Content uploaded by Anung Yoga
Author content
All content in this area was uploaded by Anung Yoga on Apr 21, 2018
Content may be subject to copyright.
Signifikan: Jurnal Ilmu Ekonomi
Volume 6 (2), October 2017
P-ISSN: 2087-2046; E-ISSN: 2476-9223
Page 247– 266
http://journal.uinjkt.ac.id/index.php/signifikan 247
DOI: 10.15408/sjie.v6i2.5210
Exchange Rate and International Trade: Case From Indonesian
Manufacturing Sector
Anung Yoga Anindhita
State Islamic University (UIN) of Sunan Ampel Surabaya
anung.yoga@uinsby.ac.id
Abstract
Exchange rate fluctuation in Floating Exchange Rate Regime is considered to have impacts on the
international trade through its adjustment to the price and its volatility to the trade risk. This paper is
aimed at estimating those impacts on the international trade of manufacturing sector in Indonesia for
period 2007 to 2014. To conduct estimation, it uses multiple regression analysis on two models: First,
the import of raw-and-auxiliary materials; Second, the export of manufacturing sector. The results
show that the exchange rate impacts both work significantly on the import of raw-and-auxiliary
materials. The finding implies that, through the import of raw-and-auxiliary materials, manufacturing
sector is very susceptible to the shock caused by exchange rate changes. Meanwhile, the export of
manufacturing sector is not able to take advantage of the depreciation of the exchange rate due to the
lack of competitiveness.
Keywords: exchange rate, international trade, manufacturing sector
Abstrak
Fluktuasi Nilai tukar dalam Rezim Nilai Tukar Mengambang Bebas dipandang akan berdampak pada
perdagangan internasional melalui penyesuaian harga dan resiko transaksi perdagangan yang
ditimbulkannya. Tulisan ini ditujukan untuk menganalisis dampak nilai tukar tersebut pada kasus
sektor industri pengolahan di Indonesia pada periode 2007 sampai 2014. Analisis dilakukan dengan
menggunakan regresi berganda melalui dua model: Pertama, impor bahan baku dan penolong; Kedua,
ekspor industri pengolahan. Hasil yang diperoleh menunjukkan bahwa kedua dampak nilai tukar
berkerja secara signifikan pada impor bahan baku dan penolong, sedangkan pada ekspor industri
pengolahan. Lebih lanjut, hasil tersebut menunjukkan bahwa melihat dampak nilai tukar terhadap
impor bahan baku dan penolong, industri pengolahan sangat rentan akan goncangan yang disebabkan
perubahan nilai tukar. Sementara, ekspor industri pengolahan tidak bisa terdorong oleh depresiasi nilai
tukar dikarenakan kurang kompetitif.
Kata Kunci: nilai tukar, perdagangan internasional, sektor industri pengolahan
Received: April 04, 2017; Revised: May 21, 2017; Approved: June 18, 2017
Exchange Rate and International Trade: Case...
Anung Yoga Anindhita
248 http://journal.uinjkt.ac.id/index.php/signifikan
DOI: 10.15408/sjie.v6i2.5210
INTRODUCTION
During the exchange rate regime, Indonesia experienced an important phase
when the release of the intervention band of the Rupiah against the US Dollar
occurred on August 14, 1997. It marked the change of the exchange rate system from
a managed floating with crawling band system to the floating exchange rate. The
background of the changes in the exchange rate system is that Indonesia, of which
foreign exchange reserves depleted and Rupiah underwent strong pressure backthen,
decided to carry out a free floating system. This system is in accordance with the
economic reform package that had been recommended by the International Monetary
Finance (IMF). However, because the policy of releasing the intervention band implied
more negative expectations, since it was done when the Rupiah was depressed and
compounded by destabilizing speculation attacks, the follow-up impact was that the
exchange rate of the Rupiah was sharply depressed (Figure 1)
Figure 1. The movement of the exchange rate of the Rupiah/US Dollar
from June 1997 to Mei 1998
Source: processed online from http://www.tradingeconomics.com/indonesia/currency, accessed
on March 2nd, 2017)
In the context of international trade, exchange rate stability is a condition
preferred by market players. This is based on the influence of the exchange rate
against trading through: First, costs and prices that eventually produce relative-price
changes in trade (Carbaugh, 2005); Second, exchange rate volatility itself that will lead
to the risk of the transaction because of the different spot value of transaction
http://journal.uinjkt.ac.id/index.php/signifikan 249
DOI: 10.15408/sjie.v6i2.5210
Signifikan Vol. 6 (2), October 2017
with spot value of payment. Even though the risk can be closed by hedging—namely
with transactions in the forward market—the premium covered by the importers will
continue to burden the transaction. (Krugman and Obsfeld, 2003).
The first impact is the price adjustment and it has been discussed in the
Marshall-Lerner Condition asserting how the demand elasticity towards import and
export will affect the balance of payment through the current transaction (with the
assumption that the capital balance remains). The Marshall-Lerner condition requires
that the absolute addition of exports and imports elasticity be greater than 1 (one) in
order to achieve improvement of current transaction (surplus). If the condition is not
met, depreciation will not improve the current transaction or even exacerbate it.
The second impact is the exchange rate risk which occurs every time cause the
investor, the company or the bank, to have to face payment obligations in the future in
foreign currency. The condition means that the payers have foreign exchange risk or
what is referred to as the "open position" (Salvatore, 1997). Related to the existence
of the risk, the traders then can make an effort to avoid the risk of the exchange rate
by hedging the exchange rate that can be done in the forward market. However,
hedging the risk of the exchange rate also causes the cost if there are premiums for its
transactions. The emergence of the cost that must be borne in order to hedge the risk
of the exchange rate will result in the cost of international trade that has a tendency to
further reduce trade volumes. Suardhini also supports this and Goeltom (1997) stated
that doing transaction in the forward market could reduce the fluctuation risk of the
benefits of trade in the short term. However, this will, unfortunately, lead the cost of
international trade to increase and finally generate the anti-trade bias.
The failure to achieve the relationship and the adjustment direction of Marshall-
Lerner condition can be seen from the discontinuation of the adjustment direction of
export or import of a country through the depreciation or appreciation. In the context
of Indonesian export on which there was a sharp depreciation after the monetary
crisis in 1997, this condition has been described by Siregar and Rajan (2003) in the
introduction of his study. When the monetary crisis began in August 1997 which is
marked by shock on export growth, it shows that, in the following year, Indonesia was
not able to undergo the adjustment of export through depreciation of exchange value
Exchange Rate and International Trade: Case...
Anung Yoga Anindhita
250 http://journal.uinjkt.ac.id/index.php/signifikan
DOI: 10.15408/sjie.v6i2.5210
as other South East Asian countries that also affected by the crisis such as Malaysia, the
Philippines and Thailand. (Figure 2).
Figure 2. The Exports Growth Rate in some countries that experienced the
Currency Crisis in Asia in 1997
Source: Siregar and Rajan (2003)
In this regard, one of the causes of the economic crisis, which is an extension of
the monetary crisis in Indonesia, can be explained by the fragility of the real sector.
Before the crisis, the performance of Indonesian import was characterized by the
domination of the import of raw-and-auxiliary materials, especially by the
manufacturing sector. Such domination happened because of its dependence on raw-
and-auxiliary materials, which were imported from abroad to perform its production.
The next impact from this dependence is resulted in manufacturing sector production
vulnerabilities to overseas externality. This has predicted before the crisis by Dumairy
(1996). This also has been reviewed as well as observed by Tambunan (1998)
particularly when the crisis occurred.
In the Figure 3, it can be seen that the contribution of the import of raw-and-
auxiliary materials are far above the capital goods and consumer goods. When
monetary crisis occurred in 1997-1998 –in the time when the exchange rate sharply
depreciated–the import of raw-and-auxiliary materials had an upward trend tendency
since 1986 until 1996. This suddenly experienced a shock that was marked by a sharp
drop in the years 1997-1998. It can also be seen from the graph that the shock tends
to occur only in the import of raw-and-auxiliary materials and capital goods while the
import of consumer goods tends not to experience a shock.
Year
Ex
po
rt
Gr
ow
th
(%)
http://journal.uinjkt.ac.id/index.php/signifikan 251
DOI: 10.15408/sjie.v6i2.5210
Signifikan Vol. 6 (2), October 2017
Figure 3. The development of non-oil and gas import based on the use of
Goods 1996 – 2004
Source: www.bps.go.id, accessed on March 2nd, 2017, processed
The vulnerability to shock that comes from the externality of the exchange rate
or the production of the manufacturing sector is not accompanied by the reduction of
the dependency on the import of raw-and-auxiliary materials. The trend in Figure 3
above shows the comparison of the import of raw-and-auxiliary materials compared to
consumer goods and capital goods until the post-crisis of 1998 that has no tendency
to shift and to be very dominating up to above 70 percent.
On the other hand, the export of manufacturing sector has contributed to the
Indonesian economy which contributes more than 80 percent of non-oil and gas
export and more than sixty percent of the total export in the period of the 1990s
before the monetary crisis in 1997-1998. The large proportion, if associated with no
corrected export when the Rupiah depreciated during the financial crisis in 1997-
1998 as illustrated in figure 2 above, can be said that the manufacturing sector has
greatly contributed to the failure of exports correction that was actually expected to
rise. The studies of the impact of exchange rate on international trade in Indonesia
were conducted before the 1998 crisis, as have been done by Suardhini and Goeltom
(1997) in period of 1979 to1991, as well as Siregar and Rajan (2003) in period of 1997
to 2007. In those periods, The exchange rate system used was Managed Floating.
Implementation of study when exchange rates fluctuate larger (with floating exchange
rate system), and using the object of manufacturing sector as the largest contributing
sector in Indonesia is necessary.
Exchange Rate and International Trade: Case...
Anung Yoga Anindhita
252 http://journal.uinjkt.ac.id/index.php/signifikan
DOI: 10.15408/sjie.v6i2.5210
This paper is intended to estimate and analyze the impact of the exchange rate
on the international trade indutries sector in Indonesia in different periods long after
the economic crisis in 1998. Through this research, it is expected that the decision
maker can formulate policy in manufacturing sector within the international trade by
its measurement on the impact of exchange rate adjustment. Based on the
assumption of the problem identification that has been described above, the research
will be specified through the model of the import of raw-and-auxiliary materials and
the export of manufacturing sector as a model that describes the condition of the
international trade in Indonesian manufacturing sector.
METHOD
In drawing up the import of raw-and-auxiliary materials model, the functional
form using natural logarithm is as follows:
(1)
Where: M is real value of import of raw-and-auxiliary materials, PROD is real
production value of domestic manufacturing sector, REER is real effective
exchange rate from the import of raw-and-auxiliary materials, VOL is exchange-rate
volatility and u is an error term.
Meanwhile, the export of manufacturing sector model is written as follow:
(2)
Where: XMAN is real value of export of manufacturing sector, YLN is real income of
destination countries, REER is the real effective exchange rate, VOL is exchange rate
volatility and u is an error term.
From the explanatory variables identified in the model, REER and VOL are
explanatory variables that exist to represent exchange rate impact. The model also
accommodates time lag that is possible to happen because of the difference in the
transaction time (decision-making) with the receipt of goods due to the required
process. The formation of the lag is also strengthened by what is known as the J curve,
which describes the relationship between the trade balance with the currency
depreciation (Krugman and Obsfeld, 2005: 464). The J curve, as stated by Gujarati and
Porter (2012), was made as the basis of the lag in the international trade model.
http://journal.uinjkt.ac.id/index.php/signifikan 253
DOI: 10.15408/sjie.v6i2.5210
Signifikan Vol. 6 (2), October 2017
The model that has been arranged also illustrates an aggregation model. It
creates the aggregation effect problem, which might appear and therefore the use of
panel data model can be used as an alternative. However, the absence of complete
data such as the data segregation of dependent variable, namely: 1) imports data of
raw-and-auxiliary materials based on the country of origin, and 2) manufacturing
exports data based on the destination country which is arranged on a monthly or
quarterly basis, makes the panel data model difficult to be used.
The formulas to calculate REER for import of raw-and-auxiliary materials and
export of manufacturing sector in sequence are written as follow:
(3)
Where: REERM is real effective exchange rate for import of raw-and-auxiliary
materials, NEER is nominal effective exchange rate, PM is import price of raw-and-
auxiliary materials, PY is domestic price of raw-and-auxiliary materials
(4)
Where: REERXMAN is real effective exchange rate for export of manufacturing
sector, NEER is nominal effective exchange rate, PXMAN is export price of manufacturing
sector, PYLN is price of manufacturing sector in abroad
To get proxy for volatility measurement or the exchange rate risk, using
nominal valuef exchange rate (NEER) is preferable. This is due to the fact that
the nominal value tends to be more able to describe the volatility that leads to
uncertainty faced directly by the international traders. The volatility then measured by
using moving average standard deviation as has been used by Kenen and Rodrik (1986).
The formula is written as follows:
(3)
Where: Xi is NEER, is the average of 4 quarterly NEER, and n= 4
The secondary data used is quarterly time series data between 2007 to 2014. It
is collected online through the official publication published on the sites of
Exchange Rate and International Trade: Case...
Anung Yoga Anindhita
254 http://journal.uinjkt.ac.id/index.php/signifikan
DOI: 10.15408/sjie.v6i2.5210
the institutions providing the needed data. Most of the data are obtained from Bank
Indonesia. The data are: 1) the import of goods according to the economic category in
thousand USD; 2) GDP according to the field of business in billion USD; 3) The index
of nominal exchange rate of the rupiah towards the major trading partner countries; 4)
non-oil and gas export based on the category of sectors in thousand USD. The next
data source is the official site of Organization for Economics and Co-operation and
Development (OECD). The data obtained are: 1) The quarterly Index of GDP
countries that become members of OECD; and 2) The producer price index of
industrial manufacturing.
RESULT AND DISCUSSION
Result for Model 1: Import of Raw-and-Auxiliary Materials
The preliminary result through a scatter plot test to the import of raw-and-
auxiliary materials model produces the specified independent variable (in the form of
natural logarithm) which tend to have a relationship with the dependent variables.
Some of them are: 1) real production value of domestic manufacturing sector with
one period lagged (LNPRODt-1); 2) real effective exchange rate with one period lagged
(LNREERt-1), and 3) exchange rate volatility (LNVOLt). From these results, the mapping
of the variable relationship shows a tendency of the presence of one period lagged in
real production value of domestic manufacturing sector and exchange rate in
explaining the import of raw-and-auxiliary materials in Indonesia.
Table 1. The Test Result of Augmented Dicky Fuller on The Variables of
Import of Raw-and-auxiliary Materials Function
The variables
ADF t value
t table (error level 5%)
LNMBt
-2.164747
-3.562882
D(LNMB) t
-4.874066
-3.568379
LNPRODt-1
-3.352311
-3.562882
D(LNPROD t-1)
-5.976503
-3.568379
LNREER t-1
-2.588685
-3.562882
D(LNREER t-1)
-5.976503
-3.568379
LNVOLt
-2.112516
-3.562882
D(LNVOL)t
4.881631
-3.568379
http://journal.uinjkt.ac.id/index.php/signifikan 255
DOI: 10.15408/sjie.v6i2.5210
Signifikan Vol. 6 (2), October 2017
Description: The ADF tests include intercept elements and trends (the largest testing
power). The number of lag based on SIC.
The follow-up step is then conducting Two Steps Engle-Granger Test in order
to identify stationarity and cointegration. In the step one, table 1 shows that the entire
variable is not stationary at level, but stationary on the first difference. The result is
seen from the absolute number on ADF t value which is entirely smaller if compared
with the absolute numbers of t table using 5 percent error level. Nevertheless, on the
first difference, the absolute numbers on ADF t value are entirely greater if compared
with the absolute numbers of t table using 5 percent error level.
Table 2. The ADF Tests on Residual of The Import of Raw-and-auxiliary
Materials Model
Residual
ADF t Value
t table (error level of 5%)
RESID 01
-2.316621
-1.952066
RESID 02
-2.569497
-1.952066
RESID 03
-3.882789
-1.953381
Description: The ADF tests for residuals, the selection of lag number based on SIC; RESID
01 is a residual of regression result of lnmbt with lnprodt-1; RESID02 is a residual of regression
result of lnmbt with lnreert-1; and RESID03 is a residual of regression result of lnmbt with lnvolt
With the stationary data of all variables on the first difference, there is a
possibility of a co-integrated regression occurring. The result from ADF test on
Residuals (Table 2) indicates the regression is cointegrated. This can be seen from the
results of ADF tests against all the stationary residuals at level indicated by the
absolute value on ADF t value that is greater than the absolute value of t table (5
percent error level). By referring to this result, the explanatory variables have a long-
term correlation with the dependent variable.
From the results of the regression calculations presented in table 3, we can see
the description of all significant parameters direction which is in accordance with what
is expected. The real production value of domestic manufacturing sector with one
period lagged (LNPROD t-1) and the real effective exchange rate with one period lagged
(LNREERt-1) are positively related, while the volatility (risk) of the exchange rate is
negatively related. The three variables also provide simultaneous influence and are able
to explain the variation of dependent variables equal to 83.3 percent.
Exchange Rate and International Trade: Case...
Anung Yoga Anindhita
256 http://journal.uinjkt.ac.id/index.php/signifikan
DOI: 10.15408/sjie.v6i2.5210
Result for Model 2: Export of Manufacturing Sector
Different from the results of the import of raw-and-auxiliary materials model,
independent variables which are specified on the export of manufacturing sector
model do not experience a lag. Clear relationships are only visible on real income of
destination countries (LNYLN) and exchange rate volatility (LNVOL). On the other
hand, the real effective exchange-rate variable (REER) does not produce a distinct
relationship either in the same period or in lag of the previous periods.
Table 3. The Results of Estimation for The Import Demand of Raw-and-
auxiliary Materials Model (LNM)
Variable
Coefficient
Std. Error
t-Statistic
Prob.
C
5.014709
1.657527
3.025416
0.0053
LNPRODT_1
0.637567
0.119681
5.327202
0.0000
LNREERT_1
0.616217
0.140348
4.390647
0.0001
LNVOL
-0.132039
0.037342
-3.535959
0.0014
R-squared
0.833467
Mean dependent var
15.04843
Adjusted R-squared
0.815624
S.D. dependent var
0.203179
S.E. of regression
0.087243
Akaike info criterion
-1.923772
Sum squared resid
0.213117
Schwarz criterion
-1.740555
Log likelihood
34.78036
Hannan-Quinn criter.
-1.863041
F-statistic
46.71155
Durbin-Watson stat
1.737541
Prob(F-statistic)
0.000000
The ADF test which is summarized in Table 4 shows that the entire
variables are not stationary at level, but stationary on the first difference. It is
clear that the absolute numbers on ADF t value are entirely smaller if
compared with the absolute numbers of t table (5 percent error level). Yet, on the
first difference, the absolute numbers on ADF t value are entirely greater if
compared with the absolute numbers of t table (5 percent error level). With the
production of stationary data of all variables on the first difference, thus in the model
of the export of manufacturing sector there is an open possibility of a regression is co-
integrated.
Then, from the results of the Table 5, it can be seen that the two relationships
that cointegrated between the dependent variable with explanatory variables is
resulted in the relationship between export of manufacturing sector (LNXMANt) with
real income of destination countries (LNYLNt) and the relationship between export of
http://journal.uinjkt.ac.id/index.php/signifikan 257
DOI: 10.15408/sjie.v6i2.5210
Signifikan Vol. 6 (2), October 2017
manufacturing sector (LNXMANt) with exchange-rate volatility variable (LNVOLt). On
the other hand, the real effective exchange-rate variable (LNREERt), which has a
tendency to not having a relationship with the exports variable of manufacturing
indutries (LNXMANt)—does not become co-integrated.
Table 4. The Test Result of Augmented Dicky Fuller on The Variables of
Export of manufacturing Sector Function
Description: The ADF tests include intercept elements and trends (the largest testing
power). The number of the lag based on SIC.
After regression calculations, positive autocorrelation disturbance is still
found. This is marked by the low value of Durbin Watson (DW) statistics,
which is only equal to 1.16. Therefore, the improvement of the model is treated by
using the method of weighted least square utilizing Durbin’s Two-Steps method. By
transforming the model in the first difference using estimated from DW statistics
, the result appears to have experienced changes of the value of the
DW statistics of 1.72 which is greater than the value of the crisis dl of 1.17.
The value—eventhough it is still slightly below the value du of 1.73 which means
it is still in the area without a conclusion—can avoid the coefficient parameter
produced in the area concluding the existence of positive autocorrelation. The
regression results after the improvement of the positive autocorrelation are
explained in Table 7.
The variables
ADF t Value
t table (error level of 5%)
LNXMANt
-1.874144
-3.562882
D(LNXMAN) t
-4.797623
-3.568379
LNYLN t
-3.259710
-3.568379
D(LNYLN) t
-3.682288
-3.574244
LNREERt
-2.463036
-3.562882
D(LNREER t)
-5.351709
-3.568379
LNVOL t
-2.139627
-3.562882
D(LNVOL t)
-4.709830
-3.568379
Exchange Rate and International Trade: Case...
Anung Yoga Anindhita
258 http://journal.uinjkt.ac.id/index.php/signifikan
DOI: 10.15408/sjie.v6i2.5210
Table 5. The ADF Tests on Residual of The Export of
Manufacturing Sector Model
The variables
ADF t Value
t table (error level of 5%)
RESID 01
-2.050862
-1.952066
RESID02
-1.605562
-1.952066
RESID 03
-3.695243
-1.952066
Description: The ADF test for residuals, the selection of the number of lag based on SIC;
RESID 01 is a residual regression result of LNXMANt with LNYLNt; RESID02 is a residual
regression result of LNXMANt with LNREERt; and RESID03 is a residual regression result of
LNXMANt with LNVOLt
The results in Table 7 show only two parameters of the variables that
are significant with an error level of 5 percent, namely real income of destination
countries (LNYLNt), and the exchange rate volatility (LVOLt) is in accordance with the
expected direction. Meanwhile, the Real Effective Exchange Rate (LNREERt) does not
partially affect the exports of manufacturing sector significantly. The independent
variables in the model can simultaneously explain the dependent variable variation of
67 percent.
Table 6. The Results of Estimation for The Export of Manufacturing Sector
(LNMAN)
Variable
Coefficient
Std. Error
t-Statistic
Prob.
C
2.696003
4.331887
0.622362
0.5387
LNYLN
2.766920
0.842875
3.282715
0.0028
LNREER
-0.027381
0.129037
-0.212192
0.8335
LNVOL
-0.040746
0.009749
-4.179609
0.0003
R-squared
0.794037
Mean dependent var
15.34448
Adjusted R-squared
0.771969
S.D. dependent var
0.136954
S.E. of regression
0.065399
Akaike info criterion
-2.500150
Sum squared resid
0.119757
Schwarz criterion
-2.316933
Log likelihood
44.00240
Hannan-Quinn criter.
-2.439419
F-statistic
35.98219
Durbin-Watson stat
1.160819
Prob(F-statistic)
0.000000
http://journal.uinjkt.ac.id/index.php/signifikan 259
DOI: 10.15408/sjie.v6i2.5210
Signifikan Vol. 6 (2), October 2017
Discussion
To simplify in interpreting the result described in Table 3 and Table 7, the
models are written as follow:
LNM = 5.014 + 0.637 LNPRODt-1 + 0.616 LNREER t-1– 0.132 LNVOLt + u
value = (0.005) (0.0000) (0.0001) (0.0014)
R2 = 0,833 F-stat = 46,71
Adj. R2 = 0,815 Prob(F-statistic) = 0,000000
LNXMANt
* = -1.085 + 3.61 LNYLNt
* + 0.095 LNREERt
*
– 0.023 LNVOL t *+ u*
value = (0.695) (0.005) (0.541) (0.024)
R2 = 0,67 F-stat = 18,34
Adj. R2 = 0,63 Prob(F-statistic) = 0,000001
Starting form the first model, the result of the estimation also shows that the
two variables which are specified as the impact of the exchange rate, namely: 1) real
effective exchange rate with one period lagged (LNREERt-1); and 2) exchange rate
volatility (LNVOLt), have significant impact the imports of the raw-and-auxiliary
materials. In other words, the dependence on import of raw-and-auxiliary materials, as
indicated earlier, is vulnerable to externality shock of the exchange rate.
Table 7. The Results of Estimation for The Export of Manufacturing Sector
(LNXMAN) with Durbin’s Two-Steps Method
Variable
Coefficient
Std. Error
t-Statistic
Prob.
C
-1.085426
2.738874
-0.396304
0.6950
LNYLN*
3.617675
0.919555
3.934158
0.0005
LNREER*
0.095515
0.154319
0.618942
0.5411
LNVOL*
-0.023816
0.010030
-2.374482
0.0249
R-squared
0.670922
Mean dependent var
8.905810
Adjusted R-squared
0.634358
S.D. dependent var
0.095243
S.E. of regression
0.057592
Akaike info criterion
-2.750962
Sum squared resid
0.089554
Schwarz criterion
-2.565932
Log likelihood
46.63992
Hannan-Quinn criter.
-2.690647
F-statistic
18.34915
Durbin-Watson stat
1.724718
Prob(F-statistic)
0.000001
Exchange Rate and International Trade: Case...
Anung Yoga Anindhita
260 http://journal.uinjkt.ac.id/index.php/signifikan
DOI: 10.15408/sjie.v6i2.5210
Description: the numbers in the brackets are probability values; The symbol (*) is the result
of the iteration of Durbin Two-Steps method, i.e.:
LNXMANt* = LNXMANt - (0.4195905 × LNXMANt-1)
LNYLNt* = LNYLNt - (0.4195905 × LNYLNt-1)
LNREERt* = LNREERt - (0.4195905 × LNREERt-1)
LNVOLt* = LNVOLt - (0.4195905 × LNVOLt-1)
In relation to the variable in the first exchange rate impacts, namely real
effective exchange rate with one period lagged (LNREERt-1), there is an indication that
the depreciation in 1 percent real effective exchange rate with one period lagged
(LNREERt-1) will correct the declining real value of import of raw-and-auxiliary
materials in the next period (LNMt) in amount 0.616. This situation also illustrates that
the exchange rate shock will cause the double impact if there is a dependence
(causality) on the production of the manufacturing sector toward imports of raw-and-
auxiliary materials. In such condition, there is a risk that a high depreciation in a long
period will likely cause the deindustrialization. Due to the fact that the specified model
is a structural from, further analysis is needed to test the relationship causality.
However, the limitations of the research with the structural model result in the
analysis of VAR with the causality test Engle-Ganger is not being used.
Meanwhile, the second impact of the effects of the exchange rate, namely the
large amount of volatility faced (LNVOLt)—even if the impact is relatively smaller (the
elasticity of 0.132)—shows that the more the amount of exchange rate volatility is
there the more potential of negative impact will occur on domestic manufacturing
indutries producers in their production activities. Thus, the risks and costs of hedging
caused by the exchange rate volatility reduce the number of imports of raw-and-
auxiliary materials just as expected.
From here, moving to the second model of the manufacturing export sector,
the price effect through changes in the exchange rate (LNREERt) on the result does
not have a significant parameter coefficient in explaining the variation of export
changes. Thus, the exports correction because of the exchange rate changes, as
described in the condition of Marshall-Lerner, does not occur. Instead, the correction
of export elastically comes from the income effect that is shown by the first variable,
namely the real income of destination countries (LNYLN) in amount 3.61.
http://journal.uinjkt.ac.id/index.php/signifikan 261
DOI: 10.15408/sjie.v6i2.5210
Signifikan Vol. 6 (2), October 2017
The last variable, the exchange rate volatility (LNVOLt), produces the
coefficient in accordance with the expected direction (negative). This is even though
the coefficient value is small or inelastic (-0.023). Thus, the increased volatility of the
exchange rate gives a bad impact on the export of manufacturing sector. The results
are in line with what is produced in the model of the import of raw-and-auxiliary
materials that means the risk of fluctuations of the benefits of trade in the short
term—although it can be reduced by doing transaction in the forward market namely
hedging—will cause the cost of international trade to increase. This eventually leads to
the reduction of export.
In various studies lately conducted in different countries, the impact of the
exchange rate has been identified in the aggregate trade model of a country. The
object of the studies may have been conducted varies, whether it is to the trade
balance, or partially to export or import, and also implemented in multilateral or
bilateral trade. The studies also implemented varied model using structural model or
developed models of time series analysis. In spite of that, the variables of the exchange
rate effect basically is not different, that is the change of the exchange rate and its
volatility. In brief, previous studies are summarized in Table 8.
Table 8. Some of the Previous Studies Results
References
Object and Sample Period
Result of Exchange Rate Impact
towards Trade
Baharumsha
h (2001)
Bilateral trade balances of
Malaysia and Thailand with the
US and Japan.
Data from 1980:
I to 1996: IV
The real effective exchange rate is an
important variable in the trade balance
equation and devaluation improves the
trade balances of both economies in the
long-run.
2.1.1 Siregar and
Rajan (2003)
Export and import of
Indonesian.
The period of the first quarter
of the year 1980 until the
second quarter of 1997.
Export Model:
Real effective exchange rate (REER) does
not has a significant effect. Exchange rate
volatility impact significantly reduces
exports.
Import Model:
REER significantly has significant impacts,
which is related positively to import.
Exchange rate volatility is not a significant
effect, but will be significant if imposed
import segregation of raw materials and
capital goods.
2.1.2 Fang, Lai and
Miller (2005)
2.1.3
Export of eight Asian countries
(Indonesia, Malaysia, Singapore,
Thailand, the Philippines, South
Korea, Taiwan and Japan) to the
Depreciation encourages exports, as
expected, for most countries, but its
contribution to export growth is weak.
Exchange rate risk generates a negative
Exchange Rate and International Trade: Case...
Anung Yoga Anindhita
262 http://journal.uinjkt.ac.id/index.php/signifikan
DOI: 10.15408/sjie.v6i2.5210
References
Object and Sample Period
Result of Exchange Rate Impact
towards Trade
United States.
The period of January 1979 until
April 2003.
effect for trade in most studied countries
Marquez and
Schindler
(2007)
China’s export and import in
world trade.
The period of January 1997 to
July 2006
Appreciation of 10 percent Renminbi
(Yuan) will lower aggregate Chinese
export by nearly one percentage point.
The response of import for appreciation
is negligible and lacks precision.
2.1.4 Prusty
(2008)
2.1.5
Export Growth of India
(multilateral).
The period of March 1992 until
April 2007.
Two-way causality occurs between
export growth and the growth of the
exchange rate.
2.1.6 Oskooee
and Wang
(2008)
Bilateral trade between
Australia and the US using
disaggregate data in 107
industries
Exchange rate volatility to have short run
effects on trade flows of most
industries. However, the short-run
effects last into long run, only in limited
cases, though more in export
commodities than import ones
2.1.7 Oskooee,
Bahmani and
Hegerty
(2009)
Export and import in 102
different industries of bilateral
trade between Mexico and the
United States in 1962 - 2004.
Volatility has worse impacts on the
bilateral trade between the industries
examined.
2.1.8 Appuhamilag
e and Alhayk
(2010)
Trade between Sri Lanka and
China with the period of the
study was from the first quarter
of 1993 to the fourth quarter of
2007.
The variables Real exchange rate with 1
period lagged (RERt-1) shows
the depreciation effect significantly
encourages export and reduces import.
Exchange rate volatility negatively impacts
on both export and import.
2.1.9 Bilquees,
Mukhtar and
Maliq (2010)
The export of three Asian
countries namely India, Pakistan
and Sri Lanka. Research period
from 1960 to 2007.
Exchange rate volatility gives a negative
impact on the export of the countries
that are examined in both the short term
and long term.
The variable term of trade also produces
the expected relationship, namely the
decline in the REER (depreciation) will
provide an increase on export.
Bethune,
Thaver and
Plante
(2012)
Trade (export and import) of
South African countries and the
European Union countries
during the period from 1980 to
2009.
REER negatively relates to exports, while,
exchange rate volatility has a bad impact
both in the short term and long term.
(Note: REER is not included as a variable
descriptor of import because South
Africa’s economic openness is considered
small and replaced by the variables of
foreign exchange reserves)
In general, most of the previous studies indicate a relationship between export-
import with price adjustment due to changes in the exchange rate, except the result
from Siregar and Rajan (2003) that shows REER has no significant effects to Indonesia
http://journal.uinjkt.ac.id/index.php/signifikan 263
DOI: 10.15408/sjie.v6i2.5210
Signifikan Vol. 6 (2), October 2017
export in the 1990s, as well as the result from Marquez and Schindler (2007) that
shows appreciation is negligible and lacks precision to China’s Import in 1997 to 2007.
The other results confirmed the relationship and the direction in line with what is
described in the condition of Marshall-Lerner regarding trade adjustment comes from
exchange rate.
The previous studies also show that exchange rate volatility concluded has
worse impact both on export and import. Thus, the stability of the exchange rate in
the short term is preferred by the international traders in running the export and
import activities compared to the presence of a high level of volatility that will cause
uncertainty and raises the cost of hedging.
Different results that need to be underlined with the most of the previous
studies above—namely, from the result of export of Indonesian manufacturing sector
model—is the ineffectiveness of price correction through this exchange rate. This
result, can refer to the competitiveness and the types of products of manufacturing
sector’s export. The result of Baharumshah (2001), conclude Malaysia and Thailand get
the benefit from the depreciation because it will improve trade balance because of
their competitiveness. Compared with case in Indonesia, export of manufacturing
sector has so far been supported especially by the products of unfinished goods that
have less added values, namely crude palm oil around 15 percent and crumb rubber
around 12 percent (average from 2007 to 2014). This circumstance, expected reduces
the competitiveness and later does not support the price adjustment effectiveness
through the exchange rate in encouraging export of manufacturing sector.
In the other hand, the elasticity of real income of destination countries
(LNYLNt) shows that the export of manufacturing sector relies very much on the
income increase of its export destination countries. This result demonstrates
that shock to the real income of exports destination countries of major manufacturing
sector will impact more on the number of Indonesian manufacturing sector
export. The distribution of the export destination countries with the largest
percentage is only spreading across 6 countries: China, Japan, the United States,
Singapore, India and Malaysia. With such distribution, the intended shock will illustrate
the decline of the export value of manufacturing sector simultaneously in 2009 in
Exchange Rate and International Trade: Case...
Anung Yoga Anindhita
264 http://journal.uinjkt.ac.id/index.php/signifikan
DOI: 10.15408/sjie.v6i2.5210
amount 15 percent along with the decrease of main destination countries’ income after
the global crisis.
The inability of the export of the manufacturing sector to respond to
depreciation needs to be cautioned given the results on imports of raw-and-auxiliary
materials is corrected by depreciation. Further studies are needed to assess the causes
of the dependence of the manufacturing sector on import of raw-and-auxiliary
materials, as well as the possibility of low value of linkage in supporting the production
of the manufacturing sector. From the follow-up study, expected will answer how the
steps in overcoming the dependence on import of raw-and-auxiliary materials, and
further encourage the export of manufacturing sector to be competitive.
CONCLUSION
As the results obtained, the direction of the exchange rate impacts produce
different results on both models specified. From both the impacts of the specified
exchange rate—namely the price effect (REER) and risk effect (volatility)—are
applicable to the model of the import of raw-and-auxiliary materials. While, on the
export of manufacturing sector model, it is only found that the risk effect is as the
impact of a significant influence.
The results show the need in the long term to seek substitution indutries of
imports of raw-and-auxiliary materials from domestic production in the manufacturing
sector, or in this case to strengthen ties (linkage) between the input sectors for
manufacturing sector. For export manufacturing sector, efforts are needed to open the
export market to non-traditional countries to reduce the risk of simultaneous
recession in the main destination countries. In addition, it is important to improve
the quality of the competitiveness and added value on the export goods of
manufacturing sectors as the main problems. These efforts will prevent the impact of
the exchange rate externality causing shock on the import prices of raw-and-auxiliary
materials and later threatening the continuation of the manufacturing sector
production.
http://journal.uinjkt.ac.id/index.php/signifikan 265
DOI: 10.15408/sjie.v6i2.5210
Signifikan Vol. 6 (2), October 2017
REFERENCES
Appuhamilage, K.S.A. & A.A.A. Alhayk. (2010). Exchange Rate Movements Effect on Sri
Lanka-China Trade. Journal of Chinese Economic and Foreign Trade Studies. Vol. 3.
No. 3: 254-267.
Baharumshah, A.Z. (2001). The Effect of Exchange Rate on Bilateral Trade Balance:
New Evidence from Malaysia and Thailand. Asian Economic Journal. Vol. 15 (3):
291-312.
Bethune, E. M., R.L. Ekanayake., Thaver, & D. Plante. (2013). “The Effects of Exchange
Rate Volatility on South Africa’s Trade With The European Union”. The
International Journal of Business and Finance Research. Vol. 6. No. 3: 13-26.
Bilquees, Faiz, T. Mukhtar & S.J. Malik. (2010). Exchange Rate Volatility and Export
Growth: Evidence from Selected South Asian Countries. Zagreb International
Review of Economics & Business. Vol. 13. No. 2: 27-37.
Carbaugh, R.J. (2005). International Economics. 10th edition. Mason: Thomson South-
Western
Dumairy. (1996). Perekonomian Indonesia (The Indonesian Economy). Jakarta: Erlangga.
Fang, W.S., L. Yi-Hao, & S.M. Miller. (2005). Export Promotion through Exchange Rate
Policy: Exchange Rate Depreciation or Stabilization? Economics Working Papers.
Paper 200507.
Gujarati, D & D.C Porter. (2012). Dasar-Dasar Ekonometrika (The Basic Econometrics).
Jakarta: Salemba Empat.
Huchet - Bourdon, M. & J. Korinek. (2011). To What Extent Do Exchange Rates and
their Volatility Affect Trade? OECD Trade Policy Papers. No. 119. OECD
Publishing.
Kenen, P. T., & D. Rodrik. (1986). Measuring and Analyzing the Effects of Short-term
Volatility in Real Exchange Rates. Review of Economics and Statistics. Vol. 68. No.
2: 311-315.
Krugman, P & M. Obsfeld. (2003). International Economics: Theory and Policy. 6th Edition,
New York: Addison Wesly.
Learmer, E.E & R.M. Stern. (1976). Quantitative International Economics. Chicago: Adline
Publishing Company.
Exchange Rate and International Trade: Case...
Anung Yoga Anindhita
266 http://journal.uinjkt.ac.id/index.php/signifikan
DOI: 10.15408/sjie.v6i2.5210
Marquez, J. & J. Schindler. (2007). Exchange-rate Effects on China’s Trade. Review of
International Economics, 15(5): 837–853.
Oskooee, M.B. & S.W. Hegerty. (2009). The Effects of Exchange-Rate Volatility on
Commodity Trade between the United States and Mexico. Southern Economic
Journal. Vol. 75. No. 4: 1019–1044.
Oskooee, M.B. & Y. Wang. (2008). Impact of Exchange Rate Uncertainty on
Commodity Trade Between the US and Australia. Australian Economic Papers,
Vol. 47, No. 3: 235-258.
Prusty, S. (2008). An Analysis of Exchange Rate and Export Growth in India. The
Business Review. Cambridge. Vol. 9. No. 2: 139-144.
Salvatore, D. (1997). Ekonomi Internasional (International Economics). Jakarta: Erlangga
Siregar, R. & S.R. Ramkishen. (2003). Impact of exchange rate volatility on Indonesia’s
tradeperformance in the 1990s. Journal of Japaneese and International Economics.
Vol. 18: 218–240
Suardhini, M. & M.S. Goeltom, (1997). Analisis Dampak Intervensi Bank Sentral dalam
Penetapan Nilai Tukar terhadap Ekspor-Impor Indonesia (The Impact Analysis
of Central Banks’ Intervention to Exchange Rate Determination of Indonesian
Export-Import). Jurnal Ekonomi dan Keuangan Indonesia. Vol. XLV. (1): 97-212.
Tambunan, T. (1998). Krisis Ekonomi dan Masa Depan Reformasi (Economic Crisis and The
Future of Reformation). Jakarta: LPFE-UI.
