the extensive margin of trade and monetary policy · 2016-11-08 · the extensive margin of trade...
TRANSCRIPT
The Extensive Margin of Trade and Monetary Policy
Yuko Imura Malik Shukayev
Bank of Canada University of Alberta
June 20, 2016
The views expressed in this presentation are our own, and do not represent those of the Bank of Canada.
Exporter dynamics: long-run versus short-run
Persistence in long-run export participation
Export continuation rate = 87.4%/yr
Non-participation rate = 86.1%/yr
(U.S. manufactures 1984-1992, Bernard and Jensen, 2004)
Little evidence of growth in extensive margin of trade amongadvanced economies (Kehoe and Ruhl, 2013)
Different picture at the business cycle frequency (Naknoi, 2015)
Extensive margin of exports is three times as volatile as output
Imura (BoC) and Shukayev (UAlberta) Extensive margin of trade June 2016 1 / 17
New dimension for monetary policy and trade
The high volatility of the extensive margin over business cycles raisesnew questions for policy makers of open economies.
What are the channels through which monetary policy mightaffect intensive and extensive margins of international trade?
Does monetary policy affect the two margins in the same way?
Imura (BoC) and Shukayev (UAlberta) Extensive margin of trade June 2016 2 / 17
This paper
develop a two-country DSGE model with
nominal rigidities
state-dependent decisions on whether to enter/exit export market
firm heterogeneity in productivity, export costs, prices
calibrate the model to match micro-level exporter characteristics
examine the effects of monetary policy on the intensive andextensive margins of trade
Imura (BoC) and Shukayev (UAlberta) Extensive margin of trade June 2016 3 / 17
Main Findings
Exporter entry/exit is sensitive to firms’ price competitivenessrelative to other exporters and firms in the destination market
Expansionary monetary policy shocks support the intensivemargin of trade, but can deter export participation.
Currency depreciation and lower interest rates are favorable toexport sales and export participation.
However, higher expected inflation discourages exportparticipation
Monetary policy that is more aggressive toward inflation reducesfluctuations in export participation.
Imura (BoC) and Shukayev (UAlberta) Extensive margin of trade June 2016 4 / 17
Related Literature
Export hysteresis in partial equilibrium
Baldwin (1988), Baldwin and Krugman (1989), Dixit (1989)
Firm heterogeneity and export decisions
Melitz (2003), Bernard et al. (2003), Das et al. (2007),Chaney (2008)
Business cycles and exporter entry/exit in general equilibrium
Ghironi and Melitz (2005), Alessandria and Choi (2007),Ruhl (2008), Imura (2016)
Optimal monetary policy with exporter entry and exit
Cooke (2015)
Imura (BoC) and Shukayev (UAlberta) Extensive margin of trade June 2016 5 / 17
Model overview
Two symmetric countries, each with
Representative household
maxEt
∞∑t=0
βt [εct logCt + χ2(1− Lt)]
Competitive final-good producers
maxyHt (i),yFt (i)
PtDt −∫ 1
0
PDt (i)yHt (i)di−∫i∈Θt
PX∗t (i)yFt (i)di
Monopolistically competitive intermediate-good producers
Probability of price adjustment increasing in the age of price
Entry and exit in the export market, subject toentry/continuation costs
Monetary authority
ipt = ρiipt−1 + (1− ρi)
[φππt + φY Yt + φQQt
]+ µt
Imura (BoC) and Shukayev (UAlberta) Extensive margin of trade June 2016 6 / 17
Intermediate-good firms
Each producing a differentiated product
yt(i) = zt(i)AtKt(i)νLt(i)
1−ν
zt(i) = firm-specific productivity, At = country-specific productivity
All intermediate-good producers sell in their own country.
Export participation
To enter the export market, a firm pays entry cost, η ∼ GE(η).
Upon entering, an entrant sets a new price for its exports.
To continue exporting, a firm pays continuation cost, ξ ∼ G(ξ).
All export costs are paid in advance of production.
Probability of price adjustment increases as price gets older
Imura (BoC) and Shukayev (UAlberta) Extensive margin of trade June 2016 7 / 17
Potential entrant
Potential entrant with productivity zc and entry cost η solves
V Et (zc, η) = max
{maxPX0,t(zc)
[QtPX0,t(zc)
P ∗tyX0,t(zc)− wtLX0,t(zc)− rtKX
0,t(zc)− iptηwt
+βEtλt+1
λt
nz∑c=1
πccH1,t+1
(zc, zc, ξ
′) ], βEtλt+1
λt
nz∑c=1
πccVEt+1(zc, η
′)
}
where
H1,t (zc, zs, ξ) = α1V0,t(zc, ξ) + (1− α1)V1,t (zc, zs, ξ)
⇒ Maximum entry cost this firm would pay to start exporting, ηEt (zc),equates the value of entry and the value of no entry.
Imura (BoC) and Shukayev (UAlberta) Extensive margin of trade June 2016 8 / 17
Price-adjusting incumbent
Price-adjusting incumbent exporter with current productivity zcdrawing export cost ξ solves
V0,t(zc, ξ) = max
{maxPX0,t(zc)
[QtPX0,t(zc)
P ∗tyX0,t(zc)− wtLX0,t(zc)− rtKX
0,t(zc)− ipt ξwt
+βEtλt+1
λt
nz∑c=1
πccH1,t+1
(zc, zc, ξ
′) ], βEtλt+1
λt
nz∑c=1
πccVEt+1(zc, η
′)
}
⇒ Max export cost this firm would pay to continue exporting, ξ0t (zc),equates the value of continuation and the value of exit.
Imura (BoC) and Shukayev (UAlberta) Extensive margin of trade June 2016 9 / 17
Non-price-adjusting incumbent
Value of non-price-adjusting incumbent of type (zc, j, zs) drawingcontinuation cost ξ
Vj,t (zc, zs, ξ) = max
[QtPXj,t(zs)
PXtyXj,t(zc, zs)− wtLXj,t(zc, zs)− rtKX
j,t(zc, zs)− ipt ξwt
+βEtλt+1
λt
nz∑c=1
πccHj+1,t+1
(zc, zs, ξ
′) , βEtλt+1
λt
nz∑c=1
πccVEt+1(zc, η
′)
]
⇒ Maximum export cost for this firm to continue exporting, ξjt (zc, zs),equates the value of continuation and the value of exit.
Imura (BoC) and Shukayev (UAlberta) Extensive margin of trade June 2016 10 / 17
Export participation decisions
Export decisions depend directly on
V Et (zc, η) = max
{maxPX0,t(zc)
[Qt
PX0,t(zc)
P ∗tyX0,t(zc)− wtLX0,t(zc)− rtKX
0,t(zc)− ipt ηwt
+βEtλt+1
λt
nz∑c=1
πccH1,t+1
(zc, zc, ξ
′) ], βEtλt+1
λt
nz∑c=1
πccVEt+1(zc, η
′)
}
Exchange rate
Relative export price
Production costs
Interest rate (export cost)
Demand for home exports, yH∗t (i) = (1− ω)ρ(PXt (i)
PXt
)−γ (PXtP∗t
)−ρD∗t
Imura (BoC) and Shukayev (UAlberta) Extensive margin of trade June 2016 11 / 17
Export participation decisions
Export decisions depend directly on
V Et (zc, η) = max
{maxPX0,t(zc)
[Qt
PX0,t(zc)
P∗t
yX0,t(zc)− wtLX0,t(zc)− rtKX0,t(zc)− ipt ηwt
+βEtλt+1
λt
nz∑c=1
πccH1,t+1
(zc, zc, ξ
′) ], βEtλt+1
λt
nz∑c=1
πccVEt+1(zc, η
′)
}
Exchange rate
Relative export price
Production costs
Interest rate (export cost)
Demand for home exports, yH∗t (i) = (1− ω)ρ(PXt (i)
PXt
)−γ (PXtP∗t
)−ρD∗t
Imura (BoC) and Shukayev (UAlberta) Extensive margin of trade June 2016 11 / 17
Export participation decisions
Export decisions depend directly on
V Et (zc, η) = max
{maxPX0,t(zc)
[QtPX0,t(zc)
P ∗tyX0,t(zc)−wtL
X0,t(zc)− rtKX
0,t(zc)− ipt ηwt
+βEtλt+1
λt
nz∑c=1
πccH1,t+1
(zc, zc, ξ
′) ], βEtλt+1
λt
nz∑c=1
πccVEt+1(zc, η
′)
}
Exchange rate
Relative export price
Production costs
Interest rate (export cost)
Demand for home exports, yH∗t (i) = (1− ω)ρ(PXt (i)
PXt
)−γ (PXtP∗t
)−ρD∗t
Imura (BoC) and Shukayev (UAlberta) Extensive margin of trade June 2016 11 / 17
Export participation decisions
Export decisions depend directly on
V Et (zc, η) = max
{maxPX0,t(zc)
[QtPX0,t(zc)
P ∗tyX0,t(zc)− wtLX0,t(zc)− rtKX
0,t(zc)− ipt ηwt
+βEtλt+1
λt
nz∑c=1
πccH1,t+1
(zc, zc, ξ
′) ], βEtλt+1
λt
nz∑c=1
πccVEt+1(zc, η
′)
}
Exchange rate
Relative export price
Production costs
Interest rate (export cost)
Demand for home exports, yH∗t (i) = (1− ω)ρ(PXt (i)
PXt
)−γ (PXtP∗t
)−ρD∗t
Imura (BoC) and Shukayev (UAlberta) Extensive margin of trade June 2016 11 / 17
Export participation decisions
Export decisions depend directly on
V Et (zc, η) = max
{maxPX0,t(zc)
[QtPX0,t(zc)
P ∗tyX0,t(zc)− wtLX0,t(zc)− rtKX
0,t(zc)− ipt ηwt
+βEtλt+1
λt
nz∑c=1
πccH1,t+1
(zc, zc, ξ
′) ], βEtλt+1
λt
nz∑c=1
πccVEt+1(zc, η
′)
}
Exchange rate
Relative export price
Production costs
Interest rate (export cost)
Demand for home exports, yH∗t (i) = (1− ω)ρ(PXt (i)
PXt
)−γ (PXtP∗t
)−ρD∗t
Imura (BoC) and Shukayev (UAlberta) Extensive margin of trade June 2016 11 / 17
Calibration
Home bias, ω
Entry costs, U (0, ηU )
Continuation costs, U (0, ξU )
Price adjustment probabilities, αj
Firm-specific productivity process
log z′ = ρz log z + ε′, ε ∼ N(0, σε)
Data Model
Mass of exporters 0.21 0.23 Bernard et al. (2003)
Continuation rate 0.97 0.87 Bernard & Jensen (2004)
Entry rate 0.04 0.04 Bernard & Jensen (2004)
Imports/GDP 0.12 0.13 Drozd & Nosal (2011)
Productivity relative to 1.12-18 1.13 Bernard & Jensen (1999)nonexporters
Mean price adjustment 1.07-3.27 2.66 Bils & Klenow (2004)frequency (qtr) Nakamura & Steinsson (2008)
Imura (BoC) and Shukayev (UAlberta) Extensive margin of trade June 2016 12 / 17
Negative home aggregate TFP shock
0 5 10 15 20-1
-0.5
0Exports (H)
% d
evia
tion
0 5 10 15 20-3
-2
-1
0Mass of exporters (H)
0 5 10 15 200
0.5
1
Export price index (H)
0 5 10 15 20-0.4
-0.2
0
0.2
Real exchange rate
% d
evia
tion
0 5 10 15 200
0.2
0.4
0.6Interest rate (H)
0 5 10 15 20-0.8
-0.6
-0.4
-0.2
Consumption (F)
0 5 10 15 20-1
-0.5
0
% d
evia
tion
TFP shock (H)TFP shock (F)
0 5 10 15 20-1
-0.5
0
0.5
1
% d
evia
tion
Monetary policy shock (H)Monetary policy shock (F)
0 5 10 15 20-1
-0.5
0
0.5
1
% d
evia
tion
Demand shock (H)Demand shock (F)
Imura (BoC) and Shukayev (UAlberta) Extensive margin of trade June 2016 13 / 17
Negative aggregate TFP shock and monetary stimulus
0 5 10 15 20-1
-0.5
0 Exports (H)
% d
evia
tion
0 5 10 15 20-6
-4
-2
0 Mass of exporters (H)
0 5 10 15 200
1
2
Export price index (H)
0 5 10 15 20
0
1
2 Real exchange rate
% d
evia
tion
0 5 10 15 20
0
0.2
0.4
0.6Interest rate (H)
0 5 10 15 20-0.8
-0.6
-0.4
-0.2
Consumption (F)
0 5 10 15 20-1
-0.5
0TFP shock (H)
0 5 10 15 20-1
-0.5
0
% d
evia
tion
Monetary policy shock (H)
TFP shockTFP shock + Monetary stimulus
Monetary stimulus to counter negative TFP shock is effective onintensive margin, but worsens initial decline in extensive margin.
Imura (BoC) and Shukayev (UAlberta) Extensive margin of trade June 2016 14 / 17
Monetary policy rule and extensive margin of trade
0 5 10 15 20-1
-0.5
0 Exports (H)
% d
evia
tion
0 5 10 15 20-3
-2
-1
0 Mass of exporters (H)
0 5 10 15 200
0.5
1
Export price index (H)
0 5 10 15 20-0.5
0
0.5 Real exchange rate
% d
evia
tion
0 5 10 15 200
0.2
0.4
0.6Interest rate (H)
0 5 10 15 20
-0.2
-0.1
0 Consumption (F)
Benchmark (=2)
More aggressive on inflation stabilization (=4)
Response of extensive margin is dampened when monetary policy ismore aggressive on inflation
Imura (BoC) and Shukayev (UAlberta) Extensive margin of trade June 2016 15 / 17
Without extensive margin adjustment
0 5 10 15 20-1
-0.5
0 Exports (H)
% d
evia
tion
0 5 10 15 20-1
0
1 Mass of exporters (H)
0 5 10 15 200
0.5
1
Export price index (H)
0 5 10 15 20-0.5
0
0.5 Real exchange rate
% d
evia
tion
0 5 10 15 200
0.2
0.4
0.6Interest rate (H)
0 5 10 15 20
-0.2
-0.1
0 Consumption (F)
Benchmark (=2)
More aggressive on inflation stabilization (=4)
Aggressive inflation stabilization increases volatility of exports
Imura (BoC) and Shukayev (UAlberta) Extensive margin of trade June 2016 16 / 17
Conclusion
Monetary stimulus may have different implications for intensivemargin and extensive margin of trade.
Currency depreciation and lower interest rates are favorable toexport sales and, to some extent, export participation.
However, inflationary effects deter entry of new firms and erodecompetitiveness of some incumbent exporters.
Monetary policy that is more aggressive toward inflationstabilization reduces fluctuations in extensive margin.
Imura (BoC) and Shukayev (UAlberta) Extensive margin of trade June 2016 17 / 17
Appendix
Imura (BoC) and Shukayev (UAlberta) Extensive margin of trade June 2016 0 / 2
Parameter Values
Discount factor β 0.99 4% annual interest rate
Weight on leisure in utility χ2 1.8 s.s. labor = 0.33
Elasticity of substitution γ 3.8 Ghironi & Melitz (2005)
Armington elasticity ρ 1.5 Backus et al. (1995)
Labor income share 1− ν 0.6 Cooley & Prescott (1995)
Depreciation rate of capital δ 0.025 10% depreciation/year
# of firm-specific productivity nz 2
Monetary policy rule (Clarida, Gali, Gertler, 1998)
inflation φπ 2
output φY 0.5
real exchange rate φQ 0.1
persistence ρi 0.8
Imura (BoC) and Shukayev (UAlberta) Extensive margin of trade June 2016 1 / 2
Household
Representative household chooses Ct, Lt, Kt+1, Bt+1(st+1), BDt+1
maxEt
∞∑t=0
βt [εct logCt + χ2(1− Lt)]
subject to
Ct + It +∑st+1
q(st+1|st)B(st+1)
Pt+BDt+1
Pt≤ wtLt + rtKt + dt +
B(st)
Pt+ ipt
BDtPt
Kt+1 = (1− δ)Kt + It −κ
2
(ItKt− δ)2
Kt
where
B(st+1) = state-contingent international bond
q(st+1|st) = price of B(st+1) in units of home currency in state st
BDt = non-contingent domestic bonds
εct = demand shock
Imura (BoC) and Shukayev (UAlberta) Extensive margin of trade June 2016 2 / 2