ERIA Disc ussion Paper S eries
The Effect of Exchange Rate Volatility
on International Trade in East Asia
Inter-Disciplinary Studies Center, Institute of Developing Economies, Japan
Faculty of Economics, Keio University, Japan
Economic Research Institute for ASEAN and East Asia, Indonesia
Abstract: In this paper, we empirically investigate the relationship between exchange
rate volatility and international trade, focusing on East Asia. Our findings are
summarized as follows: first, intra-East Asian trade is discouraged by exchange rate
volatility more seriously than trade in other regions. Second, one important source for
the discouragement is that intermediate goods trade in international production
networks, which is quite sensitive to exchange rate volatility compared with other types
of trade, occupies a significant fraction of East Asian trade. Third, the negative effect
of the volatility is greater than that of tariffs and smaller than that of distance-related
costs in East Asia.
Keywords: Exchange rate volatility; Trade; East Asia
JEL Classification: F10; F31; N75
Since the Asian currency crisis in 1997, the debate on the exchange rate regime has
taken center stage in East Asia. Because the rigid dollar-pegged rate regime was alleged
to be a direct cause of the crisis, East Asian countries began to fear an excessive
dependency on the US dollar. At the same time, exchange rate stability came to be
seen as a key issue. Particularly in East Asia, international production/distribution
networks in machinery industries have developed vigorously and have established their
significance in each economy with extensive country coverage and structural
sophistication. This development of international networks has led to a rise in the
share of intra-regional trade in East Asia and has necessitated a stable exchange rate
environment. One of the natural consequences of this has been the commencement of
vigorous discussion on the possibility of a basket currency among East Asian countries.
The academic literature neither theoretically nor empirically concludes whether the
stability of the exchange market enhances international trade or not. There are a large
number of theoretical and empirical studies that analyze the relationship between
exchange rate volatility and international trade (see, for example, McKenzie, 1999;
Clark et al., 2004). As presented in McKenzie (1999), there are theoretical models
supporting both negative and positive relationships between them. Empirical studies
do not provide clear-cut results, either. Most of the empirical results present a negative
relationship, but this relationship is not always robust. The studies often find
insignificant negative or positive relationships when employing other estimation
methods such as instrument variable estimation or the introduction of fixed country
Previous empirical studies have investigated various hypotheses and subjected them
to robustness checks. Some of the studies perform long-time series analyses and
employ samples involving a large number of countries. Various kinds of volatility
measures are employed in the literature, and furthermore the volatility is sometimes
decomposed into its anticipated part and unanticipated part by using a GARCH model.
The endogeneity between exchange rate volatility and trade is addressed in the use of
instruments. The studies also compare the impact of volatility on trade among
developed countries with that among developing countries. These studies aim to
examine the differences in the currency/exchange system, or the availability of hedging
instruments across countries, through investigating the impact of exchange rate
volatility on international trade. Recently, moreover, Clark et al. (2004)1
The fourth element is especially important in the context of East Asia. The
seminal paper in the fragmentation theory, Jones and Kierzkowski (1990), illustrates the
mechanics of fragmentation in a static setting. It claims that fragmentation of
production processes takes place when (i) production cost per se in fragmented
production blocks can be substantially reduced and (ii) service link cost for connecting
remotely located production blocks is not prohibitively high. If a reduction in
production cost by fragmentation overweighs service link cost incurred thereby, the firm
breaks apart some of its production blocks to other remote locations, so as to attain a
total cost reduction. In dynamic consideration, however, we must explicitly take into
account the cost of network set-ups and network restructuring. Apart from pure
spot-market-type transactions, transactions in production networks are relation-specific.
the impact of volatility on trade in differentiated goods with that in homogenous goods.
Our study intends to contribute to the literature by clarifying differences in the
impact of exchange rate volatility among traded products or across trade structures.
Particularly in the context of East Asia, we conduct the following analysis: Firstly, we
examine whether volatility has a greater discouraging impact on trade in East Asia than
in other regions. Secondly, we try to quantify the degree to which volatility impedes
international trade in East Asia compared with tariffs and distance-related costs (e.g.,
transportation costs). Thirdly, we construct an unanticipated volatility measure
different from those used in previous literature and examine its impact on international
trade. Different from the volatility measures employed in the previous studies, this
unanticipated volatility measure is constructed by using not only the past exchange rates
but also the prospect of countries held by exchange market players (bankers). Fourthly,
we examine whether machinery parts trade is more sensitive to volatility than finished
1 Their finding of a larger impact on differentiated goods trade indicates that exchange rate volatility
occupies a significant fraction of fixed entry costs since, from the theoretical point of view,
differentiated goods trade is more sensitive to such costs.
Only a slight defect or delay of one single part may cause a serious malfunctioning of
the whole production system, and thus firms carefully select credible business partners
hooked up with reliable service links. Exchanges rates are one of the crucial elements
that generate uncertainty in the competitiveness of business partners as well as service
link costs. As a consequence, firms or production plants located in countries with high
volatility in exchange rates are less likely to be incorporated with production networks.
Indeed, some Japanese firms report that exchange rate stability is essential for
back-and-forth transactions of intermediate goods in international production networks
(Ito et al., 2008)2. On the other side of the coin, once production networks are in place,
transactions become stable, which suggests the existence of network restructuring cost.3
2 The Research Institute of Economy, Trade and Industry (RIETI) conducted hearings on strategies
for exchange risk management with a number of Japanese machinery firms as a part of their project
on “The Optimal Exchange Rate Regime for East Asia”. Ito et al. (2008) summarizes their results.
3 Obashi (2008) rigorously verifies that machinery parts and components trade in East Asia are more
stable over time than machinery finished products by employing the method of the survival analysis.
There are also the previous studies with special attention on East Asia: for example,
Bénassy-Quéré and Lahrèche-Révil (2003), Poon et al. (2005), Chit el al. (2008), and
Thorbecke (2008). While these paper employs different approaches such as an
error-correction model and panel data techniques, different sample period, and different
sample countries, all these papers found the negative relationship between exports and
exchange rate volatility in East Asia. Our paper is in particular closely related to
Thorbecke (2008). He investigates how exchange rate volatility affects electronic
parts and components exports within East Asia and finds that the volatility does reduce
trade in electronic parts and components within the region. On the other hand, this
paper examines whether there are differences in the impacts of the volatility between
finished machinery goods and machinery parts. This investigation contributes to
enhancing our understanding on the mechanics of international production/distribution
The reminder of this paper is organized as follows: Section 2 explains our empirical
methodology and an overview of our volatility measure. Section 3 reports on our
regression results, and Section 4 concludes our argument.
2. Empirical Issues
This section offers an outline of our empirical methodology for testing the
relationship between trade and exchange volatility. Data issues and their overview
2.1. Empirical Methodology
It is well known that gravity equations can be supported by various kinds of
theoretical models. Taking advantage of this property, a large number of researchers
have performed a gravity analysis in order to carry out empirical investigations on
correlations between international trade and the variables concerned. Following the
recent literature on exchange volatility, this paper also employs a gravity equation
The baseline equation is shown by:
ln Tij = β0 +β1 ln GDPi + β2 ln GDPj + β3 ln distanceij + β4 volatilityij
+ β5 languageij + β6 adjacencyij + β7 colonyij + β8 comcolij + εij.
The time subscript t is omitted in this equation. Tij represents real export values of
country i to country j. GDPi denotes real gross domestic product in country i.4
The literature has applied various kinds of variables for exchange rate volatility,
volatilityij. In this paper, following Rose (2000), we primarily use a widely-used
indicator, the real exchange rate volatility, which is constructed as the standard
deviation of the first-difference of the monthly natural logarithm of bilateral real
distanceij is the geographical distance between countries i and j. languageij is an
indicator variable taking the value unity if a common language is spoken by at least 9%
of the population in both countries i and j, and zero otherwise adjacencyij takes the
value of one if the two countries are contiguous and zero otherwise. εij is a disturbance
4 To our best knowledge, it is an irreconcilable issue which is better in gravity analyses, nominal
variable or real variable. Although the real variables are employed in the reported tables, we have
confirmed that the results are qualitatively unchanged even in the case of the nominal variables.
exchange rates in the five years preceding period t. A number of other indicators are
also introduced in our robustness checks.
We first estimate the above gravity equation for bilateral trade values in the world
by the ordinary least squares (OLS) method. Then, by introducing an East Asia
dummy interacting with the real exchange rate volatility, we examine the impact of
exchange rate volatility on trade in East Asia relative to that in the other regions. The
East Asia dummy takes the value unity if both countries i and j are East Asian countries
and zero otherwise. Next, by restricting our sample to intra-East Asian trade, we
investigate more closely the impact of exchange rate volatility on trade. By
introducing tariffs as an independent variable, we quantify the degree of significance of
the effect of exchange rate volatility on East Asian trade compared with that of tariffs.
In addition, we decompose the real exchange rate volatility into an anticipated volatility
that is predicted from economic and social conditions and an unanticipated volatility as
the residual, both of these being introduced as explanatory variables. Finally, to verify
the importance of stable transactions of intermediate goods in the formation of
production networks, we regress the gravity equation for trade in finished machinery
goods and trade in machinery parts separately.
2.2. Data Overview
Our sample includes bilateral trade between 60 countries (see Appendix) from 1992
to 2005. Data on international trade values are obtained from UN Comtrade. The HS
code list of parts and components is drawn from Ando and Kimura (2005). Data on
GDP come from the World Development Indicator (World Bank). GDP is deflated by
the U.S. wholesale price index, which is also from the World Development Indicator.
The source of distanceij, languageij, and adjacencyij is the CEPII website. The
nominal exchange rate (monthly) is drawn from IFS (af) and is deflated by the monthly
consumer price index, which is also from IFS.
Figure 1 depicts changes in the simple average of real exchange rate volatilities
among countries in each region. Large volatility is apparent in Latin America in the
first half of the 1990s. While the volatility there subsequently declined rapidly,
volatility in East Asia began to rise in 1998. As a result, by around 2000, East Asia
had the largest degree of volatility in the world. Volatility in Africa has been relatively
large since the first half of the 1990s, while that in Europe has been relatively small.
Figure 1. Changes in Real Exchange Rate Volatility by Region
1992 19931994 1995 19961997 1998 1999 2000 2001 20022003 20042005
East Asia EuropeLatin America AfricaEA (Unanticipated)
Source: Authors’ calculation
In Figure 1, “unanticipated” indicates unanticipated exchange volatility in East Asia.
The unanticipated exchange volatility is constructed as follows: Firstly, we regress the
following equation by the OLS method:
volatilityij,t = α0 +α1 ln Riski,t-5 +α2 ln Riskj,t-5 + ςij, (1)
where Riski denotes country risk in country i. Secondly, by using estimates of α0, α1,
and α2, we can obtain the residual of each observation. Finally, we define the
unanticipated volatility as the absolute value of the residual. That is, the unanticipated
volatility is defined as the mass of exchange volatility not predicted by the country risk
for each of the countries. As a proxy for the country risk, we use the country risk
index, which is drawn from Institutional Investor (Institutional Investor, various issues).
This index is the aggregate of bankers’ evaluations on the risk of default, and a larger
index indicates that the country has a smaller risk of default. As shown in Figure 1,
together with the “total” volatility, the unanticipated volatility in East Asia rose from
1998. This rise seems to reflect the currency crisis.
3. Empirical Results
In the following, we first present baseline results regarding the several hypotheses
listed in section 2.1. Following this, the results of various kinds of robustness checks
3.1. Baseline Results
3.1.1. East Asia versus the World Countries
Table 1 reports the regression results obtained using our full sample. Columns
two and three show the values for all manufactured goods and machinery products,
respectively. Almost all coefficients are estimated to be significant with the expected
signs. Large GDP of importers and exporters, and short distances between trading
countries encourage international trade. Our main interest in this paper lies in the
results concerning the exchange rate volatility, for which coefficients are significantly
negative in both columns. The negative coefficients imply that large volatility
discourages international trade in both manufactured goods and machinery products in
the world. This result may reflect the fact that exchange rate volatility generates a
significant fraction of the fixed costs for trading activities.
The East Asian slope dummy is introduced into our equation, as shown in columns
four and five of Table 1. The results for most of the usual coefficients are unchanged
from the previous results. The slope coefficients are significantly negative, implying
that intra-East Asian trade is more seriously discouraged by exchange rate volatility
than trade in other regions. The immaturity of the international exchange market and
of hedging instruments may account for the creation of this more serious impact on East
Asia. We will observe later that, in addition to the immaturity of the financial sector,
the mechanics of machinery trade contribute to this result.
Table 1. Results of Full Sample Regressions
importer's GDP 1.412***
-1.203*** -1.037*** -1.203*** -1.038***
-1.977*** -1.626*** -1.771***
* East Asia
Notes: Heteroskedasticity-consistent standard errors (White) are in
parentheses. ***, **, and * show 1%, 5%, and 10% significance,
3.1.2. The Impact of Unanticipated Volatility on East Asian Trade
In Table 2, we narrow down our sample to intra-East Asian trade. Looking at the
results in columns two and three, tariffs, of which data are available in the Handbook of
Statistics (UNCTAD), are introduced as an independent variable, ln (1+tariffs(%)/100),
and their coefficients are estimated to be negative. The coefficient for adjacency turns
out to be positive and significant, though that for colony is estimated to be negative.
The coefficients for exchange rate volatility are significantly negative, and their
magnitude is at almost the same level as that of the results in Table 1. The results in
the other variables are qualitatively unchanged from the previous ones in Table 1,
though the magnitude of the coefficients is slightly decreased.
Table 2. Regression Results for Intra-East Asian Trade
-0.400*** -0.432*** -0.439*** -0.468*** -0.332*** -0.491*** -0.393*** -0.524***
[0.047][0.066] [0.044] [0.059]
-4.429*** -5.012*** -2.660*** -3.248***
unanticipated volatility-11.989*** -17.844***
colony-0.444*** -0.249*-0.616*** -0.458***
Year dummy YesYes Yes
Disc (distance)19% 22%7%
Wald test (p-value)
R-sq 0.70680.6223 0.7722
Notes: See notes to Table 1. The null hypothesis in the Wald test states that the coefficient for
volatility/unanticipated volatility between the final goods and parts equations is identical.
0.601*** 0.419*** 0.319*** 0.119***
0.960*** 0.788*** 0.668*** 0.524***
[0.076][0.078] [0.064] [0.070]
0.629*** 0.623*** 0.501*** 0.542***
1.464*** 1.103*** 0.638*** 0.443***
How large is the trade impediment caused by exchange rate volatility, compared
with other trade impediments such as tariffs? Our volatility measure is a form of
standard deviation, and thus direct interpretation of the magnitude of its coefficients is
difficult. In order to estimate its magnitude intuitively, we quantify the seriousness of
the effect of exchange rate volatility on East Asian trade, compared with the effects of
distance-related costs and tariffs, by calculating the following measures
where mean (i) denotes the mean value of variable i. The results are reported in the
rows, Disc (distance) and Disc (tariff). We can conclude here that exchange rate
volatility on average discourages international trade by a factor of 0.2 and has twice the
impact of distance-related costs and tariffs, respectively. The finding that exchange
rate volatility penalizes East Asian trade more seriously than one of the most
well-known impediments, tariffs, is important, even though tariffs have already been
lowered substantially in East Asia.
Columns four and five in Table 2 report regression results obtained using the
unanticipated volatility measure.5
The equations in the columns also include importer
Here we regress the gravity equation for trade in finished machinery goods and
trade in machinery parts and components separately in East Asia.
and exporter risk indices. The coefficients for both risk indices are significantly
positive, implying that the lower the risk of default in trading countries, the more
international trade occurs between them. The coefficients for unanticipated volatility
are also estimated to be significant in both columns. The negative coefficients here
indicate that, in addition to country risk, the existence of exogenous factors creating
exchange rate volatility reduces manufacturing and machinery trade. Furthermore,
unanticipated volatility on average has a slightly larger discouraging impact on trade
than tariffs. Comparing with the results in columns two and three, we conclude that a
significant portion of the negative impact of exchange rate volatility is induced by its
3.1.3. Finished Goods Trade versus Intermediate Goods Trade
5 Exporters would predict future volatility by using a lot more information. To examine the impact
of more purely random shocks in exchange rates, we also estimated the expected volatility by
introducing not only country risk but also fixed importer and exporter effects and year dummies.
But, the results were qualitatively unchanged.
6 The tariff rates are dropped from the regression equations because the ready-made tariff data are
not available for machinery final goods and machinery parts separately.
To formally test
whether exchange rate volatility has a different impact on finished products and parts,
we conduct the Wald test using the null hypothesis, which states that the coefficients are
identical in both equations. These results are reported in columns six to nine in Table
There are three points to note. First, as above, standard gravity variables such as
GDPs are estimated with the expected signs. Second, coefficients for total volatility
and unanticipated volatility are again significantly negative. Exchange rate volatility is
a significant impediment to trade in both finished machinery goods and machinery parts.
Third, and most interestingly, the Wald tests reject the null hypothesis at the 1% level of
significance. This implies that trade in machinery parts is more sensitive to exchange
rate volatility than trade in finished machinery goods, which indicates that stable
transactions of parts are crucial to the formation of production networks.
3.1.4. Simulation Analyses
Here we perform simple simulation analyses by using the results in Table 2. We
simulate the average growth of exports by an East Asian country by reducing its sample
mean level of real exchange rate volatility (0.037) to the ECU level and the EURO level.
The mean level in the ECU countries during the period 1992-1998 (0.019) is used as the
ECU level, and that in the EURO countries during the period 1999-2005 (0.010) as the
EURO level. Although those levels are not necessarily achieved only by the
introduction of the ECU and EURO, we simply apply those levels for East Asia, which
can possibly be interpreted as the effect of introducing an East Asian basket currency or
an East Asian common currency, respectively. The case of complete elimination of
mean volatility in East Asia (0.037) is also simulated. These hypothetical scenarios are
compared with the case of a complete reduction of the sample means for tariffs (0.066
for manufacture; 0.060 for machinery).
The simulation results are reported in Table 3. For example, the simulation result
of “ECU” in “Manu” is derived from the following calculation: 100*[exp((0.019 –
0.037) * (-15.960)) - 1] %. “-15.960” is the estimate of volatility shown in column two
of Table 2. In all scenarios, the magnitude of the effects is not small at all. The
introduction of a common currency almost doubles machinery exports. In the case of
complete elimination of exchange rate volatility, which would be impossible in the real
world though, the increase in machinery exports is more than twice. In addition, an
introduction of a common currency has a larger impact than the achievement of free
trade (zero tariffs). We can thus conclude that the introduction of a basket currency or
a common currency would substantially contribute to enhancing the magnitude of
international trade in East Asia.
Table 3. Simulation: Rise of Exports in Each East Asian Country
Notes: The simulation scenario involves the reduction of exchange rate
volatility to its level in ECU countries and in EURO countries respectively. In
addition, the simulation result of complete elimination of mean volatility in East
Asia is also reported (ALL). The mean of volatility in ECU countries (92-98)
and EURO countries (99-05) is 0.0194559 and 0.0101294, respectively. The
mean of volatility in East Asia (92-05) is 0.037187.
3.2. Robustness Checks
In almost all previous studies, the negative effect of exchange rate volatility on
trade is not found to be robust, which is quite contrary to our results. To confirm our
strong results, we perform various kinds of robustness checks, which are reported in
Tables 4 to 7.
Firstly, in Tables 4 and 5, we introduce importer-year and exporter-year dummy
variables in order to control time-variant country-specific characteristics. In this
specification, the effects of multilateral resistance terms, which is one of the serious
issues in gravity analysis, are also controlled. With our full sample (Table 4), the
coefficients for volatility are still negatively significant. However, those for East Asia
slope dummy turn out to be insignificant while those are still negatively estimated. In
East Asian sample (Table 5), moreover, volatility coefficients themselves are not
significant. As a result, controlling time-variant country-specific elements, e.g., the
immaturity of international exchange market and hedging instruments, in each East
Asian country, we cannot observe the significant negative impacts of exchange rate
volatility in East Asia. To put it the other way around, we can say that the sources of
its negative impacts in East Asia are time-variant country-specific elements. In
addition, the absolute magnitude of the volatility coefficient is larger in parts equation,
though that is insignificant in both equations and the Wald test does not reject the null
Table 4. Robustness Checks of Full Sample Regression:
Importer-year and Exporter-year Dummies
-1.922*** -1.686*** -1.922*** -1.686***
-8.495*** -6.293*** -8.533*** -6.365***
* East Asia
-2.027*** -1.154*** -1.961*** -1.031***
-1.782*** -1.088*** -1.782*** -1.089***
Note: See notes to Table 1.
Table 5. Robustness Checks of East Asian Sample Regression:
Importer-year and Exporter-year Dummies
-0.371*** -0.324*** -0.270*** -0.393***
-0.466*** -0.362*** -0.454*** -0.376***
Wald test (p-value)
Note: See notes to Table 2.
Secondly, to further explore volatility measures, we attempt to employ two kinds of
volatility measures: GARCH volatility and nominal volatility. In the subsequent
regressions, we restrict our sample to East Asia. By following Clark et al. (2004), the
former is the conditional volatilities of the exchange rates estimated using a GARCH
(1,1), and the underlying equation for the model is an ARIMA (0,1,0) process of the log
of real exchange rates. The last estimated conditional standard deviation of each
country pair is used as the approximation of the conditional volatility at the beginning
of next period.7
The latter is the standard deviation of the first-difference of the
7 For instance, the conditional volatility of 2000 equals the estimated conditional standard deviation
for December 1999 in the GARCH regressions.
monthly natural logarithm of bilateral nominal exchange rates in the five years
preceding period t. The results are reported in Table 6. Similar to Tables 4 and 5,
time-variant country characteristics are controlled. However, contrary to them, there
are the estimators of volatility coefficients to be negatively significant at ten percent
level. Thus, we can conclude that, in some volatility measures, the source of the
negative impacts of exchange volatility still remains even if we control time-variant
Table 6. Robustness Checks of East Asian Sample Regression:
Two Volatility Measures
-0.369*** -0.326*** -0.270*** -0.394***
MachineFinal Parts Manu
-0.378*** -0.329*** -0.276*** -0.398***
0.201*** 0.425*** 0.362*** 0.469***
0.581*** 0.547*** 0.559*** 0.501***
-0.475*** -0.368*** -0.462*** -0.383***
0.360*** 0.292*** 0.420*** 0.277***
-0.483*** -0.385*** -0.464*** -0.409***
[0.129] [0.127] [0.128]
Wald test (p-value)
Note: See notes to Table 2.
GARCH Volatility Nominal Volatility
Finally, here we regress the first-difference logarithmic form of the gravity equation
with time-variant importer and exporter dummy variables. In this paper, as in almost
all gravity studies, we have completely disregarded potential stationarity issues. If our
variables are integrated of order one, such a first-difference logarithmic form of the
gravity equation would be appropriate. In the equation, to control effects of exchange
rate changes per se on trade, we introduce the first-difference logarithm of real
exchange rates between trading countries: the larger this variable is, the more rapidly
the exporter’s currency is devaluated.8
The results are reported in Table 7. The
8 To assure comparability, we introduced the percentage change in exchange rates. The
internationally comparable exchange rates are available in the CEPII-CHELEM database.
coefficients for exchange rate change have significantly positive sign in manufacturing
and parts equations. The coefficients for volatility change are no longer estimated to
be significantly negative. Notice that not only time-variant country characteristics but
also time-invariant country pair characteristics are controlled in these regression
equations. Thus, we can say that time-variant pair elements are not a significant part
of the negative impacts of exchange volatility on trade, though we cannot well interpret
the positive coefficients.9
exchange rate change
Wald test (p-value)
Table 7. Robustness Checks on East Asian Sample Log-difference Regression:
Note: See notes to Table 2.
4. Concluding Remarks
In this paper, we empirically investigated the relationship between exchange rate
volatility and trade, focusing on East Asia. Our findings are summarized as follows:
first, intra-East Asian trade is discouraged by exchange rate volatility more seriously
than trade in other regions. Second, one important source for the discouragement is
that intermediate goods trade in international production networks, which is quite
sensitive to exchange rate volatility compared with other types of trade, occupies a
significant fraction of East Asian trade. Third, the negative effect of the volatility is
greater than that of tariffs and smaller than that of distance-related costs in East Asia.
Fourth, the sources of such negative impacts of the volatility are time-variant
country-specific elements. Last, our simulation analysis shows that the introduction of
a basket currency or a common currency would have a larger positive impact on
international trade than free trade.
In interpreting our results, we may need to consider a link between the reduction of
exchange rate volatility and foreign direct investment (FDI). Kiyota and Urata (2004)
9 Conducting dynamic least squares as a further robustness check, we obtain basically the same
show a significant negative relationship between exchange rate volatility and Japanese
FDI to East Asian countries. The introduction of an East Asian basket currency or an
East Asian common currency may induce a substantial increase in international goods
trade, together with a further encouragement of FDI in East Asia.
Appendix. Sample Countries
AfricaAlgeria, Ghana, Kenya, Madagascar, Mauritius, Morocco, Senegal, Seychelles
East Asia China, Hong Kong, Indonesia, Japan, Korea, Malaysia, Philippines, Singapore, Thailand
Austria, Denmark, Finland, France, Greece, Hungary, Iceland, Italy, Malta, Netherlands, Norway,
Portugal, Spain, Sweden, Switzerland, United Kingdom
Argentina, Barbados, Bolivia, Brazil, Chile, Colombia, Costa Rica, El Salvador, Guatemala,
Honduras, Jamaica, Mexico, Panama, Paraguay, St.Lucia, Trinidad and Tobago
Canada, United States, Fiji, Cyprus, India, Israel, Jordan, Nepal, Pakistan, Saudi Arabia, Sri
Ando, M., Kimura, F., 2005. The formation of international production and distribution
networks in East Asia. In: Ito, T. and Rose, A. (Eds.). International Trade in East
Asia, Chicago: University of Chicago Press: 177-213.
Bénassy-Quéré, A., Lahrèche-Révil, A., 2003. Trade linkages and exchange rates in
Asia: The role of China. Working paper No. 2003-21, CEPII.
Chit, M., Rizov, M., Willenbockel, D, 2008. Exchange rate volatility and exports: New
empirical evidence from the emerging East Asian Economies. Middlesex
University Economics and Statistics Discussion Paper No. 127.
Clark, P., Tamirisa, N., Wei, S., Sadikov, A., Zeng, L., 2004. A new look at exchange
rate volatility and trade flows. International Monetary Fund, Washington DC,
Occasional Paper, No. 235.
Ito, T., Koibuchi, S., Sasaki, Y., Sato, K., Shimizu, J., Hayakawa, K., Yoshimi, T., 2008.
Choice of invoice currency and exchange risk management: Case studies of
Japanese firms. RIETI Discussion Paper Series 08-J-009 (in Japanese).
Jones, R. W., Kierzkowski, H., 1990. The role of services in production and
international trade: a theoretical framework. In: Jones, R. W. and Krueger, A. O.
(Eds.). The Political Economy of International Trade: Essays in Honor of R. E.
Baldwin, Oxford: Basil Blackwell.
Kiyota, K., Urata, S., 2004. Exchange rate, exchange rate volatility and foreign direct
investment. The World Economy 27(10), 1501-1536.
McKenzie, M., 1999. The impact of exchange rate volatility on international trade flows.
Journal of Economic Surveys 13(1), 71-106.
Obashi, A., 2008. Stability of production networks in East Asia: duration and survival of
Poon, W., Choong, C., M. Habibullah, M., 2005. Exchange rate volatility and export for
selected East Asian countries: Evidence from error-correction model. ASEAN
Economic Bulletin 22(2), 144–159.
Rose, A., 2000. One money, one market: The effect of common currencies on trade.
Economic Policy, April, 9-45.
Thorbecke, W., 2008. The effect of exchange rate volatility on fragmentation in East
Asia: Evidence from the electronics industry. Journal of the Japanese and
International Economies 22(4), 535-544.
ERIA Discussion Paper Series
No. Author(s) Title Year
and Ayako OBASHI
International Production Networks:
Comparison between China and ASEAN
The Effect of Exchange Rate Volatility on
International Trade in East Asia
Ikumo ISONO, and
Predicting Long-Term Effects of Infrastructure
Development Projects in Continental South East
Asia: IDE Geographical Simulation Model
Firm-level Analysis of Globalization: A Survey