1.1 China-US trade agreement and its divergent effect on trading partners

The China-US Economic and Trade Agreement was recently signed in 2020, with China agreeing to halve tariffs on 1,717 US items such as soybeans and crude oil that were imposed in 2018 [6]. The US is also expected to lower duties on certain Chinese imports in lockstep. Tables 1 and 2 illustrate China’s imports and exports of products from its major trade partners in 2018.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. China imports (2018) in US billion $.

https://doi.org/10.1371/journal.pone.0275859.t001

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. China exports (2018) in US billion $.

https://doi.org/10.1371/journal.pone.0275859.t002

China clearly imports a lot of products from these countries, such as crude oil worth $20.67 billion from Russia, industrial machinery worth $20.83 billion from Germany, and oilseeds totaling $20.32 billion from Brazil [7]. Since these products are also covered by the Economic and Trade Agreement, there must be trade diversion impacts on foreign trading partners if China redirects its imports to the US to meet purchasing commitments. Due to reduced soybean shipments, Brazil will be the most affected among China’s key trading partners in the agriculture sector. Brazil is the world’s largest exporter of soybeans, with China accounting for roughly 80% of those shipments in 2017. Other developing nations, in addition to Brazil, are impacted by the trade agreement between the US and China [8].

1.2 China’s major trading partners

Table 3 depicts China’s trade relations with its trading partners. According to Chinese trade data from 2018 (total trade volume), the mentioned trading partners include 28 European Union (EU) countries, ten Association of Southeast Asian Nations (ASEAN) nations (Japan, South Korea, Hong Kong, and Taiwan), and the United States.

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. China’s major merchandize trading partners in 2018 ($ billions).

https://doi.org/10.1371/journal.pone.0275859.t003

According to trade statistics, the United States, the EU, and ASEAN were China’s top three export markets, while the EU, ASEAN, and South Korea were its main import sources. China has large trade surpluses with the US ($282 billion), Hong Kong ($274 billion), and the European Union ($129 billion), but large trade deficits with Taiwan ($112 billion) and South Korea ($74 billion). However, China’s trade statistics differ greatly from those of the United States [9]. Further, Fig 1 depicts China’s total volume of trade and its split into exports and imports with the major trade partners.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. China’s major merchandize trading partners in 2018.

According to trade statistics, the United States, the EU, and ASEAN were China’s top three export markets, while the EU, ASEAN, and South Korea were its main import sources. China has large trade surpluses with the US ($282 billion), Hong Kong ($274 billion), and the European Union ($129 billion), but large trade deficits with Taiwan ($112 billion) and South Korea ($74 billion). However, China’s trade statistics differ greatly from those of the United States.

https://doi.org/10.1371/journal.pone.0275859.g001

1.3 China’s exchange rate and foreign trade policy

For the past few years, China has primarily fixed the RMB against the US dollar [10], with the central bank publishing a central parity rate (CPR) of exchange between the RMB and the US dollar on a daily basis. The Chinese government’s central exchange rate with the US dollar was 8.28 RMB per dollar on average in 1998, and this rate stayed stable until 2005 [3, 11]. However, under pressure from trading partners such as the United States, China promised to increase the RMB’s value by 2.1 percent over the next three years by pegging its currency to a basket of currencies [4]. Due to the global economic crisis, China suspended RMB appreciation in 2008 in order to reduce the impact of financial problems by expanding exports. The RMB appreciated 35.3 percent against the US dollar between 2005 and 2015. The Chinese central bank declared in 2015 that the daily RMB central parity rate would become more "market-oriented." The RMB, on the other hand, fell by 4.4 percent against the dollar in the three days that followed, and it continued to fall against the dollar for the rest of 2015 and into 2016. Between August 2015 and December 2016, the RMB lost 8.8% of its value against the U.S. dollar. Between January and December 2017, the RMB appreciated by 4.6 percent against the USD. The RMB’s value versus the dollar has been trending lower since 2018.

Many experts believe the RMB’s recent depreciation is partly due to China’s weakening economy and the uncertainty created by the ongoing US-China trade war. Due to the RMB’s 8.2 percent depreciation from March 2018 to May 2019, the US administration boosted tariffs on Chinese imports from 10% to 25% in May 2019, which will have a detrimental impact on China’s economy. Others claim that China has interfered in currency markets to depreciate the RMB in order to mitigate the economic impact of US tariff rises. China surpassed Germany in 2009 to become the world’s largest merchandise exporter as well as the world’s second-largest merchandize importer after the United States. However, China became the world’s largest merchandise trading (exports plus imports) economy in 2012, surpassing the United States. China’s proportion of worldwide merchandize exports increased from 2.0% in 1990 to 14.1 percent in 2015, but then dropped to 13.4% in 2016 and 13.2% in 2017.

In a study, Xu [12] examines in detail the fundamentals that determine the exchange rate in China and comes up with two important findings in this regard. The first finding indicates that during the last few decades, China has determined adequate local market prices that are connected to global market prices. As a result, the exchange rate can act as a useful nominal anchor. In terms of exchange rate stability, domestic prices remain steady. Another finding shows that changes in the exchange rate have only a limited impact on trade flows. For current account balances, exchange rate flexibility is not required. As a result, the data suggests that China will stick to its strategy of maintaining exchange rate stability. Galesi and Lombardi [13] investigated the influence of China’s exchange rate policies on trade with Asian countries. The long-run relationship between variables was found by using the cointegration technique. In addition, the reduced form export and import equations were applied to avoid the simultaneous equation bias that would occur if the supply and demand functions were estimated independently. The study indicates that appreciation of the RMB had a negative influence on overall exports to Asian countries. Ahmed [14] investigates the sensitivity of Chinese exports due to changes in exchange rates for the period 1997 to 2004. The author developed two types of export models, one for processed export and the other for non-processed export. The findings show that devaluation reduces exports of both processed and non-processed goods. Non-processing exports are notably negative, whereas processing exports are positive but insignificant. As a result, an increase in China’s exchange rate reduces the country’s trade surplus. For quarterly data from 1995Q1 to 2011Q4, Cardoso and Duarte [15] evaluated the influence of Chinese exchange rate policy on world trade with the European Union. The findings indicate that Chinese exports increased as a result of a currency intervention, which occurs when a government or central bank buys or sells foreign currency in exchange for its own domestic currency.

The literature review revealed that China’s exchange rate is complex, necessitating deeper investigation into trade linkages with currency rates as well as an examination of risk and uncertainty for its trading partners. Several studies on China’s currency depreciation and trade policies, as well as their impact on trading partners, are presented in Table 4, with mixed results, indicating a lack of consensus.

Download:

• PNG
  larger image
• TIFF
  original image

Table 4. Literature on China exchange rate and trade.

https://doi.org/10.1371/journal.pone.0275859.t004

It is critical to comprehend China’s interactions with its trading partners in order to build effective trade policies to limit the impact of Chinese economic shocks. Therefore, the current study is designed to respond to shocks in China’s macroeconomic variables and their influence on its trade partners.

2.1 Structure of the standard GVAR model

As discussed earlier, the GVAR model is not estimated at once but estimated on the basis of each individual country-specific VAR model. The standard VAR model is a general form of the univariate autoregressive model (AR model) by allowing more than one variable. It is the stochastic process used to capture the linear interdependencies among a set of a country’s i′s macroeconomic variables x_it. The VAR (p) for the country i is presented as: (1)

In Eq (1), α and β are the coefficient vectors and ε_it is assumed independent identically distributed i.e. ε_it~i.i.d. (0, ∑ii). x_it is a vector of endogenous domestic variables in our case it is consisting of the real exchange rate (EX_it), trade volume (T_it), real equity price (Q_it), and real GDP (Y_it). In vector form x_it = (T_it, Q_it, EX_it, Y_it,)′.

2.2 Country specific VARX models

VARX is the extension of the VAR (p) model that includes the country-specific foreign variables . The external variables are calculated by the weighted sum of specific variables. The weight w_ij is the bilateral trade weight of country i in country j, therefore w_it = 0 and . There weights will also be used for the computation of the foreign variables w_ij is separating the share of country j in the trade (export plus imports) of country i. The country level trade share is constructed by dividing the total trade of each country i by the amount of trade with country j, such the i_th row sums to one, for all the i. , the vector of weak exogenous external variables is expressed as:

Where:

The VARX (p, q) for country i is presented as, (2)

Where;

β_i = k×k Matrix of coefficient associated with lagged domestic variables.

γ_i0 = k×k* Coefficient matrix of foreign specific variables

ε_i = k×1 Vectors denotes concerned country oriented disturbances shocks

In order to obtain the VECM for country i, for simplicity consider VARX (1, 1) as:

The above VAR (1, 1) model is then written in error correction form as: (3)

Where;

ECM_i,t−1 is a vector of long run co-integrating relations, also known as error correction terms corresponding to the r_i cointegrating relations of the i^th country model.The rank of is determined using the maximum Eigen value or the trace statistics. is estimated by imposing suitable exact or possibly over identifying restrictions on the elements of .

2.3 Testing for weak exogeneity of foreign variables

The GVAR methodology assumes that in country-specific VARX (p, q) models the external variables perform weakly exogenous. For an individual country, foreign variables and global variables are constructed along the lines described by Johansen [22] and Harbo, Johansen, Nielsen, and Rahbek [23] for the test of focused countries. To verify the weak exogeneity of with respect to the domestic variables, following auxiliary regression for each l^th, the element of foreign variables will be estimated: (4)

Test the joint significance of the coefficients on the estimated error correction terms as:

If the F-test cannot reject the null hypothesis so the assumption of weak exogeneity is satisfied. Weak Exogeneity test guides the specification of country-specific models.

2.4 The GVAR model

The second step of GVAR is to arrange the estimated country models in order to get the global VAR model. Therefore, rewriting the country-specific VARX equation as: (5)

Where Y_it is a vector of both x_it and Having dimension I×(k×k*) and π_i = (β_i, 0) matrix having dimension: π_i(k×k*).

As we know that the Y_it can be written as:

w_i Is a k(k×k*) link matrix which is consists of ones and country specific weight i.e.:

In terms of link matrix Eq (5) can be written as: (6)

Rewrite Eq 6 for every country the global VAR model are as follow: (7)

Whereas

Matrix H is full ranked so the global solution can be written as: (8)

Eq 8 is a reduced form of a global solution.

2.5 Impulse response analysis

The estimates of generalized impulse response functions (GIRFs) are based on the estimated GVAR model. The concept of GIRFs was suggested by Koop, Pesaran, and Potter [24] and further applied to the analysis of vector autoregressive (VAR) by Pesaran and Shin [25]. GIRF is different from the orthogonalized impulse response function (OIRF) proposed by Sims [26] for the dynamic analysis of the VAR model, which assumes shocks using Cholesky decomposition of the covariance matrix. This reduces the likelihood of errors. If a VAR consists of two or three variables, we can use orthogonalized impulse response functions (OIRFs). But the GVAR model has a large number of variables, so the OIRFs are not suitable. The advantage of GIRF is that it does not affect variables and countries. Thus, GIRF is determined by how shocks are transferred to variable “j" in the model, but the economic interpretation of these shocks is meaningless because variable “j" affects other different shocks. GIRF is useful for checking the GVAR model and the transmission of different shocks. Now in order for the interpretation of shocks here the generalized impulse response to a one standard error shock to the ‘j’ variable in a country ‘i’at time ‘t’ on the expected values of x at time t+n are given.

Mathematically: (9)

Where ∞_ii,ll is the corresponding diagonal element of the residual, variance-covariance matrix ∑u; is the information set at time t−1.

2.6 Data and variables

This section describes the data and variables used in this study. Variables’ definitions, their measurements, and data sources have been reported in Table 5.

Download:

• PNG
  larger image
• TIFF
  original image

Table 5. Variables’ definitions and data source.

https://doi.org/10.1371/journal.pone.0275859.t005

3.1 Trade weight matrix

While estimating the GVAR model, a trade weight matrix is required and constructed from the trade flow data given in Table 6. In such a case weights w_ij is the country i share in the trade of country j. Trade share matrix of the each country is constructed such that w_ii = 0 and ∑w_ij = 1.

Download:

• PNG
  larger image
• TIFF
  original image

Table 6. Trade weight matrix.

https://doi.org/10.1371/journal.pone.0275859.t006

Table 6 shows that four major trading partners have a 4×4 trade matrix. The trade of a country is the sum of its exports and imports. We have constructed a fixed trade weight matrix based on the average of three years’ trade (such as 2014, 2015, and 2016) between two countries. In the individual country, shares are constructed by dividing the sum of total trade in the given period of each country i, by the amount of trade share with country j, such that the i_th row sums to one for all i. Considering China, we can see that the Chinese trade share with Germany was 16.1%, Japan had 27.3%, and the remaining 56.8% was traded to the US.

3.2 Unit root test

The unit root test is applied to test the null hypothesis of non-stationarity at both the level and first difference of variables [27]. Table 7 shows that the domestic real GDP of Germany, Japan, and the US is non-stationary I (1), while in the case of China, the real GDP is also non-stationary even at the first difference. For Japan, the variable of equity prices is stationary at level i. e. I(0), and for the other countries, the series is stationary at first difference i. e. I(1). The domestic trade volume of China is found stationary at a level, i. e., I (0), while the others are I (1). In the case of the real exchange rate, China, Japan, and US data series are non-stationary, and the first-difference non-stationary in the case of Germany. The foreign real GDP for all countries is I (1) and so are real equity prices, also I (1). The foreign trade volume of China and Japan is I (1), while the trade volume of Germany and the United States is non-stationary at the 5% level of significance at the first difference. The exchange rate for all countries remains stationary at first difference that is I (1).

Download:

• PNG
  larger image
• TIFF
  original image

Table 7. Unit root test for domestic variables.

https://doi.org/10.1371/journal.pone.0275859.t007

3.3 Selection of lag length for the VARX* model

In the VARX* model, the selection of the lag order of domestic and foreign variables is not balanced. This study used Schwartz Bayesian criteria (SBC) for the selection of the lag order of domestic variables p_i and also the selection of the lag order of foreign variables (q_i) The results are displayed in (Table 8), in which the optimal VARX* orders for China, Germany, Japan, and the US are (2, 1), (2, 1), (2, 1), and (2, 1), respectively.

Download:

• PNG
  larger image
• TIFF
  original image

Table 8. Order of VARX* model.

https://doi.org/10.1371/journal.pone.0275859.t008

3.4 Co-integration test

Individual country co-integrating space rank is computed using trace statistics developed by Dickey and Fuller [28], which require integrated of order one and weakly exogenous independent variables. In Table 9, the trace test indicates one co-integrating relationship among the variables for all countries except China, which has two co-integrating vectors.

Download:

• PNG
  larger image
• TIFF
  original image

Table 9. Trace statistic test for Co-integration.

https://doi.org/10.1371/journal.pone.0275859.t009

3.5 Testing weak exogeneity

As previously stated, the key estimating approach for in terms of the long run parameter is weak exogeniety. Table 10 indicates that, with the exception of Japan’s real GDP, trade volume, and real exchange rate, the assumption of weak exogeniety is met for all nations’ models. It means that the real GDP of Japan is influenced by the GDP of its trade partners but that the GDP of its trading partners is unaffected by Japanese GDP. Furthermore, while Japanese trade volume is influenced by other trade volumes, its trading partners’ trade volumes are unaffected. Other real exchange rates have little effect on the Japanese real exchange rate.

Download:

• PNG
  larger image
• TIFF
  original image

Table 10. Weak exogeneity of the country-specific foreign variables.

https://doi.org/10.1371/journal.pone.0275859.t010

3.6 Structural stability test

In the context of a co-integrated model, structural stability is applicable to both short- and long-term coefficients and also to the error variance. As our interest is in discovering the transmission mechanisms between China and its major trading partners, we focus on the stability of short-run coefficients. Here we consider the country-specific error correction models of the residuals of the individual equations, which are based on structural stability tests. These residuals do not depend on the identification of co-integrating relations but depend on the rank of the co-integrating vectors. In our empirical analysis, different tests are included, such as the maximal OLS cumulative sum (CUSUM) statistic suggested by Ploberger and Krämer [29] and denoted by PK_sup and its mean square variance by PK_msq. The PK_sup statistic is similar to the CUSUM test originated by Brown, Durbin, and Evans [30], although the latter is based on recursive rather than OLS residuals to test the null hypothesis of parameter stability. As per the PK_sup test which is reported in Table 11, the US real GDP is not stable at the 5% level of significance.

Download:

• PNG
  larger image
• TIFF
  original image

Table 11. PK_sup test statistics.

https://doi.org/10.1371/journal.pone.0275859.t011

However, according to the PK_msq test (reported in Table 12), all parameters are stable at a 5% level of significance except the US real GDP. The statistics based on parameter instability are computed through critical values of PK_msq, test, which are calculated from the GVAR (p) model solution using a boot-strap sample. Both versions of the tests of stability in parameters are rejected in at most 7 out of 34 cases. The tests have little statistical evidence to reject the null hypothesis that the coefficients are stable in the case of 90% boot-strap in our GVAR model.

Download:

• PNG
  larger image
• TIFF
  original image

Table 12. PK_msq test statistics.

https://doi.org/10.1371/journal.pone.0275859.t012

3.7 Contemporaneous effects of foreign variables on their domestic counterpart

Table 13 shows that the contemporaneous effects of a foreign variable on domestic variables are collectively provided by the computed t-ratio using the Newey-West variance estimator. These estimates interpret the foreign variables’ impact on domestic variables. Most of the expected elasticities have a positive sign, which is significant and informative regarding international transmission from foreign to domestic variables. For each country, positive values indicate that the domestic variables are over-affected by foreign variables and negative values indicate that domestic variables are unaffected by foreign variables.

Download:

• PNG
  larger image
• TIFF
  original image

Table 13. Contemporaneous effect of foreign variables on domestic variables.

https://doi.org/10.1371/journal.pone.0275859.t013

To show the variables’ relationship, we have seven assumptions, which are as follows:

Firstly, the foreign GDP has a significant impact on domestic GDP in the whole sample period. The hypothesis will be; H₀ = γ_i0,G,G* = 0 VS H_a = γ_i0,G,G* ≠ 0.

Rejection of the null hypothesis leads to interdependence relationship between foreign GDP and domestic GDP.

Secondly, the foreign equity prices have a significant impact on domestic equity prices in the whole sample period; H₀ = γ_i0,eq,eq* = 0 VS H_a = γ_i0,eq,eq* ≠ 0.

Rejection of the null hypothesis leads to interdependence relationship between foreign equity prices and domestic equity prices.

Thirdly, the foreign trade volume has a significant impact on domestic trade volume in the whole sample period; H₀ = γ_i0,tv,tv* = 0 VS H_a = γ_i0,tv,tv* ≠ 0.

Rejection of the null hypothesis leads to interdependence relationship between foreign trade volume and domestic trade volume.

Fourthly, the foreign exchange rate has a significant impact on the domestic exchange rate in the whole sample period, H₀ = γ_i0,ex,ex* = 0 VS H_a = γ_i0,ex,ex* ≠ 0

Rejection of the null hypothesis leads to interdependence relationship between foreign exchange rate and domestic exchange rate.

These elasticities are very informative about the international linkages between the domestic and foreign variables. Table 13 shows that, focusing on the Chinese economy, a 1% change in foreign real GDP in a given period leads to an increase of 0.5% in Chinese real GDP, but the effect is insignificant within the same period. Similarly, foreign real GDP elasticities are obtained from the different regions because the effect is slightly weaker for Germany, Japan, and the US. In the case of the Chinese economy, the elasticity estimate is relatively large but insignificant. More importantly, the contemporaneous elasticity of real equity prices in China is greater than the elasticities of other economies. Hence, the Chinese stock market seems to overreact to foreign stock prices, but the effect is insignificant. Similar results were obtained for Germany and Japan, but they remained insignificant. The effects of US financial linkage are strong and significant. We can also observe the strong co-movement between China and foreign trade volume. And the remaining economies, like Germany, Japan, and the US, are also positively affected by foreign trade volume, which is significant for all countries. In contrast, Germany’s exchange rate is positively affected by the foreign exchange rate and the remaining economy, Japan’s exchange rate is not much affected by the foreign exchange rate, and China’s exchange rate is slightly negatively affected by the foreign exchange rate.

3.8 Generalized impulse response functions

To study the dynamic properties of the global model and to measure the time profile of the Chinese economy’s effects on macroeconomic variables of its major trading partners’ economies, we investigate the effects of a one-standard-error positive shock on Chinese real GDP, stock prices, trade volume, and the exchange rate. Here we use the generalized impulse response to the estimated GVAR model, which was proposed by Ploberger and Krämer [29] and further added a new option by Koop et al. [24]. The GIRF is the more appropriate dynamic for the transfer of shocks from China to its major trading partners. In Fig 2, a one-standard-error positive shock to Chinese real GDP is given to trace the profile of its impact on the macroeconomic variables of the trading partners. The red line shows the significant impact. It can be seen from Fig 2.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Shocks to Chinese real GDP and its impact on trading partners.

A shock to Chinese real GDP has an insignificant impact on the macroeconomic variables of Germany. Japan will enjoy the shock of Chinese GDP by improving its own. However, it appreciates the US currency only among its macroeconomic variables and will worsen its GDP.

https://doi.org/10.1371/journal.pone.0275859.g002

China’s GDP shock boosts Germany’s GDP and total trade while also appreciating its currency over time. However, it has a negligible impact on its stock prices. The same results appeared for Japan. In the case of the US, it worsens the US’s GDP and its stock prices but appreciates its currency. Therefore, it is concluded that the shock to Chinese GDP has an insignificant impact on the macroeconomic variables of Germany. Japan will enjoy the shock of Chinese GDP by improving its own. However, it appreciates the US currency only among its macroeconomic variables and will worsen its GDP.

The findings of our study are consistent with Chen, Huang, and Huang [31], who found that a negative shock in Japan’s real output causes an immediate fall in Taiwan’s and Korea’s real GDP. However, these impacts quickly become positive from quarters 3 through 9. Furthermore, following quarter 9, Taiwan’s real GDP falls to a negative value, while Korean real GDP remains positive. Japan is not just Taiwan’s largest trading partner but also a major supplier of materials and components to important sectors in Taiwan. If Japan’s real output suffers an external negative shock, Taiwan’s real GDP also suffers as a result of this strong economic connection. A decline in US real GDP may imply a decrease in the real output of many countries around the world, as well as fewer job opportunities in countries whose economies are heavily dependent on trade with the US, such as Taiwan. As shown in the middle-right panel of Fig 2, a decline in US real GDP reduces output in almost all selected foreign countries, with the effects being especially pronounced in US neighboring countries such as Mexico and Canada. Furthermore, the impacts are severe for some countries, particularly the smaller Asian economies (e.g., Taiwan, Hong Kong, and Singapore). Large Asian countries like Japan, Korea, and China appear to be relatively unscathed.

China’s output shock, on the other hand, has no effect on macroeconomic activity in Taiwan, Korea, or the rest of the world. Taiwan and the majority of selected countries have limited and insignificant effects on real output, real equity prices, and interest rates. These findings could be explained by the fact that China has recently experienced rapid growth. Our study only used data up to 2008: QIV, which may not be long enough to capture the impact of a shock to macroeconomic factors in China on the world as significantly as one in the United States. According to Dinda’s [32] research, the People’s Republic of China’s (PRC) slowdown has a greater impact on emerging BRICS economies than on developed economies. The PRC’s GDP growth shock has had a limited impact on other developing economies. In terms of the global linkage variable, the Chinese economic slowdown has a direct impact on emerging and developed economies.

In Fig 3, a one-standard-error positive shock to Chinese trade volume is given. It can be seen that the shock to Chinese trade a little bit improves the GDP and total trade volume of Germany and has an insignificant impact on its currency and its stock prices. It deteriorates Japan’s GDP and total trade volume. Whereas the US is more vulnerable to shocks from Chinese trade because it decreases its trade volume and depreciates its currency permanently, Germany’s economy is resilient to the trade shock from China. According to Zhang and Gui [33], China’s trade production has strong network effects as a result of Asia’s growing integrated regional production system. Such an integrated system has boosted exports from these countries to China as well as overall exports to third markets. With the exception of Mexico, China’s economy has no significant impact on trade in Latin America and Africa. Countries with export structures similar to China’s have experienced competitive pressure from China, whereas countries with import structures similar to China’s have experienced complementary effects from China’s economic development.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. Shocks to Chinese trade volume and its impact on its trading partners economies.

A shock to Chinese trade volume has a little bit improved the GDP and total trade volume of Germany and has had an insignificant impact on its currency and its stock prices. It deteriorates Japan’s GDP and total trade volume. Whereas the US is more vulnerable to shocks from Chinese trade because it decreases its trade volume and depreciates its currency permanently, Germany’s economy is resilient to the trade shock from China.

https://doi.org/10.1371/journal.pone.0275859.g003

Fig 4 shows that a shock to Chinese equity prices reduces GDP, trade volume, and strengthens Germany’s currency while having little effect on stock prices. In the case of Japan, this shock deteriorates its GDP, which is insignificant to its trade volume, increases equity prices, and depreciates the currency. It significantly appreciates the US currency and lowers its GDP. Therefore, it is concluded that the shock to Chinese equity prices is more alarming for the German economy than the Chinese one because it permanently reduces its GDP, deteriorates its trade volume, appreciates its currency, and improves its stock prices. The Japanese economy seems to be doing well in terms of its GDP improvement. It also significantly appreciates the US currency. According to Shi [34], for the group of developed partners, the Chinese stock market is more correlated with Asian markets than with Pacific markets, and for the group of emerging partners, it is more integrated with East Asian markets. Although different pair-wise correlations reveal different evolving patterns, most pair-wise co-movements show increasing trends between mid-2015 and early-2018. Second, contagion episodes suggest that turbulent events amplify stock market co-movements, with levels of contagion frequency increasing during the Shanghai stock market crash, US-China tariff wars, and COVID-19 pandemic. However, the findings for developed and emerging economies differ slightly.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. Shocks to Chinese stock prices and its impact on its trading partners economies.

A shock to Chinese equity prices is more alarming for the German economy than the Chinese one because it permanently reduces its GDP, deteriorates its trade volume, appreciates its currency, and improves its stock prices. The Japanese economy seems to be doing well in terms of its GDP improvement. It also significantly appreciates the US currency.

https://doi.org/10.1371/journal.pone.0275859.g004

There are signs that stock market co-movement between China and its developed partners is more sensitive to US-China trade frictions than stock market co-movement between China and its emerging partners, which is more influenced by the Shanghai stock market crash. Third, the regression results reveal several regularities in the drivers of stock market co-movement. According to Shu, He, Wang, and Dong [35], who examined China’s impact on Asian-Pacific financial markets, the analysis shows that China has a significant influence on regional equity and currency markets. In normal or non-stress times, China’s influence on Asian stock markets approaches that of the US, though spillovers from the US to Asian financial markets tend to be stronger than China’s influence in stressful times. Following China’s transition to a managed float regime in 2005, the RMB has also gained the ability to move regional currencies. Nonetheless, China’s bond market remains isolated from those of the rest of Asia and the US. China’s growing influence, combined with that of the United States, has significant implications for Asian financial markets and capital flows. Then the United States’ influence on global markets remains dominant, driving global liquidity conditions and risk appetite, especially during times of stress.

Fig 5 indicates the impact of a shock to the Chinese exchange rate on the macroeconomic variables of its trading partners. It is clearly evident that the shock to the Chinese exchange rate is good for Germany as it improves its GDP and trade volume, but bad for the US as it worsens its equity prices. However, Japan’s currency will be stronger in terms of appreciation due to the shock in the Chinese exchange rate.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 5. Shocks to Chinese real exchange rate and its impact on its trading partners economies.

A shock to the Chinese exchange rate is good for Germany as it improves its GDP and trade volume, but bad for the US as it worsens its equity prices. However, Japan’s currency will be stronger in terms of appreciation due to the shock in the Chinese exchange rate.

https://doi.org/10.1371/journal.pone.0275859.g005

According to Ahmed et al. [14], an appreciation of the Chinese currency should unambiguously reduce both Chinese processing and non-processing exports to its trade partners with advanced economies. Our empirical results are generally consistent with these predictions. We found a significant negative effect of RMB appreciation on China’s exports to advanced economies. However, the impact of appreciation was found to be positive on the exports of processed goods to emerging Asian economies. But this finding was statistically insignificant. Furthermore, the impact of global imbalances is dependent on how they are defined, which is not always clear. There is substantial disagreement over the proxies used to quantify global imbalances, and the findings may vary as these proxies change. Our findings suggest that if China’s large current account surplus or bilateral current account surplus with the US, along with the US bilateral current account deficit with China, contributes to global imbalances, a greater degree of Chinese currency appreciation would significantly help mitigate global imbalances. If, on the other hand, the US overall current account deficit and the emerging market world’s current account surplus are the major components of global imbalances, then it is less evident that further Chinese currency appreciation would put a significant dent in global imbalances.

The existing literature shows that the exchange rate of China has a direct influence on its trading partners. China’s trading partners’ exports have suffered as a result of the currency’s devaluation. Since China has a managed float exchange rate system, it allows a sovereign central bank to regularly intervene in the market to modify the direction of the currency float and boost its balance of payments in completely impulsive phases. As a result, Chinese exports are rapidly rising. Besides this, other initiatives in the form of BRI may further expand the volume of trade with major trade partners. Aside from this, additional BRI initiatives may boost the volume of trade with major trading partners. The GVAR approach is preferred for capturing impacts in such global dynamic settings [31]. As a result, we investigated the influence of Chinese macroeconomic performance, particularly the exchange rate, on global trade. The survey of literature also indicated that China’s exchange rate is complicated, necessitating a more in-depth inquiry of trade linkages with currency rates as well as an evaluation of risk and uncertainty for its trading partners. Table 4 presents the findings of various studies on China’s currency devaluation and trade policies, as well as their influence on trading partners, with conflicting results showing a lack of consensus. Understanding China’s interactions with its trading partners is crucial for developing successful trade policies that reduce the effects of Chinese economic shocks. As a result, the current study is intended to respond to shocks in China’s macroeconomic variables and their impact on its trading partners.