Tác động của tỷ giá đến thương mại song phương của việt nam với trung quốc, mỹ và nhật bản nhìn từ góc độ ngành tt tiếng anh

1 INTRODUCTION Rationale for research Vietnam’s fledgling and open economy is currently going through the process of modernization and industrialization and export is considered one of the key engines of economic growth for the country Indeed, Vietnam has accomplished a number of significantly fruitful achievements with its export-oriented industrialization strategy Particularly, the annual average growth rate of GDP remained stable at 7% while the growth rate of export constantly reached 18% per year over the period 1990-2016 However, the growth rate of imports was also reported to be greater than 22% in the same period Accordingly, Vietnam experienced a trade deficit in most of the period from 1990 to 2016, although it gained a trade surplus in the recent years – 2017 and 2018 – which was still modest in value Many studies have investigated and tried to explain the root causes of Vietnam's trade deficit and exchange rate is considered one of the important determinants of the trade balance, along with other macroeconomic variables Especially, the country has undergone a period of macroeconomic instability since 2007 Therefore, exchange rate management policies designed to promote exports and stabilize the macro-economy are regarded a matter of great concern in the current context of Vietnam For Vietnam, so far, there have been very few quantitative studies examining the impacts of exchange rates on trade at the industrial level In addition, these studies have not taken into account the assumption of the asymmetric response of the trade balance to exchange rate movements Therefore, this study will examine the impacts of exchange rate movement variability on Vietnam's trade Specifically, the main purpose of the research is to explore the relationship between exchange rates and bilateral trade performance between Vietnam and its major trading partners including China, the US, and Japan by using industry-level data These three countries were exclusively chosen because they have been reported as the three leading trading partners of Vietnam with a total import-export turnover accounting for up to 40% of the total import turnover of Vietnam in 2015 According to the General Department of Customs, the US, Japan, and China are the three largest export markets of Vietnam with a total export value of over $US10 billion, making up 42.5% the country's total export value Meanwhile, the three largest import markets are China, South Korea, and Japan with a total import value of over $US 10 billion, occupying 55% of Vietnam's total export value This study seeks to provide a more comprehensive picture of the effects of exchange rates on Vietnam's trade balance with its major partners by using disaggregated data to consider whether the exchange rates are related to the causes of Vietnam’s trade deficit with China and its surplus with other major trading partners (the USA, Japan) From the thesis's findings, the author will draw some implications for the exchange-rate policy of Vietnam Research Objectives The objectives of the thesis are as follows: - Investigating the impacts of exchange rates on the trade balance of Vietnam - Proposing some recommendations regarding the exchange-rate policy framework to improve the trade balance of Vietnam Research questions What are the impacts of exchange rates on Vietnam's trade balance with China, the US, and Japan as a whole? What are the impacts of exchange rates on Vietnam's trade balance with China, the US, and Japan for each commodity group? Is the theory of the J-curve effect supported in Vietnam's trade relationship with its major partners? Do exchange rates have asymmetric effects on Vietnam's trade balance? Subjects and Scope of the thesis - Research subjects: Vietnam's trade balance with the three partners, bilateral exchange rates, Vietnam's import and export performance in each industry or for each commodity group - Research scope: The study mainly uses data in the period 2007-2018 for the 21 industries in which trade between Vietnam and the three partners took place In addition, due to the limited access to unofficial trade data, the models used to examine the relationship between exchange rates and trade balance for Vietnam and China will ignore the issues related to cross-border trade Research methods The study employs both qualitative and quantitative research methods Regarding the quantitative method, this study uses the statistical analysis method, an autoregressive distribution lag (ARDL) model and a nonlinear autoregressive distribution lag (NARDL) model to estimate the impacts of exchange rates on trade balance in both the long run and the short run New contributions of the doctoral thesis New contributions to existing academic literature and theories Firstly, an overview of previous empirical studies has shown that the use of either aggregate trade data or disaggregate trade data at the bilateral level may suffer from the aggregation bias problem as exchange rates may exert significant impacts on some particular industries or commodities It might not bring about any/or just less robust effects, sometimes even resulting in undesirable outcomes, to some other commodities or industries Hence, using aggregate trade data may lead to mixed results, depending on which commodity groups are given the dominant positions In the case of Vietnam, there is a limited number of studies employing disaggregate data for modelling at the industrial level Secondly, a prevalent assumption usually adopted in previous studies is that exchange rates will exert symmetrical impacts on the trade balance (i.e the reactions of trade balance to changes in exchange rates in the case of currency depreciation would be the same as that of currency appreciation) Recently, there are some arguments claiming that commercial producers and traders may respond differently to currency depreciation and appreciation, or, exchange rates may have asymmetrical impacts on trade balance To date, there has been no research using an asymmetric non-linear model to examine asymmetric impacts of exchange rate on Vietnam's trade balance CHAPTER THEORETICAL ISSUES ON EXCHANGE RATES AND THE RELATIONSHIP BETWEEN EXCHANGE RATES AND TRADE BALANCE Recommendations based on the study findings 1.1 Definition of exchange rates and trade balance This study employs a set of autoregressive distribution lag (ARDL) and non-linear ARDL (NARDL) models to examine the relationship between exchange rates and the trade balance of Vietnam with its biggest trading partners including the United States of America, Japan and China for 21 key commodity groups classified by the Harmonized System (HS) In this study, the data is collected from many credible sources from January 2008 to December 2017 The empirical results have revealed some issues as follows: 1.1.1 Exchange rates Firstly, the estimation results of the ARDL and NARDL models reported a degree of bias to the regression estimates when using aggregate data and the asymmetric approach Secondly, the impacts of exchange rates on trade balance have differed across trading partners A depreciation will tend to improve Vietnam's trade balance with Japan while worsening Vietnam's trade balance with the US Meanwhile, the trade balance between Vietnam and China appears to be insensitive to real bilateral exchange rates In conclusion, the use of exchange rates as a tool to improve the trade balance of Vietnam should be considered with great caution and take into account different priority targets as well as the coordination between different policies More importantly, in the context of a developing country as Vietnam the manufacturing sector depends heavily on imported intermediate products and the import replacement capacity is relatively weak, a depreciation may lead to undesirable effects of exchange rates on trade balance Contents of the thesis There are five chapters in this study as follows: Chapter 1: Theoretical issues on exchange rates and the relationship between exchange rates and trade balance Chapter 2: Literature review and theoretical framework for analysis Chapter 3: The practice of exchange rates and trade relations between Vietnam and the three partners Chapter 4: Results and discussion Chapter 5: Conclusion and policy recommendations 1.1.2 Trade balance 1.2 The theoretical relationship between exchange rates and trade balance 1.2.1 Relationship between exchange rates and trade balance in the standard theory of international trade Standard Trade Theory demonstrates the relationship between real exchange rates and trade balance following a simple common-sense approach A depreciation of domestic currency will lead to a decrease in the consumption of imported goods of home country's households At the same time, foreign households tend to purchase relatively more domestic goods Ultimately, the higher the real exchange rate for the home country is, the more the trade surplus the country obtains (Wei-Bin, 2013) 1.2.2 Relationship between exchange rates and trade balance: The Elasticities Approach 1.2.2.1 Bickerdike-Robinson-Metzler condition The essence of the elasticities approach is to investigate the substitution effects derived from a change in relative prices caused by currency devaluation on consumption and production Trade balance adjustment path is examined on the basis of the price elasticities of demand for imports and exports in demand functions Bickerdike (1920), Robinson (1947), Metzler (1948) were the first authors who developed and build this approach by modeling nominal import and export prices as functions of import and export quantities, and provided BRM conditions The BRM condition implies that the effect of exchange rate movements on trade balance depends on import and export supply, demand elasticity, and the initial volume of trade The sufficient condition for trade balance improvement is provided by the BRM condition as follows: ∗ ∗ ∗ ∗ ∗ ∗ > (1.9) Where and denote the price elasticities (in absolute values) of domestic demand for imports and supply of exports Analogously, ∗ and ∗ denote the respective foreign price elasticities 1.2.2.2 Marshall-Lener condition Marshall-Lerner condition was derived from the BRM condition (Marshall, 1923; Lerner, 1944) This condition (referred to here as the ML condition) comes from letting ε→∞ and ε^*→∞ In other words, it is implied that if domestic and foreign supply elasticities are strictly elastic and if income remains constant, the left side of the condition (2.9) will become θ+θ*-1 Thus, the trade balance will improve after a depreciation if θ+θ*>1 holds Or, in the standard presentation of the ML condition, a depreciation will lead to s an improvement of the trade balance when the domestic plus the foreign import demand elasticities for imports, in absolute value, exceeds one 1.2.2.3 J curve effect Houthakker and Magee (1969) proposed a reduced-form equation for the import and export demand functions that has since become the basis for studies on the relationship between exchange rates and trade balance employing the ML condition The theory of the J-curve effect suggests that the depreciation of the domestic currency will improve the trade balance in the long run However, the trade balance will deteriorate in the short-run, and trade balance movement, if graphically represented, will look like the letter J More recently, empirical studies associated with the ML condition have used the cointegration technique in analysis of exchange rate impacts on trade balance, such as those of Andersen (1993), Rose (1991), Johansen and Juselius (1990), Bahmani-Oskooee and Niroomand (1998), Upadhyaya et al (1999), Kemal and Qadir (2005)… 1.2.3 The relationship between exchange rates and trade balance: the absorption approach Other studies using single equation methods (including the ARDL model) based on the cointegration technique also give more detailed estimation results to consider the ML condition The absorption approach proposed by authors such as Harberger (1950), Meade (1951), and Alexander (1952) is a combination of the elasticities approach and Keynesian macroeconomic school of thought The core assumption of this approach is that any improvement in trade balance would require an increase in income over total domestic expenditures (Malimi, 2013) This approach indicates that a devaluation of the domestic currency will bring about the improvement of trade balance if the substitution effect of relative price changes boosts output more than absorption 1.2.4 Relationship between exchange rates and trade balance by monetary approach The monetary approach assumes that the balance of payments should be considered from the perspective of money supply and demand Accordingly, the devaluation of the currency leads to an increase in the prices of traded goods and services People will then reduce their spending relative to income in order to restore the real value of their holdings of money In brief, a reduction in consumption will ultimately result in a decline in absorption and the improvement of trade balance In general, the three theoretical approaches are complementary for each other rather than being contradictory, and each approach has its own advantages and disadvantages According to Cooper (1971), these three approaches can be considered in a temporal sequence The elasticities approach is regarded as a short-term approach; the absorption approach is applicable to medium-term analyses, and the monetary approach can be used in the case of long-term analysis CHAPTER OVERVIEW OF EMPIRICAL STUDIES ON RELATIONSHIP BETWEEN EXCHANGE RATES AND TRADE BALANCE AND RESEARCH METHODOLOGY 2.1 Overview of international studies on the relationship between exchange rates and trade balance 2.1.1 Studies on the elasticities approach 2.1.1.1 Studies employing the ML condition a) Studies employing aggregate trade data Early studies on trade elasticities to exchange rates paid little attention to exchange rates, but rather focused on export and import price indicators to examine the "substitution effects" Recent studies have used relative price elasticities to test the ML condition Some other studies estimated a single elasticity (export or import elasticity) if the elasticity value is greater than 1,the ML condition is then satisfied In addition, some studies focused on trade balance as the variable of interests rather than looking at exports and imports separately to test the ML condition b) Studies employing bilateral data A considerable number of empirical studies investigated the impact of exchange rates on trade balance at the bilateral level based on the ML condition such as Marquez (1990), Shirvani and Wilbratte (1997), Bahmani-Oskooee and Goswami (2004), Sek and Har (2014), Arunachalaramanan and Ramesh (2011), and Memarian and Shanei (2016) etc 2.1.1.2 Studies testing the J-curve effect Studies using the aggregate approach that supported the existence of the J-curve effect include Bahmani-Oskooee (1985), Rosensweig and Koch (1988), Himarios (1989)…etc Meanwhile, the J-curve effect was not accepted by research of Felrningham (1988), Demirden and Pastine (1995), Felmingham and Divisekera (1986) 2.1.1.3 Studies employing industry/commodity trade data Recent studies examining the impacts of exchange rates on trade balance have employed disaggregated trade data for individual industries These studies suggest that the use of aggregate or bilateral trade data can cause bias problems when a number of significant exchange rate impacts could be offsetted by a number of insignificant effects and vice versa (Baek, 2013) See, for example, Bahmani-Oskooee and Ardalani (2006), Baek (2013), Baek (2014), Huchet-Bourdon et al (2012) etc 2.1.2 Studies adopting thef monetary approach The literature review shows that there are relatively few empirical studies adopting these two approaches This may be due to the fact that both of the two approaches have not been developed to cope with changes in the nature of the current account balance in the postBretton Woods period Therefore, the two approaches are still quite primitive and underdeveloped (Abbas Ali et al., 2014) In regard to these different views, recent studies have developed an alternative reduced form model which incorporates the three approaches 7 2.1.3 New studies on the nonlinear relationship or asymmetric reaction of trade balance to exchange rate movements predicted to be positive REX is seen to rise when the VND depreciates, thus d should be positive A number of recent studies state that the effects of exchange rate changes on trade balance are asymmetric based on some emerging theories ssuch as the hysteresis behavior, the price stickiness, or the pricing-to-market behavior Several studies chose to employ the asymmetric approach in the recent three years such as Bahmani-Oskooee and Fariditavana (2015), Bahmani-Oskooee and Fariditavana (2016), Shin et al (2014) 2.2.3 The Econometric Procedures 2.2 Empirical studies in Vietnam and research gaps 2.2.1 Empirical research in Vietnam 2.2.3.1 The ARDL model The equation (9) is applied to estimate only long-run coefficients To estimate shortrun impacts of the variables on trade balance, especially the short-run impacts of exchange rates (to consider the J-curve effect), following Pesaran et all (2001), the equation (9) will be rewritten in the form of an error correction model as follows: ∆ ! = - + ∑203 /! ∆ln ! + ∑2034 1! ∆ 5!"#0 + ∑203 6! 5!780 + 0∆ The employment of quantitative models to study the role and impacts of exchange rates on trade in Vietnam is increasingly drawing attention from Vietnamese academics A series of recent papers by Pham (2012), Pham and Nguyen (2013), Hoang (2013), Vu (2013), Phan and Jeong (2015), Mai (2016), Nguyen Cam Nhung Tran Thi Thanh Huyen (2019), Nguyen cộng (2018), Hoang (2016) etc show different impacts of exchange rates on the trade balance of Vietnam by employing various approaches Pesaran, Shin, and Smith (2001) recommended a standard F-test for the hypotheses: I4 : 91 =92 =93 =94 =0 If the F-statistic is statistically significant, the null hypothesis I4 will be rejected or in other words, there does exist cointegration among variables 2.2.2 Research gaps 2.3.3.2 The NARDL model The research gap related to the topic of the relationship between exchange rates and trade balance in Vietnam associates with two main issues: (i) Firstly, there are very few studies examining the effects of exchange rates on trade balance using disaggregated trade data for individual industries; (ii) Secondly, no studies have considered the possibility and assumed that the effect of exchange rates on trade balance could be asymmetric In order to test the hypothesis of an asymmetric reaction to exchange rate changes, studies of Bahmani-Oskooee and Fariditavana (2016) or Arize, Malindretos, and Igwe (2017) have included new variables to describe the movements of the variable LNREX which originally denotes logarithm of real bilateral exchange rate Of which, the variable NEG specifies the decrease of the exchange rate In this study, it denotes the appreciation of the VND, and the variable POS signifies the increase of the exchange rate or the depreciation of the VND Specifically, POS and NEG are defined as follows: 2.3 Research methodology 2.3.1 Autoregressive distribution lag model (ARDL) and bound test ∑2034 1! ∆ '()! 2.3.1.1 An introduction to the autoregressive distribution lag model (ARDL) and bound test + ln 2.3.1.3 Requirements for the application and advantages of the ARDL Approach 2.3.2 The regression model ln = + "# ! +$ %! ! + & '()! + ! (9) Of which *+! represents the value of imports to Vietnam from its partners (Japan, China, US) for the commodity group/industry i, )+! represents the value of exports from Vietnam to its partners (Japan, China, US) for the commodity group/industry I "# %! ! ( ! ) is the real income of Vietnam (its partner) which are measured by the industrial production index, '()! is the bilateral real exchange rate between Vietnam Dong (VND) and the foreign currency Accordingly, b is expected to be negative while c is ! ! N3 N3 ! ! N3 N3 + μ: JKL! = M ∆ '()N = M max ∆ '()N , 2.3.1.2 The steps of the ARDL Cointegration Approach Based on the work conducted by Bahmani-Oskooee and Baek (2016), we built a model to research the relationship between exchange rates and the trade balance of Vietnam with its three partners by commodity group/industry The model is illustrated as follows: + φ< lnY:>? + φ@ lnY:AB + φC lnREX : : (S! = M ∆ '()N = M ∆ '()N , Shin, Yu, and Greenwood-Nimmo (2014) suggested to replace the variable lnREX in the error correction model (9) by the two new variables POS and NEG Thus, the equation (9) will be rewritten and named NARDL or the non-linear autoregressive distributed lag model ∆ ∑2034 W! φC POS: ! = - + ∑203 /! ∆JKL! 0 ∆ln + ∑2034 X! + φa NEG: + μ: ! 0∆ (11) + ∑2034 1! (S! 0 + ln "# ! ∆ : + ∑2034 6! 0∆ UV ! \] + φ< lnIND>? + φ@ lnIND: : + + 10 The authors proved that the bounds F-test employed in the ARDL model might be used in the NARDL model in a similar way If the estimation coefficients φC and φa have the same signs and equal in magnitude, it means exchange rates have symmetrical impacts on trade balance in the long run 2.3.3.3 Steps for estimating the models 2.3.4 Data sources In this study, the monthly trade data over the period 2005M1-2017M12 are drawn from the International Trade Centre Based on the one-digit Harmonized System, trade data is divided into 21 industries/commodities Data on real income (proxied by the industrial production index), price level and exchange rates are collected from the International Financial Statistics database of the International Monetary Fund Real exchange rate is defined as follow: J '() = ∗ × (' J In which NER is the nominal exchange rate between VND and JPY, P (J level of Vietnam (the partners) ∗ is the price (99 chapters - or 99 commodity groups) As a result, five export commodities are selected as follows: Articles of chapter 85 (including electrical machinery and equipment and parts thereof; sound recorders and reproducers, television image and sound recorders and reproducers, and parts and accessories of such articles) of the HS table code is denoted HS85 The export turnover of phones and components carried 80% of total export turnover of the articles of HS85 in 2017 Articles of chapter 84 of the HS table (Including nuclear reactors, boilers, machinery and mechanical appliances; parts thereof) is denoted HS84 The export turnover of computer and components carried 70% of total export turnover of HS84 in 2017 The textile products (HS61) accounting for 10.2% of Vietnam's exports are goods of chapter 61 and 62 in the HS code table The footwear products (HS64) are goods of chapter 64 Exports of these articles ranked third in Vietnam's key export products Articles of chapter 03 (HS03) include fish and crustaceans, molluscs and other aquatic invertebrates The export value of these commodities reached over 6,9 billion USD in 2017 3.3 The general trade performance between Vietnam with each of the three trading partners 3.3.1 Bilateral trade between Vietnam and China CHAPTER THE PRACTICE OF BILATERAL TRADE AND EXCHANGE RATES BETWEEN VIETNAM AND CHINA, JAPAN AND THE US 3.1 VND exchange rate and exchange rate policies of Vietnam 3.2 Vietnam’s export and import performance 3.2.1 Summary of Vietnam export and import records Vietnam has accomplished certain achievements in export activities and improved the trade balance in recent decades In the period of 1990-2017, total import-export turnover increased by 82 times In particular, export value increased 89 times with an average annual growth of 17.4%, import value rose 76 times with an average annual growth of 16.7% However, Vietnam still ran a trade deficit from 1990 to 2015 In the last two years, the country has managed to post a modest trade surplus Regarding the structure of import and export goods, Vietnam also experienced positive changes The country’s commodity export has been moving away from raw and semiprocessed goods to processed and higher-tech products However, enterprises with foreign direct investment (FDI) are dominating the export sector, and trade surplus is mainly contributed by FDI enterprises, while domestic enterprises are suffering deficit 3.2.2 Vietnam's major import-export markets and key export products In this section, the study presents the situation of the import and export of key export commodities of Vietnam and their major markets Key export commodities will be identified based on the data collected by UN Comtrade in 2017 and the 2-digit HS harmonization system 3.3.3.1 Trend of merchandise trade between Vietnam and China In the last decade, China has risen and become the largest trading partner of Vietnam Export and import turnover both increased sharply in the period from 1990 up to now However, it can be seen that import is growing faster than export, so the trade balance between Vietnam and China has always been in deficit for many years 3.3.1.2 Structure of Vietnam's exports to China The structure of Vietnam's exports to the Chinese market has experienced positive changes Currently, technology-intensive manufacturing products are occupying a bigger share than the others and have been growing dramatically in the market 3.3.1.3 Structure of imported goods from China The structure of Vietnam's imports from China has changed in the last 10 years, but the change is not as significant as that of exports 3.3.2 Bilateral trade between Vietnam and Japan 3.3.2.1 Trend of recent bilateral trade between Vietnam and Japan Bilateral trade between Vietnam and Japan accounts for about 10% of Vietnam's trade value As of 2017, Japan is the fourth largest partner of Vietnam From 2009 to 2017, Vietnam's exports to this market increased by 160%, while imports increased equally at approximately 140% Regarding the trade balance of the two countries, between 1995 and 2017, Vietnam achieved trade surplus for most of the period, except for 2009-2010 and the most recent three years 2015-2017 During these three years, imports have outpaced exports, causing a trade deficit in Vietnam’s bilateral trade relations with Japan 11 3.3.2.2 Structure of goods exported to Japan In 2007-2008, Vietnam's exports to Japan largely involved fuel and agricultural products which is similar to its trade with China There is a minor change in the export structure over the 10-year period from 2008 to 2018 Japan is also a potential market for some other agricultural products such as cashew, tea, handicrafts if the exported goods meet the standards of food hygiene, safety and quarantine 3.3.2.3 Structure of imported goods from Japan There has been a little change in the structure of Vietnam's imports from Japan from 2007-2008 to 2007-2017 The most important import items still include machinery, spare parts, steel, computers and components, cars…etc 3.3.3 Bilateral trade between Vietnam and US 3.3.3.1 Trend of merchandise trade between Vietnam and US Trade and investment relations of the two countries were strengthened after the two countries normalized their relations in October 1995 Vietnam's export to this market has increased by more than times, while import has increased by times However, during this period, trade between the two countries was found to grow below its potential A bilateral trade agreement between the two countries was signed in 2011 This agreement provides the needed foundation for addressing bilateral issues As a result, the US has become Vietnam's second-largest export partner since 2003 Vietnam's exports to the US reached 41.6 billion USD, increasing39 times from 2001 to 2017, while imports grew slower with 9.2 USD billion, up 22 times Accordingly, the trade surplus has increased from 654 million USD to 32.4 billion USD in the period 3.3.3.2 Structure of export goods In 2007-2008, the top Vietnamese export products to the US with a trade value of over billion USD are textiles (5 USD billion), footwear (1,075 USD billion) and wood and wood products (1,063 USD billion) There are six products with an export value of over billion USD in 2017-2018 including textiles, paper, food wear, phones and components, computers and components, wood and wood products, machinery and other spare tools These commodity groups accounted for 74% of Vietnam's total exports to the US in 2018 3.3.3.3 Structure of import goods 3.4 Real bilateral exchange rate and trade balance of Vietnam with the three trading partners 3.4.1 USD/VND exchange rate and trade balance between Vietnam and the US The real exchange rate (USD/VND) and the nominal exchange rate have been moving in opposite directions over more than a decade The real exchange rate line was above the nominal exchange rate line before 2011 After that, the nominal exchange rate line overcame the real exchange rate line and this gap has been widened Between January 2005 and January 2017, the nominal exchange rate consistently decreased by an average of 2.8% per month 12 Bui et al (2017) show that the Vietnamese currency is being overvalued The main reason is that the monetary policy of the anchor currency is not developed and implemented based on the basic principles of macroeconomics However, trade balance increased in the same direction as the nominal exchange rate, but in the opposite direction with the real exchange rate This shows that the real exchange rate and Vietnam's trade balance with the US are less likely to experience a theoretically strong relationship Therefore, we can come up with a conclusion similar to the work of Phan Thanh Thanh (2018), that is, Vietnam's trade balance is not affected by nominal factors but is determined by real and structural factors Specifically, domestic production and export structure largely depend on imported inputs as there is a serious lack of supporting industries in Vietnam, not to mention that these industries have suffered from pro-long underdevelopment 3.4.2 CNY/VND exchange rate and trade balance with China The nominal exchange rate of CNY/VND increased continuously over the period 2000-2013 The nominal exchange rate of CNY/VND depreciates by 80% from January 2005 to December 2017 Meanwhile, the real exchange rate increased by 8.3% The trade balance of the two countries during this period was quite unstable It seems to bring a similar conclusion like the case of the US, that is, Vietnam's trade balance is less affected by nominal factors or real exchange rates 3.4.3 The rate of JPN/VND and the Japanese trade balance The figures show that nominal exchange rate depreciated about 31% in the period 2005-2017, and the real exchange rate appreciated by about 50% The strengthening of the JPY since the end of 2015 has partly explained the slowdown of Vietnam's exports to this market and caused Vietnam's trade deficit to this market in 2016 and 2017 CHAPTER ESTIMATION RESULTS AND DISCUSSIONS 4.1 Estimation results of the models of 21 commodities This section demonstrated the estimation results of 21 models for 21 commodity groups classified by the Harmonized Systems The data of some commodities is not available, so their models cannot be performed 4.1.1 Diagnostic tests for the ARDL and NARDL models The results of stationary, cointegration, and other diagnostic tests are presented as follows: - - The results of the unit root test showed that most of the independent variables are stationary of order zero or order one according to the ADF standard with a 1% significant level and the PP standard with a 5% significant level The results of Lagrange multiplier (LM) statistics show that the null hypothesis cannot be rejected at the 10% level in all cases, indicating that there is no serious problem of serial correlation in our models 13 - The results also confirmed that there is no existence of autocorrelation or heteroscedasticity in the models and the models are correctly specified The CUSUM and CUSUMSQ tests showed the constancy of the coefficients in most models 4.1.2 Estimation results of long-run coefficients 4.1.2.1 Results for China The results show that there are commodities (accounting for 18% of total import value and 9% export value for China's market) where real exchange rate coefficients are significant at 5% or 10% level Most of the coefficients have positive sign, meaning thata depreciation will positively affect the trade balance of these commodities The coefficient of the real exchange rate variable in the aggregate model of China is not statistically significant, implying that exchange rate does not affect the total trade balance between Vietnam and China in the long run An important assumption adopted in the ARDL model is that all independent variables will exert symmetrical impacts on the trade balance Therefore, the conclusions regarding the movements of the trade balance will be reversed in the case of VND appreciation 4.1.2.2 Results for the US 14 For Japanese, there are 11 commodities that have at least one statistically significant short-term coefficient (accounting for 42% of and 48% of exports of Vietnam in this market in 2017) 4.1.3.2 The J curve effect According to the results of the models of China, the J-curve effect is found in mineral products (SC05) and raw hides and skins (SC08) For the US, the J-curve effect is spotted in products of chemical or allied industries (SC06) For Japan, the J-curve effect is found in mineral products (SC05) In general, the trade balance between Vietnam with China, Japan, the US are less sensitive to exchange rate movements However, looking at the three models that measure the impacts of exchange rates on Vietnam’s total trade balance (aggregate models) with the three trading partners, it is possible to conclude that exchange rate is not sensitive to trade balance The results justified the conclusion of Baek (2013), saying that the use of aggregate data will cause bias in the estimates of the coefficients Apparently, it is not reasonable to conclude how differently exchange rate affects the trade balance with different partners and how it affects different commodity groups/industries just by looking at the results of the aggregate model There are product categories (accounting for about 59% of Vietnam's import value, and 52% of Vietnam's export value to the US market) where the balance trade is affected by exchange rates However, the estimation results of the long-run coefficient in the aggregate model indicate that trade balance is not sensitive to exchange rate movements For the income variable, all three partners have share one thing in common, that is, the trade balance is more sensitive to Vietnam's income than the income of the partners Most of Vietnam's income coefficients are found positive 4.1.2.3 Results for Japan 4.1.4 Estimation results of long-run coefficients in the NARDL models The estimation results of the long-run coefficients included in the models showed that there are only commodity groups/industries (accounting for 28% of imports and 34% of exports of Vietnam for this market) which are affected by exchange rate The income coefficient of Vietnam suggested there are commodity groups/industries (making up 32% of imports and 31% of exports in the Japanese market) which have statistically significant estimation results The coefficient of the variable denoting exchange rate in the aggregate model (Sctotal) of Japan is not statistically significant, implying that exchange rate does not affect the total trade balance between Vietnam and Japan 4.1.4.1 Results for China 4.1.3 Estimation results of short-run coefficients of the ARDL models 4.1.3.1 Results of short-run coefficients For China, there are commodities that have at least one statistically significant shortrun coefficient These groups make up 54% of imports and 56% of exports of Vietnam in this market in 2017 For the US, there are commodities that have at least one statistically significant shortrun coefficient, (accounting for 68% of imports and 54% of exports of Vietnam in this market in 2017) Results of the NARDL models pointed out that NEG and POS are statistically significant in models of commodities Of which, commodities have been specified already by the ARDL models and other commodities that have been discovered by the NARDL models This finding is consistent with many previous studies, highlighting the problem of bias that distressed previous works when assuming that the impacts of exchange rate on trade balance are always symmetrical (Bahmani-Oskooee and Baek, 2016; Arize et al., 2017) Clearly, without the adoption of the non-linear regression model (NARDL), these commodity groups/industries might be mistaken that their real exchange rate coefficients were not statistically significant Or in other words, their trade balance did not respond to the change of exchange rate 4.1.4.2 Results for the US Similar to the conclusion of the models of China, in the NARDL model, trade balance is found to be more sensitive to exchange rate movements than in the ARDL model There are 11 commodities where trade balance is sensitive to exchange rate changes These commodities make up 71% of imports and 81% of exports of Vietnam for this market In 15 16 addition, these results also indicate the asymmetric responses of trade balance to exchange rate movements the US For Japan, the J-curve effect is found in some commodities, including SC04, SC05, SC21 Regarding the aggregate model, the estimated coefficients of SCtotal indicate that, in the long run, the depreciation of VND would not exert a positive impact on the trade balance of Vietnam with the US In fact, the estimation results show that when the VND depreciated by 1%, the trade balance with the US would worsen by 2.5% In addition, there is no significant impact on trade balance if the VND is appreciated (estimated coefficients turn out statistically insignificant) Regarding the aggregate model (Sctotal), J-curve effect is spotted in the trade balance between Vietnam and Japan 4.1.4.3 Results for Japan Results of the NARDL models pointed out that there are 15 commodity groups (accounting for 46% of imports and 63% of exports between Vietnam and Japan) where their trade balances are sensitive to exchange rate changes Of which, there are commodities that were specified by the ARDL models, and the NARDL models listed out more commodities Concerning the aggregate model, the results indicated that exchange rate does affect the trade balance between Vietnam and US in the case of VND depreciation If the VND depreciates by 1%, Vietnam’s trade balance with Japan will improve by 1.2% In case of VND appreciation, the trade balance will not be affected In general, there is a lack of a support for a J curve in the relationship between exchange rate and the trade balance of Vietnam with the three trading partners 4.2 Results from the NARDL models for key export products The results of the ARDL and NARDL models have shown and proved that in the case of Vietnam, the assumption about the symmetric reaction of the trade balance to exchange rate changes also causes a bias in the results Therefore, in this section, the author examines the impact of exchange rates on the trade balance of Vietnam's key export products based on the NARDL model 4.2.1 Long-term coefficient results 4.1.5 Estimation results of short-run coefficients in the NARDL models The results of the ARDL and NARDL models have shown and proved that the assumption about the symmetric reaction of the trade balance with the exchange rate changes also causes a bias in the results in the case of Vietnam Therefore, in this section, the author examines the impact of exchange rates on the trade balance of Vietnam's key export products based on the NARDL model 4.1.5.1 Results for China 4.2.1.1 Electric equipment, machinery, components and components (HS85) The results show that there are 11 commodities where the estimated coefficients of POS (specifying the increase of the exchange rate or the depreciation) and NEG (specifying the decrease of the exchange rate or the appreciation) are statistically significant The sign and magnitude of estimated coefficients vary across commodity groups The results of the short-term coefficients of POS and NEG also indicate the asymmetric effect of exchange rate on trade balance Similar to the results of the long-term coefficient, trade balance in the NARDL model is more sensitive to exchange rate movements than in the ARDL model The HS85 model results show that the trade balance of this commodity group is not affected in the case of Vietnam-US and Vietnam-China when the VND depreciated Regarding Vietnam-Japan trade, the trade balance of this commodity group decreases by 0.7% when the VND depreciates by 1% In the case of VND appreciation, only the Vietnam-China trade balance is affected, and the trade balance of this commodity group will decrease by 6% when the VND appreciates by 1% Thus, the devaluation of the domestic currency obviously does not improve the trade balance of Vietnam with the three partners The situation is even worse for the Vietnam-Japan trade balance Furthermore, the non-linear approach reveals that, in most cases, exchange rate changes have asymmetric effects on the trade balance 4.1.5.3 Results for Japan 4.2.1.2 Machinery, mechanical equipment, nuclear reactors, boilers; parts thereof (HS84) The results of the Wald test statistic indicate that we can reject the null hypothesis of symmetry in the short run for 11 commodity groups Of which, there are commodities that are sensitive to exchange rate when the VND depreciates Meanwhile, the trade balances of 12 commodities are sensitive to exchange rate when the VND appreciates Overall, it is apparent that using the NARDL model can detect more commodity groups/industries that respond to exchange rate than using the ARDL model The results indicate that a VND depreciation does not have an effect on the trade balance between Vietnam and China for this commodity group Meanwhile, a VND depreciation of 1% will lead to a decrease of 8% in the Vietnam-US trade balance, and an increase of 1.79% in the Vietnam-Japan trade balance These commodities are mainly exported by FDI enterprises and deeply involved in global value chains Therefore, they are less likely to be affected by exchange rates 4.1.5.4 The J Curve effect 4.2.1.3 Footwear (HS64) For China, the J-curve effect is only found in industry SC08 The J curve effect is not supported by the relationship between exchange rate and the trade balance of Vietnam with The results show that when the VND depreciates by 1%, the Vietnam-China trade balance of footwear products is not affected, while the Vietnam-US trade balance deteriorates by 4%, but the Vietnam-Japan trade balance improves by 1.75% The production of this 4.1.5.2 Results for US 17 commodity group depends on the imports of raw materials, especially those from the Chinese market Therefore, the exchange rate movements did not have much effect on the trade balance of this commodity group 4.2.1.3 Textile products (HS61) The results show that when the VND depreciates by 1%, the trade balance of textile products in the case of the Vietnam-US and Vietnam-China improve by 4.9% and 19% respectively Meanwhile, the Vietnam-Japan trade balance is not affected 4.2.1.4 Fish and crustaceans, molluscs, and other aquatic invertebrates (HS03) The results indicate the VND depreciation does not affect the Vietnam-US and ChinaUS trade balance of the HS03, while the Japan-US trade balance decreases by 4.8% in case the VND depreciates by 1% In general, it is suggested that the export and trade balance of this commodity group depend substantially on production capacity and "the capacity to dominate the market" rather than exchange rates 4.2.2 Result of short-run coefficients In the short run, the commodity group of HS85 (electrical goods and electrical appliances ) will see its trade balance respond to the depreciation of the VND (POS) in the case of trade between Vietnam and the US as well as trade between Vietnam and Japan The short-run coefficient of POS is not statistically significant in the case of trade between Vietnam and China Specifically, the VND depreciation improves the trade balance of textile and apparel products in the case of trade between Vietnam and the US as well as between Vietnam and Japan When the VND appreciates, the trade balance deteriorates significantly for the Vietnam-US trade, but it improves in case of Vietnam-China trade Meanwhile, the Vietnam-Japan trade balance did not respond to the appreciation of VND The coefficient results of the commodity group HS84 (machinery, equipment ) are similar to those of the HS85 For the HS64 (footwear), the results of short-run coefficients were similar to those of the HS85 Specifically, when VND depreciates, the trade balance of this commodity group is not affected in the case of China-Vietnam trade, improves in the case of Vietnam-Japan trade and worsen in the case of Vietnam-US trade The model results also show that the HS61 (textile products) has better trade balance in the short-run in the case of Vietnam-US trade if the VND depreciates It, however, becomes worse when it comes to Vietnam-China trade and does not show any reaction when trading with the Japanese counterpart An VND appreciation will lead to the improvement of the trade balance in the case of Vietnam-China trade and Vietnam-Japan trade The HS03 trade balance is less responsive to exchange rate changes in the short term Meanwhile, a VND devaluation causes the Vietnam-Japan trade balance to worsen AVND appreciation will improve the Vietnam-China trade balance In general, the results of the POS coefficients in the Vietnam-China models indicate that only the HS61 sees the trade balance respond to exchange rate depreciation in the short 18 run In contrast, most of the NEG short-run coefficients have at least one coefficient that is statistically significant and most of them are negative (excepting the HS61) This implies that, in the short run, a VND depreciation will improve the trade balance of the commodity groups (HS85, H64, HS84, HS03) Meanwhile, the Vietnam-US trade balance responds to changes in exchange rate movements for most of the commodities (excepting seafood products) The Vietnam-Japan trade balance of key export commodities is more sensitive to exchange rates in the case of VND devaluation (4/5 commodities) There are only commodities where the Vietnam-Japan trade balances are sensitive to an exchange rate appreciation (HS84 and HS61) CHAPTER CONCLUSION AND POLICY RECOMMENDATIONS 5.1 Results The study uses the linear ARDL and nonlinear ARDL (NARLD) models to examine the effect of exchange rate changes on Vietnam's trade balance with China, the US, and Japan employing the disaggregated data of 21 industries (1-digit Harmonized System, HSS) and key export products data (2-digit HS) In particular, the paper seeks to overcome the shortcomings of previous works in assessing the relationship between trade balance and exchange rates in Vietnam by: (1) using industry data to address any limitations raised by the utilization of aggregate data, and (2) employing a non-linear model to avoid the bias problem of the symmetric assumption 5.1.1 Conclusions from the estimation results for 21 commodities Given the estimation results of the short-term and long-term coefficients included in the models, several conclusions can be drawn as follows: First, the estimation results of the ARDL and NARDL models for all industries revealed a degree of bias to the regression estimates when using aggregate data and the asymmetric approach Therefore, the author’s approach of using the NARLD model with industry-level data would provide a more comprehensive and complete picture of the effects of exchange rate changes on trade balance between Vietnam and Japan Second, the theory on the J-curve effect was not supported by the estimation results of the model examining the relationship between exchange rates and trade balance in the case of Vietnam-US trade and Vietnam-China trade Results from models using industry-specific data implied that only a few industries have experienced the J-curve effect while the models employing aggregate data did not find any signs of the J-curve effect Third, Vietnam’s income had a positive impact on trade balance in the case of Vietnam-US trade and Vietnam-China trade The Vietnam-China trade balance is less sensitive to exchange rate changes than that of the Vietnam-US trade and Vietnam-Japan trade This result confirms Vietnam's heavy dependence on the Chinese import-export market The exchange rate is the most important tool in regulating trade policies he nullification of this tool in trade relations with China shows that Vietnam has no preventive measures against bilateral exchange rate shocks 19 For the Vietnam-US trade, the ARDL and NARLD model results show that Vietnam's trade balance is quite sensitive to exchange rates The NARDL models indicate that some commodity groups will see an improvement in trade balance, but others' trade balances will get worse when the VND depreciates The aggregate model shows that a VND depreciation will lead to a decrease in trade balance between Vietnam and the US Therefore, a currency's devaluation should not be used as a tool to improve the trade balance of Vietnam in any case In addition, the results showed that the Vietnam-Japan trade balance for these commodities is also sensitive to exchange rate movements (16/19 commodities) The results of the NARDL models using aggregate data show that areal exchange rate depreciation will improve trade balance Meanwhile, trade balance does not respond to an exchange rate appreciation Thus, the exchange rate is likely to be an effective tool to improve the VietnamJapan trade balance 5.1.2 Conclusions from the estimation results for key export products (5 groups of commodities) Some conclusions can be drawn from considering the effect of exchange rate changes on the trade balance of Vietnam's key commodities with its three trading partners as follows: Firstly, Vietnam's leading export industries are of two types: (i) firstly, hightechnology commodities (HS85 and HS84) which are mainly manufactured by FDI enterprises and deeply involved in global value chains which has somewhat reduced the impact of policies with exchange rate tools; (ii) secondly, the commodities that Vietnam has competitive advantages with cheap labour costs (textile and footwear product)), but these groups mainly depend on imported raw materials Secondly, trade balance seems to be more sensitive to exchange rate depreciation in the case of Vietnam-US trade and Vietnam-Japan trade and less responsive to exchange rate depreciation in the case of Vietnam-China trade in both the long run and short run In case of VND appreciation, trade balance is more sensitive to exchange rate movements in the case of Vietnam-China trade (the trade balance of most commodities worsens when the VND appreciates) Meanwhile, the appreciation of the VND affects only out of commodities in the case of Vietnam-US trade, and out of commodities in the case of Vietnam-Japan trade In general, the use of exchange rates as a tool for improving trade balance is unlikely to be effective for trade between Vietnam and China 5.1.3 General conclusions The results have shown that currency devaluation is unlikely to make Vietnam's trade balance better and even become worse in the case of Vietnam-US trade For the key export commodities, trade balance is less sensitive to an exchange rate depreciation Meanwhile, an exchange rate appreciation is found to negatively affect the trade balance of most commodity groups Based on the theory of trade and the results in chapters and 4, I will point out a number of reasons why devaluation is unlikely to make Vietnam trade balance better as follows: 20 Firstly, chapter shows the relationship between the nominal exchange rate and the real exchange rate of Vietnam Recently, these two exchange rates have tended to move in opposite directions Vietnam's bilateral exchange rate with the three partners has appreciated in recent years Therefore, the problem here is that the relatively rigid exchange rate regime currently adopted by Vietnam was followed by a currency appreciation As a result, Vietnam's exports tended to lose competitiveness In the long term, the appreciation of the VND will make it difficult for Vietnam to improve and maintain a trade surplus The highly appreciated currency also reduces the effectiveness of the devaluation policy on trade balance Secondly, Vietnam’s exports remain heavily dependent on imported goods On the other hand, the possibility of import substitution is quite low The dependence on imports makes the trade balance less sensitive to currency depreciation, especially in the case of Vietnam-US trade Thirdly, the foreign direct investment (FDI) sector contributed 72% to Vietnam's total exports turnover Key export items include high-tech products such as phones, computers, electrical and other mechanical This contribution also leads to an increase in the import of materials and components The intensive participation in the global value chains of FDI enterprises will lead to a decrease in the effectiveness of the exchange rate tool (especially in the case of currency devaluation) Fourthly, Vietnam’s export products have relatively low competitiveness The reason is that the current level of technological endowment of Vietnam is fairly low and modest In addition, the poor quality of human resources remains a huge problem whichleads to the poor performance of Vietnamese labour productivity It is the low competitiveness that makes it difficult to find new markets or expand markets and seize available opportunities for improving the country’s export performance when the depreciation exchange rate policy is adopted In other words, the weak competitiveness of export products reduces the effectiveness of the exchange rate depreciation 5.2 Policy implications for each trading partner For the Vietnam-China, the results indicate that trade balance is less sensitive to exchange rate changes Therefore, it can be affirmed that there are no feasible measures for Vietnam to limit trade deficit with China (provided that Vietnam's exchange rate management policy does not change much) In order to reduce the dependence on imports from China, Vietnam needs to increase the manufacturing production capacity of supporting industries while pushing the search for alternative import markets such as ASEAN markets For Vietnam-US trade, the result models show that an exchange rate depreciation does not improve trade balance but even makes the situation worse Therefore, it is not easy to use the exchange rate as a tool to improve the Vietnam-US trade balance For Vietnam-Japan trade, the NARDL model results indicates that trade balance will be improved in the case of VND depreciation Meanwhile, trade balance does not react to an exchange rate appreciation 21 22 CONCLUSION Thus, the exchange rate is likely to be an effective tool to improve Vietnam's trade balance with this trading partner 5.3 Policy implications for each commodity 5.4 General policy recommendations Firstly, the study indicates that the use of exchange rates as a tool to promote exports and improve the trade balance of Vietnam should be considered based on various priority objectives and policy coordination Secondly, trade balance is a macroeconomic variable that depends on many different factors: (i) the influential factors of the quality of export products, (ii) relative prices including exchange rates The analyses in chapter and chapter show that Vietnam is facing structural problems in production and export Moreover, the dependence of import performance on foreign markets, especially China, also reduces the positive effects of VND devaluation on trade balance Therefore, the study proposes some policy implications in addition to the exchange rate policy to boost exports, improve trade balance, and improve the efficiency of exchange rate instruments in regulating Vietnam's trade as follows: - Developing auxiliary industries, enhancing self-control capacity to ensure raw materials source of exported products, increasing the localization rate of export products, especially for key industries such as electrical equipment, mechanical machinery and equipment, footwear, textile - Vietnam should also be cautious about the quality of many FDI projects while proactively supporting and protecting domestic manufacturing enterprises,or tightening the conditions for FDI enterprises The country should develop appropriate sanctions for FDI enterprises in case of violations to actively push for technology transfer and engage domestic enterprises in the supply of input materials - Developing human resources, improving labour productivity in manufacturing enterprises, improving the competitiveness of Vietnam's export products Conclusion Based on the analysis of the relationship between exchange rates and trade balance, the thesis has achieved some results as follows: Firstly, the thesis presented and introduced the theoretical basis for the relationship between exchange rates and trade balance As a result, we have generalized the development history of theoretical literature on the relationship between exchange rates and trade balance Secondly, the thesis has reviewed both international and Vietnamese empirical studies on the relationship between exchange rates and trade balance Thereby, the study has pointed out the research gap about the relationship between exchange rates and the trade balance of Vietnam Thirdly, the combination of qualitative and quantitative analysis in the thesis has led to the following conclusions: (i), the exchange rate’s impact on trade balance varies across trading partners and commodities, (ii) trade balance is more sensitive to exchange rates in case of Viet-US trade and Vietnam-Japan trade than in Vietnam-China trade (iii) the exchange rate’s impact on trade balance is asymmetric From the findings, the thesis also gave some policy recommendations as follows: - In the Vietnam's context, the use of currency devaluation should be considered based on priority objectives and the coordination between different policies An exchange rate depreciation can lead to undesirable outcomes for a developing economy that is dependent on imported raw material inputs with a low possibility of import substitution like Vietnam - The implementation of fiscal policy instruments and economic restructuring, the promotion of auxiliary industries, and the enhancement of the competitiveness of export firms should be performed in practice to improve trade balance rather than using the exchange rate tool Recommendations for future research directions - Adding the impact of the global value chains when examining the relationship between exchange rates and trade balance - Considering cross-border trade in the relationship between exchange rates and trade balance in the case of Vietnam-China trade 23 THE LIST OF AUTHOR'S PUBLICATION RELATED TO THESIS Tran Thi Ha (2018), "The impact of exchange rate on trade balance between Vietnam and US in 2008-2017", Review of World Economic and Political Issues, 10 (270)/2018 Tran Thi Ha (2016), "The impact of exchange rate on imports of key products from China to Vietnam", Journal of Economics and Development, 9/2016 Tuan Anh Pham, Ha Tran Thi, Nguyen Manh Hung (2019), "The impact of exchange rates on Vietnam-China bilateral trade", JATI-Journal of Southeast Asian Studies', 24.1 (2019): 70-96 Ha Thi Tran (2018), Impact of exchange rate on China-Vietnam bilateral trade: findings from ARDL approach, Hội thảo "Vietnam Economics Annual Meeting"  ... Vietnam and Japan Bilateral trade between Vietnam and Japan accounts for about 10% of Vietnam's trade value As of 2017, Japan is the fourth largest partner of Vietnam From 2009 to 2017, Vietnam's... exchange rate and exchange rate policies of Vietnam 3.2 Vietnam’s export and import performance 3.2.1 Summary of Vietnam export and import records Vietnam has accomplished certain achievements in... between Vietnam and the US as well as between Vietnam and Japan When the VND appreciates, the trade balance deteriorates significantly for the Vietnam-US trade, but it improves in case of Vietnam-China