T OULOUSE S CHOOL OF E CONOMICS , M1, 2013-2014

Macroeconomics – Patrick Feve

P ROBLEM SET 1 - C ONSUMPTION AND DYNAMIC FACTOR D EMANDS

Exercice I: The Permanent Income Model with Capital Accumulation

We consider an economy in which an homogenous good Yt can be either consumed Ct or invested It :

Yt = Ct + It

The good Yt is produced with the following technology

Yt = Zt + aKt

where a > 0 and Kt is the capital stock in period t. Zt is a stochastic variable. This stochastic variable
evolves according to
Zt = ρZt−1 + εt
where |ρ| ≤ 1. The random variable εt satisfies Et−1 εt = 0, where Et−1 is the expectation operator
conditional on the information set in period t − 1. The capital stock evolves according to the following law
of motion:
Kt+1 = (1 − δ)Kt + It
where δ ∈ (0, 1) is a constant depreciation rate. The representative household seeks to maximize the
following intertemporal expected utility function:
∞
∑
Et β i u(Ct+i )
i=0

where
α1
u(Ct ) = α0 Ct − (Ct )2
2
and α0 , α1 are positive real numbers. The parameter β ∈ (0, 1) is a subjective discount factor. Et is the
expectation operator conditional on the information set in period t (i.e. when consumption decisions are
made).

1. Determine the FOCs of the social planner optimization problem.

2. We impose β(1 + a − δ) = 1. Interpret this restriction. Show that the Euler equation on consumption
has the following form
Et ∆Ct+1 = 0

3. Compute the solution, i.e. express the the choice variable Ct in terms of the pre-determined variable
Kt and the exogenous variable Zt .

4. Show that the solution illustrates the Lucas critique.

5. Determine the value of the ratio σ∆c /σ∆y when ρ = 1.

1
 6. Compute the dynamic responses of consumption after a positive shock to the technology.

Exercice II: News Shocks and Stochastic Permanent Income

This exercise is an adaptation of Hall (1978) to news shocks on disposable labor income. A given
household seeks to maximize
∞
∑
Et β i U (Ct+i ) where the discount factor satisfies β ∈ (0, 1)
i=0

under the budget constraint

At+1 = (1 + r)At + Ytd − Ct
where A is the financial wealth, Y d the disposable labor income and C the real consumption. The real
interest rate r is strictly positive and constant. Et is the conditional expectations operator. The instantaneous
utility U (Ct ) is given by:
α1 2
U (Ct+i ) = α0 Ct − C
2 t
where the parameters α0 and α1 are strictly positive. In addition, we assume β(1 + r) = 1

1. Determine the optimal decision on consumption.

2. Define the intertemporal budget constraint and thus determine the consumption decisions in terms of
financial At and non–financial wealth (the expected discounted sum of net disposable labor income).

3. Now assume that the disposable labor income follows the process

∆Ytd = εt−1

where ∆Ytd = Ytd − Yt−1 d . The random term satisfies E(ε 2

t−1 ) = 0 and V (εt−1 ) = σεy . This shocks
is also iid, i.e. Et εt+i = 0 for all i ≥ 1. Determine the consumption function. Compare to the case
where
∆Ytd = εt

4. Show that the model implies V ar(∆Ct ) < V ar(∆Ytd ).

5. Discuss the quantitative properties in terms of Granger causality

(see Wikipedia http://en.wikipedia.org/wiki/Granger causality).

Exercice III: Habits in Consumption

This exercise is an adaptation of Hall (1978) to external habits in consumption. A given household
seeks to maximize
∞
∑
Et β i U (Ct+i ) where the discount factor satisfies β ∈ (0, 1)
i=0

under the budget constraint

At+1 = (1 + r)At + Ytd − Ct

2
where A is the financial wealth, Y d the disposable labor income and C the real consumption. The real
interest rate r is strictly positive and constant. Et is the conditional expectations operator. The instantaneous
utility U (Ct ) is given by:
α1 ⋆ 2
U (Ct+i ) = α0 Ct⋆ − C
2 t
The variable Ct⋆ is given by
Ct⋆ = Ct − bC̄t−1
where b ∈ [0, 1) is the habit parameter (b = 0 corresponds to Hall, 1978). The variable C̄t−1 denotes the
aggregate (or the reference for the group) consumption, which is considered as given at the individual level.
At the aggregate level (or the group level), we have C̄t−1 = Ct−1 . So, C̄t−1 acts as an externality (optimal
decisions at individual level does not internalize this past consumption, but the aggregate consumption is
a function of past consumption). The parameters α0 and α1 are strictly positive. In addition, we assume
β(1 + r) = 1

1. Determine the optimal decision on consumption at the individual level and then at the aggregate level.
2. Now assume that the disposable labor income follows the process

∆Ytd = εt

where ∆Ytd = Ytd − Yt−1 d . The random term satisfies E(ε ) = 0 and V (ε ) = σ 2 . This shocks is
t t εy
also iid, i.e. Et εt+i = 0 for all i ≥ 1. Define the intertemporal budget constraint and thus determine
the consumption function.
3. From the reduced form, compute the dynamic responses of the consumption to an innovation εt in
period t.
4. Show that the model can solve the excess smoothing puzzle (remind that with industrialized countries,
we have σ∆c /σ∆yd ≃ 0.7). What is the critial value on b that allows to solve the puzzle?

Exercice IV: Quadratic Adjustment Costs on Labor with a Linear Technology

The value of the firm Vt is the expected discounted sum of profit.

∞ (
∑ )i
1
Vt = Et Πt+i
1+r
i=1

where Et is the conditional expectations operator. r > 0 (the constant real interest rate) and the instanta-
neous profit is given by
b
Πt = Yt − wLt − (Lt − Lt−1 )2
2
where w is a constant real wage and b ≥ 0 is the adjustment cost parameter.
The production function is given by
Yt = (a + at )Lt
where a > 0 is a scale parameter and at a technology shock. This stochastic variable evolves according to

at = ρat−1 + εt

where |ρ| ≤ 1. The random variable εt satisfies Et−1 εt = 0, where Et−1 is the expectation operator
conditional on the information set in period t − 1.

3
 1. Determine the FOC of the firm’s optimization problem.

2. Compute the solution, i.e. express the the choice variable Lt in terms of the pre-determined variable
Lt−1 and the exogenous variable Zt . First use successive forward substitutions. Second, use the
method of undetermined coefficients.

3. Show that the solution illustrates the Lucas critique.

4. Compute the dynamic responses of labor after a positive shock to the technology.

5. Assume now a new process for at

at = ρat−1 + εt−1
Comment this specification for the technology shock. Redo the exercise (FOC, solution).

Exercice V: Dynamic Factor Demand

Let a representative firm, endowed with a linear-quadratic production function and subject to quadratic
adjustment costs, choose the capital stock so as to maximize the present discounted value of its profits
∞
∑ [ ]
α 2 b
max Et β i
(a + at ) kt+i − kt+i − (kt+i − kt+i−1 )2
kt 2 2
i=0

where Et is the expectation operator conditional on the information set in period t. β ∈ (0, 1) is a constant
discount factor. The technology parameters a and α are positive. The adjustment cost parameter is b ≥ 0.
We assume that the productivity shock at follows an autoregressive process of order one,

at = ρa at−1 + εt

where |ρa | ≤ 1. The random variable εt satisfies Et−1 εt = 0.

1. Determine the FOC of the firm’s optimization problem.

2. Compute the solution, i.e. express the choice variable kt in terms of the pre-determined variable kt−1
and the exogenous variable at .

3. Show that the solution illustrates the Lucas critique.

4. Compute the dynamic responses of capital after a positive shock to the technology.

5. Assume now a new process for at

at = ρa at−1 + εt−1
Comment this specification for the technology shock. Redo the exercise (FOC, solution).