Download Demand Imbalances, Exchange-Rate Misalignments, and Monetary Policy

Demand Imbalances, Exchange-Rate
Misalignments, and Monetary Policy
Giancarlo Corsetti
IUE, Roma III & CEPR
Luca Dedola
European Central Bank & CEPR
Sylvain Leduc
Federal Reserve Bank of San Francisco
Reserve Bank of New Zealand
December 16-17 2010
Exchange Rate misalignments and demand/trade
imbalances
I
Concerns about ER misalignments and associated imbalances
are recurrent in the policy debate — e.g. recent G-20 meeting
I
Motivation for cooperative (monetary) policy to mitigate them
I
Can misalignments in asset prices like the ER and ensuing
imbalances cause trade-o¤s with traditional stabilization
objectives?
I
NOEM literature has focused on the impact of nominal price
rigidities in goods markets:
I
I
Emphasize e¤ects of ER misalignments on relative price of
goods
What about misalignments arising from the ER dual role as
good and asset price in incomplete markets economies?
What does this paper do?
I
Understand trade-o¤s when asset markets support allocation
far from e¢ cient one
I
Standard incomplete markets model yielding misalignments
and demand imbalances even with ‡exible price
I
I
I
Misalignments re‡ect suboptimal wealth movements across
countries
Ine¢ cient adjustment, although ER is driven by fundamentals
only
Allowing for alternative pricing mechanisms (PCP vs LCP),
characterize optimal policy trade-o¤s
Monetary policy trade-o¤s with incomplete markets
I
Analytic quadratic loss functions under incomplete markets
I
Numerical analysis of Ramsey policy implications for in‡ation,
output gap, real exchange rates, net exports,...
I
Focus on "news” about future productivity to stress ER dual
role in goods and asset markets — But also look at standard
AR shocks
Main results
I
I
I
Welfare trade-o¤s between domestic as well as external
distortions:
I
Output gaps and in‡ation, but also misalignment and
cross-country demand (consumption) gaps
I
External distortions depend on openness, trade elasticities
Low trade elasticity:
I
Sizable ine¢ ciencies as currency misalignments and large
demand imbalances
I
Optimal policy delivers close to …rst-best allocation under LCP
Policy leans against misalignment, reducing domestic and
trade imbalances
Workhorse NOEM model
I
2-country, cashless economy with sticky prices
I
All goods are traded
I
Both producer currency pricing (PCP) and local currency
pricing (LCP)
I
No capital, technology linear in labor
I
Complete markets domestically, incomplete markets
internationally: only one short-term, risk-free, nominal bond
I
Encompass Clarida-Gali-Gertler, Obstfeld-Rogo¤,
Devereux-Engel, Benigno-Benigno, and other NOEM
contributions
Preferences
1 σ
+ κ (1
L t )1 + η
1 +η
I
Per-period utility: ζ C ,t C1t
I
Consumption consists of Home and Foreign tradables:
C =
CH,t
I
Z 1
0
"
φ 1
1/φ
φ
aH CH
σ
φ
φ 1 #φ 1
1/φ
φ
+ aF CF
θ
θ 1
θ 1
θ
Ct (h)
dh
Trade elasticity: φ
CF,t
φ>0
,
Z 1
0
Ct (f
θ 1
) θ df
θ 1
θ
Welfare-relevant gaps in open economy: ER misalignments
I
I
As in closed economies, output gap relative to …rst-best
allocation
bH ,t Y
eHfb,t .
Y gap = Y
International-price gaps provide welfare-based concept of
misalignments:
bt
RER gap = Q
TOT gap = Tbt
I
b H,t = Ebt + P
bH ,t
∆
e fb
Q
t
fb
e
Tt
bH ,t
P
A “misaligned” real exchange rate is one that is not equal to
its e¢ cient level:
I
No guarantee that e.g. ‡exible exchange rates will deliver an
e¢ cient real exchange rate
Welfare-relevant gaps in open economy: Demand
imbalances
I
Deviations from e¢ ciency condition under complete markets:
bt = σ C
bt
Q
b
ζ C ,t
bt
C
result in cross-country demand imbalances:
b tgap = σ(C
bt
D
I
bt )
C
bt
Q
b
ζ C ,t
b
ζ C ,t
b
ζ C ,t
?0
General feature of multi-agent economies, sounder foundations
than using current account or trade balance measures
Gaps enter the Open-Economy Phillips curves...
I
NKPC is a function of output gaps, markups, misalignment,
and imbalances ...:
π H ,t
I
I
= βEt π H ,t +1
8
>
>
>
>
>
>
>
<
+κ
(1
>
>
>
>
>
>
>
:
b gap
(η + σ) Y
H ,t
h
aH ) (σφ
(1
aH )
h
gap
b gap
1) Tbt
+Q
t
b H,t
∆
b tgap
D
b H,t =0 under PCP, but not under LPC
∆
b tgap =0 under complete markets, but not under IM
D
i
9
>
>
>
>
>
>
=
i >
>
>
>
>
>
>
>
;
... and characterize the loss function
under cooperation and LCP
1
b gap
f(σ + η ) Y
H ,t
2
∝
+
(1
2
+Ψ 4
θα
αβ) (1
(σφ
where Ψ =
α)
1) σ
4aH (1
2
b gap
+ (σ + η ) Y
F ,t
aH π 2H ,t + (1
bH ,t
Y
2
aH ) π 2F ,t + aH π F2,t + (1
e fb
Y
H ,t
bt
bt + D
φ ∆
2aH (1 aH )
aH ) (σφ 1) + 1
bF ,t
Y
2
e fb
Y
F ,t
2
+
3
5g
aH ) π H2,t
CM and PCP: Divine coincidence in open economy
Optimal policy in two targeting rules: global (sum)
h
i
b gap Y
b gap
b gap Y
b gap
0= Y
+
Y
+ θ π H ,t + π F ,t
H ,t
H ,t 1
F ,t
F ,t 1
and relative (di¤erence)
h
b gap Y
b gap
0= Y
H ,t
H ,t 1
b gap
Y
F ,t
b gap
Y
F ,t
1
i
+ θ π H ,t
π F ,t
I
International variables matter only if they in‡uence output
gap and in‡ation
I
Same as in closed economy: strict in‡ation targeting is
optimal against all – current and anticipated future – e¢ cient
shocks (preferences technology)
I
correcting output gaps coincides with closing misalignments
(as in DE)
But divine coincidence breaks with incomplete market
With …nancial autarky and PCP, global targeting rule is the same
as with CM, but relative rule is
h
i
b gap Y
b gap
b gap Y
b gap
0 =
Y
Y
(σ + η )
H ,t
H ,t 1
F ,t
F ,t 1
+θ (σ + η ) + 2
+4aH (1
+
I
( 1 aH ) ( 1
1 2aH (1
aH ) (σφ
1)
4aH (1 aH ) (φ 1)
1 2aH (1 φ)
σ)
φ)
gap
[t
TOT
b tgap
D
π H ,t
π F ,t
gap
1
[t
TOT
b gap
D
t 1
Trade-o¤ between stabilizing relative Y gap and π, and
gap
b tgap .
[ t and D
correcting TOT
Trade-o¤s and e¤ectiveness of optimal policy
I
By making the ‡exible price allocation ine¢ cient, incomplete
markets may induce relevant policy trade-o¤s
I
Relevance of misalignment/imbalances in trade-o¤s depends
on size of incomplete market distortions:
I
I
Structure of …nancial markets, nature and number of the
shocks, economic features like trade elasticity
E¤ectiveness of optimal policy also depends on type of
nominal distortions:
I
I
b gap and relative output all
With PCP terms of trade (RER), D
t
proportional
b t , but LOOP deviations
With LCP no longer true because of ∆
costly
Trade-o¤s and e¤ectiveness of optimal policy: PCP
I
With PCP, terms of trade (RER) and relative output are
proportional to each other:
Tbt =
bH ,t
Y
bF ,t
Y
bt =
D
bH ,t
Y
bF ,t
Y
1
1
2aH (1
φ)
σ (2aH φ 1) (2aH
1 2aH (1 φ)
1)
Trade-o¤s and e¤ectiveness of optimal policy: LCP
1
1 2aH (1
2aH φ
2aH (1 φ)
Tbt =
bH ,t
Y
bF ,t
Y
bt =
D
bH ,t
Y
bF ,t
Y
bt
∆
1
b t aH
+∆
2aH
φ)
σ (2aH φ 1) (2aH 1)
1 2aH (1 φ)
1 + φ (2σ (1 aH ) 1)
1 2aH (1 φ)
Technology: News shocks vs Autoregressive AR
Values of key parameters
I
θ = 0.75
I
θ = 6 : implying a 20% markup
I
aH set to generate a 10% import share
I
κ set so that households spend 1/3 of their time working
I
φ = f0.45, 6g
Understanding misalignment and demand imbalances:
‡ex-price allocation and CM
Anticipated Home productivity increase with High trade elasticity / good substitutability
Understanding misalignment and demand imbalances:
‡ex-price allocation, gaps between CM and IM
Anticipated Home productivity increase with High trade elasticity / good substitutability
Understanding misalignment and demand imbalances:
larger gaps with low elasticity
Anticipated Home productivity increase with low trade elasticity / good complementarity
Large imbalances: role of low trade elasticity
I
Following an increase in Home tradable output, with a low
trade elasticity, income e¤ects are large
I
If TOT worsen (as in the e¢ cient case):
I
I
Home demand falls (large negative income e¤ect)
With enough home bias: world demand for Home goods falls
I
TOT must improve to bring the goods market in equilibrium
(and RER appreciates)
I
These movements amplify cross-country (relative) demand
imbalances
Optimal cooperative policy
I
Jointly maximize welfare under commitment and timeless
perspective
I
High elasticity: near divine coincidence, PCP versus LCP
I
Low elasticity: optimal policy corrects misalignment and
imbalances
I
I
moderate success (short run) with PCP
higher degree of success with LCP
‘Near divine coincidence’with high elasticity and PCP
Optimal policy arbitrarily close to strict in‡ation targeting
(‡ex-price)
Low trade elasticity and PCP
Correcting Dgap via consumption: e¤ective in the short-run
Low trade elasticity and PCP
Correcting Dgap via consumption: e¤ective in the short-run
Low trade elasticity and LCP
Leaning against misalignment has persistent e¤ects on imbalances
Optimal policy with low trade elasticity and LCP
Considerable improvement over ‡ex-price allocation: 0.36% of
steady state consumption
Anticipated AR shocks: low trade elasticity and LCP
Conclusion
I
Asset-price misalignments pose relevant trade-o¤s for
monetary policy when they lead to wealth and demand
imbalances even for reasons independent of nominal frictions
I
Optimal policy should generally use monetary stance to
redress misalignments in these cases
I
The success of optimal monetary policy in doing so depends
on the structure of the economy, including the form of
nominal rigidities – PCP vs. LCP in our case
Conclusions
Further exercises/directions:
I
Implementation
I
Domestically and international incomplete markets
I
Cooperative vs Nash
Conclusions
I
Misalignments and wealth/demand imbalances —
ine¢ ciencies that generally arise independently of monetary
and nominal distortions — raise relevant policy trade-o¤s
I
Examples of economies where inward-looking strict in‡ation
targeting, rather than correcting, results in signi…cant
misalignments
I
I
Even when the latter only re‡ects fundamental-based
valuations
Generate suboptimal demand and current account imbalances
I
Monetary policy should address misalignments in the foreign
exchange market
I
Degree of success of optimal monetary policy depends on the
structure of the economy, including the form of nominal
rigidities – PCP vs. LCP in our case
Conclusions
Further exercises/directions:
I
Implementation
I
Domestically and international incomplete markets
I
Cooperative vs Nash
Welfare and optimal (cooperative) policy
I
Planner chooses world allocation that maximizes:
Welfare =
W0 + W0
2
subject to households and …rms FOCs and resource constraints
I
Commitment and timeless perspective
Experiments: focus on news shocks
Linear technology with standard (ξ) shocks and news (ξ ν ) shocks:
Yt (h) = Zt Lt (h) ,
Ut = 0.95Ut
Zt = Ut Vt
1
Vt = 0.9Vt
I
+ ξ t + ξ tν
1
+ ξ tν
Focus on a positive shock in the Home country
Households
V (st ) = max
C H ,C F,
L,B H,t +1
ζ C ,t
Ct1 σ
(1 Lt )1
+κ
1 σ
1 η
η
+ β Ct , Lt EV (st +1 )
subject to:
PH,t CH,t + PF,t CF,t + BH,t +1
Wt Lt + (1 + it )BH,t +
Z 1
0
Π(h)dh
BH,t : risk-free nominal bond that pays in unit of domestic currency
‘Near divine coincidence’with high elasticity and LCP
Some correction, but ine¢ cient gaps remain open