Will Substantial Revaluation of Chinese Currency Be Helpful in Attenuating the United States’ Trade Deficit with China? New Evidence from the Dynamic Threshold Kink Model

Abstract

This study explores the possible existence of threshold kink
effects in the impact of the real dollar-renminbi exchange rate
misalignment on the bilateral trade flows between the United
States and China. Empirical results show a significant threshold
effect, with differential effects of undervaluation and overvaluation. Specifically, our findings suggest that while an overvaluation of the Chinese currency over 2.83% will attenuate the United States’ trade deficit with China, an undervaluation
greater than 4.04% will deepen it. This suggests that a revaluation of the renminbi against the dollar would be helpful in improving the United States trade balance with China.

Full text

Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=uitj20
The International Trade Journal
ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/uitj20
Will Substantial Revaluation of Chinese Currency
Be Helpful in Attenuating the United States’
Trade Deficit with China? New Evidence from the
Dynamic Threshold Kink Model
Sabrine Ferjani, Sami Saafi & Ridha Nouira
To cite this article: Sabrine Ferjani, Sami Saafi & Ridha Nouira (2023): Will Substantial
Revaluation of Chinese Currency Be Helpful in Attenuating the United States’ Trade Deficit with
China? New Evidence from the Dynamic Threshold Kink Model, The International Trade Journal,
DOI: 10.1080/08853908.2023.2236220
To link to this article: https://doi.org/10.1080/08853908.2023.2236220
View supplementary material
Published online: 23 Jul 2023.
Submit your article to this journal
View related articles
View Crossmark data

Will Substantial Revaluation of Chinese Currency Be Helpful
in Attenuating the United States’ Trade Decit with China?
New Evidence from the Dynamic Threshold Kink Model
Sabrine Ferjani
a
, Sami Saa
b
, and Ridha Nouira
c
a
LAMIDED, ISG Sousse, Sousse University, Sousse, Tunisia;
b
LAMIDED, ISG Sousse, Sousse University and
FSEG Mahdia, Monastir University, Mahdia, Tunisia;
c
LAMIDED, ISG Sousse and ISFF Sousse, Sousse
University, Sousse, Tunisia
ABSTRACT
This study explores the possible existence of threshold kink
eects in the impact of the real dollar-renminbi exchange rate
misalignment on the bilateral trade ows between the United
States and China. Empirical results show a signicant threshold
eect, with dierential eects of undervaluation and overvalua-
tion. Specically, our ndings suggest that while an overvalua-
tion of the Chinese currency over 2.83% will attenuate the
United States’ trade decit with China, an undervaluation
greater than 4.04% will deepen it. This suggests that
a revaluation of the renminbi against the dollar would be help-
ful in improving the United States trade balance with China.
KEYWORDS
China; dynamic threshold
kink model; exchange rate
misalignment; nonlinearity;
the United States; trade
balance
I. Introduction
China’s bilateral trade surplus with the rest of the world in general, and with
the United States in particular, has increased substantially over the few past
decades, resulting in an ongoing trade war between the two countries (Moosa
et al. 2020; Wan 2020). According to United States Census Bureau trade
statistics, China’s trade surplus with the United States jumped in value from
$10.43 billion in 1990 to $335.3 billion in 2021. China has been criticized
intensely by a considerable part of American policymakers and academic
economists for this trade surplus, and it has been commonly argued that the
Chinese government should reevaluate further its undervalued currency, the
renminbi (RMB).
In the economic literature, however, the relationship between the Chinese
exchange rate and the United States-China trade imbalance remains an unre-
solved debate, and it is still undecided whether a revaluation of the RMB
against the dollar would be helpful in improving the United States trade
balance with China (see, for instance, Tu and Zhang 2019; Wang 2020;
Weber and Shaikh 2021). While studies such as those by Chiu, Lee, and Sun
CONTACT Sami Saafi samisaafifsegm@gmail.com LAMIDED, ISG Sousse, Sousse University and FSEG Mahdia,
Monastir University, Sidi Messaoud - Mahdia - Hiboun - 5111, Mahdia, Tunisia
Supplemental data for this article can be accessed online at https://doi.org/10.1080/08853908.2023.2236220
THE INTERNATIONAL TRADE JOURNAL
https://doi.org/10.1080/08853908.2023.2236220
© 2023 Taylor & Francis Group, LLC

(2010), Wang, Lin, and Yang (2012), Cheung, Chinn, and Qian (2016), Hurley
and Papanikolaou (2021), and Islam (2022) suggest that the United States’
bilateral trade deficit with China can be reduced through a revaluation of the
RMB, other studies such as those by Zhang and Sato (2012), Kim and Kim
(2016), Shi and Li (2017), Nasir and Jackson (2019), Moosa et al. (2020), and
Weber and Shaikh (2021) claim that such a RMB revaluation would only have
limited effects.
1
One of the possible reliable explanations for divergent findings in the
literature could be that previous studies have implicitly maintained that the
relationship between real exchange rate (RER) misalignment and the trade
balance is linear, which means that they so far neglect the possible threshold
effects in the response of trade flows to RER misalignment. Indeed, an
important number of recent empirical studies provides some evidence that
the connections between the RER misalignment and the economic activity in
general, and between the RER misalignment and the trade flows in particular,
are unlikely to be linear and may involve threshold effects (Béreau,
Villavicencio, and Mignon 2012; Couharde and Sallenave 2013; Cuestas,
Mourelle, and Regis 2020; Fišera and Horváth 2022; Grekou 2015; Tipoy,
Breitenbach, and Zerihun 2017). Along with the above empirical evidence,
there is also a deeper theoretical support for this possible threshold effect. In
fact, the theoretical hysteresis models formulated by Baldwin (1988), Baldwin
and Krugman (1989), and Dixit (1989) predict that in the presence of sunk
costs, the decision of a trader to enter or leave the foreign market depends
nonlinearly on the magnitude of exchange rate fluctuations. According to
Verheyen (2013), the logic underlying such nonlinearity could be that traders
do not respond to every change in the exchange rate since any adjustment
involves sunk costs. In other words, traders’ decisions are expected to be more
prone to large exchange rate fluctuations than to small ones. Therefore, there
may be a “band of inaction” and traders do not have enough incentives to
make any trade adjustment if exchange rate fluctuations still remain in that
band. They would adjust only when fluctuations reach a certain critical size
called “pain threshold” (Belke and Kronen 2019; Belke, Göcke, and Günther
2013).
Kilian and Taylor (2003) and Arize, Malindretos, and Igwe (2017)
provide an alternative economic explanation for the threshold behavior
in the response of the trade balance to the exchange rate. They argue
that in addition to trade hysteresis, a nonlinear threshold effect could
emanate because market participants have heterogeneous beliefs con-
cerning the accurate equilibrium level of the nominal exchange rate
1
Yu and Zhang (2019) and Weber and Shaikh (2021), for instance, argue that the main factors driving the persistent
trade deficit between the United States and China are, inter alia, the decline in the United States’ saving, the
difference in factor endowments, the use of the dollar as an international reserve currency, and not the extent of
the Chinese currency misalignment.
2S. FERJANI ET AL.

and the degree of RER misalignment. Wu, Liu, and Pan (2013) empha-
size that nonlinearity may also arise from the changes in monetary
policies. Various levels of change in the monetary supply would induce
distinctive impacts on the relationship between the RER and the trade
balance. Furthermore, one can argue that since it has been shown that
the relationship between exchange rate or exchange rate volatility and
the trade balance is subject to threshold effects (see, e.g., Ben Doudou
et al. 2022; Chang et al. 2020; Chen 2012; Hashmi, Chang, and Shahbaz
2021; Loermann 2021; Wu, Liu, and Pan 2013), it is possible that the
same holds for RER misalignment.
Therefore, in this article, we aim to make an empirical contribution to
the current debate and revisit the impact of the real dollar-RMB rate
misalignment on the trade flows between the United States and China. As
such, the current research study distinguishes itself from the extant litera-
ture in several ways. First, unlike most prior studies that have relied
exclusively on the conventional linear framework, this study makes use of
a more realistic and suitable framework allowing for the possible existence
of potentially nonlinear threshold effects. Second, to the best of the authors’
knowledge, the present study is the first to apply an innovative dynamic
threshold kink model recently developed by Hansen (2017) to assess the
potential nonlinearities in the relationship between RER misalignment and
trade flows. One of the appealing advantages of Hansen’s kink model is that
it allows us to determine endogenously misalignment threshold levels. This
may offer a way to reconcile the divergent findings reported in the related
literature. Third, instead of using aggregate trade data, we employ highly
disaggregated industry-level data in order to control for possible hetero-
geneity across industries. As pointed out by Diop, Goujon, and Niang
(2018) and Wong (2019), different industries might have various levels of
sensitivity to RER misalignment. Such an industry-specific analysis would
help us, therefore, to check whether there exists or not a common threshold
effect in the RER misalignment-trade flows nexus across the investigated
industries. Fourth, the measures of RER misalignment employed in pre-
vious studies are mostly based on annual data. In contrast, in this article, we
rely on monthly time-series data. This is motivated by the findings of
Giordano (2021), who provides some evidence that misalignment estimates
might be affected by data frequency. Fifth, we decompose our measure of
RER misalignment into undervaluation and overvaluation as they might not
have the same threshold effects on trade flows (Allegret and Sallenave 2014;
Ferjani et al. 2022; Fišera and Horváth 2022).
The remainder of the article is organized as follows. Section 2 is devoted to
the estimates of the magnitude of the RER misalignment, while Section 3
presents the econometric methodology. Section 4 discusses the empirical
results, and Section 5 concludes.
INTERNATIONAL TRADE JOURNAL 3

II. Estimation of the real exchange rate misalignment
The main objective of this study is to examine empirically whether there have
been nonlinear threshold effects in the response of bilateral trade flows
between the United States and China to dollar-RMB exchange rate misalign-
ment. Accordingly, the first step of our analysis is to measure the degree or
magnitude of that misalignment. In doing so, we follow the recent empirical
contributions of Comunale (2017, 2019), Fidora, Giordano, and Schmitz
(2021), Giordano (2021), and Razek and McQuinn (2021), among others,
and rely on the behavioral equilibrium exchange rate (BEER) approach
2
which computes directly the equilibrium value of RER through the assessment
of a long-run relationship between RER and its economic fundamentals (see
Clark and MacDonald 1998). The BEER approach presents five main appeal-
ing features: (i) it does not need assumptions concerning the internal and
external balances, (ii) it considers short-run cyclical/temporary determinants
that may contribute strongly to medium-to-long-run changes of the equili-
brium exchange rate, (iii) it accounts for possible stock impacts via the net
foreign asset position, (vi) it is quite robust to alternative econometric speci-
fications, and (v) it is particularly more suitable in the case of small data
samples (Bénassy-Quéré, Béreau, and Mignon 2009; Ca’zorzi et al. 2020;
Carrera et al. 2021; Couharde et al. 2018; Schnatz 2011).
In this article, we follow Lòpez-Villavicencio, Mazier, and Saadaoui (2012),
Couharde et al. (2018), Comunale (2019), and Ferjani et al. (2022), among
others, and retain the three following variables as economic fundamentals
determinants of RER: net foreign assets scaled by trade volume (NFAT),
a proxy of the Balassa-Samuelson effect (BS), and the terms of trade (TOT).
3
Therefore, we assume that the long-run BEER equation takes the following
form:
4
ln RERt
ð Þ ¼ α0þα1ln NFATt
ð Þ þ α2ln BSt
ð Þ þ α3ln TOTt
ð Þ þ εt(1)
where RER is the real bilateral exchange rate between the dollar and RMB. It is
defined as RERit ¼CPICH
t
NEXtCPIUS
t
ð Þ, where NEXt is the nominal exchange rate defined as
the number of units of renminbi per American dollar; CPIUS and CPICH are,
2
A potential candidate alternative approach, used for instance by You and Sarantis (2011), is the fundamental
equilibrium exchange rate (FEER) approach. Nevertheless, Lòpez-Villavicencio, Mazier, and Saadaoui (2012) show
that the two approaches are, in fact, closely related and converge toward the same equilibrium exchange rate in
the long run. Of course, there are also other techniques such as the purchasing power parity (PPP) model, the
natural real exchange rate (NATREX) model, the Penn effect model, and the ratio model. In estimating the RMB’s
misalignment vis-à-vis the US dollar, Zhang and Chen (2014) rely on three models (the BEER, Penn effect, and ratio
models) and show that the misalignment measure from the BEER model is the best.
3
In addition to those fundamentals, the literature on equilibrium exchange rates also discusses other potential
regressors, such as fiscal variables or interest rate differentials (see Clark and MacDonald 1998; Fidora, Giordano,
and Schmitz 2021; Miyamoto, Nguyen, and Sheremirov 2019). We have decided to restrict our regression to the
three main economic fundamentals variables not only because data for these variables are available at the monthly
frequency, but also because it is shown in a recent study by Ca’zorzi and Rubaszek (2023) that the BEER model with
only three fundamentals outperforms alternative BEER specifications with more regressors.
4
For more details on the theoretical background of this model, see Corden (1994) and Lane and Milesi-Ferretti (2004).
4S. FERJANI ET AL.

respectively, the consumer price index of the United States and China. Thus,
an increase in RERit reflects a real depreciation of the US dollar against the
RMB. NFAT is the ratio of China’s net foreign assets to its trade volume
(export plus import) relative to the corresponding measure for the US.
A debtor country will need a depreciated exchange rate to generate the trade
surpluses necessary to service its external liabilities. Conversely, a country with
a large and positive NFA position can afford a more appreciated exchange rate,
and the associated trade deficit remains solvent (Lane and Milesi-Ferretti
2002). Thus, α1is expected to be positive. The variable BS, which is included
to capture the Balassa-Samuelson effect (Balassa 1964; Samuelson 1964), is
calculated by dividing the CPI/PPI ratio of China by the same ratio of the US.
Therefore, α2 is expected to be positive. TOT refers to the relative terms of
trade, that is, the ratio of terms of trade of China to that of the US.
5
A country’s
terms of trade consist of the ratio of its export unit value to its import unit
value. As noted by Baak (2012), the terms of trade may have a positive or
negative effect on the equilibrium exchange rate. Hence, α3 may be either
negative or positive.
In computing the equilibrium real exchange rate (ERER) and the degree of
misalignment, in line with Nouira, Plane, and Sekkat (2011) and Nouira and
Sekkat (2012), we make use of the cointegration framework which involves the
following steps. In the first step, the well-known augmented Dickey-Fuller
(ADF) unit root test is used to examine the integration order of the variables
involved in Model (1). Once it is established that these variables are integrated
of the same order, the second step would be to test for cointegration among
those variables. For this purpose, we perform the Johansen and Juselius (1990)
technique. As already noted by Gan et al. (2013), this approach can lead to
consistent estimates of the long-run equilibrium parameters. Finally, if the
variables are cointegrated, the third step of our cointegration analysis can be
implemented by estimating the long-run equilibrium relationship. To this end,
as suggested by Comunale (2022), we employ the fully modified ordinary least
squares (FMOLS) method proposed by Phillips and Hansen (1990).
As can be seen in Tables A1 and A2 in the online Appendix, all the variables
are integrated of order 1 (i.e., I (1)) and they are also co-integrated. Thus, the
FMOLS can be implemented to get the long-run estimates of Equation (1). The
results are presented in Table 1.
Based on the estimated coefficients displayed in Table 1, we can derive the
level of the RER misalignment. It should be noted that misalignment is
commonly defined as the gap between the RER and its equilibrium level, the
5
All three explanatory variables included in Equation 1 are constructed in relative terms between China and the US
since RER used here is a bilateral exchange rate which cannot be explained merely by a country’s own
characteristics but should also reflect “foreign characteristics” (Phillips et al. 2013, 8). Qin and He (2011) and
Fidora, Giordano, and Schmitz (2021), among others, have also used relative variables in their estimates of
equilibrium exchange rates.
INTERNATIONAL TRADE JOURNAL 5

ERER. The latter is given by the fitted values using together the estimates in
Table 1 and the long-run equilibrium values of each of the explanatory
variables involved in Equation (1) (i:e:; ln NFATt
ð Þ;ln BSt
ð Þ;ln TOTt
ð ÞÞ. To
obtain such equilibrium values, we follow Clark and MacDonald (1998) and
use the Hodrick-Prescott filter to neutralize the impact of any possible tem-
porary fluctuation in these variables on the ERER. We define misalignment as:
Mis ¼RER=ERERð Þ  1ð Þ100 (2)
A positive misalignment rate corresponds to an overvaluation of the American
dollar (undervaluation of the Chinese RMB) while a negative rate refers to an
undervaluation. Figure 1 illustrates the RER misalignment of the Chinese
RMB calculated as the deviation of the actual RER from its equilibrium values.
As reflected in Figure 1, the misalignment of the Chinese real exchange rate
ranged between –9.27% and 10.18% from January 2008 to September 2021.
The RMB was overvalued at the beginning of the period. As of
November 2010, the Chinese currency has been significantly undervalued.
The extent of undervaluation is relatively modest, reaching a maximum of less
than 10%. In December 2016, however, the Chinese currency began to reveal
overvaluation until November 2020. The RMB was undervalued for the rest of
our sample. Our results are in line with several studies. For instance, Banerjee
and Goyal (2021) estimate RER misalignment for a panel of eight large
Table 1. Equilibrium exchange rate estimates.
Coefficient t-statistic P-value
Net foreign assets 0.006*** 3.063 .002
Balassa-Samuelson 0.253*** 3.064 .002
Terms of trade 0.011 0.177 .859
Intercept 3.380*** 7.22 .000
Note: ***, **, and * indicate the significance level of 1%, 5%, and 10%, respectively.
-15
-10
-5
0
5
10
15
01-01-2008
01-08-2008
01-03-2009
01-10-2009
01-05-2010
01-12-2010
01-07-2011
01-02-2012
01-09-2012
01-04-2013
01-11-2013
01-06-2014
01-01-2015
01-08-2015
01-03-2016
01-10-2016
01-05-2017
01-12-2017
01-07-2018
01-02-2019
01-09-2019
01-04-2020
01-11-2020
01-06-2021
Renminbi misalignment
Renminbi Undervaluation
Renminbi Overvaluation
Figure 1. Misalignment rates (%). Source: Authors’ own calculations.
6S. FERJANI ET AL.

emerging market economies from 1995 to 2017. Their FMOLS estimates show
that the Chinese currency was overvalued over 2008 to 2010 and undervalued
over 2010 to 2013 and in 2015 to 2017. Similarly, Yue, Qiang, and Kai (2016)
show that the RMB was overvalued from 2008 to 2010 and undervalued during
the period of 2011 to 2012.
6
III. Econometric methodology
The economic model
We specify the bilateral trade balance as a function of the level of economic
activity in any two trading partners, the real exchange rate and the exchange
rate misalignment. As such, the reduced form of the equation is as follows (see
Mamun, Akça, and Bal 2021; Rivera-Batiz and Rivera-Batiz 1994):
ln TBt
ð Þ ¼ β0þβ1ln IPUS
t
 þβ2ln IPCH
t
 þβ3ln RERt
ð Þ þ β4MIStþint(3)
where TBt is the trade balance between the United States and China. It is
defined as the ratio of the United States’ imports from China over its exports to
China TB ¼MUS=CH
XUS=CH
 . IPUS
t is the level of economic activity in the United
States, while IPCH
t is the measure of economic activity in China. Since the
data are monthly, the only available measure of economic activity is the
industrial production index (IPI) (see Bahmani‐Oskooee and Ardalani
2006). An increase in the economic activity of the United States is expected
to improve its imports; we expect an estimate of β1 to be positive. Similarly,
since an increase in economic activity in China is expected to boost the United
States’ exports, we expect an estimate of β2 to be negative. The RERt in (3)
denotes the real exchange rate and is defined in a way that an increase reflects
a real depreciation of the dollar against the RMB. If the dollar depreciation is
expected to discourage the United States’ imports and stimulate its exports, we
expect an estimate of β3 to be negative. Lastly, as mentioned above, RER
misalignment (MIStÞcould have positive or negative effects on exports and
imports. This involves that the estimate of β4 might be either positive or
negative. A popular argument in the US policymaking circle is that the US
trade deficit with China is attributed to the undervaluation of the RMB vis-
a-vis the US dollar (Moghaddam and Duan 2017; Moosa et al. 2020; Weber
and Shaikh 2021). If this argument is validated, then the expected sign of β4 is
positive. Indeed, a positive value of misalignment (i.e., undervaluation of the
RMB) gives a boost to China’s exporters (America’s imports) because the
dollars they earn convert into more RMB than they would otherwise and
worsens the deficit in the balance of trade.
6
For other studies that try to assess the magnitude of RMB’s misalignment, see the review article by Cheung and He
(2022).
INTERNATIONAL TRADE JOURNAL 7

As discussed in the Introduction, the relationship between RER misalign-
ment and the trade balance may be nonlinear and may involve threshold
effects. In the next section, we will explain the approach used to test the
existence of a threshold effect in the relationship between the trade balance
and the misalignment.
Regression kink with unknown threshold
To examine the nonlinear relationship between RER misalignment and the
trade balance, we adopt the threshold kink model with an unknown threshold
proposed by Hansen (2017). This model can be seen as an interesting advance-
ment of the threshold regression models (Bentour 2021). Distinguishing from
the conventional regression kink models, the most striking feature of Hansen’s
(2017) regression kink model is that the threshold is treated as an unknown to
be estimated. According to Hansen (2017), Equation (3) can be rewritten as
follows:
ln TBt
ð Þ ¼ β1MIStγð Þþβ2MIStγð Þþþβztþet(4)
where TBt, as mentioned above, refers to the bilateral trade balance between
the United States and China, MISt represents the misalignment of the RMB,
and zt denotes a k-vector of other explanatory regressors (i.e.,
ln IPUS
ð Þ;ln IPCH
ð Þ;and ln RERð ÞÞthat includes an intercept. γ is the threshold
parameter or the “kink point” at which a possible regime switching occurs. In
this study, we are distinguishing between two distinct regimes, one with a “high
degree of misalignment” and the other with a “small degree of misalignment.”
In line with Hansen (2017), we assume that the threshold level of misalignment
is unknown and, therefore, needs to be endogenously estimated by the model.
The variables (yt, xt, zt) are observed for t¼1; :::; n. The parameters are
β1;β2;β;and γ, with β¼β0;β3;β4;β5
 . As in Hansen (2017), we use
MIStγð Þ¼min MIStγ;0½  and MIStγð Þþ¼max MIStγ;0½  to
denote, respectively, the “negative part” and “positive part” of MIStγ.
A notable feature of this type of kink modeling approach outlined by
Equation 4 is that even though the regression function is continuous in all
variables, the slope of the misalignment variable carries a kink or discontinuity
at MISt¼γ. That is, the slope of MISt equals β1 for values of MISt less than γ,
and equals β2 for values of MISt greater than γ.
The economic intuition of our threshold kink model is straightforward.
The large and increasing US-China trade imbalance since the 2000s has
been commonly attributed to China’s exchange rate policy (Cheung and
He 2022). A dominant explanation was that the Chinese currency was
artificially misaligned or undervalued to boost exports (Dooley, Folkerts-
Landau, and Garber 2004). This increase in China’s exports and trade
8S. FERJANI ET AL.

surplus has also attracted major academic interest in what policy initia-
tives are required to reduce the trade deficit (Zhang 2012). For instance,
Islam (2022) points out that the real RMB-dollar exchange rate plays
a strong role in the US trade deficit with China. He states that the RMB
revaluation could be an effective policy to attenuate the persistent and
growing trade imbalance between the two largest economies. In this
article, we try to determine the level of revaluation needed to address
such a trade imbalance. In other words, is there such a thing as
a threshold level or the kink point of misalignment above which the effect
of RER misalignment on the trade balance is reversed? For these reasons,
a kink regression model is more justified than a linear regression one.
We apply Hansen’s (2017) method to estimate the unknown threshold (the
kink point). The methods in Hansen (2017) are briefly described as follows.
For a given γ, let ^
β γð Þ be the least-squares coefficients from a regression of
ln TBt
ð Þ on the variables xtγð Þ ¼ 1MIStγð ÞMIStγð Þþzt
 0. The
corresponding concentrated sum-of-squared errors function is:
S
nγð Þ ¼ Sn^
β γð Þ;γ
 ¼1
nXn
t¼1yt^
β γð Þ0xtγð Þ
 2(5)
An estimator of γ can be obtained by ^
γ¼arg min
γ2ΓS
nγð Þ, where Γ is the
parameter space for the threshold parameter. Once γ is identified by a grid
search over γ2Γ, the parameters ^
β can be determined by conventional least
squares of ln TBt
ð Þ on the variables xtγð Þ:The asymptotic distribution of b
β and
^
γ has been derived in Hansen (2017), and a bootstrap method is suggested to
obtain their confidence intervals. Hansen (2017) also suggests an F-test on
whether there exists an unknown threshold effect. The null hypothesis of the
kink effect and the alternative hypothesis of no kink effect (linear model) are as
follows:
H0:β1¼β2;H1:β1�β2(6)
The test statistic is of the F-type:
Tn¼n~
σ2^
σ2
 
^
σ2(7)
where ~
σ2 and ^
σ2 are the error variances of linear regression and kink regres-
sion, respectively. As mentioned in Hansen (2017), the distribution of the
F-test statistic might not be standard, and that is why he proposed a bootstrap
method to stimulate asymptotically first-order corrected p-values.
INTERNATIONAL TRADE JOURNAL 9

IV. Empirical ndings
In this section, we first present and discuss the aggregate estimation results.
Then, we proceed to the estimation results that allow for sectoral
heterogeneity.
Aggregate analysis
As a preliminary step of our aggregate analysis, we first estimate Model (4)
using our measure of misalignment, Mis, derived from Equation (2) without
distinguishing RMB over- and undervaluation. Aggregate bilateral trade data
between the United States and China with monthly frequency over the
January 2008 to September 2021 period are used to carry out the estimation.
The definitions and sources of all variables used in our analysis are given in the
online Appendix. The estimated results of Model (4) are shown in Table 2.
From the F-test results, it can be clearly deduced that the null hypothesis of
the no-kink threshold effect can be rejected at the 10% significance level,
suggesting the presence of RER misalignment–kink effects on the trade bal-
ance. The point estimate of the kink value of misalignment is 5.4% with an
associated 90% confidence interval [–7; 8]. This finding is further reinforced
by the evidence provided in Figures 2 and 3. Indeed, from Figure 2 which
depicts a scatter plot of (ln TBt
ð Þ;MIStÞalong with the fitted kink regression
line and pointwise 90% confidence intervals, it can be seen that the kink point
or threshold (red point) occurs effectively at around 5.4%. It can be concluded
from Figure 3, that plots the F-statistic against the threshold parameter γ;that
this kink value is statistically significant since the bootstrap critical value (in
red) is larger than the asymptotic critical value (in blue).
7
7
For more details, please see Hansen (2017).
Table 2. Results from estimation of the kink regression model with unknown threshold γ of
misalignment value.
Dependent variable: Trade balance
Threshold level γð Þ: 5.4 [-7; 8]
F-Test: 3.93**[0.04]
Coefficient Std Error Lower bound Upper bound
Misalignment below threshold –0.009 0.0066 –0.008 .027
Misalignment above threshold 0.018** 0.009 0.01 .02
US Industrial Production Index 1.055*** 0.24 0.64 1.464
Chinese Industrial Production Index –0.57 0.54 –1.45 .308
Real Exchange Rate –2.01*** 0.55 –2.92 –1.103
Intercept 8.72** 4.5 1.42 16.033
Note: ***, **, and * indicate the significance level of 1%, 5%, and 10%, respectively. p-value of the F-test for
a threshold is in square brackets. Lower bound and upper bound are for 95% confidence intervals.
10 S. FERJANI ET AL.

Figure 2. Scatter plot of trade balance and RMB misalignment with the estimated regression kink
model, with 90 confidence intervals.
Figure 3. Confidence interval construction for threshold in the case of RER misalignment and the
trade balance.
INTERNATIONAL TRADE JOURNAL 11

Once it is established that the relationship between RER misalignment and
the trade balance is subject to threshold kink effects, the next relevant question
would be whether the magnitude of misalignment affects differently the trade
balance in the two detected regimes of high and low degree of misalignment.
The regression kink results displayed in Table 2 show that in the first regime,
where the misalignment rate is less than the threshold level γ¼5:4ð Þ, the
estimated coefficient of MIStβ1¼  0:009
 is negative but not significant.
However, in the second regime, where the misalignment rate is greater than
5.4, the estimated coefficient of MIStβ2¼0:018
 is positive and significant at
the 5% level. This result implies that, when the Chinese currency is under-
valued, any additional increase in the RER misalignment above the threshold
will lead to a rise in the deficit of the trade balance between China and the
United States. The changes in directions of the RER misalignment impact over
the two regimes described above put forward strong evidence for the existence
of the threshold kink effect proposed at the beginning of the study. Indeed, the
results suggest that there is a threshold level above which RER misalignment,
and more specifically RER undervaluation, begins to deepen the US-China
trade deficit. This finding is consistent with those of Chiu, Lee, and Sun (2010),
Wang, Lin, and Yang (2012), Cheung, Chinn, and Qian (2016), Hurley and
Papanikolaou (2021), and Islam (2022), which suggest that the US trade deficit
with China can be reduced through a revaluation of the RMB. What is new,
however, in the current study compared to all prior studies on this subject is
that we show that such a revaluation needs to reach or exceed a certain
threshold level to trigger its effect on the bilateral trade balance between the
United States and China. Therefore, one possible explanation for the growing
US trade deficit with China, despite the RMB revaluation since 2005, could be
that this revaluation may not be enough to attenuate this deficit, and so further
revaluation would reduce it.
8
Regarding the control variables, we gather that all explanatory variables
have the expected sign. The coefficient of the real exchange rate (RER) is
statistically significant at the 1% level and has the expected negative sign.
Indeed, a 1% increase in RER (real depreciation of the dollar) leads to
approximately a 2.01% decrease in the trade balance. The results also indicate
that the estimated coefficient of the United States’ economic activity (IP
us
) is
statistically significant at the 1% level and has the expected positive sign. A 1%
increase in IP
us
leads to approximately a 1.055% increase in the trade balance.
8
There might be, of course, other additional reasons for the persistent and growing United States trade deficit with
China. These include, according to Yue and Zhang (2013, 80), “relocation of exports to China from elsewhere in
Asia, measurement differences, overcounting Chinese exports to the United States while undercounting US exports
to China, American consumption without saving, and US restriction of high-tech exports to China.” Other factors,
such as the low saving and investment rates in the United States (Feldstein 2017; Stiglitz 2018), the use of the US
dollar as an international money and reserve currency (Tu and Zhang 2019), and the shift in labor-intensive
factories from other Asian countries to China (Lin and Wang 2018; Reinbold and Wen 2018), have also been
advanced as possible causes for China’s huge bilateral trade surplus with the United States.
12 S. FERJANI ET AL.

However, the estimated coefficient of China’s economic activity (IP
ch
) is not
statistically significant.
As already mentioned in the Introduction, the impact of the RER misalign-
ment on the trade balance may depend on the sign of misalignment (i.e.,
overvaluation or undervaluation). For this reason, it is interesting to decom-
pose our measure of RER misalignment into undervaluation and overvalua-
tion, and we try to highlight its effects on the trade balance. The results are
given in Table 3.
According to the F-test results, since the p-values are less or equal to
10%, a significant threshold kink effect is present in the undervaluation-
trade balance and overvaluation-trade balance nexuses. Hence, the non-
linear link between RER misalignment and the trade balance still exists in
the two models for under and overvaluation. In the case of overvaluation,
the results show a threshold level at –2.83%, while in the case of under-
valuation, the threshold value appears to be at 4.04%. The same conclu-
sion can be drawn from Figures 4 to 7, which confirm the existence of
a threshold effect in the relationship between under/overvaluation and the
trade balance.
Once the threshold levels are detected, we turn now to discuss the
effect of under/overvaluation on the trade balance in the two identified
regimes. In the case of overvaluation, the results displayed in Table 3
indicate that in the first regime, i.e., when the overvaluation rate is less
than –;2.83% (in absolute value, greater than 2.83), the estimated coeffi-
cient of overvaluation β1¼  0:046
 is negative and significant at the 1%
level which is consistent with Abbas et al. (2020), Bonga-Bonga (2021),
and Mamun, Akça, and Bal (2021). In the second regime, i.e., when the
Table 3. Results from estimation of the kink regression model with unknown threshold γ of
undervaluation and overvaluation values.
Dependent variable: Trade balance
Model with overvaluation Model with undervaluation
Threshold level γð Þ: –2.83 [–8; –;1] Threshold level γð Þ: 4.04 [2.3; 6.7]
F-Test: 2.95*[0.07] F-Test: 3.86**[0.05]
Coefficient Std Error Coefficient Std Error
Misalignment below threshold –0.046*** 0.002 0.037 0.024
Misalignment above threshold 0.06 0.07 0.01** 0.005
US Industrial Production Index 0.92*** 0.3 0.45 0.81
Chinese Industrial Production Index –0.22 0.71 –2.27*** 0.75
Real Exchange Rate –1.84** 0.91 –2.376*** 0.76
Intercept 6.79 6.36 21.2*** 5.61
Note: ***, **, and * indicate the significance level of 1%, 5%, and 10%, respectively. p-value of the F-test for
a threshold is in square brackets.
INTERNATIONAL TRADE JOURNAL 13

Figure 4. Scatter plot of trade balance and RMB undervaluation with the estimated regression kink
model, with 90 confidence intervals.
Figure 5. Scatter plot of trade balance and RMB overvaluation with the estimated regression kink
model, with 90 confidence intervals.
14 S. FERJANI ET AL.

Figure 6. Confidence interval construction for threshold in the case of undervaluation and the
trade balance.
Figure 7. Confidence interval construction for threshold in the case of overvaluation and the trade
balance.
INTERNATIONAL TRADE JOURNAL 15

undervaluation rate is greater than –2.83% (in absolute value, less than
2.83), the estimated coefficient of overvaluation β2¼0:06
 is positive and
insignificant. Thus, our results suggest that the RMB overvaluation would
improve the United States’ bilateral trade deficit with China only when its
magnitude exceeds, in absolute value, 2.83%. This implies that the RMB
revaluation could be an effective policy to attenuate the persistent and
growing trade imbalance between the US and China.
In the case of undervaluation, when the undervaluation rate is less than
4.04%, the estimated coefficient of, undervaluation β1¼0:037
 is positive and
insignificant. Hence, in this regime, there is no significant link between RER
undervaluation and the trade balance. In the second regime, when the under-
valuation rate is greater than 4.04%, the estimated coefficient of
undervaluation β2¼0:01
  is positive and significant at the 5% level.
Therefore, RER undervaluation also needs to reach or exceed a certain thresh-
old level to trigger its effect on the trade balance.
In summary, the change in signs and significance of the RER under/over-
valuation effect over the two regimes described above confirm the existence of
the threshold effect. This finding reveals that misalignment effects change over
different degrees of misalignment. Therefore, a linear regression tends to
obscure the underlying relationship.
Table 3 also proves that the estimated coefficients for the real exchange rate
are statistically significant and negative in both models. The income coeffi-
cients are statistically significant and have the expected signs.
Disaggregate analysis
The industries’ trade balance may respond differently to RMB under/over-
valuation. Indeed, there might be industries that are more sensitive to mis-
alignment than others. In this section, we adopt the kink test of Hansen (2017)
for testing the existence of the nonlinear effect between RER under/over-
valuation and the trade balance at the industry level. The results of the kink
effect test are shown in Table 4. The F-test results reveal that the nonlinear
relationship between RMB overvaluation (dollar undervaluation) and the
trade balance holds in 50 out of 61 industries. From the F-test results, we
also gather that the kink effect of the RMB undervaluation (dollar overvalua-
tion) on the trade balance holds in 48 industries.
9
Then, we turn to estimate the effect of under/overvaluation on the industry’s
trade balance. The results of the kink regression model are reported in Tables A4
and A5 in the online Appendix. From Table A4, the results reveal that Hansen’s
(2017) hypothesis holds in 38 out of 61 industries. The 38 industries that were
9
In Figures A1 and A2 in the online Appendix, we have also plotted, for each industry, the graphs of confidence
interval construction for threshold in the case of under/overvaluation.
16 S. FERJANI ET AL.

Table 4. Threshold effect test results.
The case of overvaluation The case of undervaluation
Industries F-test
Threshold
(γ) Interval F-test
Threshold
(γ) Interval
00LIVE ANIMALS 2.78* –4.00 [–8; –;1] 3.62* 7.10 [2; 8]
01MEAT AND MEAT PREPARATIONS 2.29 –1.10 [–2.4; –;1] 2.89* 4.00 [2; 8]
02DAIRY PRODUCTS AND BIRDS’ EGGS 4.56** –3.50 [–8; –;1] 3.77* 8.00 [2.1; 8]
03FISH (EXCEPT MARINE MAMMAL) 3.01* –6.80 [–8; –;1] 3.42* 6.70 [2.5; 8]
04CEREALS AND CEREAL PREPARATION 4.10** –2.00 [–8; –;1] 2.87* 8.00 [2; 8]
05VEGETABLES AND FRUIT 5.05** –4.60 [–5.8; –;1.8] 2.71* 2.30 [2; 8]
06SUGARS, SUGAR PREPARATIONS 2.59 –3.20 [–8; –;1] 2.63 6.10 [2; 8]
07COFFEE, TEA, COCOA 3.59* –2.60 [–5.3; –;1] 3.08* 7.50 [6; 8]
08FEEDING STUFF FOR ANIMALS 3.25* –1.00 [–1.5; –;1] 3.30* 2.40 [2; 8]
09MISCELLANEOUS EDIBLE 2.73* –1.00 [–8; –;1] 5.16** 4.20 [2; 6]
11BEVERAGES 3.83** –7.20 [–8; –;1] 2.90* 2.90 [2; 8]
22OIL SEEDS AND OLEAGINOUS 4.95** –4.50 [–6.8; –;1.2] 3.27* 5.70 [3.8; 6.3]
23CRUDE RUBBER 3.88** –6.50 [–8; –;1] 2.46 7.70 [6; 8]
24CORK AND WOOD 2.89* –2.5e + 00 [–8; –;1] 2.57 5.40 [2; 8]
25PULP AND WASTE PAPER 3.42* –2.20 [–8; –;1] 3.83* 8.00 [2; 8]
26TEXTILE FIBERS 2.61 –3.20 [–8; –;1e +
0]
2.72* 4.40 [2; 8]
27CRUDE FERTILIZERS 3.19* –6.80 [–8; –;1] 2.42 3.30 [2; 8]
28METALLIFEROUS ORES 3.33* –5.60 [–8; –;1] 3.40* 6.80 [5.5; 8]
29CRUDE ANIMAL AND VEGETABLE
MATERIALS
3.40* –2.60 [–8; –;1] 3.47* 6.60 [2; 8]
32COAL, COKE AND BRIQUETTES 2.88* –1.00 [–1.70; –;1] 3.23* 3.90 [2; 8]
33PETROLEUM, PETROLEUM PRODUCTS 3.53* –6.00 [–8; –;1] 3.16* 6.20 [2.9; 8]
41ANIMAL OILS AND FATS 1.94 –2.40 [–8; –;1] 3.54* 6.10 [4.1; 8]
42FIXED VEG. FATS & OILS 2.98* –1.00 [–8; –;1] 3.37* 3.10 [2; 5.7]
43ANML/VEG FATS/OILS PROCESS/
WASTE
2.66 –1.00 [–8; –;1] 3.50* 2.00 [2; 3.4]
51ORGANIC CHEMICALS 2.87* –1.00 [–8; –;1] 4.92** 8.00 [2; 8]
52INORGANIC CHEMICALS 3.59* –7.60 [–8; –;1] 2.62 5.50 [2; 8]
53DYEING, TANNING AND COLORING
MATERIALS
3.26* –4.10 [–8; –;1.0e
+ 00]
3.02* 4.60 [2; 8]
54MEDICINAL AND PHARMACEUTICAL
PRODUCTS
3.47* –5.20 [–8; –;1] 3.07* 3.90 [2; 8]
55ESSENTIAL OILS 2.78* –1.90 [–8; –;1] 2.79* 2.00 [2; 8]
56FERTILIZERS 3.14* –1.70 [–8; –;1] 2.25 6.3e + 00 [2; 8]
57PLASTICS IN PRIMARY FORM 4.66** –3.30 [–8; –;1] 2.90* 2.60 [2; 5.8]
58PLASTICS IN NONPRIMARY FORM 3.62* –5.20 [–8; –;1] 2.54 2.10 [2; 3.3]
59CHEMICAL MATERIALS 3.40* –6.00 [–8; –;1] 2.33 8.00 [2; 8]
61LEATHER, LEATHER MFR 2.84* –1.00 [–8; –;1] 2.86* 7.10 [2; 8]
62RUBBER MANUFACTURES 2.85* –1.00 [–8; –;1] 3.17* 4.00 [2; 8]
63 CORK AND WOOD MANUFACTURES 4.08** –7.60 [–8; –;1] 2.45 8.00 [5.9e + 00; 8]
64PAPER, PAPERBOARD 3.19* –2.90 [–8; –;1] 2.87* 2.10 [2; 8]
65TEXTILE YARN, FABRICS 2.57 –1.10 [–8; –;1] 2.03 7.30 [6.1; 8]
66NONMETALLIC MINERAL 2.34 –2.90 [–8; –;1] 2.73* 3.20 [2; 8]
67IRON AND STEEL 3.85** –3.50 [–8; –;1] 3.10* 4.10 [2; 7.7]
68NONFERROUS METALS 2.41 –1.10 [–2.9; –;1] 2.82* 5.90 [2; 8]
69MANUFACTURES OF METALS 5.26** –7.60 [–8; –;1] 4.00** 2.10 [2; 8]
71POWER GENERATING MACHINERY 2.64 –1.00 [–8; –;1] 3.74* 6.50 [5.5; 8]
72MACHINERY SPECIALIZED 3.62* –3.20 [–8; –;1] 2.468 2.10 [2; 4.4]
73METALWORKING MACHINERY 3.18* –1.70 [–8; –;1] 2.243 7.10 [3.6; 8]
74GENERAL INDUSTRIAL MACHRY 3.11* –1.70 [–8; –;1] 3.11* 6.80 [2; 8]
75OFFICE MACHINES AND ADP
EQUIPMENT
3.79* –2.70 [–8; –;1] 2.87* 2.10 [2; 8]
76TELECOMMUNICATIONS EQUIPMENT 2.87* –1.00 [–8; –;1] 2.32 2.00 [2; 8]
77ELECTRICAL MACHRY, APPARATUS &
APPLIANCES
3.23* –2.60 [–8; –;1.3] 3.37* 2.00 [2.0e + 00; 8]
78MOTOR VEHICLES 3.94** –4.10 [–8; –;2.9] 3.46* 4.30 [2; 5.7]
79TRANSPOR*T EQUIPMENT 3.61* –6.00 [–8; –;1] 3.15* 6.80 [2; 8]
(Continued)
INTERNATIONAL TRADE JOURNAL 17

found to have a statistically significant turning point of their RER overvaluation-
trade balance nexus are 00, 01, 02, 03, 04, 05, 06, 07, 08, 09, 11, 23, 27, 28, 33, 52,
54, 57, 62, 63, 67, 68, 69, 71, 72, 73, 77, 78, 81, 82, 83, 84, 85, 87, 88, 89, 93, and 97.
Table A4 shows also that in the low RER overvaluation regime, i.e., below the
threshold level, the relationship between RER overvaluation and the trade
balance is significant in 46 industries coded 00, 01, 02, 03, 04, 05, 06, 07, 08,
09, 11, 22, 23, 25, 27, 28, 33, 42, 43, 51, 52, 54, 55, 57, 61, 62, 63, 64, 66, 68, 69, 71,
73, 75, 76, 77, 78, 81, 82, 83, 84, 85, 88, 89, 93, and 97. However, in a high RER
overvaluation regime, i.e., higher than the threshold level, the link between RER
overvaluation and the trade balance becomes significant in only 17 industries
coded 06, 08, 22, 28, 52, 54, 63, 71, 77, 81, 83, 84, 85, 88, 89, 93, and 97. The
suggestion that the United States’ bilateral trade deficit with China can be
reduced through a revaluation of the RMB still holds for 39 industries. Indeed,
for these industries, the coefficient of overvaluation (in the lower regime or in
the upper regime) is negative and significant.
From Table A5, the findings indicate a statistically significant threshold
effect of RER undervaluation on the trade flows of 44 industries coded as 00,
01, 02, 03, 04, 06, 07, 09, 22, 23, 24, 28, 29, 33, 41, 42, 43, 51, 52, 53, 54, 55, 57,
59, 61, 62, 63, 65, 66, 67, 68, 69, 71, 73, 74, 77, 78, 79, 82, 83, 87, 88, 89, and 97.
Table A5 shows also that in a low RER undervaluation regime, the relationship
between RER undervaluation and the trade balance is significant in 41 indus-
tries coded 00, 01, 02, 03, 07, 08, 09, 11, 22, 23, 24, 26, 27, 28, 29, 32, 33, 41, 42,
52, 54, 55, 57, 61, 62, 63, 65, 66, 67, 71, 72, 73, 75, 77, 78, 83, 84, 87, 88, 89, and
97. As for a high RER undervaluation regime, the relationship between RER
undervaluation and the trade balance appears significant in 31 industries
coded 00, 01, 02, 03, 07, 08, 09, 11, 22, 23, 24, 26, 27, 32, 33, 41, 55, 58, 61,
62, 63, 65, 66, 67, 71, 73, 75, 77, 84, 88, and 97. Notably, our results suggest that
the RMB undervaluation has been worsening the trade balance deficit of 35
Table 4. (Continued).
The case of overvaluation The case of undervaluation
Industries F-test
Threshold
(γ) Interval F-test
Threshold
(γ) Interval
81PREFAB BUILDINGS; SANITARY,
PLUMBING, ETC.
2.38 –8.00 [–8; –;1.4] 3.39* 2.10 [2; 4.3]
82FURNITURE & BEDDING 3.56* –1.00 [–8; –;1] 3.12* 6.20 [2; 8]
83TRAVEL GOODS, HANDBAGS 3.35* –1.00 [–2.2; –;1] 3.01* 8.00 [2; 8]
84ARTICLES OF APPAREL AND CLOTHING 2.59 –6.00 [–8; –;1] 3.12* 2.00 [2; 8]
85FOOTWEAR 3.27* –6.00 [–8; –;1] 3.79* 2.00 [2; 8]
87PROFESSIONAL SCIENTIFIC
INSTRUMENTS
2.81* –1.50 [–8; –;1] 3.65* 2.10 [2; 4.1]
88PHOTO APPT, EQUIPMENT & OPTICAL
GOODS
3.13* –3.50 [–8; –;1] 2.78* 6.10 [2; 8]
89MISCELLANEOUS MANUFACTURED
ARTICLES
3.25* –1.50 [–2.1; –;1] 3.24* 2.00 [2; 2.9]
93SPECIAL TRANSACTIONS 3.12* –7.30 [–8; –;1] 3.14* 4.5 [2; 8]
97GOLD, NONMONETARY 3.09* –5.80 [–8; –;1] 3.61* 7.20 [2; 8]
Note: ***, **, and * indicate the significance level of 1%, 5%, and 10%, respectively.
18 S. FERJANI ET AL.

industries. Indeed, for these industries, the coefficient of undervaluation (in
the lower regime or in the upper regime) is positive and significant.
V. Conclusion
In light of the mixed results documented in the empirical literature, the ongoing
debate about the relationship between RER misalignment of the Chinese currency
RMB and the United States-China trade imbalance continues and still remains an
unresolved issue. There is, therefore, a need to shed new light on whether and how
trade flows between the two world’s largest economies respond to dollar-renminbi
exchange rate misalignment, especially in the context of the escalation of the trade
war between them which could have huge repercussions on their bilateral trade in
particular and on international trade in general (as recently demonstrated, for
instance, by Itakura (2020) or by Fusacchia (2020)). In this article, we have
attempted to contribute to this debate by allowing for possible nonlinearities in
the nexus between RER misalignment and trade flows, and by addressing the
question of whether there has been any evidence of threshold effects in such
a nexus. To that end, an innovative dynamic threshold kink model recently
developed by Hansen (2017) has been deployed. We have first performed an
aggregate analysis in which we have considered the bilateral trade flows data
between the two investigated countries over the period of January 2008 to
September 2021. We have then disaggregated their bilateral trade flows by indus-
try to control for a possible heterogeneity across industries.
Overall, our findings from the aggregate analysis revealed there are indeed
significant threshold effects of RER misalignment on the trade balance. More
importantly, when distinguishing between the impacts of undervaluation and
overvaluation, we retrieved evidence indicating that while a RMB overvaluation
(dollar undervaluation) of more than 2.83% will attenuate the United States’ trade
deficit with China, an undervaluation of 4.04% or greater will deepen it. Looking
at the industry-level results, the presence of significant threshold effects of real
dollar-renminbi exchange rate overvaluation on the trade balance was confirmed
in 38 out of 61 industries. In these industries, the overvaluation threshold level
ranged from –8% to –1%. As for real dollar-renminbi exchange rate under-
valuation, the hypothesis of threshold effects was supported in 44 industries.
The point estimate of the undervaluation threshold value ranged from 2% to 8%.
Therefore, our findings from both aggregate and disaggregate analysis provided
evidence backing the existence of the threshold kink effect in the impact of
currency under/overvaluation on the trade balance. Moreover, our results seem
to support the American viewpoint asserting that the United States’ trade imbal-
ance with China is mainly due to an undervaluation of the RMB. That is,
a substantial revaluation of the RMB against the dollar would be helpful in
improving the United States’ trade balance with China.
INTERNATIONAL TRADE JOURNAL 19

It is believed that our findings are useful for academic researchers as well as
for policymakers. From an academic standpoint, the aggregate and disaggregate
analysis performed in the present study brings into question the implicit pre-
sumption that has been commonly made in the related literature on this topic,
that the relation between RER misalignment and the trade balance is linear. In
this regard, our results may be useful in future research works, as they strongly
suggest that academic researchers should consider nonlinearity and possible
regime changes when investigating the RER misalignment-trade balance nexus.
From a policy viewpoint, our empirical findings do have some interesting
implications for trading industries as well as for policymakers. Indeed, they
provide valuable guidance to each trading industry under investigation as to
how their trade flows react to RER misalignment. Industries that are particularly
found to respond asymmetrically to dollar-renminbi rate misalignment should
not adopt the same strategies during phases of RMB undervaluation (dollar
overvaluation) and phases of RMB overvaluation (dollar undervaluation). As for
policymakers, if they are to rely on dollar devaluation as a toll to attenuate the
United States’ trade deficit with China, such a policy will be more effective only
when devaluation reaches a certain threshold value. Furthermore, the threshold
impact of RER misalignment on trade flows should be taken into consideration
in implementing monetary and industry-specific trade policies. More specifi-
cally, for industries whose trade balance is found to be hurt by dollar-renminbi
rate misalignment, any trade-adjustment program in the United States aimed at
improving the trade balance of these industries could be unsuccessful if the
dollar-renminbi exchange rate misalignment is above their threshold level. On
the other hand, for industries in which the dollar-renminbi rate misalignment
had a positive impact on their trade balance, a relatively more flexible dollar-
renminbi rate and less market intervention would be beneficial.
Acknowledgments
The authors would like to thank warmly the two anonymous referees for their helpful com-
ments and suggestions.
Disclosure statement
No potential conflict of interest was reported by the author(s).
References
Abbas, S., V. C. Nguyen, Z. Yanfu, and H. T. Nguyen. 2020. “The Impact of China Exchange
Rate Policy on Its Trading Partners: Evidence Based on the GVAR Model.” The Journal of
Asian Finance, Economics & Business 7 (8): 131–141. https://doi.org/10.13106/jafeb.2020.
vol7.no8.131 .
20 S. FERJANI ET AL.

Allegret, J. P., and A. Sallenave. 2014. “The Impact of Real Exchange Rates Adjustments on
Global Imbalances: A Multilateral Approach.” Economic Modelling 37:149–163. https://doi.
org/10.1016/j.econmod.2013.10.022 .
Arize, A. C., J. Malindretos, and E. U. Igwe. 2017. “Do Exchange Rate Changes Improve the
Trade Balance: An Asymmetric Nonlinear Cointegration Approach.” International Review
of Economics & Finance 49:313–326. https://doi.org/10.1016/j.iref.2017.02.007 .
Baak, S. 2012. “Measuring Misalignments in the Korean Exchange Rate.” Japan and the World
Economy 24 (4): 227–234. https://doi.org/10.1016/j.japwor.2012.09.001 .
Bahmani‐Oskooee, M., and Z. Ardalani. 2006. “Exchange Rate Sensitivity of U.S. Trade Flows:
Evidence from Industry Data.” Southern Economic Journal 72 (3): 542–559. https://doi.org/
10.1002/j.2325-8012.2006.tb00719.x .
Balassa, B. 1964. “The Purchasing-Power Parity Doctrine: A Reappraisal.” Journal of Political
Economy 72 (6): 584–596. https://doi.org/10.1086/258965 .
Baldwin, R. 1988. “Hysteresis in Import Prices: The Beachhead Effect.” The American Economic
Review 78 (4): 773–785.
Baldwin, R., and P. Krugman. 1989. “Persistent Trade Effects of Large Exchange Rate Shocks.”
The Quarterly Journal of Economics 104 (4): 635–654. https://doi.org/10.2307/2937860 .
Banerjee, K., and A. Goyal. 2021. “Behavioural Equilibrium Real Exchange Rates and
Misalignments: Evidence from Large Emerging Markets.” Economic Analysis and Policy
70:414–436. https://doi.org/10.1016/j.eap.2021.03.004 .
Belke, A., M. Göcke, and M. Günther. 2013. “Exchange Rate Bands of Inaction and Play-
Hysteresis in German Exports–Sectoral Evidence for Some OECD Destinations.”
Metroeconomica 64 (1): 152–179. https://doi.org/10.1111/meca.12000 .
Belke, A., and D. Kronen. 2019. “Exchange Rate Bands of Inaction and Hysteresis in EU
Exports to the Global Economy: The Role of Uncertainty.” Journal of Economic Studies
46 (2): 335–355. https://doi.org/10.1108/JES-01-2018-0022 .
Bénassy-Quéré, A., S. Béreau, and V. Mignon. 2009. “Robust Estimations of Equilibrium
Exchange Rates within the G20: A Panel BEER Approach.” Scottish Journal of Political
Economy 56 (5): 608–633. https://doi.org/10.1111/j.1467-9485.2009.00500.x .
Ben Doudou, M., R. Nouira, S. Saafi, and A. Belhadj. 2022. “Do Exchange Rate Changes Have
Threshold Effects on the Trade Balance? Evidence from Tunisia.” Economic Change and
Restructuring 55 (1): 511–537. https://doi.org/10.1007/s10644-020-09306-4 .
Bentour, E. M. 2021. “On the Public Debt and Growth Threshold: One Size Does Not
Necessarily Fit All.” Applied Economics 53 (11): 1280–1299. https://doi.org/10.1080/
00036846.2020.1828806 .
Béreau, S., A. L. Villavicencio, and V. Mignon. 2012. “Currency Misalignments and Growth:
A New Look Using Nonlinear Panel Data Methods.” Applied Economics 44 (27): 3503–3511.
https://doi.org/10.1080/00036846.2011.577022 .
Bonga-Bonga, L. 2021. “Exchange Rate Misalignments and Current Accounts in BRICS
Countries.” MPRA Paper No. 107973. University Library of Munich, Germany.
Carrera, J., B. Gnimassoun, V. Mignon, and R. Restout. 2021. “Currency Misalignments and
Exchange Rate Regimes in Latin American Countries: A Trade-Off Issue.” Annals of
Economics and Statistics 141 (141): 71–102. https://doi.org/10.15609/annaeconstat2009.
141.0071 .
Ca’zorzi, M., A. Cap, A. Mijakovic, and M. Rubaszek. 2020. “The Predictive Power of
Equilibrium Exchange Rate Models.” ECB Working Paper No. 2358. European Central
Bank (ECB), Frankfurt, Germany. https://doi.org/10.2866/927929 .
Ca’zorzi, M., and M. Rubaszek. 2023. “How Many Fundamentals Should We Include in the
Behavioral Equilibrium Exchange Rate Model?” Economic Modelling 118:106071. https://
doi.org/10.1016/j.econmod.2022.106071 .
INTERNATIONAL TRADE JOURNAL 21

Chang, B. H., S. K. O. Rajput, N. A. Bhutto, and Z. Abro. 2020. “Asymmetric Effect of Extreme
Changes in the Exchange Rate Volatility on the US Imports: Evidence from Multiple
Threshold Nonlinear Autoregressive Distributed Lag Model.” Studies in Economics and
Finance 37 (2): 293–309. https://doi.org/10.1108/SEF-03-2019-0122 .
Chen, J. 2012. “Real Exchange Rate and Economic Growth: Evidence from Chinese Provincial
Data (1992-2008).” PSE Working Paper No. 2012-05. Paris School of Economics, Paris,
France. halshs–00667467, HAL.
Cheung, Y. W., M. D. Chinn, and X. Qian. 2016. “China–US Trade Flow Behavior: The
Implications of Alternative Exchange Rate Measures and Trade Classifications.” Review of
World Economics 152 (1): 43–67. https://doi.org/10.1007/s10290-015-0232-y .
Cheung, Y. W., and S. He. 2022. “RMB Misalignment: What Does a Meta-Analysis Tell Us?”
Review of International Economics 30 (4): 1038–1086. https://doi.org/10.1111/roie.12593 .
Chiu, Y. B., C. C. Lee, and C. H. Sun. 2010. “The U.S. Trade Imbalance and Real Exchange Rate:
An Application of the Heterogeneous Panel Cointegration Method.” Economic Modelling
27 (3): 705–716. https://doi.org/10.1016/j.econmod.2010.01.011 .
Clark, M. P. B., and M. R. MacDonald. 1998. “Exchange Rates and Economic Fundamentals:
A Methodological Comparison of BEERs and FEERs.” IMF Working Paper WP/98/67.
International Monetary Fund, Washington, DC.
Comunale, M. 2017. “Dutch Disease, Real Effective Exchange Rate Misalignments and Their
Effect on GDP Growth in EU.” Journal of International Money and Finance 73 (Part B):
350–370. https://doi.org/10.1016/j.jimonfin.2017.02.012 .
Comunale, M. 2019. “Long-Run Determinants and Misalignments of the Real Effective
Exchange Rate in the EU.” Scottish Journal of Political Economy 66 (5): 649–672. https://
doi.org/10.1111/sjpe.12206 .
Comunale, M. 2022. “A Panel VAR Analysis of Macro-Financial Imbalances in the EU.”
Journal of International Money and Finance 121:102511. https://doi.org/10.1016/j.jimon
fin.2021.102511 .
Corden, W. M. 1994. Economic Policy, Exchange Rates and the International System. Oxford:
Oxford University Press.
Couharde, C., A. L. Delatte, C. Grekou, V. Mignon, and F. Morvillier. 2018. “EQCHANGE:
A World Database on Actual and Equilibrium Effective Exchange Rates.” International
Economics 156:206–230. https://doi.org/10.1016/j.inteco.2018.03.004 .
Couharde, C., and A. Sallenave. 2013. “How Do Currency misalignments’ Threshold Affect
Economic Growth?” Journal of Macroeconomics 36:106–120. https://doi.org/10.1016/j.jma
cro.2012.11.002 .
Cuestas, J. C., E. Mourelle, and P. J. Regis. 2020. “Real Exchange Rate Misalignments in CEECs:
Have They Hindered Growth?” Empirica 47 (4): 733–756. https://doi.org/10.1007/s10663-
019-09454-5 .
Diop, M., M. Goujon, and B. B. Niang. 2018. “Impact of Real Exchange Rate Misalignments on
the Performance of the Manufacturing Sector in Senegal.” Revue d’économie du developpe-
ment 26 (1): 107–138. https://doi.org/10.3917/edd.321.0107 .
Dixit, A. 1989. “Hysteresis, Import Penetration, and Exchange Rate Pass-Through.” The
Quarterly Journal of Economics 104 (2): 205–228. https://doi.org/10.2307/2937845 .
Dooley, M. P., D. Folkerts-Landau, and P. M. Garber. 2004. “The US Current Account Deficit and
Economic Development: Collateral for a Total Return Swap.” NBER Working Paper 10727.
National Bureau of Economic Research, Cambridge, MA. https://doi.org/10.3386/w10727 .
Feldstein, M. 2017. “Inconvenient Truths About the US Trade Deficit.” Project Syndicate, April
25. https://www.project-syndicate.org/commentary/america-trade-deficit-inconvenient-
truth-by-martin-feldstein-2017-04 .
22 S. FERJANI ET AL.

Ferjani, S., S. Saafi, R. Nouira, and C. Rault. 2022. “The Impacts of the Dollar-Renminbi
Exchange Rate Misalignment on the China-United States Commodity Trade: An
Asymmetric Analysis.” Journal of Quantitative Economics 20 (3): 507–554. https://doi.org/
10.1007/s40953-022-00301-2 .
Fidora, M., C. Giordano, and M. Schmitz. 2021. “Real Exchange Rate Misalignments in the Euro
Area.” Open Economies Review 32 (1): 71–107. https://doi.org/10.1007/s11079-020-09596-1 .
Fišera, B., and R. Horváth. 2022. “Are Exchange Rates Less Important for Trade in a More
Globalized World? Evidence for the New EU Members.” Economic Systems 46 (1): 100868.
https://doi.org/10.1016/j.ecosys.2021.100868 .
Fusacchia, I. 2020. “Evaluating the Impact of the US–China Trade War on Euro Area
Economies: A Tale of Global Value Chains.” Italian Economic Journal 6 (3): 441–468.
https://doi.org/10.1007/s40797-019-00109-9 .
Gan, C., B. Ward, S. T. Ting, and D. A. Cohen. 2013. “An Empirical Analysis of China’s
Equilibrium Exchange Rate: A Co-Integration Approach.” Journal of Asian Economics
29:33–44. https://doi.org/10.1016/j.asieco.2013.08.005 .
Giordano, C. 2021. “How Frequent a BEER? Assessing the Impact of Data Frequency on Real
Exchange Rate Misalignment Estimation.” Scottish Journal of Political Economy 68 (3):
365–404. https://doi.org/10.1111/sjpe.12274 .
Grekou, C. 2015. “Revisiting the Nexus Between Currency Misalignments and Growth in the
CFA Zone.” Economic Modelling 45:142–154. https://doi.org/10.1016/j.econmod.2014.11.015 .
Hansen, B. E. 2017. “Regression Kink with an Unknown Threshold.” Journal of Business &
Economic Statistics 35 (2): 228–240. https://doi.org/10.1080/07350015.2015.1073595 .
Hashmi, S. M., B. H. Chang, and M. Shahbaz. 2021. “Asymmetric Effect of Exchange Rate
Volatility on India’s Cross‐Border Trade: Evidence from Global Financial Crisis and
Multiple Threshold Nonlinear Autoregressive Distributed Lag Model.” Australian
Economic Papers 60 (1): 64–97. https://doi.org/10.1111/1467-8454.12194 .
Hurley, D. T., and N. Papanikolaou. 2021. “Autoregressive Distributed Lag (ARDL) Analysis of
U.S.-China Commodity Trade Dynamics.” The Quarterly Review of Economics & Finance
81:454–467. https://doi.org/10.1016/j.qref.2020.10.019 .
Islam, A. M. 2022. “Yuan-Dollar Real Exchange Rate and the US Real Trade Balance with
China: Long-Run Cointegration and Short-Run Dynamic Analysis.” The International
Trade Journal 36 (1): 43–66. https://doi.org/10.1080/08853908.2021.2008282 .
Itakura, K. 2020. “Evaluating the Impact of the US-China Trade War.” Asian Economic Policy
Review 15 (1): 77–93. https://doi.org/10.1111/aepr.12286 .
Johansen, S., and K. Juselius. 1990. “Maximum Likelihood Estimation and Inference on
Cointegration – with Applications to the Demand for Money.” Oxford Bulletin of
Economics and Statistics 52 (2): 169–210. https://doi.org/10.1111/j.1468-0084.1990.
mp52002003.x .
Kilian, L., and M. F. Taylor. 2003. “Why is It Difficult to Beat the Random Walk Forecast of
Exchange Rates?” Journal of International Economics 60 (1): 85–107. https://doi.org/10.
1016/S0022-1996(02)00060-0 .
Kim, S., and Y. Kim. 2016. “The RMB Debate: Empirical Analysis on the Effects of Exchange
Rate Shocks in China and Japan.” The World Economy 39 (10): 1539–1557. https://doi.org/
10.1111/twec.12417 .
Lane, P. R., and G. M. Milesi-Ferretti. 2002. “External Wealth, the Trade Balance, and the Real
Exchange Rate.” European Economic Review 46 (6): 1049–1071. https://doi.org/10.1016/
S0014-29210200160-5 .
Lane, P. R., and G. M. Milesi-Ferretti. 2004. “The Transfer Problem Revisited: Net Foreign
Assets and Real Exchange Rates.” The Review of Economics and Statistics 86 (4): 841–857.
https://doi.org/10.1162/0034653043125185 .
INTERNATIONAL TRADE JOURNAL 23

Lin, J. Y., and X. Wang. 2018. “Trump Economics and China-US Trade Imbalances.” Journal of
Policy Modeling 40 (3): 477–646. https://doi.org/10.1016/j.jpolmod.2018.03.009 .
Loermann, J. 2021. “The Impact of CHF/EUR Exchange Rate Uncertainty on Swiss Exports to
the Eurozone: Evidence from a Threshold VAR.” Empirical Economics 60 (3): 1363–1385.
https://doi.org/10.1007/s00181-019-01780-8 .
Lòpez-Villavicencio, A., J. Mazier, and J. Saadaoui. 2012. “Temporal Dimension and
Equilibrium Exchange Rate: A FEER/BEER Comparison.” Emerging Markets Review
13 (1): 58–77. https://doi.org/10.1016/j.ememar.2011.10.001 .
Mamun, A., E. E. Akça, and H. Bal. 2021. “The Impact of Currency Misalignment on Trade
Balance of Emerging Market Economies.” Organizations and Markets in Emerging
Economies 12 (2): 285–304. https://doi.org/10.15388/omee.2021.12.57 .
Miyamoto, W., T. L. Nguyen, and V. Sheremirov. 2019. “The Effects of Government Spending
on Real Exchange Rates: Evidence from Military Spending Panel Data.” Journal of
International Economics 116:144–157. https://doi.org/10.1016/j.jinteco.2018.11.009 .
Moghaddam, M., and J. Duan. 2017. “Currency Manipulations and Bilateral Trade Between
China and the USA: A Fourier Gravity Approach.” Foreign Trade Review 52 (3): 171–184.
https://doi.org/10.1177/0015732516681869 .
Moosa, N., V. Ramiah, H. Pham, and A. Watson. 2020. “The Origin of the US-China Trade
War.” Applied Economics 52 (35): 3842–3857. https://doi.org/10.1080/00036846.2020.
1722797 .
Nasir, M. A., and K. Jackson. 2019. “An Inquiry into Exchange Rate Misalignments as a Cause
of Major Global Trade Imbalances.” Journal of Economic Studies 46 (4): 902–924. https://doi.
org/10.1108/JES-03-2018-0102 .
Nouira, R., P. Plane, and K. Sekkat. 2011. “Exchange Rate Undervaluation and Manufactured
Exports: A Deliberate Strategy?” Journal of Comparative Economics 39 (4): 584–601. https://
doi.org/10.1016/j.jce.2011.08.002 .
Nouira, R., and K. Sekkat. 2012. “Desperately Seeking the Positive Impact of Undervaluation
on Growth.” Journal of Macroeconomics 34 (2): 537–552. https://doi.org/10.1016/j.jmacro.
2011.12.002 .
Phillips, S., L. Catao, L. Ricci, R. Bems, M. Das, J. Di Giovanni, D. F. Unsal, et al. 2013. “The
External Balance Assessment (EBA) Methodology.” IMF Working Paper WP/13/1.
International Monetary Fund, Washington, DC.
Phillips, P. C., and B. E. Hansen. 1990. “Statistical Inference in Instrumental Variables
Regression with I(1) Processes.” The Review of Economic Studies 57 (1): 99–125. https://
doi.org/10.2307/2297545 .
Qin, D., and X. He. 2011. “Is the Chinese Currency Substantially Misaligned to Warrant
Further Appreciation?” The World Economy 34 (8): 1288–1307. https://doi.org/10.1111/j.
1467-9701.2011.01376.x .
Razek, N. H., and B. McQuinn. 2021. “Saudi Arabia’s Currency Misalignment and
International Competitiveness, Accounting for Geopolitical Risks and the Super-Contango
Oil Market.” Resources Policy 72:102057. https://doi.org/10.1016/j.resourpol.2021.102057 .
Reinbold, B., and Y. Wen. 2018. “Understanding the Roots of the U.S. Trade Deficit.” The
Regional Economist 26 (3): 4–9. https://doi.org/10.20955/es.2018.15 .
Rivera-Batiz, F. L., and L. A. Rivera-Batiz. 1994. International Finance and Open Economy
Macroeconomics. 2
nd
ed. New York: Macmillan.
Samuelson, P. A. 1964. “Theoretical Notes on Trade Problems.” The Review of Economics and
Statistics 46 (2): 145–154. https://doi.org/10.2307/1928178 .
Schnatz, B. 2011. “Global Imbalances and the Pretence of Knowing Fundamental Equilibrium
Exchange Rates.” Pacific Economic Review 16 (5): 604–615. https://doi.org/10.1111/j.1468-
0106.2011.00567.x .
24 S. FERJANI ET AL.

Shi, C., and J. Li. 2017. “Does Dollar-Pegging Matter? A Closer Look at US Trade Deficits with
China and Germany.” The Journal of International Trade & Economic Development 26 (4):
451–472. https://doi.org/10.1080/09638199.2016.1264988 .
Stiglitz, J. E. 2018. “The US is at Risk of Losing a Trade War with China.” Project Syndicate 30:
24–35. https://www.project-syndicate.org/commentary/trump-loses-trade-war-with-china-
by-joseph-e--stiglitz-2018-07?barrier=accesspaylog .
Tipoy, C. K., M. C. Breitenbach, and M. F. Zerihun. 2017. “Exchange Rate Misalignment and
Economic Growth: Evidence from Nonlinear Panel Cointegration and Granger Causality
Tests.” Studies in Nonlinear Dynamics & Econometrics 22 (2): 1–16. https://doi.org/10.1515/
snde-2016-0117 .
Tu, X., and K. H. Zhang. 2019. “International Monetary Systems and US Trade Deficits with China.”
The Chinese Economy 52 (5): 377–386. https://doi.org/10.1080/10971475.2019.1617924 .
Verheyen, F. 2013. “Exchange Rate Nonlinearities in EMU Exports to the US.” Economic
Modelling 32:66–76. https://doi.org/10.1016/j.econmod.2013.01.039 .
Wan, J. 2020. “US-China Trade War: Speculative Saving Perspective.” The Chinese Economy
54 (2): 107–123. https://doi.org/10.1080/10971475.2020.1830230 .
Wang, Y. 2020. “Yuan’s Valuation Under Managed Floating Exchange Rate Regime.” The
Chinese Economy 53 (6): 455–464. https://doi.org/10.1080/10971475.2020.1792067 .
Wang, C. H., C. H. A. Lin, and C. H. Yang. 2012. “Short-Run and Long-Run Effects of Exchange
Rate Change on Trade Balance: Evidence from China and Its Trading Partners.” Japan and the
World Economy 24 (4): 266–273. https://doi.org/10.1016/j.japwor.2012.07.001 .
Weber, I., and A. Shaikh. 2021. “The U.S.–China Trade Imbalance and the Theory of Free
Trade: Debunking the Currency Manipulation Argument.” International Review of Applied
Economics 35 (3–4): 432–455. https://doi.org/10.1080/02692171.2020.1814221 .
Wong, H. T. 2019. “Real Exchange Rate Misalignment and Economy.” Journal of Economic
Studies 46 (1): 211–227. https://doi.org/10.1108/JES-07-2017-0181 .
Wu, P. C., S. Y. Liu, and S. C. Pan. 2013. “Nonlinear Bilateral Trade Balance-Fundamentals
Nexus: A Panel Smooth Transition Regression Approach.” International Review of
Economics & Finance 27:318–329. https://doi.org/10.1016/j.iref.2012.10.010 .
You, K., and N. Sarantis. 2011. “Structural Breaks and the Equilibrium Chinese Yuan/US
Dollar Real Exchange Rate: A FEER Approach.” Review of International Economics 19 (5):
791–808. https://doi.org/10.1111/j.1467-9396.2011.00986.x .
Yue, L. H., J. Qiang, and T. Y. Kai. 2016. “Determination of Renminbi Equilibrium Exchange
Rate, Misalignment, and Official Intervention.” Emerging Markets Finance and Trade 52 (2):
420–433. https://doi.org/10.1080/1540496X.2016.1110448 .
Yue, K., and K. H. Zhang. 2013. “How Much Does China’s Exchange Rate Affect the U.S. Trade
Deficit?” The Chinese Economy 46 (6): 80–93. https://doi.org/10.2753/CES1097-1475460605 .
Yu, M., and R. Zhang. 2019. “Understanding the Recent Sino-U.S. Trade Conflict.” China
Economic Journal 12 (2): 160–174. https://doi.org/10.1080/17538963.2019.1605678 .
Zhang, J. 2012. “Will RMB Appreciation Reduce Trade Deficit in the US?” Journal of the Asia
Pacific Economy 17 (1): 171–187. https://doi.org/10.1080/13547860.2012.640021 .
Zhang, Z., and L. Chen. 2014. “A New Assessment of the Chinese RMB Exchange Rate.” China
Economic Review 30:113–122. https://doi.org/10.1016/j.chieco.2014.06.001 .
Zhang, Z., and K. Sato. 2012. “Should Chinese Renminbi Be Blamed for Its Trade Surplus?
A Structural VAR Approach.” The World Economy 35 (5): 632–650. https://doi.org/10.1111/
j.1467-9701.2012.01438.x.
INTERNATIONAL TRADE JOURNAL 25