- Journal List
- Heliyon
- v.9(4); 2023 Apr
- PMC10070517
As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsem*nt of, or agreement with, the contents by NLM or the National Institutes of Health.
Learn more: PMC Disclaimer | PMC Copyright Notice
Heliyon. 2023 Apr; 9(4): e14455.
Published online 2023 Mar 28. doi:10.1016/j.heliyon.2023.e14455
PMCID: PMC10070517
PMID: 37025797
Loc Dong Truong∗ and Dut Van Vo
Author information Article notes Copyright and License information PMC Disclaimer
Associated Data
- Data Availability Statement
Abstract
This study examines the asymmetric effects of exchange rate on Vietnam’s trade balance. Data used in this study consist of monthly trade balance, exchange rate, industrial production index and foreign direct investment series over the period from January 2010 to June 2020. Using the nonlinear autoregressive distributed lag (ARDL) bounds testing approach, the empirical findings confirm that the exchange rate has asymmetric effects on trade balance in both the long-run and short-run, meaning that a decrease in exchange rate has the different effect on trade balance as an increase in exchange rate with the same size. Specifically, in the short-run, one percent increase in the exchange rate (USD/VND) is associated with 4,2607% decrease in the trade balance, while appreciation of VND have no effects on the trade balance. In the long-run, one percent increase in the exchange rate results in 0.902% increase in the trade balance. However, no evidence is found for the effect of the appreciation of VND on the trade balance in the long-run. Moreover, the results derived from the error correction model (ECM) indicate that 89.07% of the disequilibria from the previous month are converged and corrected back to the long-run equilibrium in the current month.
Keywords: Exchange rate, Trade balance, NARDL, Vietnam
1. Introduction
Vietnam started a program of comprehensive economic reforms (doi moi) in 1986 which marked the end of a central planning period of the economy. An important component of the transition to a market-oriented economy is to reform the financial system in which the role of the exchange rate policy has been considered as an important macroeconomic instrument for keeping low inflation and stability of the financial system, improving trade balance, and boosting economic growth. To achieve these goals, the government has applied a managed floating exchange rate regime under which the exchange rate of USD/VND (Vietnamese dong) has been adjusted according to the market and macro-economic signals.
The Vietnam’s economy has been increasingly integrated in the world economy during the last some decades. In this context, international trade has played a remarkable role in Vietnam’s economic development. The statistics of the General Department of Customs show that before 2012, Vietnam’s trade balance had always been in deficit, with a value of billions US. dollars per year. However, from 2012 to 2020, the trade balance has reversed and the surplus has continuously increased (except for 2015, the deficit was US$3.55 billion). In addition, the total value of import and export for nearly 20 years (from 2000 to 2019) of Vietnam has reached US$3,995 billion. In particular, in only 5 years from 2015 to 2019, Vietnam’s import and export reached 2,106 billion USD, higher than these of all 15 years ago (the period of 2000–2014). It is expected that the improvements in Vietnam’s trade balance could be derived from the changes of exchange rate during the period from 2010 to 2020.
Exchange rate is theoretically an important determinant of trade balance of a country. The Marshall-Lerner condition states that a currency depreciation only results in improvements in trade balance of a country if the sum of demand elasticity for imports and exports is greater than one. However, many scholars have argued that the Marshall-Lerner condition is only satisfied in the long-run, but not held in the short-run because demand in the long-run is more elastic than in the short-run. In other words, devaluation of domestic currency worsens trade balance in the short-run but improves it in the long-run. It is known as the J-curve effect in economic theory. Empirically, many studies have been conducted to test for the effects of exchange rate on trade balance in both developed and developing countries over the last decades. However, the debate on the effects of exchange rate on trade balance has been still ongoing, especially for developing countries. Several studies concluded that currency devaluation leads to an improvement in trade balance [[1], [2], [3], [4], [5]]. Contrary to the first category, some studies confirmed that currency devaluation has negative effects on trade balance [[6], [7], [8]]. Especially, some recent studies found that exchange rate has asymmetric effects on trade balance, meaning that a decrease in exchange rate has the same effects on trade balance as an increase in exchange rate with the same size [[9], [10], [11], [12], [13], [14]].
Although the effects of exchange rate on trade balance have substantially studied in emerging or advanced economies, a few studies have investigated this effect for Vietnam, a transitional economy. Several studies have shown that the institutional framework in transition economies differs from that of emerging or advanced economies. However, the knowledge about the effects of exchange rate on international trade is limited within the context of a transition economy. Carrying out research on this issue in a transition economy is crucial because the manner of transition economies is characterized as a multitude of different, and possibly conflicting institutional pressures, there is, thus, a need to explore how they differ. Based on the nature of the transition economy, it is anticipated that underdeveloped market mechanisms, and insufficient legal and regulatory conditions relating to exchange rate policies are likely to impede linkage or integration activities between foreign and domestic counterparts [15,16], thereby are likely to affect international trade. For this reason, it is believed that investigating in such issue in the context of a transition economy would offer new insights for understanding how the exchange rate volatility is likely to affect international trade of Vietnam - a transition economy consisting of high institutional uncertainties [16].
The aim of this paper is to explore the asymmetric effects of the exchange rate on the trade balance of Vietnam during the post-global financial crisis period. The contribution of this paper is to enrich the literature covering the effect of exchange rate on trade balance as follows. First, Vietnam is an interesting case for studying because Vietnam’s economy has been in a transitional period with deep and wide integration in the world economy during the last two decades and the managed floating exchange rate regime has been applied. Second, while most of studies investigated the effects of exchange rate on Vietnam’s trade balance with an assumption that the effects of exchange rate on trade balance are symmetric [8,17] or examined the asymmetric impact of exchange rate on Vietnam’s bilateral trade balance [18], this study is the first to investigate the asymmetric effects of exchange rate on the aggregate trade balance of Vietnam for the post-global financial crisis (2010–2020). By using a nonlinear autoregressive distributed lag (NARDL), this study measures the effects of exchange rate on the trade balance of Vietnam more effectively than the conventional ARDL model because under the managed floating exchange rate regime in the context of transition economy, an increase in the exchange rate (depreciation) could have different effects on trade balance compared to a decrease in the exchange rate (appreciation) with the same magnitude. The rest of the paper is organized as follows. Section 2 presents the review of literature. Section 3 describes the data collection and methodology employed in the study. Section 4 discusses the empirical results. Finally, conclusions are presented in Section 5.
2. Literature review
Numerous empirical studies have been undertaken to determine the relationship between exchange rate and trade balance in developing countries. However, the findings from these studies have not been consensus. Some studies have found that exchange rate is an important determinant of trade balance, whereas others have found no relationship between exchange rate and trade balance. Specifically, the Marshall-Lerner condition and J-curve have been tested by many studies recently. Some studies empirically support the hypothesis that currency devaluation leads to an improvement in trade balance. Bahmani-Oskooee and Kantipong [1] used ARDL cointegration technique to examine the effects of exchange rate on trade balance between Thailand and its five major trading partners (Germany, Singapore, Japan, the UK and the US). They found out the existence of the J-curve only in bilateral trade of Thailand with the US and Japan. Besides, by applying the Engle-Granger and Johansen-Juselius cointegration methodology, Bahmani-Oskooee [2] investigated the long-run response of the trade balance of Middle Eastern countries to currency devaluation. The results of this study indicate that there is a positive long-term impact of a real depreciation on the trade balance for all seven nations (Algeria, Bahrain, Egypt, Jordan, Morocco, Tunisia, and Turkey). Similarly, Baharumshah [19] determined the impact of currency devaluation on the bilateral trade balances of Malaysia and Thailand with the US and Japan. The study found that the currency devaluation leads to an improvement of the trade balances in the long-run for both cases, but the J-curve effect does not exist. Moreover, Gomes and Paz [3] tested the relationship between real exchange rate and trade balance for Brazil during the period from 1990 to 1998. The findings from this study confirmed that currency depreciation leads to improvements of Brazil’s trade balance in the long-run. Using the FMOLS model, Yol and Baharumshah [4] found that currency depreciation is associated with improvements of trade balance in six African countries. Recently, Hunegnaw and Kim [5] investigated the effects of real exchange rate on trade balance in East African countries. Using the ARDL approach, the main findings of the study indicate that real exchange depreciation has significantly positive effects on trade balance in the long-run for four countries. However, this study also concludes that the J-curve effect does not exist for all countries in the sample.
Contrary to the first category, other studies assert that currency devaluation has negative effects on the trade balance. Upadhyaya and Dhakal [6] examined the effects of currency depreciation on the trade balance of eight developing countries in Asia, Europe, Africa and Latin America. They found that the devaluation, in the long-run, has negative effects on the trade balance in Cyprus, Greece and Morocco. In addition, Yol and Baharumshah [4] explored the effects of the exchange rate changes on the bilateral trade balance of ten African countries. The results of the study indicate that currency devaluation has negative effects on trade balance for Tanzania. Similarly, Shahbaz et al. [7] investigated the long-term relationship between the real exchange rate and the trade balance in Pakistan. Using quarterly data for the period from 1980 to the end of 2006, the findings from the autoregressive distributed lag (ARDL) model revealed that the effects of currency devaluation on trade balance are negative and statistically significant. In other words, currency devaluation has an unfavorable impact on trade balance. Moreover, Phan and Jeong [8] investigated the effects of the real exchange rate on bilateral trade balance between Vietnam and its sixteen trading partners during the period from 1999 to 2012. They assert that the real exchange rate has a significantly negative effect on the trade balance.
Moreover, some recent studies found that exchange rate has asymmetric effects on the trade balance. Bahmani-Oskooee and Halicioglu [9] measured the effects of exchange rate changes on Turkey’s trade balance and found the asymmetric effects of exchange rate changes. Specifically, lira depreciation has significantly positive effects on Turkey’s trade balance while lira appreciation does not have any significant effects on the trade balances. In addition, Bahmani-Oskooee and Aftabb [10] employed a Nonlinear Autoregressive Distributed Lag (NARDL) model to examine asymmetric effects of exchange rate changes on trade balance of 59 industries between Malaysia and China. They found that exchange rate changes have asymmetric effects on the trade balance for some industries. Similarly, Bahmani-Oskooee and Kanitpong [11] investigated the asymmetric effects of exchange rate changes on trade balance of 45 industries between Thailand and China. This study reported short-run asymmetric effects of exchange rate in 27 industries and long-run asymmetric effects in 15 industries. In addition, Bao and Le [12] investigated asymmetric effects of the real effective exchange rate and vehicle currency exchange rate on the total trade balance between ASEAN and the EU. The main findings of the study indicate that the vehicle currency exchange rate has the asymmetric effects on the trade balance in the short-run. Moreover, Bao and Le [13] determined the influences of exchange rate on trade balance at the industry levels between ASEAN and the EU-28. They found the asymmetric effects of exchange rate on trade balance in both short-run and long-run for most of the industries. Recently, Bao et al. [14] employed a NARDL approach to investigate asymmetric effects of exchange rate on China’s trade balance with the EU. This study documented that the real effective exchange rate and vehicle currency exchange rate has long-run and short-run asymmetric effects on the trade balance.
The last category consists of studies that find no significant effect of exchange rate on trade balance. Specifically, Ogbonna [20] measured the effect of currency devaluation on Nigeria’s balance of payments and concluded that the balance of payments is not improved by currency devaluation. Additionally, Upadhyaya and Dhakal [6] found that the devaluation has no impact on the trade balance in the long-run for Colombia, Guatemala, Singapore and Thailand. Moreover, Wilson [21] investigates the relationship between the real exchange rate and trade balance for bilateral trade between Singapore, Korea, Malaysia and the USA, Japan during the period from 1970 to 1996. The findings derived from the study indicate that the real exchange rate has a statistically insignificant effect on the real trade balance for Singapore and Malaysia. Similarly, Kwalingana et al. [22] measured the long-run and short-run effects of real exchange rate on the trade balance in Malawi. The results from the study indicate that currency devaluation does not have statistically significant effects on the trade balance in the long-run.
In summary, most of the previous studies have examined the effects of exchange rate on international trade in emerging or advanced economies. A few studies have investigated this effect for Vietnam with an assumption that the effects of exchange rate on trade balance are symmetric. However, the effects of the exchange rate on the trade balance could be asymmetric, meaning that a decrease in the exchange rate has different effects on the trade balance compared to an increase in the exchange rate with the same magnitude. Therefore, this study is poised to make a valuable contribution to the literature by using the NARDL approach to investigate the asymmetric effects of the exchange rate on the trade balance for the transition economy of Vietnam.
3. Data and methodology
3.1. Data sources
The data used in this study are monthly real exchange rate (RER), trade balance (TB), industrial production index (IPI) and foreign direct investment (FDI) series over the period from January 2010 to June 2020. The data are obtained from various reliable sources, namely the General Statistics Office (GSO), the World Bank (WB) and the International Monetary Fund (IMF). Specifically, the data sources are presented in Table 1.
Table 1
Data sources.
| Data | Unit of measurement | Data source |
|---|---|---|
| Export volumes | USD (millions) | General Statistics Office (GSO) |
| Import volumes | USD (millions) | General Statistics Office (GSO) |
| Nominal exchange rate | USD/VND | International Monetary Fund (IMF) |
| Consumer price index (CPI) | Compared to CPI2010 (CPI2010=100) | International Monetary Fund (IMF) |
| Industrial production index (IPI) | Compared to IPI2010 (IPI2010=100) | World Bank (WB) |
| Foreign direct investment (FDI) | USD (millions) | General Statistics Office (GSO) |
Open in a separate window
3.2. Methodology
To measure the effects of exchange rate on trade balance of Vietnam, the regression model with the form as equation (1) is obtained:
TBt=β0+β1LRERt+β2IPIt+β3LFDI+ut
(1)
where:
-
TB: Trade balance of Vietnam, measured by the exports to imports ratio;
-
LRER: The natural logarithm of real exchange rate of USD/VND. The real exchange rate (RER) is calculated by equation (2) as follow:
(2)
where:
•
NER: Nominal exchange rate of USD/VND.
•
: Vietnam consumer price index.
•
: The United States consumer price index.
-
IPI: Industrial production index of Vietnam.
-
LFDI: The natural logarithm of FDI.
In this study, the nonlinear Autoregressive Distributed Lag (NARDL) bounds testing approach which was proposed by Shin et al. [23] as an extended model of the ARDL model of Pesaran et al. [24] is used to estimate the long-run and short-run asymmetric effects of the exchange rate on trade balance of Vietnam. In this model, the exchange rate is decomposed in positive and negative partial sum series. Because the NARDL is based on the ARDL, this approach also has some main advantages compared to other co-integration methods. Firstly, the NARDL does not require that all variables in the model have the same integration order. They can be integrated of purely order zero, purely order one or a combination of both. Secondly, the NARDL test is relatively more efficient and reliable than other approaches in the case of small and limited sample sizes. The final strong point of this approach over other alternative cointegration methods is that an error correction model (ECM) can be obtained from the ARDL model, so the short-run and the long-run impacts of the independent variables on the dependent variable can be assessed at the same time.
3.2.1. Unit root test
As mentioned above, the NARDL bound test of cointegration requires that all variables are I(0) or I(1). For this purpose, the unit root test should be conducted before proceeding the NARDL bound test. In this study, the widely used ADF and Phillips-Perron tests are employed to investigate whether the studied variables are stationary.
3.2.2. NARDL bound test for cointegration
Before estimating the short-run and long-run relationship among the variables in the model, cointegration tests should be conducted as a required condition. In order to determine the co-integration between variables, the bound test is employed in this study. The bound test of co-integration is performed by the following equation:
(3)
where,
-
Δ represents the first difference of the variables.
-
LRER+ is positive exchange rate change.
-
LRER– is negative exchange rate change.
The null hypothesis (H0) of the bound test estimated by equation (3) is λ1=λ2=λ3=λ4=λ5=0 (no co-integration in the long-run between variables). If the results of the bound test show that the F-statistic value is greater than critical value with the significance level of 5%, the null hypothesis is rejected. In this case, it can be concluded that there is a long-term relationship (co-integration) between the variables in the model. If the long-run equilibrium relationship between trade balance and other regressors is discovered, the long-run and short-run asymmetric effects of the exchange rate on trade are estimated by equations (4), (5), respectively.
(4)
(5)
3.2.3. Diagnostic tests
In order to make the NARDL model robust and unbiased, it is appropriate to conduct some diagnostic tests. As other studies [17,25], the results will be further tested to ensure that the results are statistically robust. Thus, tests for stability, serial correlation and heteroscedasticity in the residuals will be utilized in this step. If none of these biases are found in the model, the results are reliable and can be used for analysis.
4. Empirical results
4.1. Descriptive statistics of the sample
On the basis of the collected data, the descriptive statistics of the variables used in the model are calculated and summarized in Table 2. Overall, the monthly trade balance of Vietnam, defined as the exports to imports ratio, fluctuated highly over the period from January 2010 to June 2020. Specifically, the trade balance ranges from 0.727 to 1.188, with an average of 0.988. The results indicate that the trade balance of Vietnam is lightly deficit for the period of 2010–2020. In addition, it is shown in Table 2 that the mean of real exchange rate of USD/VND is 27,792 with a standard deviation of 4,011. The exchange rate has increased significantly for the research period. In other words, the Vietnamese Dong has been devalued throughout the period. Specifically, the lowest rate of USD/VND was 17,322, which was recorded in the first month of the period (January 2010), whereas the highest exchange rate was 33,465 in January 2020. In addition, Table 2 reveals that the mean of IPI for the period from 2010 to 2020 is 90.19%, ranging from 51.97% to 122.41%, with a standard deviation of 16.24. Moreover, statistics presented in Table 2 show that the mean of FDI during the period of 2010–2020 is 952 million USD.
Table 2
Descriptive statistics of the variables.
| Variables | Observations | Mean | Minimum | Maximum | Standard deviation |
|---|---|---|---|---|---|
| TB | 126 | 0.988 | 0.727 | 1.188 | 0.819 |
| RER (USD/VND) | 126 | 27,792 | 17,322 | 33,465 | 4,011 |
| IPI (%) | 126 | 90.19 | 51.97 | 122.41 | 16.24 |
| FDI (million USD) | 126 | 952 | 619 | 1,340 | 265 |
Open in a separate window
4.2. Unit root tests
As mentioned above, before using the NARDL bound test for cointegration, the unit root test must be performed as a required condition to check whether the variables used in the model have unit roots. In other words, the unit root test helps to verify whether variables are stationary or not at different orders. The unit root tests are estimated for both level and the first difference of studied series. In addition, the test is also performed for both cases with and without time trends. The results of ADF and Phillips-Perron tests are presented in Table 3.
Table 3
Results of unit root tests.
| Variables | ADF test | Phillips-Perron test | ||
|---|---|---|---|---|
| Constant | Constant and linear trend | Constant | Constant and linear trend | |
| TB | ||||
| Level | −6.36*** | −8.24*** | −6.31*** | −8.38*** |
| LRER+ | ||||
| Level | −3.44** | −4.04*** | −4.56*** | −3.41** |
| LRER− | ||||
| Level | −0.37 | −2.86 | −0.60 | −2.57 |
| First difference | −8.77*** | −8.81*** | −8.54*** | −8.58*** |
| IPI | ||||
| Level | −3.93*** | −4,29*** | −3.84*** | −4.22*** |
| LFDI | ||||
| Level | −0.42 | −2.76 | −0.36 | −2.77 |
| First difference | −11.43*** | −11.38*** | −11.47*** | −11.43*** |
Open in a separate window
***and ** indicate significance at 1% and 5% levels respectively.
The results derived from the ADF unit root test indicate that the null hypothesis of a unit root is significantly rejected at one percent level for TB, LRER+ and IPI at the level, indicating that TB, LRER+ and IPI series are integrated to the order zero or I(0). In addition, Table 3 reveals that the null hypothesis of a unit root cannot be rejected at the conventional significant level (5%) for LRER− and LFDI series because the t-statistic is smaller than their corresponding critical value (MacKinnon’s critical value). However, when the first differences are taken and tested for a unit root, the null hypothesis is significantly rejected for these series, indicating that they are stationary. With the evidences, it is concluded that LRER− and LFDI series are integrated of order 1, denoted as I(1). Besides, the results of Phillips-Perron test confirm that TB, LRER+ and IPI series are I(0) and LRER− and LFDI series are I(1). Based on these results, it is concluded that all the variables in the model satisfy the conditions of the NARDL bound test (with the order of integration less than 2) even with or without trend.
4.3. NARDL bound test for cointegration
As discussed above, this study uses the bounds test to examine the long-term relationship among variables in the model. Using Akaike Information Criterion to find optimal lags, the best model used in this study is ARDL (1,1,0,4,2). The results of the bounds test are summarized in Table 4.
Table 4
Results of the bounds test.
| Model | k | F-statistic | Significance level | Critical value | |
|---|---|---|---|---|---|
| Lower bounds I(0) | Upper bounds I(1) | ||||
| ARDL (1,1,0,4,2) | 4 | 20.35*** | 10% | 2.45 | 3.52 |
| 5% | 2.86 | 4.01 | |||
| 1% | 3.74 | 5.06 | |||
Open in a separate window
k represents the number of regressors. The critical value of bounds are from Pesaran et al. *** indicates significance at 1% level.
The results of the bounds test show that the F-statistic (20.35) is higher than the upper-bound critical value at the 1% level (5.06). Therefore, the null hypothesis of no co-integration among variables can be rejected. The rejection of the null hypothesis means that there is a long-run equilibrium relationship between trade balance and the regressors. Hence, the NARDL approach can be applied to estimate the long-run and short-run asymmetric effects of the exchange rate on trade balance of Vietnam.
4.4. Long-run and short-run asymmetric effects of exchange rate on the trade balance
The long-run and short-run coefficients derived from the NARDL model are reported in Table 5. The results indicate that in the long-run, the exchange rate (LRER) has asymmetric effects on trade balance. It is observed that, in the long-run, increases in the exchange rate (LRER+) have significantly positive effects on the trade balance of Vietnam at the one percent level. Specifically, one percent increase in the exchange rate is associated with 0.902% increase in the trade balance. The implication of this finding is that in the long-run the depreciation of VND leads to an improvement in the Vietnamese trade balance. However, in the long-run, the appreciation of VND has statistically no effect on the trade balance of Vietnam. It is important to stress that this evidence is different from previous findings of [[1], [2], [3],5]. Specifically, these studies documented positively symmetric effects of the exchange rate on the trade balance, meaning that a depreciation has positive effects on the trade balance and an appreciation has negative effects on it with the same magnitude. The explanation for the asymmetric effects of the exchange rate on Vietnam’s trade balance is that expectation and reaction of traders to a depreciation of VND are different from an appreciation. In fact, Vietnam has adopted the managed float exchange rate regime since 1990 with the goal to support for export. With this goal, the appreciation of VND has not often applied. Besides, when VND is appreciated, it usually lasts for a relatively short time. Therefore, the appreciation of VND does not impact on the trade balance of Vietnam in the long-run.
Table 5
The estimated long-run and short-run coefficients.
| Variables | Coefficients | t-statistics |
|---|---|---|
| Panel A: Long-run estimates | ||
| Constant | 0.6729 | 0.40 |
| LRER+ | 0.9020 | 4.19*** |
| LRER− | 3.2822 | 1.10 |
| IPI | −0.0001 | −0.28 |
| LFDI | 0.0218 | 0.11 |
| Panel B: Short-run estimates | ||
| ΔLRER+ | −4.2607 | −2.81*** |
| ΔLRER− | 2.9234 | 1.10 |
| ΔIPI | −0.0006 | −0.94 |
| ΔIPI(−1) | 0.0008 | 1.19 |
| ΔIPI(−2) | −0.0014 | −2.20** |
| ΔIPI(−3) | 0.0016 | 3.00*** |
| ΔLFDI | −0.0041 | −0.01 |
| ΔLFDI(−1) | −1.1118 | −2.39** |
| ECM(−1) | −0.8907 | −10.00*** |
| R2 | 0.5118 | |
| F-statistic | 9.52*** | |
Open in a separate window
***and ** indicate significance at 1% and 5% levels respectively.
In addition to estimating the long-term relationship, the NARDL cointegration model also allows for the estimation of the short-term relationship between the dependent variable and explanatory variables. The main findings derived from the model indicate that, in the short-run, depreciation of VND has significantly negative impact on the trade balance of Vietnam at the one percent level. Specifically, one percent increase in the exchange rate (USD/VND) results in 4.2607% decrease in the trade balance. However, the findings confirm that, in the short-run, appreciation of VND have no effect on the trade balance. Taken all together, it is concluded that the J-curve effect caused by the depreciation of VND exists in Vietnam. This finding is different as compared to previous studies. It is important to note that the J-curve effect has been found in many previous studies [1,3,17,26,27]. However, these studies are based on the assumption that the exchange rate has symmetric effects on trade balance. Besides, while Bao and Le [18] investigated the asymmetric effects of exchange rate on bilateral trade balance between Vietnam and each of EU-27 countries and the UK and reported the J-curve effect caused by depreciation of VND for some cases, our findings also show the J-curve effect, but at the aggregate trade balance level between Vietnam and the rest of the world.
Moreover, the results presented in Table 5 confirm that, in the short-run, the industrial production index has a statistically negative effect on the trade balance at the five percent level for the 2-month lag, but it has positive impact on the trade balance at the one percent level for the 3-month lag. Furthermore, it is shown in Table 5 that foreign direct investment (FDI) has a significantly negative effect on the trade balance for the 1-month lag at the five percent level statistically.
It is important to note here that the error correction term is statistically significant at the one percent level, meaning that the causality exists in at least one direction. The coefficient of error correction term is −0.8907 implying that 89.07% of the disequilibria from the previous month is converged and corrected back to the long-run equilibrium in the current month. The adjustment speed in this case is so high. In other words, the system will quickly get back to the long-run equilibrium after a short-run shock.
4.5. Diagnostic and structural stability tests
This study uses different diagnostic tests to examine the validity and reliability of the estimated results. Specifically, the diagnostic tests comprise Breach-Godfrey test for serial correlation and ARCH test for heteroscedasticity. The results of these diagnostics tests are summarized in Table 6. The results of Breusch-Godfrey test indicate that serial correlation does not exist among the residuals. Besides, the ARCH test for heteroscedasticity confirms that the residuals are free from heteroskedasticity phenomenon. Therefore, these diagnostics tests further strengthen and verify the reliability and validity of the research model as well as the estimated results.
Table 6
Results of diagnostic tests.
| Diagnostic test | Statistics | P-value | Conclusions |
|---|---|---|---|
| Autocorrelation (Breusch-Godfrey test) H0: No serial correlation | 1.85 | 0.14 | Fail to reject H0 |
| Heteroskedasticity (ARCH test) H0: No ARCH effects | 0.04 | 0.83 | Fail to reject H0 |
Open in a separate window
The stability of the coefficients needs to be examined because the ARDL model is quite sensitive to structural breaks and the financial time series used in this study are also sensitive to global events. To assess the long-term stability of the relationships between the trade balance and explanatory variables, the cumulative sum of the recursive residuals (CUSUM) and the cumulative sum of squared recursive residuals (CUSUMSQ) tests that were proposed by Brown et al. [28] are performed in this study. The results of the tests presented in Fig. 1, Fig. 2 reveal that the plots of CUSUM line within the critical bounds at the significance level of 5%. Therefore, it can be concluded that the research model used is stable over the studied period.
Fig. 1
Plots of cumulative sum of recursive residuals.
Fig. 2
Plots of cumulative sum squares of recursive residuals.
5. Conclusion
This study devotes to investigate the asymmetric effect of real exchange rate on the Vietnam’s trade balance over the period from January 2010 to June 2020. The empirical results derived from the NARDL bounds testing approach indicate that the exchange rate has significantly asymmetric effects on the trade balance of Vietnam in both the long-run and short-run. Specifically, the exchange rate devaluation leads to improvements of the trade balance in the long-run. However, in the short-run, real exchange rate depreciation has significantly negative effects on the trade balance. It is concluded that the J-curve effect caused by the depreciation of VND is present in Vietnam. In addition, the results derived from the error correction model (ECM) reveal that 89.07% of the disequilibria from the previous month is converged and corrected back to the long-run equilibrium in the current month. In other words, the system will quickly get back to the long-run equilibrium after a short-run shock.
The policy implication that can be drawn from the empirical findings of the study is that Vietnamese government should pursue the managed float exchange rate policy to improve sustainably the trade balance. By adopting this regime, the government can actively adjust the exchange rate in order to stimulate exports, control inflation at home and absorb most of the foreign shocks or policy changes from opened “big” economies like Euro Zone countries, the US, or China. Although this study has enriched our understanding of the asymmetric effects of exchange rate on trade balance in the context of transition economies, it still has some limitations that should be addressed in future studies. Firstly, due to limitations of the data, this study omits some factors that can impact on the trade balance. The second limitation of the study is that the Covid-19 pandemic has not been considered while measuring the effect of the exchange rate on trade balance of Vietnam. These limitations await further research.
Author contribution statement
Loc Truong: Conceived and designed the experiments; Performed the experiments; Analyzed and interpreted the data; Wrote the paper.
Dut Van Vo: Contributed reagents, materials, analysis tools or data; Wrote the paper.
Funding statement
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Data availability statement
Data will be made available on request.
Declaration of interest’s statement
The authors declare no conflict of interest.
References
1. Bahmani-Oskooee M., Kantipong T. Bilateral J-curve between Thailand and her trading partners. J. Econ. Dev. 2001;26(2):107–117. [Google Scholar]
2. Bahmani-Oskooee M. Nominal and real effective exchange rates of Middle Eastern countries and their trade performance. Appl. Econ. 2001;33:103–111. [Google Scholar]
3. Gomes F.A.R., Paz L.S. Can real exchange rate devaluation improve the trade balance? The 1990–1998 Brazilian Case. Appl. Econ. Lett. 2005;12(9):525–528. [Google Scholar]
4. Yol M.A., Baharumshah A.Z. Estimating exchange rate and bilateral trade balance relationships: the experience of Sub-Saharan African countries. S. Afr. J. Econ. 2007;75(1):35–51. [Google Scholar]
5. Hunegnaw F.B., Kim S. Foreign exchange rate and trade balance dynamics in East African countries. J. Int. Trade Econ. Dev. 2017;26(8):979–999. [Google Scholar]
6. Upadhyaya K.P., Dhakal D. Devaluation and the trade balance: estimating the long run effect. Appl. Econ. Lett. 1997;4:343–345. [Google Scholar]
7. Shahbaz M., Jalil A., Islam F. Real exchange rate changes and the trade Balance: the evidence from Pakistan. Int. Trade J. 2012;26(2):139–153. [Google Scholar]
8. Phan T.H., Jeong Y. Vietnam trade balance and exchange rate: evidence from panel data analysis. J. Appl. Econ. Bus. Res. 2015;5(4):220–232. [Google Scholar]
9. Bahmani-Oskooee M., Halicioglu F. Asymmetric effects of exchange rate changes on Turkish bilateral trade balances. Econ. Syst. 2017;41(2):279–296. [Google Scholar]
10. Bahmani-Oskooee M., Aftab M. Asymmetric effects of exchange rate changes on the Malaysia-China commodity trade. Econ. Syst. 2018;42(3):470–486. [Google Scholar]
11. Bahmani-Oskooee M., Kanitpong T. Asymmetric effects of exchange rate changes on Thailand-China commodity trade: evidence from 45 industries. Chin. Econ. 2019;52(3):203–231. [Google Scholar]
12. Bao H.H.G., Le H.P. The role of vehicle currency in ASEAN-EU trade: a double-aggregation method. J. Asian Fin., Econ. Bus. 2021;8(5):43–52. [Google Scholar]
13. Bao H.H.G., Le H.P. ASEAN’s trade balance with the whole EU-28 at industry level: the role of vehicle currency. J. Econ. Asymmetries. 2021;24:1–17. [Google Scholar]
14. Bao H.H.G., Tran T.H.L., Le H.P. The roles of vehicle currency and real effective exchange rate in China’s trade with the whole EU. Cogent Econ. Fin. 2022;10:1–16. [Google Scholar]
15. Vo V.D., Le H.D.P. External linkages and product innovation: theory and empirical evidence from subsidiaries in Vietnam. Int. J. Econ. Manag. 2017;11:3–47. [Google Scholar]
16. Vo V.D., Akbar Y.H., Truong D.L. The moderating role of subsidiary size on the association between institutional distance and subsidiary's access to complementary local assets: evidence from Vietnam. J. Asia Bus. Stud. 2022;16(6):963–987. [Google Scholar]
17. Trinh P.T.T. The impact of exchange rate fluctuation on trade balance in the short and long run: the case of Vietnam. J. Southeast Asian Econ. 2014;31(3):432–452. [Google Scholar]
18. Bao H.H.G., Le H.P. Asymmetric impact of exchange rate on trade between Vietnam and each of EU-27 countries and the UK: evidence from nonlinear ARDL and the role of vehicle currency. Heliyon. 2021;7 [PMC free article] [PubMed] [Google Scholar]
19. Baharumshah A.Z. The effect of exchange rate on bilateral trade balance: new evidence from Malaysia and Thailand. Asian Econ. J. 2001;15(3):291–311. [Google Scholar]
20. Ogbonna B.C. Inflation, exchange rate and efficacy of monetary policy in Nigeria: the empirical evidence. Acad. Int. Sci. J. 1982;13:40–53. [Google Scholar]
21. Wilson P. Exchange rates and the trade balance for dynamic Asian economies - does the J-curve exist for Singapore, Malaysia, and Korea? Open Econ. Rev. 2001;12:389–413. [Google Scholar]
22. Kwalingana S., Simwaka K., Munthali T., Chiumia A. The short-run and long-run trade balance response to exchange rate changes in Malawi. J. Dev. Agric. Econ. 2012;4(8):221–232. [Google Scholar]
23. Shin Y., Yu B., Greenwood-Nimmo M. In: The Festschrift in Honor of Peter Schmidt.: Econometric Methods and Applications. Horrace W., Sickles R., editors. Springer; New York, NY, USA: 2014. Modelling asymmetric cointegration and dynamic multipliers in a nonlinear ARDL framework; pp. 281–314. [Google Scholar]
24. Pesaran M.H., Shin Y., Smith R.J. Bounds testing approaches to the analysis of level relationships. J. Appl. Econom. 2001;16(3):289–326. [Google Scholar]
25. Bahmani-Oskooee M., Gelan A. Is there J-curve effect in Africa? Int. Rev. Appl. Econ. 2012;26(1):73–81. [Google Scholar]
26. Lal A.K., Lowinger T.C. Nominal effective exchange rate and trade balance adjustment in South Asia countries. J. Asian Econ. 2002;13(3):371–383. [Google Scholar]
27. My N.Q., Sayim M., Rahman H. The impact of exchange rate on market fundamentals: a case study of. J-curve Effect Vietnam’ Res. Appl. Econ. 2017;9(1):45–68. [Google Scholar]
28. Brown R.L., Durbin J., Evans J.M. Techniques for testing the constancy of regression relationships over time. J. Roy. Stat. Soc. 1975;37(2):149–192. [Google Scholar]
Articles from Heliyon are provided here courtesy of Elsevier
