A Gains from Trade Perspective on Macroeconomic ...cli2/index_files/discussion_ilut...Discussion of...

$: A Gains from Trade Perspective on Macroeconomic ...cli2/index_files/discussion_ilut...Discussion of \A Gains from Trade Perspective on Macroeconomic Fluctuations" by Paul Beaudry and$
Discussion of “A Gains from Trade Perspective onMacroeconomic Fluctuations” by Paul Beaudry and

Franck Portier

Cosmin Ilut

Duke Univ.

International Conference on “Macroeconomic Modeling in Times ofCrisis”, Paris, 2012

Cosmin Ilut (Duke Univ.) Discussion of Beaudry and Portier Oct. 2012 1 / 20

Paper

Motivation:

1 Data: business cycles feature comovement and stable inflation.2 Model: depart from the stringent representative agent model.

Heterogenous agent model: main ingredients

1 some market incompleteness.2 some labor market specialization.

Main findings:

1 Model: offers a constructive way to show comovement2 Policy: revised implications for fiscal and monetary policy3 Data: ’perception’ driven cycles with stable inflation

My discussion:

1 Mechanism to get comovement.2 Nominal rigidities as an alternative.


General model

Two sectors: investment and consumption goods

Two types of agents:

I decide how much to consume C i , buy K i , work Li

Production function for sectors:

I if only one type of labor used to produce a good → specialized labormkt.

Preferences:

I per period felicity: U i (C i , 1− Li ); both normal goodsI continuation value: V i (K i ,Ω) ≡ E [βV i (K i ,S)|Ω1,Ω]I Ω1 : information used to predict exogenous state


Competitive equilibrium

Each individual i :

maxC i ,K i ,Li

U i (C i , 1− Li ) + V i (K i ,Ω)

s. t. C i + PK i = w iLi

I P relative price of capital; consumption good is numeraire; w i wage

Here assume fully specialized labor mkts: C = F (L1);K = F (L2):

maxL1

F (L1)− w1L1; maxL2

PF (L2)− w1L2

Optimality conditions: Intra Euler

U i2(C i , 1− Li ) = piF1(Li )U i

1(C i , 1− Li )

pi =

1, for i = 1P, for i = 2

and Inter Euler:V i1(K i ,Ω) = PU i

1(C i , 1− Li )


Mechanism: Intertemporal optimality

Define the change in Ω1 : such that

∂2V i (K i ,Ω)

∂K i∂Ω1> 0

UsingV i1(K i ,Ω) = PU i

1(C i , 1− Li )

then since demand for capital goes up, under general conditions priceof capital goes up.


Mechanism: Intratemporal optimality

Take agent 2:

U2(C , 1− L) = PF1(L)U1(C , 1− L)

an increase in P, acts like a production possibility frontier shifter(TFP shock)

Gains from trade: increase in the relative price for capital.

it makes possible an increase in both L2 and C 2.

This is in contrast to the standard Barro-King logic: C and L cannotcomove unless a technology shock.



−0.2 0 0.2 0.4 0.6 0.8 1 1.2−0.2

0

0.2

0.4

0.6

0.8

1

1.2

LS (λt)LD (Z

t; P

t)

Labor

Wage


Shift in labor supply: Barro-King logic

−0.2 0 0.2 0.4 0.6 0.8 1 1.2−0.2

0

0.2

0.4

0.6

0.8

1

1.2

1.4

LS (λt)LD (Z

t; P

t)

LS (λt′ ↓ )

Labor

Wage


Shift in labor supply and demand (price ↑)

−0.2 0 0.2 0.4 0.6 0.8 1 1.2−0.2

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

LS (λt)LD (Z

t; P

t)

LS (λt′ ↓ )

LD (Zt; P′ ↑ )

Labor

Wage



Agent 1:U2(C , 1− L) = F1(L)U1(C , 1− L)

no change in the price, so here standard Barro-King logic applies

What direction of change?

Negative wealth effect: from increase in relative price of capital−→ L1 increases and C 1 falls

I at most C 1 stays constant.


Overall: recall Proposition 1

1 Aggregate positive comovement is possible:

L2 ↑, Investment ↑, L1 ↑, ( C 1 + C 2) ↑

2 Individual positive comovement is not possible:

C 1, L1 and C 2, L2 cannot move all up.

Next, discuss the two main mechanisms:

1 increased demand for capital

2 comovement between C 2 and L2.


Discussion : 1) increased demand for capital

Intertemporal Euler in the standard dynamic optimization formulation

λt = βEtλt+1Rkt+1

I λt = LM on budget constraint; here:

βEtλt+1Rkt+1 = V i

1(K i ,Ω)

Start from: as Ω1 ↑, then V i1(K i ,Ω) ↑

It is the marginal ‘valuation’ of capital ↑In standard RBC: good news on TFP −→ βEtλt+1R

kt+1 ↓

Need the intertemporal substitution stronger than wealth effect.

Same logic if signals about future means or variances.


Discussion : 2) comovement between C and L

Irrespective of effect on U1(C , 1− L), need C and L to comove:

U2(C , 1− L) = PF1(L)U1(C , 1− L)

Change in P: essential; shifts labor demand

Other solutions:

1 internal habit formation: λt = f ( Ct︸︷︷︸−

,EtCt+1︸︷︷︸+

)

I can have Ct ↑, and still λt ↑, so Lt can ↑2 nominal rigidities: countercyclical markups:

U2(C , 1− L) =1

µF1(L)U1(C , 1− L)

I C ↑, Lsupply ↓, real wage ↑, markup over mg. cost µ ↓, labor demand ↑3 other shocks (eg. demand), act like TFP shocks (eg. Bai et al., 2012)


Sticky prices: Shift in supply and demand (markup ↓)

−0.2 0 0.2 0.4 0.6 0.8 1 1.2−0.2

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

LS (λt)LD (Z

t; µ

t)

LS (λt′ ↓ )

LD (Zt; µ

t′ ↓ )

Labor

Wage


Back to motivation: inflationary ’perception-driven’booms?

BP’s argument: standard New Keynesian model, ’perception’ shockswill be highly inflationary (act like ’demand’ shocks)

In the data: inflation roughly constant in last 3 cycles−→ not a priori plausible

their model, ’perception’ shock not inflationary because affectspotential output


Disinflationary ’perception-driven’ booms?

Christiano, Ilut, Motto, Rostagno (2008, 2010): news about TFP in areal and monetary model.

Document that inflation tends to be low during (stock market) booms

Table 3: Variables Over Various Sub-periods, 1919Q1-2010Q1Periods CPI Credit GNP Stock PriceBoom 1.8 5.3 4.6 13.8Other 4.0 2.3 0.2 -11.7

Whole period 2.7 4.0 2.7 2.7Notes: (1) numbers represent 100 tim es average of log rst d i¤erence of ind icated variab le over in -d icated p eriod .(2) Boom periods are the union of the trough to p eak p eriods enumerated in Tablexx.(3) O ther are p eriods that are not b oom s and that exclude World War II (1939Q4-1945Q4).Whole p eriod corresp onds to the fu ll sample, exclud ing World War II.

Table 4: Variables in Stock Market Boom EpisodesA. Non-boom, non-World War II, 1919Q1-2010Q1

CPI Credit GNP Stock Price4.0 2.3 0.2 -11.7

B. Boom episodestrough-peak CPI Credit GNP Stock Price

1921Q3-1929Q3 -0.2 5.7 5.9 19.31932Q2-1937Q2 0.6 -2.1 6.5 24.21949Q2-1968Q2 2.0 6.3 4.2 8.11982Q3-1987Q3 3.2 7.5 4.3 17.51994Q2-2000Q2 2.5 6.1 3.9 16.42003Q1-2007Q1 3.0 4.6 3.0 10.1Notes: (1) numbers represent 100 tim es average of log rst d i¤erence of ind icated variab le over ind i-cated p eriod . (2) panel A : data averaged over p eriod 1919Q1-2010Q1, sk ipp ing 1939Q4-1945Q4 andtrough to p eak years. (3) panel B : data averaged only over the ind icated trough-p eak years. (4) Panicso ccur after sto ck market p eak.

2


cli2

Oval

Disinflationary ’perception-driven’ booms

Monetary model: following positive news about future TFP

I positive comovement: C ,H, I and price of capital all ↑I inflation is low.

Why can inflation be low? Two counter effects

1 Standard, inflationary: because current marginal cost ↑2 Disinflationary: future marginal costs ↓, since expected TFP ↑


Disinflationary ’perception-driven’ booms

For the textbook CGG model, analytical results:

I perfect signal st about TFP

at+1 = ρat + ξt+1 + st

Closed form solution for output gap and inflation:

yt = ηyat + φy st ; πt = ηπat + φπst

1 Proposition: ηy , ηπ < 0

2 For wide parameterization: φy > 0, φπ < 0

Inside coefficient φπ : sum of the two counter effects

πt = λyt + βEt πt+1


5 10 15 20

0.20.40.60.8

11.2

perc

ent d

evia

tion

from

ss

consumption

Perturbed modelBaseline model

5 10 15 200

1

2

3

4

hours worked

perc

ent d

evia

tion

from

ss

5 10 15 202

2.2

2.4

2.6

2.8

net inflation (APR)

5 10 15 200

5

10

15

20

investment

perc

ent d

evia

tion

from

ss

5 10 15 200

2

4

output

perc

ent d

evia

tion

from

ss

5 10 15 20

6

8

10

12

Ann

ualiz

ed, p

erce

nt

net nominal rate of interest

5 10 15 20

-0.5

0

0.5

perc

ent d

evia

tion

from

ss price of capital

Figure 6: Response of Baseline and Perturbed Model to Signal Shock (Signal not realized); Perturbation = Ramsey

5 10 15 200

0.5

1

1.5

perc

ent d

evia

tion

from

ss labor productivity


Conclusion

Very rich paper.

Heterogeneity is important.

Labor market specialization can deliver many new insights:

1 comovement driven by perception

2 demand vs supply shocks

3 redistribution effects of fiscal policy


A Gains from Trade Perspective on Macroeconomic ...cli2/index_files/discussion_ilut...Discussion of...

Documents

Transcript of A Gains from Trade Perspective on Macroeconomic ...cli2/index_files/discussion_ilut...Discussion of...