Ramsey - Model.Advanced Macroeconomics I

Download as pdf or txt
Download as pdf or txt
You are on page 1of 19

Advanced Macroeconomics I

Mototsugu Shintani

University of Tokyo

S1/S2 2018
Outline

6. Consumption
Euler equation in Ramsey model
I Recall in Ramsey model, the Euler equation is given by
u00 (ct )
ct = f 0 (kt ) δ ρ
u 0 ( ct )
I Also, in decentralized economy (no social planner) version of
Ramsey model, from …rm’s FOC, we have f 0 (kt ) = rt + δ and
u00 (ct )
ct = rt + δ δ ρ
u 0 ( ct )
= rt ρ
c1t θ 1
I For CRRA utility u(ct ) = 1 θ , Euler equation simpli…es to

ct 1
= (rt ρ)
ct θ
or
ct
= σ ( rt ρ)
ct
where σ = 1/θ
Fisher model
I Two-period (discrete) model of consumption
max u(c1 ) + 1+1 ρ u(c2 ) s.t. c1 + s = w and c2 = (1 + r)s

1
U = u(w s) + u((1 + r)s)
1+ρ
1 u 0 ( c2 ) 1+ ρ
I FOC ∂U
∂s = u 0 ( c1 ) + 0
1+ρ u (c2 )(1 + r) = 0) u 0 ( c1 )
= 1+r
1+ ρ
I In continuous time 1 + ρ r since
1+r
(1 + r)(1 + ρ r) = 1 + ρ r + r + rρ r2 1+ρ

u 0 ( c2 ) u0 (c2 ) u0 (c1 )
ρ r 1 =
u 0 ( c1 ) u 0 ( c1 )
u00 (c2 )(c2 c1 ) u00 (c2 )∆c
=
u 0 ( c1 ) u 0 ( c1 )
u00 (ct )
) ct = r ρ
u 0 ( ct )
Diamond model
I Overlapping-generations (OLG) model developed by
Samuelson (1958) and Diamond (1965)
I Two generation at the same time (discrete time) - young and
old
I young generation at t
1
max u(c1t ) + u(c2t+1 )
1+ρ

s.t. c1t + st = wt and c2t+1 = (1 + rt+1 )st


I old generation at t
1
max u(c1t 1) + u(c2t )
1+ρ

s.t. c1t 1 + st 1 = wt 1 and c2t = (1 + rt )st 1


I All generations live only two period, in the next period former
young individuals becomes old individuals, previous old
individuals die, new young individuals are born
Euler equation in Diamond model

I From the result of Fisher model


u0 (c2t+1 ) 1+ρ
0
=
u (c1t ) 1 + rt+1

I Substitution of c1t = wt st and c2t+1 = (1 + rt+1 )st in Euler


equation yields saving function

st = s(wt , rt+1 )

I For example, for CRRA utility u(ct ) = (c1t θ


1)/1 θ, we
have u0 (ct ) = ct θ so that

(1 + rt+1 )(1 θ )/θ


s(wt , rt+1 ) = wt
(1 + ρ)1/θ + (1 + rt+1 )(1 θ )/θ
Dynamics in Diamond model
I Population growth Nt+1 = (1 + n)Nt
I Capital accumulation
Kt+1 Kt = Nt s(wt , rt+1 ) Kt ) (1 + n)kt+1 = s(wt , rt+1 )
I Firms are same as decentralized economy version of Ramsey
model so FOCs are f 0 (kt ) = rt and wt = f (kt ) kt f 0 (kt ) (here,
we use net output production F(Kt , Nt ) = F e(Kt , Nt ) δKt )
I Substitution yields

s(wt , rt+1 ) s (f (kt ) kt f 0 (kt ), f 0 (kt+1 ))


kt+1 = =
1+n 1+n
I For example, for Cobb-Douglas function f (kt ) = ktα δkt , we
have f 0 (kt ) = αktα 1 δ so that rt = αktα 1 δ and
wt = ktα δkt αktα + δkt = (1 α)ktα so that

s((1 α)ktα , αktα+11 δ)


kt+1 =
1+n
Dynamic Ine¢ ciency
I In Diamond model, command economy 6= decentralized
economy
I No population growth case (n = 0)
I r > 0 )dynamically e¢ cient (young better o¤, old worse o¤)
I r < 0 )dynamically ine¢ cient(young better o¤, old better o¤)

I Population growth case (n > 0)


I r > n )dynamically e¢ cient (young better o¤, old worse o¤)
I r < n )dynamically ine¢ cient(young better o¤, old better o¤)

I Recall the golden-rule of capital stock maximizes


consumption level
f 0 (k ) = r > n = f 0 (kgr )
or k < kgr )dynamically e¢ cient
f 0 (k ) = r < n = f 0 (kgr )
or k > kgr )dynamically ine¢ cient (over-accumulated)
Life cycle-permanent income hypothesis
I Permanent income hypothesis (PIH): consumption only
responds to the permanent component of income (permanent
income) not to the transitory component of income

I Suppose …nite life (life cycle) T, no discount ρ = 0, zero


interest rate r = 0, initial wealth A0 , endowment (labor
income) yt is deterministic

max ∑Tt=1 u(ct )

s.t. ∑Tt=1 ct = A0 + ∑Tt=1 yt (intertemporal budget constraint)

FOC from Lagrangian ∑Tt=1 u(ct ) + λ A0 + ∑Tt=1 yt ∑Tt=1 ct


yields u0 (ct ) = λ or ct = c is constant (consumption smoothing)
1
c1 = = cT = A0 + ∑Tt=1 yt = yp (permanent income)
T
Rational expectations
I Introduce randomness in endowment (labor income) so that yt
is now a random variable

I Employ quadratic utility function of the form (for certainty


equivalence)
1
u ( ct ) = ( ct c ) 2
2

I We use conditional expectation operator E1 [ ] = E[ jΩ1 ]


where Ω1 is information set available at time 1 (rational
expectations)
max ∑Tt=1 E1 u(ct )
s.t. ∑Tt=1 ct = A0 + ∑Tt=1 yt (intertemporal budget constraint,
same as before)
Certainty equivalence
I FOC implies marginal utility u0 (ct ) = ( ct c) should be
same for all t = 1, 2, ..., T, or
( c1 c) = E1 (ct c)
) c1 = E1 ct
) ∑Tt=1 E1 ct = A0 + ∑Tt=1 E1 yt = Tc1
1 p
) c1 = A0 + ∑Tt=1 E1 yt = y1 (permanent income at time 1)
T

I Random walk hypothesis (martingale hypothesis) of Hall


(1978)
ct 1 = Et 1 ct
or
ct = ct 1 + εt
where Et 1 εt = 0 (so that εt is martingale di¤erence sequence,
ct is martingale).
RE-PIH model
I Generalize the stochastic model of rational expectations (RE)
with a representative agent who lives in…nite period
(T ! ∞)
max ∑t∞=0 E0 βt u(ct )
s.t. At = (1 + r)At 1 + yt ct with assumptions of (1)
(constant) discount rate ρ =(constant) interest rate r,
1
discount factor β = 1+1 ρ = 1+ r (2) quadratic utility (certainty
A
equivalence) (3) no-Ponzi game condition lim Et (1+t rj)j = 0
j! ∞

I FOC from Lagrangian


L = E0 ∑t∞=0 βt [u(ct ) + λt (yt ct + (1 + r)At 1 At )]
yields (∂L/∂ct = 0 )) u0 (ct ) = λt and (∂L/∂At = 0 ))
λt = β(1 + r)Et λt+1 so that

ct = Et ct+1
Solution to RE-PIH model
I Rearranging terms in budget constraint
At = (1 + r)At 1 + yt ct yields
(1 + r)At 1 = yt ct At
I Dividing by 1 + r and t 1 ! t yields
1 1
At = (1 + r) ( yt + 1 ct + 1 ) (1 + r) A t+1
I Substituting second equation into …rst equation, repeated
substitution and taking (conditional) expectation yields
Et [ (1 + r)At 1]
= Et (yt ct At )
h i
1 1
= Et yt ct + ( 1 + r ) (yt+1 ct + 1 ) (1 + r) A t+1
j
1 A t+j
= ∑j∞=0 Et (yt+j ct + j ) lim Et
1+r j! ∞ (1 + r)j
j
1
= ∑j∞=0 Et (yt+j ct + j )
1+r
Permanent income in RE-PIH model

I Substitute ct = Et ct+1 = = E t ct + j
j j
1 1
(1 + r)At 1 = ∑j∞=0 Et yt+j ∑j∞=0 ct
1+r 1+r
j
1 1+r
= ∑j∞=0 Et yt+j ct
1+r r
j+1
1 1
At 1 = ∑j∞=0 Et yt+j ct
1+r r
I Closed-form solution (permanent income at t)
" #
j+1
1
ct = r At 1 + ∑j∞=0
p
Et yt+j = yt
1+r
Implications of RE-PIH model: stationary income

I AR(1) model for yt

yt = φyt 1 + εt

where jφj < 1


I j-period ahead prediction at time t

Et yt+j = φj yt

1 j+1
I Substitution in ∑j∞=0 1+r Et yt+j yields
j
1
∑j∞=0 1 1 1
φ
1+r 1+r yt = 1 + r 1 φ yt = 1 + r φ yt so that
1+r

1 p
ct = r A t 1 + yt = yt
1+r φ
Implications of RE-PIH model: stochastic trend
I AR(1) model for ∆yt

∆yt = φ∆yt 1 + εt

where jφj < 1


I j-period ahead prediction at time t

Et ∆yt+j = φj ∆yt

I Saving
j
1
st = yt + r A t 1 ct = ∑j∞=1 Et ∆yt+j
1+r
j φ
substitution yields ∑j∞=1 ∆yt = ∆yt = 1+r φ ∆yt
φ 1+r φ
1+r 1
φ
1+r
so that
φ
st = ∆yt
1+r φ
Implications of RE-PIH model: VAR analysis
I Testable implications considered by Campbell (1987)

I VAR model
xt = Bxt 1 + εt
where
∆yt
xt =
st
I j-period ahead prediction at time t

Et xt+j = Bj xt

so that
j j
1 1
∑j∞=1 Et xt+j = ∑j∞=1 Bj xt
1+r 1+r
1
B B
= I2 xt
1+r 1+r
Implications of RE-PIH model: Granger causality

I Testable restrictions 1: First element


1 j
st = ∑j∞=1 1+ r Et ∆yt+j so that

1
B B
1 0 I2 = 0 1
1+r 1+r
I Testable restrictions 2: st Granger-causes ∆yt so that in VAR
model
xt = Bxt 1 + εt
H0 : β12 = 0 in

β11 β12
B=
β21 β22
should be rejected (saving for a rainy day)
References
I Blanchard, Olivier and Stanley Fischer (1989) Lectures on
Macroeconomics, MIT Press.
I Campbell, John Y. (1987) Does saving anticipate declining
labor income? An alternative test of the permanent income
hypothesis, Econometrica 55, 1249-1273.
I Diamond, Peter (1965) National debt in a neoclassical growth
model, American Economic Review 55, 1126-1150.
I Hall, Robert, (1978) Stochastic implications of the life
cycle-permanent income hypothesis: theory and evidence,
Journal of Political Economy 86, 971-987.
I Romer, David (2011) Advanced Macroeconomics, 4th edition,
McGraw-Hill Education.
I Samuelson, Paul A. (1958) An exact consumption-loan model
of interest with or without the social contrivance of money,
Journal of Political Economy 66, 467-482.
I Uribe, Martín, and Stephanie Schmitt-Grohé (2017) Open
Economy Macroeconomics, Princeton University Press.

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy