Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
Introduction Data and Facts Model Calibration Conclusion
Intermediate Inputs, Firm Size, and Import
Content of Production
Mehmet Fatih Ulu
Central Bank of Turkey
May 12, 2014
Introduction Data and Facts Model Calibration Conclusion
Introduction
Introduction Data and Facts Model Calibration Conclusion
Introduction
This paper is about intermediate goods and their joint role inboth trade and production processes.
Introduction Data and Facts Model Calibration Conclusion
Introduction
This paper is about intermediate goods and their joint role inboth trade and production processes.
Trade in intermediate goods is big: 2/3 of all merchandiseflows.
60% - 70% of production costs is on intermediate goods.
Introduction Data and Facts Model Calibration Conclusion
Introduction
This paper is about intermediate goods and their joint role inboth trade and production processes.
Trade in intermediate goods is big: 2/3 of all merchandiseflows.
60% - 70% of production costs is on intermediate goods.
We aim to quantify the role of costs attached to importingand exporting while explaining the relationship between firmsize, intermediate input imports, and export behavior.
Introduction Data and Facts Model Calibration Conclusion
Main Preliminary Findings
Costs attached to importing and exporting are sizeable anddecisive in domestic value-added creation.
Introduction Data and Facts Model Calibration Conclusion
Main Preliminary Findings
Costs attached to importing and exporting are sizeable anddecisive in domestic value-added creation.
Iceberg costs
Adaptation costs
Introduction Data and Facts Model Calibration Conclusion
About the Data
Combine two distinct datasets:
Trade transactions of Turkish manufacturing firms (NACE15-37) in 2008
Introduction Data and Facts Model Calibration Conclusion
About the Data
Combine two distinct datasets:
Trade transactions of Turkish manufacturing firms (NACE15-37) in 2008
Industry Census of Turkish manufacturing firms in 2008
Introduction Data and Facts Model Calibration Conclusion
Some Regularities
Figure : Intermediate Import Ratio By Exporter Status0
.1.2
.3.4
Ave
rage
Inte
rmed
iate
Impo
rt R
atio
0 20 40 60 80 100Size Percentile (in # Value of Production)
Exporters Non_Exporters
Introduction Data and Facts Model Calibration Conclusion
010
020
030
040
0N
umbe
r of
Firm
s
0 20 40 60 80 100Size Percentile (in # Value of Production)
nooffirmsEXP nooffirmsALLnooffirmsNONEXP
Number of Fims by Size and Exporter Status
Figure : Number of Firms By Size and Exporter Status
Introduction Data and Facts Model Calibration Conclusion
Figure : Intermediate Import Ratio by Size (Within)
0.1
.2.3
.4.5
Ave
rage
Inte
rmed
iate
Impo
rt R
atio
0 20 40 60 80 100Size Percentile (in # Value of Production)
Exporters Non_Exporters
Introduction Data and Facts Model Calibration Conclusion
Figure : Number of Imported Varieties vs. Fraction of Firms
1
10
100
1000
10000#
impo
rt v
arie
ties
(HS
12 −
cou
ntry
)
.001 .01 .1 1Fraction of Firms
Introduction Data and Facts Model Calibration Conclusion
Figure : Number of Import Varieties by Size (Within)
.01
.1
1
10
100
20 40 60 80 100
2008
Exporters Non_Exporters
log(
Ave
rage
Num
bero
f Im
port
ed V
arie
ties)
Size Percentile (in # Value of Production)
Graphs by year
Introduction Data and Facts Model Calibration Conclusion
Figure : Revenue by Size (Within)
.001
.01
.1
1
10
100
1000lo
g(A
vera
ge R
even
ue (
mill
ion
TL)
)
0 20 40 60 80 100Size Percentile (in # Value of Production)
Exporters Non_Exporters
Introduction Data and Facts Model Calibration Conclusion
Previous Studies
Imported intermediate goods and productivityHalpern, Koren, and Szeidl (2011), Kasahara and Rodrigue(2008), Gopinath and Neiman (2014)
Joint analysis of import and export decisionsKasahara and Lapham (2012)
New intermediate goods and product scopeGoldberg et al. (2010)
Motives for importing intermediate inputsSaygili et al. (2010)
Introduction Data and Facts Model Calibration Conclusion
Preliminaries
Extend Gopinath and Neiman (2014) by adding exportsmarket and demand shocks
Introduction Data and Facts Model Calibration Conclusion
Preliminaries
Extend Gopinath and Neiman (2014) by adding exportsmarket and demand shocks
Two countries: home and foreign
Introduction Data and Facts Model Calibration Conclusion
Preliminaries
Extend Gopinath and Neiman (2014) by adding exportsmarket and demand shocks
Two countries: home and foreign
Continuum of monopolistic firms in both markets
Introduction Data and Facts Model Calibration Conclusion
Preliminaries
Extend Gopinath and Neiman (2014) by adding exportsmarket and demand shocks
Two countries: home and foreign
Continuum of monopolistic firms in both markets
Production requires two types of input: labor L and a bundleof intermediate inputs X
Yi = Ai (Lp,i )1−µX
µi
Introduction Data and Facts Model Calibration Conclusion
Intermediate Goods
Zi and Mi are the bundles of domestic and importedintermediate inputs used by firm i , respectively.
Xi = [Z ρi +M
ρi ]
1ρ
Introduction Data and Facts Model Calibration Conclusion
Intermediate Goods
Zi and Mi are the bundles of domestic and importedintermediate inputs used by firm i , respectively.
Xi = [Z ρi +M
ρi ]
1ρ
where
Zi =
[∫
zθijdj
]1θ
Mi =
[∫
Ωi
(bmik)θdk
]1θ
b ≥ 1 is the quality attached to imported inputs.
Introduction Data and Facts Model Calibration Conclusion
Sales and Fixed Costs of Importing and Exporting
gi is the domestic sales while g∗i is the exports.
Yi = gi + g∗i
Introduction Data and Facts Model Calibration Conclusion
Sales and Fixed Costs of Importing and Exporting
gi is the domestic sales while g∗i is the exports.
Yi = gi + g∗i
Three kinds of fixed costs: fe , fI , and fυ.
F (|Ωi |, g∗i ) = [fI1|Ωi |6=0 + fυ|Ωi |
λ + fe1|g∗
i|>0]
where λ > 1.
Introduction Data and Facts Model Calibration Conclusion
Input Prices
PXi=
(
Pρ
ρ−1
Z + Pρ
ρ−1
Mi
)ρ−1ρ
if firm i imports
PZ if firm i does not import
Introduction Data and Facts Model Calibration Conclusion
Input Prices
PXi=
(
Pρ
ρ−1
Z + Pρ
ρ−1
Mi
)ρ−1ρ
if firm i imports
PZ if firm i does not import
PMi=
[∫
k∈Ωi
(pm
b)
θθ−1dk
]θ−1θ
=pm
b|Ωi |
θ−1θ
where 0 < θ < 1.
Introduction Data and Facts Model Calibration Conclusion
Firm’s Problem
Unit cost of production for firm i is
Ci =1
µµ(1− µ)1−µ
w1−µPµXi
Ai
.
Introduction Data and Facts Model Calibration Conclusion
Firm’s Problem
Unit cost of production for firm i is
Ci =1
µµ(1− µ)1−µ
w1−µPµXi
Ai
.
mi =
(
pmPMi
)1
θ−1(
PMi
PXi
)1
ρ−1Xi if firm i imports
0 if firm i does not import
Introduction Data and Facts Model Calibration Conclusion
Firm’s Problem Cont...
Firm i receives demand shock si in the foreign market.
g∗i (si , pi ) =
sip1
ǫ−1
i if firm i exports0 if firm i does not export
Introduction Data and Facts Model Calibration Conclusion
Firm’s Problem Cont...
Firm i receives demand shock si in the foreign market.
g∗i (si , pi ) =
sip1
ǫ−1
i if firm i exports0 if firm i does not export
pi =
Ci
ǫin the domestic market
τ Ci
ǫin the foreign market
Introduction Data and Facts Model Calibration Conclusion
Firm’s Problem Cont...
Firm i receives demand shock si in the foreign market.
g∗i (si , pi ) =
sip1
ǫ−1
i if firm i exports0 if firm i does not export
pi =
Ci
ǫin the domestic market
τ Ci
ǫin the foreign market
Firm has to decide about being an exporter and an importer,as well as, the number of varieties to import.
Ψ = maxΩi ,gi ,g
∗
i
Πi − wF (|Ωi |, g∗i )
Introduction Data and Facts Model Calibration Conclusion
F.O.C.
∂Ψ
∂Ω= W (1− ε)(1 + I (g∗ > 0)sτ
εε−1 )
∂p(Ω)ε
ε−1
∂Ω− λwfνΩ
λ−1 = 0
Introduction Data and Facts Model Calibration Conclusion
F.O.C.
∂Ψ
∂Ω= W (1− ε)(1 + I (g∗ > 0)sτ
εε−1 )
∂p(Ω)ε
ε−1
∂Ω− λwfνΩ
λ−1 = 0
=⇒ κµε
ε− 1
θ − 1
θ
(pm
b
)ρ
ρ−1P
µεε−1
+ ρ1−ρ
X = λwfνΩλ− θ−1
θρ
ρ−1
where
κ = W (1− ε)(1 + I (g∗ > 0)sτε
ε−1 )εε
1−ε
(
w1−µ
Aµµ(1−µ)1−µ
)ε
ε−1.
Introduction Data and Facts Model Calibration Conclusion
Some Discussions
Relative expenditures on imported and domestic intermediateinputs
Em
EZ
=(pm)
ρρ−1 (b)
1θ−1
− 1ρ−1
(PZ )ρ
ρ−1
Ωθ−1θ
ρρ−1
Responses
∂(Em/EZ )
∂Ω> 0,
∂(Em/EZ )
∂pm< 0,
∂(Em/EZ )
∂PZ
> 0
Introduction Data and Facts Model Calibration Conclusion
Calibration
Each firm is a tuple of shocks (Ai , si )
Targeting moments from the data calibrate the vector ofparameters Θ.
Θ = θ, ρ, b, µ, λ, fe , fυ, fI , τ,w , pm, σs , corr ,W ,PZ , ǫ
Introduction Data and Facts Model Calibration Conclusion
Simulation Parameters
Param. Val. Desc.
θ 0.67 elasticity of substitution within intermediate input groupsρ 0.52 elasticity of substitution between intermediate input groupsb 2 quality attached to imported intermediate varietiesµ 2/3 cost share of intermediate inputsλ 2.33 curvature of the convex adjustment costfe 0.3 entry cost for the export marketfv 0.0003 scale parameter for the adjustment costfI 0.0001 entry cost for the import marketτ 1.2 iceberg costw 60 wagepm 20 unit price of imported intermediate varietiesσ 0.5 std. dev. for the demand shockscorr 0.8 correlation between the demand and productivity shocksW 1000 demand shifterPZ 2 price of the domestically produced intermediate inputsǫ 0.75 elasticity of substitution between intermediate input groups
Introduction Data and Facts Model Calibration Conclusion
Export Decisions
5
10
15
20
25
30
35 1020
3040
5060
7080
0
0.2
0.4
0.6
0.8
1
s
Exports Decisons over the State Space
A
Exp
ort D
ecis
ion
Introduction Data and Facts Model Calibration Conclusion
Import Decisions
510
1520
2530
35
20
40
60
80
0
0.2
0.4
0.6
0.8
1
A
Import Decisons over the State Space
s
Impo
rt D
ecis
ion
Introduction Data and Facts Model Calibration Conclusion
Model Fit
Figure : Intermediate Import Ratios, Exporters
0 10 20 30 40 50 60 70 80 90 1000
0.1
0.2
0.3
0.4
0.5
0.6
0.7
Size Percentile
Ave
rage
Inte
rmed
iate
Impo
rt R
atio
DataModel
Introduction Data and Facts Model Calibration Conclusion
Model Fit
Figure : Intermediate Import Ratios, Non-Exporters
0 10 20 30 40 50 60 70 80 90 1000
0.02
0.04
0.06
0.08
0.1
0.12
Size Percentile
Ave
rage
Inte
rmed
iate
Impo
rt R
atio
DataModel
Introduction Data and Facts Model Calibration Conclusion
Model Fit
Figure : Average Number of Imported Varieties by Firm Size, Exporters
0 10 20 30 40 50 60 70 80 90 1000
20
40
60
80
100
120
140
160
Size Percentile
Ave
rage
Num
ber
of Im
port
ed V
arie
ties
DataModel
Introduction Data and Facts Model Calibration Conclusion
Model Fit
Figure : Average Number of Imported Varieties by Firm Size,Non-Exporters
0 10 20 30 40 50 60 70 80 90 1000
2
4
6
8
10
12
14
16
18
Size Percentile
Ave
rage
Num
ber
of Im
port
ed V
arie
ties
DataModel
Introduction Data and Facts Model Calibration Conclusion
Model Fit
Figure : Number of Imported Varieties by the Fraction of Firms - Data
10−5
10−4
10−3
10−2
10−1
100
101
102
103
104
Actual Data
Fraction of Firms
Num
ber
of Im
port
ed V
arie
ties
Introduction Data and Facts Model Calibration Conclusion
Model Fit
Figure : Number of Imported Varieties by the Fraction of Firms - Model
10−4
10−3
10−2
10−1
100
100
101
102
103
Model
Fraction of Firms
Num
ber
of Im
port
ed V
arie
ties
Introduction Data and Facts Model Calibration Conclusion
Revenue Decomposition
0 200 400 600 800 1000 1200 1400 1600 1800 20000
0.5
1S
hare
in R
even
ueRevenue and Its Decomposition across Markets (Ranked by Revenue Size)
Firms
0 200 400 600 800 1000 1200 1400 1600 1800 2000−20
0
20
Log(
Rev
enue
)
Export RevenueDomestic Revenuelog(Revenue)
Introduction Data and Facts Model Calibration Conclusion
The Impact of Imported Intermediate Varieties on Revenue
0 200 400 600 800 1000 1200 1400 1600 1800 2000−15
−10
−5
0
5
10
15
Firms
Log(
Rev
enue
)
The Impact of Imported Intermediate Varieties on Revenue
Importing not AllowedImporting Allowed
Introduction Data and Facts Model Calibration Conclusion
How Sizeable are the Trade Costs?
Table : Sunk and Fixed Costs of Trade
tau f e f I f v lambda F e/R F v/R T/R1.2 0.3 0.0001 0.0003 2.33 0.000175 0.1928 0.14521.08 0.3 0.0001 0.0003 2.33 0.007794 0.07472 0.04541.2 0.27 0.0001 0.0003 2.33 0.000164 0.1928 0.14521.2 0.3 0.00009 0.0003 2.33 0.000175 0.1928 0.14521.2 0.3 0.0001 0.00027 2.33 0.000163 0.19902 0.14521.2 0.3 0.0001 0.0003 2.1 0.007825 0.053921 0.14Notes: This table shows the magnitudes of trade-related costs where Fe =∑
iI(Ei = 1)wfe , FI =∑
iI(Ii = 1)wfI , Fυ =∑
iI(Ii = 1)wfυ |Ωi |λ, T =∑
i (τ − 1)pig∗
i , R =∑
ipiYi
Introduction Data and Facts Model Calibration Conclusion
Next Step and Further Research
Counterfactual experiments regarding cost items fe , fI , fυ, τand input prices
Behavioral transitions of firms in response some costalleviations
Level effects of alleviating some fixed or sunk cost elements
Introduction Data and Facts Model Calibration Conclusion
Next Step and Further Research
Counterfactual experiments regarding cost items fe , fI , fυ, τand input prices
Behavioral transitions of firms in response some costalleviations
Level effects of alleviating some fixed or sunk cost elements
Further Research
Getting closer to a general equilibrium analysis
Studying the nature of the adjustment costs
Introduction Data and Facts Model Calibration Conclusion
Conclusion
How the import and export decision interact with each other?
Introduction Data and Facts Model Calibration Conclusion
Conclusion
How the import and export decision interact with each other?
The impact of imported intermediates on economic activity
Introduction Data and Facts Model Calibration Conclusion
Conclusion
How the import and export decision interact with each other?
The impact of imported intermediates on economic activity
Quantification of the trade costs
Introduction Data and Facts Model Calibration Conclusion
Conclusion
How the import and export decision interact with each other?
The impact of imported intermediates on economic activity
Quantification of the trade costs
Introduction Data and Facts Model Calibration Conclusion
Conclusion
How the import and export decision interact with each other?
The impact of imported intermediates on economic activity
Quantification of the trade costs
Policy suggestions
Introduction Data and Facts Model Calibration Conclusion
Thank You.