∞
PO = {{ct1 , ct2 }t =0 : ct1 = 2α; ct2 = 2(1 − α) for some α ∈ [0, 1]}
The Lagrange multipliers are given by
αβt βt
µt = =
ct1 2
How does this help us in finding the competitive equilibrium (CE) of
this economy?
Compare the FOCs of the social planner’s problem (SP) for agent 1
with that of the CE in Equation (9).
βt µt
1
= S.P.
ct α
βt
= λ1 pt C.E.
ct1

Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 24 / 37
 A Simple Dynamic Economy An Example Economy

Negishi’s Method to Compute Equilibria

1
Pick λ1 = α and µt = pt , these first order conditions are identical.
1
Similarly pick λ2 = 1− α the same is true for agent 2.
Hence for the appropriate choices of the individual Lagrange
multipliers λi and prices pt , the optimality conditions for the social
planner’s problem and the household maximization problem coincide.
Resource feasibility is required in both CE and SP problem.
We found a unique CE equilibrium but a lot of Pareto efficient
allocations, there must be an additional requirement imposed by the
CE but not by the SP problem.
In the CE, households are restricted by the budget constraint.
The planner is only concerned with resource constraint in the
economy.

Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 25 / 37
 A Simple Dynamic Economy An Example Economy

Negishi’s Method to Compute Equilibria

We single out the CE allocations from the set of Pareto efficient

allocations by asking which Pareto efficient allocations are affordable
to HH if they face they were to face Lagrange multipliers of the SP
problem as the market prices.
To achieve this we define transfer functions t i (α), i = 1, 2 as

t i (α) = ∑ µt [cti (α) − eti ]

t

The number t i (α)

is the amount of the numeraire good that agent i
would need as transfer in order to afford the Pareto efficient
allocation.

Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 26 / 37
 A Simple Dynamic Economy An Example Economy

Negishi’s Method to Compute Equilibria

Computing t i (α) for the current economy yields,

t 1 (α) = ∑ µt [ct1 (α) − et1 ]

t
βt
= ∑ [2α − et1 ]
t 2
α 1
= −
1 − β 1 − β2
1−α β
t 2 (α) = −
1 − β 1 − β2

To find the CE, we need to find the Pareto weight α such that
t 1 (α) = t 2 (α) = 0.

Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 27 / 37
 A Simple Dynamic Economy An Example Economy

Negishi’s Method to Compute Equilibria

This yields
1
α=
1+β
the corresponding allocations are
2
ct1 =
1+β
2β
ct2 =
1+β

Hence we have solved for the eqm allocations and equilibrium prices
βt
are given by the Lagrange multipliers µt = 2

Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 28 / 37
 A Simple Dynamic Economy An Example Economy

Summary of the Negishi’s Method

To summarize to compute competitive equilibria using Negishi’s method

one does the following
Solve the social planners problem for Preto efficient allocations
indexed by the Pareto weights (α, (1 − α)).
Compute the transfers indexed by α, necessary to make the efficient
allocation affordable.
As prices use Lagrange multipliers on the resource constraint in the
planner’s problem.
Find Pareto weights α̂ that make the transfer functions 0.
The Pareto efficient allocations corresponding to α̂ are equilibrium
allocations and the supporting equilibrium prices are the Lagrange
multipliers from the planners problem.

Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 29 / 37
 A Simple Dynamic Economy An Example Economy

Summary of the Negishi’s Method

Solving the competitive equilibria directly involves solving an infinite

number of equations in an infinite number of unknowns.
The Negishi method converts this problem to a finite dimension.
This is why solving the social planner’s problem instead of the
competitive equilibrium simplifies the problem.

Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 30 / 37
 A Simple Dynamic Economy An Example Economy

Sequential Markets Equilibrium

The market structure of Arrow-Debreu equilibrium is one in which all

agents meet once at the beginning of time and trade claims to future
consumption.
This may look empirically implausible.
In fact the same allocations would arise if we let agents trade
consumption and one period bonds in each period.
The market structure is such that the market for consumption goods
and bonds open each period in this case.
This structure is called sequential markets and this type of equilibrium
is called sequential markets equilibrium. (SE)

Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 31 / 37
 A Simple Dynamic Economy An Example Economy

Sequential Markets Equilibrium

Solving the competitive equilibria directly involves solving an infinite

number of equations in an infinite number of unknowns.
The Negishi method converts this problem to a finite dimension.
This is why solving the social planner’s problem instead of the
competitive equilibrium simplifies the problem.
Let rt +1 denote the interest rate on one period bond from period t to
t + 1.
A one period bond is a promise to pay 1 unit of consumption good in
period t + 1 in exchange for qt = 1+1rt +1 units of consumption good
in period t.
We can interpret qt as the relative price of one unit of consumption
good in period t + 1 in terms of period t consumption goods.

Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 32 / 37
 A Simple Dynamic Economy An Example Economy

Sequential Markets Equilibrium

Let ati +1 denote the amount of such bonds purchased by agent i in

period t and carried over to t + 1.
Household i’s budget constraint in period t is

ati +1
cti + = eti + ati (19)
1 + rt +1
or
cti + qt ati +1 = eti + ati

Agents start out life with initial bond holdings aoi .

We will mostly focus on the situation where aoi = 0.

Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 33 / 37
 A Simple Dynamic Economy An Example Economy

Sequential Markets Equilibrium

We then have the following definition
Definition
A sequential markets equilibrium is allocations (ĉti , âti +1 ) and interest rate
{rˆt +1 }t∞=0 such that
1 For i = 1, 2 given interest rates {rˆt +1 }t∞=0 {ĉti , âti +1 }t∞=0 solves
∞
max ∞ ∑ βt ln(cti )
{cti ,ati +1 }t =0 t =0
(20)

s.t.
a i
t +1
cti + ≤ eti + ati (21)
1 + rˆt +1
cti ≥ 0 (22)
ati +1 ≥ −Ā i
(23)

Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 34 / 37
 A Simple Dynamic Economy An Example Economy

Sequential Markets Equilibrium

2 ∀t ≥ 0
2 2
∑ ĉti = ∑ eti
i =1 i =1
2
∑ âti = 0 (24)
i =1

The constraint in Equation (23) on borrowing is necessary to

guarantee the existence of equilibrium.
Suppose that this constraint did not exist and assume that there is a
SM eq.m {rˆt +1 }t∞=0 {ĉti , âti +1 }t∞=0

Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 35 / 37
 A Simple Dynamic Economy An Example Economy

Sequential Markets Equilibrium

Without the constraint agent i can always do better by setting

ε
coi = ĉoi +
1 + rˆ1
cti = ĉti ∀t > 0.
a1i = â1i − ε; a2i = â2i − (1 + rˆ2 )ε... ati +1 = âti +1 − Πtτ =1 (1 + rˆτ +1 )ε

i.e. by borrowing ε > 0 more in period 0, consuming it and rolling

over the debt forever, by borrowing more and more.
This is called a Ponzi scheme. Hence without a limit on borrowing no
SM equilibrium can exist because agents would run Ponzi schemes
and augment their consumption without bound.

Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 36 / 37
 A Simple Dynamic Economy An Example Economy

Arrow Debreu and Sequential Markets Equilibrium

Under complete markets, Arrow-Debreu equilibrium coincides with

sequential markets equilibrium.
Complete markets: Markets in which you can trade all possible
contingencies in the future.
All possible state-contingent claims are available which provides the
individuals a perfect insurance option.
In other words, you can guarantee you consumption for each possible
scenario in the future under complete markets.

Econ 602 Spring 2020 (Ibn Haldun University) Lecture 5 April 13, 2020 37 / 37