AM Lecture 1
AM Lecture 1
AM Lecture 1
Benedetto Molinari
Universidad de Málaga
Advanced Macroeconomics
DEFINITION OF MODEL:
Concept of model: Street map, ...
A macroeconomic model can be described as a system of differential
equations. Those equations include a number of dynamic relationship
among a set of endogenous variables Xt ∈ R n and a set of exogenous
variables Zt ∈ R m .
Similar to the model for a physic system, but with an important
difference: The behavior of the economic variables depends on
expectations about the future derived from human thought.
Xt = Et [F (Xt +1 , Zt , ut )]
Zt = G (Zt −1 , vt )
where Xt is a vector of endogenous variables, Zt is a vector of exogenous
variables, Et is the expectation operator, and ut and vt are i.i.d. random
shocks.
F : Economic Theory.
G : Economic Policy.
Solution: A sequence of probability distributions.
Xt = Et [F (Xt +1 , Zt , ut )]
Xt = [F (Xt +1 , Zt )]
xt = ln Xt (3)
This way the time derivative is equivalent to the growth rate of the
variable:
dXt
d ln Xt Ẋ
ẋt = = dt = t (4)
dt Xt Xt
xt = ln Xt (7)
This way the time derivative is equivalent to the growth rate of the
variable:
Xt + 1
∆xt = xt +1 − xt = ln Xt +1 − ln Xt = ln (8)
Xt
Xt + 1 − Xt Xt + 1 − Xt
= ln 1 + ≃ (9)
Xt Xt
Stability:
Det [A − λI ] = 0 (16)
a11 a12 λ 0
Det − =0 (17)
a21 a22 0 λ
Solution:
Eigenvalues:
p
(a11 + a22 ) ±(a11 + a22 )2 − 4(a11 a22 − a12 a21 )
λ1,2 = (19)
2
Three alternative solutions. In continuous time:
Eigenvalues Case 1 Case 2 Case 3
λ1 <0 <0 >0
λ2 <0 >0 >0
Case 1 (λ1 < 0, λ2 < 0): Global stability. All dynamic paths tend to
the Steady State.
Case 2 (λ1 < 0, λ2 > 0): Saddle point. Some dynamic trajectories
tend to the Steady State but other trajectories move away from the
Steady State.
Case 3 (λ1 > 0, λ2 > 0): Global instability. All trajectories move
away from the Steady State.
✲
❍
❍❥
❍
❅ ✿ ❍❍
✘
✘✘ ❍
❥
❅❘
❅ ✠ ❅
✻ ❅❘
❅
• •
❍
❨
❍❍
✒ ❅
■
❅ ✠
❍
❨
❍❍ ✠
❅
✛
✒
❅
❅❘
❅
•
❅
■
❅
✠ ❅
✲
✟✯
✟ ❍
✟ ❍❥
❍
❅
■
❅ ✒
❅ PP
q
P
✻ ❅
• • ❅❘
❅
❅ ❍
❨
❍❍ ✛ ✠
✠ ❅❘
❅
In order to solve the model numerically, we will work with system of two
difference equations.
∆x1,t x1,t
=A + Bzt (20)
∆x2,t x2,t
a11 a12
A= (21)
a21 a22
Eigenvalues:
p
(a11 + a22 ) ±(a11 + a22 )2 − 4(a11 a22 − a12 a21 )
λ1,2 = (22)
2
Three alternative solutions. In discrete time:
Eigenvalues (Modulus) Case 1 Case 2 Case 3
λ1 + 1 <1 <1 >1
λ2 + 1 <1 >1 >1