42
DSGE Models for Monetary Policy Lawrence Christiano

DSGE Models for Monetary Policy

  • Upload
    others

  • View
    8

  • Download
    1

Embed Size (px)

Citation preview

Page 1: DSGE Models for Monetary Policy

DSGE Models for Monetary Policy

Lawrence Christiano

Page 2: DSGE Models for Monetary Policy

Outline• Basic New Keynesian (NK) model

– Two fundamental frictions in the model.

• Implications of NK model for monetary policy:– Clarifying the concepts of ‘excess and inadequate aggregate demand’.

– The Taylor principle and inflation targeting.– Cases where ‘overzealous inflation targeting’ can go awry:

• News shocks and the relationship between monetary policy and stock market volatility

• The working capital channel and the Taylor principle.

Page 3: DSGE Models for Monetary Policy

Outline, cnt’d• Adjustments to NK model to enable it to account for empirical estimates of the monetary transmission mechanism (the ‘David Hume puzzle’).

• Costly State Verification model of financial frictions: microeconomics.

• Introducing csv model into otherwise standard NK DSGE model– The importance of a ‘risk shock’ in economic fluctuations.– Using the model to assess a monetary policy that reacts to the interest rate premium.

Page 4: DSGE Models for Monetary Policy

Basic New Keynesian Model• Description of the economic environment

– Preferences and technology.– The best possible allocation of labor and consumption, regardless of how they are brought about.

• How ideally functioning markets can bring about the best allocations (the ‘invisible hand’).

• What happens with markets that resemble actual real world markets where prices don’t adjust instantly.

Page 5: DSGE Models for Monetary Policy

Model

• Household preferences:

E0∑t0

t logCt − exp tNt

1

1 ,

t t−1 t, t~iidN0,2

Page 6: DSGE Models for Monetary Policy

Production• Final output requires lots of intermediate inputs:

• Production of intermediate inputs:

• Constraint on allocation of labor:

0

1Nitdi Nt

Yt 0

1Yi,t

−1 di

−1

, 1

Yi,t eatNi,t, Δat Δat−1 ta, ta~iidN0,a2

Page 7: DSGE Models for Monetary Policy

Efficient Allocation of Total Labor• Suppose total labor,      , is fixed.

• What is the best way to allocate       among the various activities,                ? 

• Answer: – allocate labor equally across all the activities

0 ≤ i ≤ 1

Nit Nt, all i

Nt

Nt

Page 8: DSGE Models for Monetary Policy

Suppose Labor Not Allocated Equally

• Example:

• Note that this is a particular distribution of labor across activities:

Nit 2Nt i ∈ 0, 1

2

21 − Nt i ∈ 12 , 1

, 0 ≤ ≤ 1.

0

1Nitdi 1

2 2Nt 12 21 − Nt Nt

Page 9: DSGE Models for Monetary Policy

Labor Not Allocated Equally, cnt’d

Yt 0

1Yi,t

−1 di

−1

0

12 Yi,t

−1 di

12

1Yi,t

−1 di

−1

eat 0

12 Ni,t

−1 di

12

1Ni,t

−1 di

−1

eat 0

12 2Nt

−1 di

12

121 − Nt

−1 di

−1

eatNt 0

12 2

−1 di

12

121 −

−1 di

−1

eatNt 12 2

−1 1

2 21 − −1

−1

eatNtf

Page 10: DSGE Models for Monetary Policy

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

0.88

0.9

0.92

0.94

0.96

0.98

1

f

Efficient Resource Allocation Means Equal Labor Across All Sectors

6

10

f 12 2

−1 1

2 21 − −1

−1

Page 11: DSGE Models for Monetary Policy

Economy with Efficient N Allocation• Efficiency dictates

• So,  with efficient production:

• Resource constraint:

• Preferences:

Yt eatNt

Ct ≤ Yt

E0∑t0

logCt − exp tNt

1

1 , t t−1 t, t~iid,

Nit Nt all i

Page 12: DSGE Models for Monetary Policy

Efficient Determination of Labor• Lagrangian:

• First order conditions:

• or:

maxCt ,Nt

E0∑t0

t

logCt−exp t Nt

1

1

uCt,Nt, t teatNt − Ct

ucCt,Nt, t t, unCt,Nt, t teat 0

un,t uc,teat 0

marginal cost of labor in consumption units− dudNtdudCt

dCtdNt

−un,tuc,t

marginal product of laboreat

Page 13: DSGE Models for Monetary Policy

Efficient Determination of Labor, cont’d• Solving the fonc’s:

• Note:– Labor responds to preference shock, not to tech shock

−un,tuc,t eat

Ct exp tNt eat

eatNt exp tNt eat

→ Nt exp − t1

→ Ct exp at − t1

Page 14: DSGE Models for Monetary Policy

Response to a Jump in a

Leisure, 1-N

consumptionHigher a

Indifference curve

Production frontier

Page 15: DSGE Models for Monetary Policy

Case Where Markets Work Beautifully(triumph of the ‘invisible hand’)

• Give households budget constraints, put them in markets and let them pursue their individual interests.

• Give the production functions to firms and suppose that they seek to maximize profits.

• There is monopoly power….extinguish the effects of that by a subsidy.

• Let markets work perfectly: prices and wages adjust instantly all the time, to clear markets.

Page 16: DSGE Models for Monetary Policy

Households• Solve:

• Subject to:

• First order conditions:

maxE0∑t0

t logCt − exp tNt

1

1 ,

Ct bonds purchases in t

Bt1 ≤

wage ratewt Nt

profits t

(real) interest on bondsrt−1 Bt

−un,tuc,t Ct exp tNt

wt ‘marginal cost of working equals marginal benefit’

uc,t Etuc,t1rt ‘marginal cost of saving equals marginal benefit’

Profits, net of government taxes

Page 17: DSGE Models for Monetary Policy

Final Good Firms• Final good firms buy       ,               ,  at given prices,        , to maximize profits:

• Subject to

• Fonc’s:   

Yt 0

1 Yi,t−1 di

−1

Yi,t i ∈ 0,1Pi,t

Yt − 0

1Pi,tYi,tdi

Pi,t YtYi,t

1

→ Yi,t Pi,t−Yt, 1 0

1Pi,t1−di

Page 18: DSGE Models for Monetary Policy

Intermediate Good Firms• For each         there is a single producer who is a monopolist in the product market and hires labor,       in competitive labor markets.  

• Marginal cost of production:

• Subsidy will be required to ensure markets work efficiently.

Yi,tNi,t

(real) marginal costst dCostdwor kerdoutputdwor ker

1 −subsidy payment to firm

wt

expat

Page 19: DSGE Models for Monetary Policy

Intermediate Good Firms

Yi,t

Pi,t Yi,t Pi,t−Yt

Marginal revenue

Marginal cost,

Demand curve

st

Page 20: DSGE Models for Monetary Policy

ith Intermediate Good Firm• Problem:

• Subject to demand for       :

• Problem:

• Note: all prices are the same, so resources allocated efficiently across intermediate good firms.

Yi,t Yi,t Pi,t−Yt

maxNit PitYit − stYit

maxNitPitPi,t−Yt − stPi,t−Yt

fonc : 1 − Pit−Yt stPi,t−−1Yt 0

Pit − 1 st ‘price is markup over marginal cost’

Pi,t Pj,t 1, because 1 0

1Pi,t1−di

Page 21: DSGE Models for Monetary Policy

Equilibrium• Pulling things together:

• If proper subsidy is provided to monopolists, employment is efficient:

1 − 1 st

− 1

1 − wtexpat

household fonc 1 −

− 1

−un,tuc,t

expatif 1−

−1

−un,tuc,t

expat.

if 1 − − 1 , then −un,tuc,t expat

=1

Page 22: DSGE Models for Monetary Policy

Equilibrium Allocations• With efficient subsidy,

• Bond market clearing implies:

Bt 0 always

−un,tuc,t

functional form Ct exp tNt

resource constraint

expatexp tNt1 expat

→ Nt exp − t1

Ct eatNt exp at − t1

Page 23: DSGE Models for Monetary Policy

Interest Rate in Equilibrium• Interest rate backed out of household intertemporal Euler equation:uc,t Etuc,t1rt → 1

Ct Et 1

Ct1rt

→ rt 1Et CtCt1

1Et expct − ct1

1Et exp at − at1 − t−t1

1

1exp Et −Δat1 − t−t1

1 12 V

, Vt a2 11

22

logrt − log Et

ct1−ct

Δat1 − t1 − t1 1

2 V

using assumptions about Δat and t − log Δat − −1t

1 12 V

Page 24: DSGE Models for Monetary Policy

Dynamic Properties of the Model

2 4 6 8 10 12 14 16 18 20 22 240

0.005

0.01

0.015

0.02

0.025

0.03

0.035

0.04

Response to .01 Technology Shock in Period 1

(at - at-1) = 0.75(at-1 - at-2) + epsatlog

Ct

t

2 4 6 8 10 12 14 16 18 20 22 24

0.008

0.009

0.01

0.011

0.012

0.013

0.014

interest rate

log,

r t

t

log rt = -log + 0.75at-1

2 4 6 8 10 12 14 16 18 20 22 24-5

-4.5

-4

-3.5

-3

-2.5

-2

-1.5

-1

-0.5

x 10-3

log

Ct

t

t = 0.5t-1 + epstaut

Response to .01 Preference Shock in Period 1

2 4 6 8 10 12 14 16 18 20 22 24

7.5

8

8.5

9

9.5

x 10-3

log,

r t

t

log rt = -log - (0.5 - 1)t/(1 + 1)

interest rate

logrt − log 0.75Δat

Page 25: DSGE Models for Monetary Policy

Key Features of Equilibrium Allocations• Allocations efficient (as long as monopoly power neutralized)

• Employment does not respond to technology– Improvement in technology raises marginal product of labor and marginal cost of labor by same amount.

• First best consumption not a function of intertemporal considerations– Discount rate irrelevant.– Anticipated future values of shocks irrelevant.

• Natural rate of interest steers consumption and employment towards their natural levels, ‘as if guided by an invisible hand’.

Page 26: DSGE Models for Monetary Policy

Introducing Price Setting Frictions(Clarida‐Gali‐Gertler Model)

• Households maximize:

• Subject to:

• Intratemporal first order condition:

E0∑t0

logCt − exp tNt

1

1 , t t−1 t, t~iid,

PtCt Bt1 ≤ WtNt Rt−1Bt Tt

Ct exp tNt Wt

Pt

Profits, net of taxes raised byGovernment to finance subsidies.

Page 27: DSGE Models for Monetary Policy

Household Intertemporal FONC• Condition:

– or

– take log of both sides:

– or

1 Etuc,t1uc,t

Rt1 t1

1 Et CtCt1

Rt1 t1

Et explogRt − log1 t1 − Δct1

≃ explogRt − Et t1 − EtΔct1 , ct ≡ logCt

0 log rt − Et t1 − EtΔct1, rt logRt

ct − log − rt − Et t1 ct1

Page 28: DSGE Models for Monetary Policy

NK IS Curve• Euler equation in two equilibria:

• Subtract: Output gap

xt −rt − Et t1 − rt∗ Etxt1

Actual equilibrium: yt −rt − Et t1 − rr Etyt1

Natural equilibrium: yt∗ −rt∗ − rr Etyt1∗

Output equals consumption

Page 29: DSGE Models for Monetary Policy

Final Good Firms• Buy                     at prices       and sell      for • Take all prices as given (competitive)• Profits:

• Production function:

• First order condition:    

Yt 0

1Yi,t

−1

−1di, 1,

PtYt − 01 Pi,tYi,tdi

Yi,t, i ∈ 0,1 Pi,t Yt Pt

→ Pt 0

1Pi,t1−di

11−Yi,t Yt Pi,t

Pt

Page 30: DSGE Models for Monetary Policy

Intermediate Good Firms• Each ith good produced by a single monopoly producer. 

• Demand curve:

• Technology:

• Calvo Price‐setting Friction

Yi,t expatNi,t, Δat Δat−1 ta,

Yi,t YtPi,tPt

Pi,t P̃t with probability 1 − Pi,t−1 with probability

Page 31: DSGE Models for Monetary Policy

Marginal Cost

real marginal cost st dCostdwor kerdOutputdwor ker

1 − Wt/Ptexpat

−1 in efficient setting

1 − Ct exp tNt

expat

Page 32: DSGE Models for Monetary Policy

Marginal Cost• Marginal cost:

• Then,

st 1 fCt exp tNt

expatCtexpat Nt holds only as a first order approximation

≃1fCt exp t Ct

expat

expat

1 fCt

1 exp t1

expat

1

1 f

CtCt∗

1

ŝta hat indicates log-deviation from steady state

≡ logst f 1 ct − ct∗ 1 xt

1f

− 1 ~ tax subsidy, 1 − , to ‘trick’ firms into

setting price equal to marginal cost

Page 33: DSGE Models for Monetary Policy

Marginal Cost• Marginal cost:

• Then,

st 1 fCt exp tNt

expatCtexpat Nt holds only as a first order approximation

≃1fCt exp t Ct

expat

expat

1 fCt

1 exp t1

expat

1

1 f

CtCt∗

1

ŝta hat indicates log-deviation from steady state

≡ logst f 1 ct − ct∗ 1 xt

=1/s

Page 34: DSGE Models for Monetary Policy

The Intermediate Firm’s Decisions• ith firm is required to satisfy whatever demand shows up at its posted price.

• Its only real decision is to adjust price whenever the opportunity arises.

• It sets its price as a function of current and expected future marginal cost.– Inflation evolves (to first order approximation) as follows:

t 1 − 1 −

ŝt Etŝt1 2Etŝt2 . . .

ŝt ≡ st − ss ≃ log sts , t Pt − Pt−1

Pt−1

Page 35: DSGE Models for Monetary Policy

NK Phillips Curve• Repeating from previous slide:

• Also,

• Then,

t 1−1−

ŝt Etŝt1 2Etŝt2 . . .

Et t1 1 − 1 −

Etŝt1 2Etŝt2 3Etŝt3 . . .

t Et t1 1 − 1 −

ŝt

Page 36: DSGE Models for Monetary Policy

Back to the Phillips Curve• Phillips curve summarizes price setting by intermediate good firms:

• or, substituting out for marginal cost:

t Et t1 1 − 1 −

ŝt

t Et t1 1−1−

1 xt

Page 37: DSGE Models for Monetary Policy

Equations of Actual Equilibrium, with Monetary Policy Rule

t Et t1 xt (Calvo pricing equation)

xt −rt − Et t1 − rt∗ Etxt1 (intertemporal equation)

rt rt−1 1 − t xxt ut (policy rule)

rt∗ Etyt1∗ − yt∗ Et Δat1 − 1

1 Δ t1 (natural rate)

yt∗ at − 11 t (natural output), xt yt − yt∗ (output gap)

also, time series representations for shocks listed above

Page 38: DSGE Models for Monetary Policy

Solving the Model• Express the equations in matrix form:

• Solution: A and Bmatrices such that (*) is satisfied and

zt Azt−1 Bst

zt

txtrtrt∗

, st tatut

, t t

ta

tu

∗ 0Etzt1 1zt 2zt−1 0Etst1 1st 0st Pst−1 t

Page 39: DSGE Models for Monetary Policy

Solving the model….• Iterating forward one period on the solution:

• Then,

zt1 Azt Bst1

A2zt−1 ABst Bst1

Et0zt1 1zt 2zt−1 0st1 1st Et0A2zt−1 ABst Bst1 1Azt−1 Bst 2zt−1 0st1 1st 0A2 1A 2 zt−1 Et0B 0 st1 0AB 1B 1st

0

0A2 1A 2 zt−1 0

0B 0P 0AB 1B 1 st 0, all st, zt−1

Finding solution for A is only hard part. Choose A that solves problem if there is only onewith eigenvalues less than unity in absolute value.If more than one A like this: multiple solutions.

Page 40: DSGE Models for Monetary Policy

Impulse Response Function• Law of motion of endogenous and exogenous variables:

• Set one element of      non‐zero and let  

• Set                            and compute

zt Azt−1 Bstst Pst−1 t

0

t 0, t 0s−1 z−1 0

s0, s1, s2, . . .z0, z1, z2, . . .

Page 41: DSGE Models for Monetary Policy

1 2 3 4 5 6 7

0.01

0.02

0.03

0.04

Inflation

1 2 3 4 5 6 7

0.05

0.1

0.15

Ouput Gap

1 2 3 4 5 6 7

0.05

0.1

0.15

0.2Nominal Rate

Natural Nominal Rate

Actual Nominal Rate

1 2 3 4 5 6 7

0.05

0.1

0.15

0.2Real Rate

NaturalActual

1 2 3 4 5 6 7

1.05

1.1

1.15

1.2

log Technology

1 2 3 4 5 6 7

1.05

1.1

1.15

1.2

Output

Natural OutputActual Output

1 2 3 4 5 6 70

0.05

0.1

0.15

Employment

Natural EmploymentActual Employment

Cost of workingBenefit of working

Dynamic Response to a Technology Shock

1 2 3 4 5 6 7

1.1

1.2

1.3

x 0, 1.5, 0.97, 1, 0.2, 0.75, 0, 0.5.

Page 42: DSGE Models for Monetary Policy

1 2 3 4 5 6 7

0.02

0.04

0.06

0.08

0.1Inflation

1 2 3 4 5 6 7

0.05

0.1

0.15

0.2

0.25

Ouput Gap

1 2 3 4 5 6 7

0.05

0.1

0.15

0.2

0.25Nominal Rate

Natural Nominal RateActual Nominal Rate

1 2 3 4 5 6 7

0.05

0.1

0.15

0.2

0.25Real Rate

NaturalActual

1 2 3 4 5 6 7

0.2

0.4

0.6

0.8

1Preference Shock

1 2 3 4 5 6 7-0.5

-0.4

-0.3

-0.2

-0.1

Output

Natural OutputActual Output

1 2 3 4 5 6 7-0.5

-0.4

-0.3

-0.2

-0.1

Employment

Natural EmploymentActual Employment

Cost of workingBenefit of working

Dynamic Response to a Preference Shock

1 2 3 4 5 6 70

0.1

0.2

0.3

0.4

0.5