Exchange rates and trade balance in West African economy and monetary union countries: does the content of the traded goods matter?

Guidime, Camille Detondji; Diaw, Adama; Biao, Barthélémy Article Exchange rates and trade balance in West African economy and monetary union countries: does the content of the traded goods matter? Cogent Economics & Finance Provided in Cooperation with: Taylor & Francis Group Suggested Citation: Guidime, Camille Detondji; Diaw, Adama; Biao, Barthélémy (2024) : Exchange rates and trade balance in West African economy and monetary union countries: does the content of the traded goods matter?, Cogent Economics & Finance, ISSN 2332-2039, Taylor & Francis, Abingdon, Vol. 12, Iss. 1, pp. 1-14, https://doi.org/10.1080/23322039.2024.2413963 This Version is available at: https://hdl.handle.net/10419/321629 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Cogent Economics & Finance ISSN: 2332-2039 (Online) Journal homepage: www.tandfonline.com/journals/oaef20 Exchange rates and trade balance in West African economy and monetary union countries: does the content of the traded goods matter? Camille Detondji Guidime, Adama Diaw & Barthélémy Biao To cite this article: Camille Detondji Guidime, Adama Diaw & Barthélémy Biao (2024) Exchange rates and trade balance in West African economy and monetary union countries: does the content of the traded goods matter?, Cogent Economics & Finance, 12:1, 2413963, DOI: 10.1080/23322039.2024.2413963 To link to this article: https://doi.org/10.1080/23322039.2024.2413963 © 2024 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group View supplementary material Published online: 13 Oct 2024. Submit your article to this journal Article views: 556 View related articles View Crossmark data Citing articles: 1 View citing articles Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=oaef20 DEVELOPMENT ECONOMICS | RESEARCH ARTICLE Exchange rates and trade balance in West African economy and monetary union countries: does the content of the traded goods matter? Camille Detondji Guidime a , Adama Diaw b and Barth el emy Biao c a Departement Economie, Universit e de Parakou, Parakou, B enin; b D epartement Economie, Universit e Gaston Berger de Saint-Louis, Saint-Louis, S en egal; c D epartement Economie, Universit e Africaine de D eveloppement Coop eratif, Cotonou, B enin ABSTRACT This paper analyzes the role of the nature of traded goods in the effect of the real exchange rate on the trade balance in West African Economy and Monetary Union countries. Empirically, the estimation of the parameters of a distributed lag autoregressive model using the technique of dynamic common correlation estimators is carried out. The results show that, firstly, the low level of intra-industry trade comfort positive real exchange rate effect on trade balance. However, WAEMU countries export raw materials used by their foreign trading partners to manufacture products, which WAEMU countries import in turn; but the raw materials percentage used in the manufacture of those products is negligible to have a beneficial impact on trade balance. Secondly, foreign income levels are less favorable to the countries’balance of trade, as partners have a preference for increasingly sophisticated goods that WAEMU countries do not produce. In terms of economic policy implications, a currency devaluation would have an expected positive effect if trade policy would encourage the consumption of goods and services containing, in their manufacturing process, a significant quantity of their exported products. This article contributes to the existing economic literature by taking into account the characteristics of traded goods in the analysis of the influence of the real exchange rate on the trade balance. A low content of exported goods in the manufacturing process of imported goods from a less advanced country could limit the rebalancing of its trade balance following a change in the real exchange rate. IMPACT STATEMENT The importance of this research is to empirically show the significant role of the quantity of domestic goods contained in the manufacture of imported goods in studying the relationship between exchange rate changes and the trade balance. The implementation of a specific trade policy coupled with an exchange rate policy in favour of the transformation of primary products into semi-manufactured goods is important for the economy of WAEMU countries. ARTICLE HISTORY Received 11 May 2024 Revised 30 September 2024 Accepted 2 October 2024 KEYWORDS Real exchange rate; nature of goods exchanged; trade balance; ARDL model; WAEMU countries SUBJECTS Economics and development; economics; finance; African studies JEL F10; F31; C23; 055 1. Introduction Rebalancing the trade balance is an issue of scientific importance for researchers and decision-makers. Economic theory establishes the basis for external rebalancing of the balance of payments through devaluation or depreciation of the national currency. Depending on the evolution of the exchange rate, countries’trade balances improve or deteriorate (Dogru et al., 2019; Bahmani-Oskooee & Aftab, 2017; Bahmani-Oskooee, 1991; Magee, 1973). For small-scale economies, such as those of the WAEMU, the exchange rate plays an crucial role in macroeconomic equilibrium under the hypothesis of MarshallLerner-Robinson critical elasticities (Keho, 2021b; Hunegnaw & Kim, 2017; Guillaumont & Guillaumont CONTACT Camille Detondji Guidime [email protected] Departement Economie, Universit e de Parakou, Parakou, B enin Supplemental data for this article can be accessed online at https://doi.org/10.1080/23322039.2024.2413963. ß2024 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. The terms on which this article has been published allow the posting of the Accepted Manuscript in a repository by the author(s) or with their consent. COGENT ECONOMICS & FINANCE 2024, VOL. 12, NO. 1, 2413963 https://doi.org/10.1080/23322039.2024.2413963 Jeanneney, 1993). The result is a mechanism whereby the trade balance has a short-term deficit and a surplus in the long term with an adjustment coefficient: this phenomenon is referred to as the J-curve. However, empirical verification of the J-curve phenomenon remains relatively observed (Ng et al., 2008). Moreover, the mechanisms describing the effects of the exchange rate on trade balance are not free of uncertainty. Therefore, the exchange rate can have positive or negative effects on a country’s trade balance (Ayele, 2019; Dogru et al., 2019; Hunegnaw & Kim, 2017; Keho, 2021b). These contrasting results can be observed particularly in developing or less advanced countries (Bahmani-Oskooee, 1991). For WAEMU countries, the work of Keho (2021b) shows the positive effect of the exchange rate on the trade balance in the presence of heterogeneity in countries’gross domestic product per capita. However, the nature of the goods traded between countries, which could slow down or accelerate the expected effect of real exchange rate evolution on the trade balance, is not taken into account. In this article, we show that the nature of the goods traded between WAEMU countries and their trading partners could limit the positive effect of the real exchange rate on the balance of trade, given the relative trade dependence of countries with intermediate incomes. The aim of this article, therefore, is to analyse the role of the nature of the goods traded between WAEMU countries and their trading partners in the expected effect of a change in the real exchange rate on their trade balances. This article contributes to the theoretical and empirical debate on the relationship between the real exchange rate and the trade balance of developing countries, especially those of the WAEMU. WAEMU countries are known for their exports of raw materials and their imports of manufactured products, made from raw materials that do not come sufficiently from them. Taking into account the shares of exported raw materials that are contained in the manufacturing of imported manufactured goods improves the understanding of the effect of the real exchange rate on the trade balance of developing countries. Furthermore, a low share of exported goods in the manufacturing process of goods imported from a less advanced country could limit or reduce the expected rebalancing of its trade balance following a variation in the real exchange rate. The rest of the paper is structured as follows: stylized facts on trade and exchange rate evolution in WAEMU countries are presented in Section (2); a literature review outlines the effects of the real exchange rate on the trade balance in Section (3); empirical analysis of exchange rate effects on the trade balance is described in Section (4); results of econometric estimations are discussed in Section (5); economic policy implications presented as conclusion in section (6). 2. Stylized facts about intra-industry trade and the real exchange rate in the WAEMU countries 2.1. Evolution of trade Trade between WAEMU countries has increased over the last ten years, rising from 6750.3 billion in 2006 to 17753.0 billion in 2020 (BCEAO, 2021,2006). Overall, imports remain higher than exports of goods and services (Cf. Figure 1). Moreover, the index of main exported products rose from 0.2% in 2019 to 1.8% in 2020 (BCEAO, 2021). Exports of goods remain dominated by gold (31.8%) and cocoa (16.8%), followed by petroleum products (6.8%), cotton (4.8%) and cashew nuts (3.9%). Imports fell from 8097.6 billion in 2006 to 18664.3 billion in 2020. The structure of the Union’s goods imports is dominated mainly by consumer goods (33.7%), capital goods (22.4%), intermediate goods (17.9%) and energy products (15.1%) (BCEAO, 2021). In short, exports are still dominated by primary products, and imports by manufactured goods. The Union’s trading partners are mainly the United States, Asian countries, Europe and, to a lesser extent, ECOWAS and African countries. 2.2. Trade balance and real exchange rate WAEMU countries are experiencing a structural imbalance in their trade balance, with the exception of C^ ote d‘Ivoire (BCEAO, 2021). The real exchange rate is relatively correlated with the trade balance of UEMOA countries (Cf. Figure 1). C^ ote d‘Ivoire did not experience a relative decline in its trade balance over the study period. 2 C. D. GUIDIME ET AL. In fact, the trade balance deficit of WAEMU countries is worsening over time, reaching -6.4% of GDP in 2021, compared with −5.1% in 2020 and −5.7% in 2019 (BCEAO, 2022). This deficit in the union’s trade balance has persisted since 1970. These trends can be explained in particular by the international economic conjuncture, which influences the economies of WAEMU countries with regard to the high dependence of their imports on the outside world, and the evolution of raw material prices and export revenues. It’s depended also by the evolution of real exchange rate. The real exchange rates of the WAEMU countries have undergone a three-phase evolution since 1994 (the year their common currency was devalued). From 1994 to 2002, the evolution is increasing with a plus at the peak of 6.6. After 2002, the real exchange rate experienced a decline until 2010 then started to appreciate again until 2020 even if the level remains lower than that of 2002 (Cf. Figure 2). The real exchange rate seems to decrease with the type of goods traded between trading partners. WAEMU countries import manufactured goods whose inputs do not come essentially from their exports of raw materials. The sensitivity of the trade balance following the movement of the real exchange rate would be low for countries with a low level of intra-industry trade (Kharroubi, 2011) because no national industry can easily and quickly replace imports which have become more expensive because of devaluation. Intuitively and following this idea, countries specialized in the production of raw materials cannot have high intra-industry trade. 3. Literature review This session presents theoretical and empirical views of the relationship between the exchange rate and the trade balance. It shows how the nature of traded goods determines the effect of exchange rate movements on the balance of trade by taking into account their manufacturing inputs. Figure 2. Real exchange rates in WAEMU countries. Source: Authors’calculations. Figure 1. Statistical relationship between the real exchange rate and the trade balance of WAEMU countries from 1994 to 2020. Source: Authors’calculations. COGENT ECONOMICS & FINANCE 3 3.1. Critical elasticity approaches and the J-curve phenomenon A reasoning that goes back to classical economists’states that a change in the real exchange rate improves the competitiveness-price of the national economy in the following ways: (i) higher import prices in domestic currency –leading to lower imports –and (ii) lower export prices in foreign currency – leading to increased exports. For the balance of trade to improve, the effect of increasing exports and decreasing imports must be sufficiently strong for the resulting balance to offset the effect of deterioration in the terms of trade (Magee, 1973). The Marshall-Lerner-Robinson critical elasticity hypothesis is therefore introduced, which states that, as long as the sum of export and import elasticities is greater than unity, a devaluation of the national currency will improve the balance of trade in the long term (BahmaniOskooee, 1991). Furthermore, the process of the theoretical positive effect of devaluation on the balance of trade is nuanced through the J-curve; an empirical observation which shows that the balance of trade initially becomes more in deficit, then becomes progressively in surplus again in a second phase. Himarios (1985), criticizing Miles’article (1979), shows that the real exchange rate is still well suited to understanding the effect of a devaluation of the national currency on the balance of trade. He again specifies the trade balance equation and emphasizes the temporal nature of the shift in the real exchange rate needed to influence the trade balance. However, it should be noted that a significant number of empirical analyses have been carried out, without any absolute agreement on the direct and positive effect of a change in the exchange rate on the balance of trade. If it has been shown that for small countries two successive devaluations are necessary to hope for a positive effect on the balance of trade (Prakash and Maiti, 2016), it remains to be shown that the quantity of goods imported by small countries, containing in their manufacture a given level of products exported upstream by the same small countries, is a key factor in the expected effect of the exchange rate on the trade balance: this is what our paper is concerned with. To the best of our knowledge, there is no research on the role of the nature of trade between countries in the effect of the exchange rate on the trade balance for WAEMU countries. 3.2. Key factors in the impact of exchange rates on the trade balance In developed countries, exogenous economic shocks can influence the expected effect of a change in the real exchange rate on the trade balance (Bahmani-Oskooee & Zhang, 2013; Hunegnaw & Kim, 2017; Kang & Dagli, 2018). In emerging countries, the favorable effect of the real exchange rate on the trade balance is observed in the long term, while in the short term, the trade balance appears to deteriorate (Bahmani-Oskooee et al., 2019; Serdar & Hakan, 2017; Prakash & Maiti, 2016; Baharumshah, 2001). Thus, it emerges that the time it takes for the effects of an exchange rate change to be realized determines the difference between the short-term and long-term results of the expected effects (Acar, 2000). On the other hand, depending on the country’s economic status –least developed, developed and developing country –the long-term outcome is mixed (Ayele, 2019). In less advanced and developing countries, the effects found relate to the consideration of the temporal factor, short-term and long-term, (BahmaniOskooee et al., 2019), the transmission mechanism of the exchange rate effect on the trade balance (Prakash & Maiti, 2016) and the consideration of real exchange rate horizons. Furthermore, Keho (2021a) performs a heterogeneous panel analysis and finds that the trade balance is negatively influenced by domestic and foreign products over the period 1975–2017, while it is positively influenced by the real exchange rate in the long term in UEMOA countries. In the short term, there is no deterioration in the trade balance. From a methodological point of view, it would be useful to take into account the heterogeneity of the countries in a panel in order to assess differently the effects of the real exchange rate on the trade balance. Moreover, an appreciation of the real effective exchange rate worsens C^ ote d‘Ivoire’s trade balance, while a depreciation improves it over the period 1975–2017 (Keho, 2021b). However, the difference in econometric tools used in the work could explain the difference in results. Nevertheless, country specificities and economic dynamics, particularly the degree of intra-industry trade with trading partners, could influence the expected effect of the real exchange rate on the balance of trade. In addition, WAEMU countries are relatively heterogeneous in the diversification of the products they export (Keho, 2021a). Furthermore, the difference in effects may stem from the nature of traded products 4 C. D. GUIDIME ET AL. (Burc¸ak & Payaslıo glu, 2016; Dogru et al., 2019; Kharroubi, 2011), industrial integration between trading countries and country dependence in the nature of exported products. Two cases are identified: the effect of the real exchange rate on the trade balance is amplified when, the part of the goods and services imported by a country from its trading partner contain, to a significant extent in their manufacture, the goods and services exported by the latter to the trading partner is high. The second case is the opposite phenomenon. In addition, the development of trade is linked to industrial development, which increases the volume of trade (sometimes in products of the same type). This increases substitutability between the types of imported and exported goods, making the trade balance more sensitive to movements in the real exchange rate. Furthermore, countries import and export different types of goods according to their revealed comparative advantages. Thus, when the normalized sum of trade balance deviations of firms in an industry relative to total trade is high, this would mean that the country is trading different types of goods with its partners. Consequently, the sensitivity of the trade balance to real exchange rate movements is low for countries with a low level of intra-industry trade, as no domestic industry can easily and quickly replace imports that have become more expensive following a devaluation (Kharroubi, 2011). Intuitively, and following this idea, countries specializing in the production of raw materials cannot have high intra-industry trade. Intra-industry trade in products would be low, and therefore the effect of a change in the real exchange rate could not significantly affect the balance of trade in the sense of improving it. 4. Methodology 4.1. Model specification and variables description The econometric modeling derives from the work of Bahmani-Oskooee (1991), Hunegnaw and Kim (2017) and Keho (2021b) and leads to the construction of a delayed autoregressive distribution model by considering the expression of the trade balance as follows: ln TB ðÞ it ¼ait þbln RERit þvln Ydit þdln Yfit þeit (1) With TB,RER,Yd,Yfrepresenting respectively the trade balance, real exchange rate, and domestic production by value and foreign production by value. The equation is in log-linearizing. Furthermore, taking into account the role of content of the traded goods on the effects of the exchange rate on the trade balance, the model to be estimated is as follows: ln TB ðÞ it ¼aitþbln RERit þvln Ydit þdln Yfit þ/ln RERitIITit þeit (2) where is the trade balance of country "i" at date "t". The trade balance is calculated as the relation between countries’exports and imports to avoid negative sign for the log expression (Bahmani-Oskooee et al., 2019). RERit is the real exchange rate of country i against the currencies of its trading partners. The real effective exchange rate controls a country’s external competitiveness. The interaction between the real exchange rate and trade dependency allows us to understand how the types of exported versus imported goods affect the expected effect of a change in the real effective exchange rate on the trade balance. We use the Grubel and Lloyd (1975) index 1 as a proxy to capture the share of product exported by a WAEMU country, contained in the manufacture of products imported by this country. And this share valued by the real exchange rate is measured by the product between the real exchange rate and the intra-industry trade index. The expected sign of the interaction between the real exchange rate and the intra-industry trade index will be negative if the degree of intra-industry trade is low. The effects of a change in the real exchange rate are observed over time. Therefore, Pesaran et al. (2001) represent an ARDL-type equation with error term correction to understand long and short term dynamics in a context where variables have an order of integration between I(0) and I(1). On the other hand, several previous works use the ARDL-type cointegration analysis approach to the detriment of error-correction vector autoregression analyses on large samples. Bahmani-Oskooee and Aftab (2017) indicate that the ARDL approach makes it possible to simultaneously assess the short-term effects –through the coefficients of the exogenous variables in first difference, and the long-term effects through the coefficients –of the influence of the real exchange rate on the trade balance (Cf. Equation (4)). The equation is as follows: COGENT ECONOMICS & FINANCE 5 Dln TBit ¼aiþX n j¼1 x0iDln TBi,t−jþX m j¼0 x1iDln RERi,t−jþX m j¼0 x2iDln Ydi,t−jþX m j¼0 x3iDln Yfi,t−j þqECTi,t−1þgln TBi,t−1þbln RERi,tþvln Ydi,tþdln Yfi,tþeit (4) In order to consider the constraint that the degree of intra-industry trade could pose on the effect of changes in the real exchange rate, Equation (5) is established: Dln TBit ¼aiþXn j¼1x0iDln TBi,t−jþXm j¼0x1iDln RERi,t−jþXm j¼0x2iDln Ydi,t−jþXm j¼0x3iDln Yfi,t−j þXm j¼0x4iDln RERi,t−jIITi,t−jþqECTi,t−1þgln TBi,t−1þbln RERi,tþvln Ydi,tþdln Yfi,t þ/ln RERitIITit þeit (5) With n 1; m 0, maximum lag of the model. Dis the first difference operator. a,xare the short-run coefficients and g,b,v,d,/are the long-run coefficients of the model. ECT the speed of adjustment between short-term and long-term dynamics; a negative and statistically significant sign is expected. eis the error term. 4.2. Technical estimation The parameters of Equation (7) are estimated to understand the short and long term dynamics of the area as a whole, as well as of the countries making up the area. The Im, Pesaran et al. (2001) unit root test is applied to study the stationary properties of the model’s variables. The distributed autoregressive in delay (ARDL) model developed by Pesaran (2006) is used to study shortand longterm relationships in the event that the variables have an integration order I (0) and I (1). To assess a cointegrating relationship between variables, the test of Pedroni (2004,2000) is applied. The test considers the heterogeneity between the panel’s cross-sectional units (taking into account the presence of diversification in the countries’economies). Thus, under the alternative hypothesis, there is a cointegrating relationship for each panel unit. Cointegration vectors are rarely identical from one panel unit to the next (Hurlin & Mignon, 2007), which means that cross-sectional dependency between error terms must be taken into account. Taking into account the heterogeneity of coefficients estimated from panel data is becoming increasingly demanding (Chudik & Pesaran, 2015). There are often unobserved dependencies between panel units, and ignoring these leads to auto-correlation of the error term and biased least squares regression results (Kao & Liu, 2000). A cross-sectional panel dependence test is proposed by Pesaran (2015) and (Chudik & Pesaran, 2015). Ditzen (2018) shows that Mean Group estimators are unbiased and consistent. On the other hand, when the ARDL model parameters suggest a cross-sectional dependence of the panel with the presence of heterogeneity, it is of interest to use the cross-sectional Mean Group estimator (CS-ARDL Mean Group estimator) of Pesaran (2015) and Chudik and Pesaran (2015) programmed by Ditzen (2018) under the name xtdcce2. Formally, the following model estimators are proposed by: yi;t¼kiyi,t−1þbixi,tþui,t(6) hence ui,t¼c0 iftþei,t(7) With U i,t , the error term including the common unobservable factor, the heterogeneous factor present and the white noise. The heterogeneous coefficients are distributed around a common mean bi¼bþ vi, vi IID(0, Xv) and ki¼kþ1i, 1iIID(0, X1). To estimate the parameters, (Chudik et al., 2016) propose two methods for estimating the coefficients: the cross-sectionally augmented estimator ARDL (CS-ARDL) and the cross-sectionally augmented and distributed lag estimator. The CS-ARDL estimator directly estimates the long-run coefficients, by adding the explanatory variables in difference and their lags. The CS-ARDL first estimates the short-term coefficients, then calculates the long-term coefficients. Optimal lags are those with a high frequency over the whole 6 C. D. GUIDIME ET AL. panel. In other words, using the unrestricted model and an information criterion, decide on the choice of lags for each country by variable, then choose the most common lag for each variable to represent the lags for the model. Here, we obtain ARDL (1,0,0,0) after testing. The performant test program of Pedroni (2000) which introduced the panel ordinary least squares dynamics (PDOLS) estimation technique, is run for robustness of results. 4.3. Data To examine the effects of changes in the real exchange rate on the trade balance of WAEMU countries, time-series data from 1994 to 2020 are provided by World Development Indicators and the Central Bank of West African States (BCEAO). Only countries whose data are available at regular frequency are included, namely Benin, Burkina Faso, C^ ote d‘Ivoire, Mali, Niger, Senegal and Togo. The dependent variable studied is the trade balance, measured by the ratio between exports and imports (Hunegnaw & Kim, 2017; Keho, 2021b). An increase in this ratio reflects an improvement in the trade balance, and a deterioration in the opposite case. Domestic and foreign incomes are approximated by the gross domestic products of WAEMU countries and the ten (10) main trading partners on the basis of traded volumes. The real effective exchange rate is obtained by transforming the nominal exchange rate from the BCEAO. The construction of the real effective exchange rate through the intersession of intra-industry trade is based on data on exports and imports of goods and services from the primary, secondary and tertiary sectors between countries and their ten (10) main trading partners. WAEMU countries have been exporting raw materials since independence in the 1960s, and importing the end products resulting from their exports. They devalued their currencies in 1994, hoping for a favourable effect on their trade balance. This effect cannot be assessed without taking into account the degree of integration of their exports with imported products. It is therefore necessary to cross-reference the effective exchange rate with the degree of intra-industry trade with the main trading partners. Table 1 describes the observations by country and at panel level. The correlation coefficients between the real effective exchange rate and the trade balance are positive and statistically significant for Benin, Senegal and Togo, but statistically insignificant for C^ ote d‘Ivoire, Mali and the panel. For Burkina Faso and Niger, the correlation coefficients are negative but not significant. As for the correlation between the trade balance and the real effective exchange rate crossed with the degree of intra-industry trade, it is negative and statistically significant for Benin, Burkina-Faso and the panel, but not significant for Mali and Niger. The correlation is positive and statistically significant for Senegal, but not significant for Togo and C^ ote d‘Ivoire over the period 1994–2020 (Cf. Table A1,Supplementary Appendix 1). The heterogeneity text shows that the space studied is not homogeneous in terms of economic dynamics and their trade balances (Cf. Figure A1 in Supplementary Appendix). Which corroborates the result of k eho (2021b). 5. Results and discussion 5.1. Data The following Table 2 shows the unit root test results. Table 1. Mean values of variables by country and correlation coefficient. Countries lntb lnyd lnyf lnrer lnreriit ґÚ B enin −1.27 22.82 27.36 6.35 6.60 0.58−0.63 Burkina faso −0.79 22.75 26.80 6.33 6.34 −0.23 −0.94 C^ ote d’Ivoire 0,30 24.26 28.47 6.33 5.91 0.37 0.12 Mali −0.45 22.93 27.43 6.32 6.22 0.26 −0.33 Niger −0.53 22.60 27.80 6.33 6.13 −0.10 −0.30 S en egal −0.98 23.31 28.56 6.31 6.22 0.400.50 Togo −0.57 21.84 26.70 6.348 6.30 0.390.25 WAEMU −0.61 22.93 27.59 6.33 6.25 0.07 −0.59 Note. lntb, lnyd, lnyf, lnreriit are trade balance, domestic and foreign income, real exchange rate, real exchange rate associated with the level of intra-industry trade, respectively. Ґand Úare respectively correlation coefficients between trade balance and real exchange rate and between trade balance and real exchange rate associated with the level of intra-industry trade. Data are sample averages. Source: Authors’calculations. COGENT ECONOMICS & FINANCE 7 Kharroubi, E. (2011). The trade balance and the real exchange rate. BIS Quarterly Review, September, 33–42. Kremers, J. J. M., Ericsson, N. R., & Dolado, J. J. (1992). The power of cointegration tests. Oxford Bulletin of Economics and Statistics,54(3), 325–348. https://doi.org/10.1111/j.1468-0084.1992.tb00005.x Magee, S. P. (1973). Currency contracts, pass-through, and devaluation. Brookings Papers on Economic Activity, 1973(1), 303. https://doi.org/10.2307/2534091 Mark, N. C., & Sul, D. (2003). Cointegration vector estimation by panel DOLS and long-run money demand. Oxford Bulletin of Economics and Statistics,65(5), 655–680. https://doi.org/10.1111/j.1468-0084.2003.00066.x Ng, Y.-L., Har, W.-M., & Tan, G.-M. (2008). Real exchange rate and trade balance relationship: An empirical study on Malaysia. International Journal of Business and Management,3(8), 130–137. https://doi.org/10.5539/ijbm.v3n8p130 Pedroni, P. (2004). Panel cointegration: asymptotic and finite sample properties of pooled time series tests with an application to the ppp hypothesis. Econometric Theory,20(03), 597–625. https://doi.org/10.1017/ S0266466604203073 Pedroni, P. (2000). Fully modified OLS for heterogeneous cointegrated panels. Advances in Econometrics, 15, 93–130. https://doi.org/10.1016/S0731-9053(00)15004-2 Pesaran, M. H. (2015). Testing weak cross-sectional dependence in large panels. Econometric Reviews,34(6–10), 1089– 1117. https://doi.org/10.1080/07474938.2014.956623 Pesaran, M. H. (2006). Estimation and inference in large heterogeneous panels with a multifactor error structure. Econometrica,74(4), 967–1012. https://doi.org/10.1111/j.1468-0262.2006.00692.x Pesaran, M. H., Shin, Y., & Smith, R. J. (2001). Bounds testing approaches to the analysis of level relationships. Journal of Applied Econometrics,16(3), 289–326. https://doi.org/10.1002/jae.616 Prakash, K., & Maiti, D. (2016). Does devaluation improve trade balance in small island economies? The case of Fiji. Economic Modelling,55, 382–393. https://doi.org/10.1016/j.econmod.2016.02.023 Saikkonen, P. (1991). Asymptotically efficient estimation of cointegration regressions. Econometric Theory,7(1), 1–21. https://doi.org/10.1017/S0266466600004217 Serdar, O., & Hakan, P. (2017). Testing the validity of the J-curve hypothesis between Brazil and the USA. Atlantic Review of Economics, Colegio de Economistas de A Coru~ na, A Coru~ na,2(2). ISSN 2174–3835 Stock, J. H., & Watson, M. W. (1993). A simple estimator of cointegrating vectors in higher order integrated systems. Econometrica,61(4), 783. https://doi.org/10.2307/2951763 Xiao, J., Juodis, A., Karavias, Y., Sarafidis, V., & Ditzen, J. (2023). Improved tests for Granger non-causality in panel data. The Stata Journal: Promoting Communications on Statistics and Stata,23(1), 230–242. https://doi.org/10.1177/ 1536867X231162034 14 C. D. GUIDIME ET AL.