*Article* **Regional Stock Exchange Development and Economic Growth in the Countries of the West African Economic and Monetary Union (WAEMU)**

#### **Babatounde Ifred Paterne Zonon**

School of Economics and Management, Southwest Jiaotong University, Jiuli Campus, Chengdu 610031, China; zifrpat@yahoo.fr

**Abstract:** This study used panel data covering 27 years to investigate the causality between regional stock exchange development and economic growth in the West African Economic and Monetary Union (WAEMU) countries. We performed a homogeneous Granger non-causality with an autoregressive distributed lag model (ARDL) and Markov-switching analysis, using six indicators for the stock and financial market and six for control. The results showed a close economic relationship between WAEMU countries and causality from the regional stock exchange, which supports the supply leading hypothesis. The causality was confirmed in the short and long run, depending on the variable. The causal relationships that support the demand-driven hypothesis were recorded from the economic growth for four market measurements.

**Keywords:** BRVM; WAEMU; regional stock exchange; economic growth; developing countries

**JEL Classification:** E44; G10; O43

#### **1. Introduction**

Financial market conditions are usually linked to a country's political and economic conditions. Developed countries have an established market or trade in well-developed markets, whereas emerging financial markets belong to developing countries. Further, pre-emerging markets, or frontier markets, are those belonging to less developed countries. Almost every developed country has many financial services provided by a multitude of institutions operating in a diversified and complex network, which together make up the financial system.

However, in developing countries, the financial system offers shorter ranges or lowerquality services. Today, in developing countries, development of the financial system is considered an excellent way to support economic growth and includes the development of financial institutions, financial markets, etc. Rao and Cooray (2011) found that for countries with lower incomes, but not for those with higher ones, there is a close relationship between future economic growth and stock exchange activity.

Most low-income countries are still at an early stage of development. With a low Gross Domestic Product (GDP) per capita, and a small population, these frontier market countries (Shum 2015) usually do not even have a domestic financial market. However, they can depend on or be influenced by stock exchange activities. Among these countries, some have established communities, such as the West African Economic and Monetary Union (WAEMU). Categorised as a frontier market (Morgan Stanley Capital International (MSCI) 2021), the WAEMU possesses a financial market known as the Regional Securities Exchange or the Bourse Régionale des Valeurs Mobilières (BRVM). As the saying goes, "United, we are stronger." WAEMU countries are neither financially stable nor economically strong enough to have different exchanges, and so by pooling resources they can safely navigate internationally.

**Citation:** Zonon, Babatounde Ifred Paterne. 2021. Regional Stock Exchange Development and Economic Growth in the Countries of the West African Economic and Monetary Union (WAEMU). *Economies* 9: 181. https://doi.org/ 10.3390/economies9040181

Academic Editor: Robert Czudaj

Received: 10 October 2021 Accepted: 8 November 2021 Published: 17 November 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

The BRVM, the common exchange of the eight states (Benin, Burkina-Faso, Cote d'Ivoire, Guinea-Bissau, Mali, Niger, Senegal, Togo) of the WAEMU, is one of the very few exchanges worldwide that is regionally integrated and serves as the central exchange for a group of regionally related countries, as noted by Ganti (2019). Despite many challenges, it should be a pioneer for the further integration of the stock exchanges in Africa (James 2018). The eight member states of WAEMU are all developing countries, which are still considered low-income countries; hence, the impact of the regional stock exchange on their economy becomes important.

Another factor that should not be overlooked is what WAEMU countries have in common: they are all post-colonial states that are still under monetary control. They use a currency called the CFA franc (Franc of the African Financial Community). Assane and Malamud (2010) pointed out that the constraints of monetary unions affect the financial depth of the CFA countries and hurt economic growth. Koddenbrock and Sylla (2019) later found that the CFA caused extreme external repression of financial and monetary policies because of its dependence on the euro and the US dollar. The reason is that France prints the CFA for the CFA countries and is, therefore, able to control the financial and monetary regulations, the money supply, the credit allocation, the banking activities, and the economic and budgetary policies of these nations.

Moreover, as suggested by Tadesse (2018), this control breeds corruption and illegal diversion of public aid between France and its former colonies. For instance, conditional French public aid has forced these African states to spend the money destined for aid on French equipment, goods or contracts with French firms, especially construction and public work firms. Internally, the design of the CFA franc strengthens the constraints that are imposed on all bank–firm relations and central bank policies in the Global South and makes it more difficult to pursue growth strategies for the benefit of the broader population. The CFA franc is a heavy economic burden on WAEMU countries. To support these observations, Sow et al. (2020) explained that various financial development policies have been unsuccessful from independence (1960) to the present. As a result, economic development is slow, not to mention the existence of lasting economic precarity. WAEMU countries are facing an unprecedented economic dilemma stemming from their common currency, whereas, in contrast, most of the low- and high-income countries studied have independent currencies.

Given the interest in financial markets, the specifics of developing markets, and WAEMU countries, this study attempts to assess the existence and nature of the relationship between BRVM development and economic growth in WAEMU countries. To this end, a panel Granger non-causality procedure is conducted for the eight WAEMU countries, followed with an Autoregressive Distributed Lag model (ARDL) and a Markovswitching analysis.

The contribution of this paper to the empirical literature on stock exchange development and economic growth covers various aspects. First, it examines the WAEMU regional stock exchange, which has not been the subject of many studies up to now. Second, awareness is created concerning whether WAEMU countries should use the BRVM as an economy booster given their current level of development. Third, given the analysis process, cross-sectional dependence is considered for the causality approach, in order to avoid inconsistency and bias in the empirical results. To the best of our knowledge, no such attempt to include the cross-sectional hypothesis in the literature on stock exchange development and economic growth in the WAEMU countries has been made up to date. Fourth, financial development is known to be of a multidimensional nature. Thus, to capture the numerous aspects of the market in the economic development process, we utilize six indicators. Finally, the dataset contains eight countries for a period of 27 years, to abide by the well-known rule of thumb in economics which stipulates that a large dataset is necessary for a relevant panel data analysis.

The rest of this paper is divided into four sections: the literature review is outlined in Section 1. Section 2 presents the data, describes the variables and the empirical strategy. Section 3 presents the analysis and results. The conclusion is given in the last part.

#### **2. Review of Literature**

Long-run securities, such as bonds and shares, are among the services provided by capital markets. Several authors, such as Ewah et al. (2009) and Oke and Adeusi (2012), stated in their contributions to the literature that the capital market (comprising bond and equity markets) is the market in which medium- to long-term financing can be obtained. The purpose is to develop companies that, as a result, promote a nation's economic growth. Therefore, the existence of the capital market seems to be an economic booster. However, such an expected positive effect should differentiate between countries to avoid scientific bias, as economic realities differ from country to country.

Seven and Yetkiner (2016), who examined 146 middle- and high-income countries from 1991 to 2011, found a significant positive correlation between economic growth and stock exchange development. However, almost a decade earlier, Ben Ben Naceur and Samir (2007) had contradicting views on developments in financial markets and economic growth in middle-income countries. By examining a sample of 11 countries in the MENA (the Middle East and North Africa) region, they reported that the development of financial systems could harm economic growth. Later, Stephen and Enisse (2012) pointed out that financial booms rarely encourage growth and suggested re-examining the relationship between real growth and finance. Yu et al. (2012) were more specific, arguing that the positive relationship between growth and finance found in many studies is a long-run relationship. Regarding underdeveloped countries, they mentioned that these countries may face slower economic growth in the short run, although the stock exchange is developing, mainly because of political instability and poorly enforced legal systems. It can be observed here that the positive effect of stock exchange developments on economic growth, especially in the middle- and low-income countries, is shadowed by the negative impact of political instability and poor enforcement. Thus, depending on the size and direction of the changes, the positive effect may prevail, or the countries may face a slowdown in their economy in the short or long run.

With underdeveloped or low-income countries, however, some authors, such as Haque (2013), who examined the countries of the South Asian Association for Regional Corporation (SAARC), could show that stock exchange developments had no significant influence on economic growth. Rioja and Valev (2014) supported these findings. Using a large cross-country panel, they observed a significant positive impact on stock exchanges and capital accumulation in banks for high-income countries; however, banks have not contributed to productivity growth or capital accumulation in low-income countries. Here, the stock exchange development seems to have a neutral effect on economic growth, and thus we can conclude that capital markets are not among the best channels for economic development in developing countries.

This idea is supported by the findings of Ewah et al. (2009). Using multiple regression and ordinary least squares (OLS) on 44 years of data, they conducted a study on Nigeria and showed that the Nigerian capital market, although endowed with the ability to induce growth, has not contributed significantly to Nigeria's economic growth because of low market capitalisation, small market size, low transaction volume, illiquidity, and few listed companies, etc. Francis and Ofori (2015), working with data on 101 countries from 1980 to 2009, showed that political instability in some underdeveloped countries can hamper the development of stock exchanges and justify their ineffectiveness regarding economic growth because the stock exchange development is positively influenced by policy scores. Karim and Chaudhary (2017) also pointed out that while stock exchange developments contribute to South Asia's economic growth to some extent, their effect is negligible. A year later, Pan and Mishra (2018) concluded that there is no significant impact on economic growth in economies where the stock exchange plays a minimal role (large economies like China).

The preceding lines reveal two major pieces of information. Firstly, the positive or negative impact of the stock exchange on economic growth is to be observed in the long or short run, depending on the income level and political characteristics of the country. Secondly, in most cases, the development of the stock exchange plays an economically neutral role and can even be a hindrance.

Instead, some authors have looked only for a connection and have concentrated on analysing the causal relationship between economic growth and stock exchange developments. In Nigeria, Adamu and Sanni (2005) used regression analysis and the Granger causality test to assess the role of the stock exchange in economic growth in Nigeria. A one-way causality between market capitalisation and GDP and a two-way causality between market turnover and GDP growth was discovered. A significantly positive relationship between turnover ratios and GDP growth was also observed. This suggests that market turnover and turnover ratios are essential stock exchange proxies that impact economic growth.

In Nigeria too, Kolapo and Adaramola (2012) carried out a Granger causality and Johansen cointegration test. They discovered a long-run positive influence of the capital market on economic growth. Demirguc-Kunt et al. (2012) also showed that the development of banks and stock exchanges shows a parallel relationship with the economic growth of countries. Therefore, even if no direct causality is shown, it is evident that the development of stock exchanges and economic growth advance as a pair. Ikikii and Nzomoi (2013) confirmed this by examining the effects of the development of the stock exchange on economic growth in Kenya. Using quarterly time-series data, they found that stock exchange developments had a positive impact on economic growth.

Mittal (2014) also found strong indications of an existing causality in his search for a causal relationship between economic growth and the stock exchange in the newly industrialised countries (NIC). He suggested that the governments of these countries should better direct monetary and fiscal policies towards promoting the growth of the financial sector. Later, Milka (2021) carried out a similar analysis in Serbia. With the Toda-Yamamoto-Dolado-Lütkepohl approach for the Granger causality test, the vector autoregression model, the forecast error variance decomposition, and the impulse response function, he demonstrated the existence of a unidirectional Granger causality from stock exchange development to economic growth.

Similarly, Maku (2020) used an autoregressive distributed lag model (ARDL) bound test and examined the relationship between the development of stock exchanges and economic growth in Nigeria. The empirical results confirmed the existence of a long-run relationship between stock exchange development and growth. Ezeibekwe (2021) used a vector error correction model and concluded that using the ratio of market capitalisation to GDP as a proxy for stock exchange development does not contribute significantly to economic growth in Nigeria in the long run. This finding implies that the Nigerian economy is still at a stage of development where the stock exchange cannot play a crucial role in economic development.

It is evident that there are still contradictory views regarding the relationship between stock exchange development and economic growth, even though the causal effect of the stock exchange on the economy, especially in underdeveloped countries, has been confirmed by most studies. This hypothesis of the causal effect of stock exchange development on the real economy belongs to supply leading theory.

This theory suggests that the accumulation of financial assets improves economic growth; therefore, the development of the financial markets will lead to positive economic growth (McKinnon 1973; Shaw 1973). As observed in some studies, the results sometimes suggest an economy to stock exchange causality or a reverse relationship between the stock exchange and economic growth. This gives rise to another hypothesis, expressed as the demand-driven hypothesis, by Friedman and Schwartz (1963). They suggested that economic growth brings with it the emergence and establishment of financial centres. Simply put, the growth of the real economy endogenously determines financial development. These two theories are part of others drawn from decades of research into the finance– growth nexus. Fink et al. (2006) pointed out that the nexus between the real economy and the financial market can be summarised in five forms: supply leading, demand-driven, no causal relationship, interdependence, and negative causality from finance to growth.

One reason for the ambiguity in the existing literature about the financial growth nexus issue could be the proxies used for the stock exchange and the real economy. Using one or two proxies is too restrictive to capture different aspects of the market. As often used, gross domestic product (GDP) also measures an annual global or individual level of production instead of the actual economic growth trend. Another reason may be the geographical limitations and economic diversity of countries. Most studies relate to specific countries and rarely to groups of countries with a financial market. Even if different countries are included in panels, they are often heterogeneous. Most of the countries examined so far have also been economically independent with a local currency, unlike WAEMU countries, which still use a common currency, often referred to as colonial money.

Given the peculiarities of the WAEMU context, this paper seeks to determine whether WAEMU countries will benefit from having a financial market. We also test demand-driven and supply-driven hypotheses

#### **3. Research Data, Materials, and Methods**

#### *3.1. Data Source and Type*

The first steps of the BRVM can be traced back to 14 November 1973 when the treaty establishing the WAEMU was signed. Members included Burkina Faso, Benin, Côte d'Ivoire, Niger, Mali, Togo, and Senegal. In 1997, Guinea-Bissau joined, and to date there are eight member countries. On 17 December 1993, the WAEMU Council of Ministers set up a regional financial market and commissioned the West African Central Bank (WACB), known as Banque Centrale des États de l'Afrique de l'Ouest (BCEAO) in French, to manage the project.

Introducing the market regime in 1997 gave investors more confidence and helped stimulate business. With e-commerce and the transition to daily basic trading (1999–2001), a steady increase has been observed since 2000. As Proshare (2010) reported, many years of reforms caused significant changes in 2011 and 2013, and saw a sharp decline in operations, with the market recovering and booming in 2012. Since then, the BRVM has experienced a stable evolution.

Pan and Mishra (2018) found that most time-series research used a short sampling period because of data limitations. This problem is more pronounced in developing countries where obtaining data is quite a challenge. This study also deals with this problem, even though the focus is on panel data. While updated time-series data can be obtained for most of the companies listed on the BRVM, countries updated market data remain a challenge. The data used in this study were obtained for each WAEMU country from the World Bank (Global Financial Development Database, World Development Indicators Database, and World Governance Indicators Database) and the International Monetary Fund (Financial Development Index Database). The first data collected covered from 1971 to 2020. Data were then filtered variably. For some variables, data were not available as early as 1989, and the last update for some variables was before 2016. Therefore, after filtering, the final data set contained observations from 1989 to 2015.

#### *3.2. Stock Exchange and Other Variables*

The model used was, like that of Mohtadi and Agarwal (2007), a modified and improved version of the well-known model by Levine and Zervos (1998). The first specificity of this improved model is to overcome the measurement and consistency problem associated with the use of two different data sources by Levine and Zervos (1998). Although the present study uses different data sources, they are filtered and combined into one single

final dataset for analysis. Moreover, the data sources are interconnected, which guarantees the consistency of our data. Second, this model uses several measures of stock exchange development to maximize the use of information extracted from the data, as opposed to a single composite measure. It enables the economic growth in year t to be estimated as a function of the stock exchange development in year t-1.

This study also used several proxies of the stock exchange development to optimally employ the information obtained from the data instead of a single composite measure. However, some financial indices were added individually to represent the development of both the financial market and of the stock market. The purpose was to capture different aspects of the market and create more space for relevant results to be achieved. The PSE (primary school enrolment) variable was also added to the model, along with CC (corruption control) and OP (oil prices). WAEMU countries face many challenges in primary and secondary education. In a 2015 education report, the Africa-America Institute (AAI) mentions that many African countries cannot keep up with rising enrolments, the results of which are that the learning outcomes have been negatively impacted (Africa-America Institute (AAI) 2015). Therefore, governments need to invest in educational innovation to improve the quality of education in schools. The variables were the following:


Three variables for financial development were taken from the financial development index database and used in this study, as follows:


Information for the remaining variables was collected from the World Development Indicators database:


as this exchange rate does not reflect the economic fundamentals of the countries it should serve (Des Adom 2012). Given also the increasing globalisation of financial markets, raw materials, and volatile oil prices, the CFA franc faces challenges, such as significant changes in export prices and a prolonged genuine appreciation of the currency (Gulde and Tsangarides 2008).

#### *3.3. Model Specification*

According to Mohtadi and Agarwal (2007), the original approach enables control for countries as a block and each country specifically. It is divided into two models: the first model tests whether the stock exchange affects economic growth, and the second examines their relationship. However, our approach was slightly different in that we only checked for the entire region. This is because the WAEMU states have a unique stock exchange that pools the countries' financial market investments.

Checking for country-specific observations would also have meant a small sample of observations over 27 years. As a result, a country-specific process in WAEMU countries could have resulted in biases. We only used one model and directly investigated the relationship between regional financial market development (with an emphasis on stock exchange development) and economic growth in WAEMU countries.

Here, the growth (as a dependent variable) at time t was a function of the other variables at time t − 1. Our analysis was based on the following panel data model:

$$\mathbf{y}\_{\rm it} = \beta\_{\rm j} + \sum\_{\rm j=1}^{k} \beta\_{\rm j} \mathbb{X}\_{\rm ji(t-1)} + \mathbf{c}\_{\rm i} + \delta\_{\rm t-1} + \varepsilon\_{\rm i(t-1)} \tag{1}$$

where i represents the unit of observation, t indicates the time, j is the observed explanatory variables, t is the time trend, and δ signifies the implicit assumption of a constant rate of change. y stands for the dependent variable and X for the independent variable (k is up to 12 in this study). c represents the unobserved effect, and ε is the error term. β are the individual intercepts or slope coefficients that can differ across the states.

We reinforced the model with a step-by-step analysis process that compensated for our dropping of model 1 by Mohtadi and Agarwal (2007) and made the results more robust. For the analysis process, the steps recommended by Menegaki (2019) to implement an ARDL were followed. This procedure was chosen because it is a step-by-step sequence that guarantees control of model misspecifications and more relevant results. The approach of Mohtadi and Agarwal (2007) has long since been proven efficient, but the economic case of the WAEMU requires more precision. With the additional variables in the model (both stock and financial market) and considering the improvement of econometric measurements in the scientific literature, traditional analysis procedures (regression, correlation tests, etc.) may not be enough. Therefore, it was applied in the ARDL procedure. The ARDL is a step-by-step procedure in which the next test to apply depends on the results from the previous test; thus, the final results are more precise. The first stage of the analysis was a cross-sectional dependency test, the second stage was a stationarity test, the third stage was a cointegration test (for non-stationarity), and the last step was a causality test. After these steps, an actual ARLD regression with structural break control was implemented to check the robustness. The control of structural breaks was important to our analysis because of many global events (such as the Afghanistan and Iraq wars, the Asian financial crisis, the attacks on the World Trade Centre, and the 2008 Global Financial Crisis), which occurred during our sample period. We completed our analysis process by performing a dynamic regression (Markov-switching regression).

#### **4. Data Analysis, Results, and Discussion**

The details of the data for each variable are shown in Table A1. Negative values can be observed for Growth, CC, INV, FDI, and INF in WAEMU countries. Figures A1 and A2 in the Appendix A Table A1 give a visual representation of the evolution of economic growth along with the stock and financial market variables in each of the WAEMU countries. A quick look at the figures does not seem to show that growth for any WAEMU country followed the same trend as other variables. However, higher stock market capitalisation (SMC) appears to be associated with higher economic growth.

We had an overall observation of 216 per variable with T = 27 and N = 8. Where T is small and N is large (the common situation when analysing panel data), a Pesaran test, a Friedman test, or a Frees test (Friedman 1937; Frees 1995, 2004; Pesaran 2004) are better adapted, but if T > N, the Lagrange Multiplier (LM) test by Breusch and Pagan (1980) is well suited to test the cross-sectional dependency (CD) (see Equation (2)).

$$\text{LMP} = \text{T} \sum\_{i=1}^{N-1} \sum\_{j=i+1}^{N} \text{\hat{P}\_i^2} \tag{2}$$

where Pˆ <sup>2</sup> <sup>i</sup> is the estimated correlation coefficient between the residuals derived from the panel model estimate. Under H0, there is an asymptotic chi-square distribution (chi2) concerning the LM statistic with a degree of freedom of N (N − 1)/2. i, j, and T are derived from the panel model equation, with t = 1, 2, ..., T.

As shown in Table A2, the results suggest the rejection of the null hypothesis; there was a correlation between the panels' attributes. This means that any shock in one of the WAEMU countries will be transmitted to others. In line with Pesaran and Yamagata (2008), we also checked the nature of our panels.

The slope homogeneity null hypothesis H0 is βi = β for each is compared to the heterogeneity hypothesis H1: βi = βj for pair-wise slopes non-zero fraction for i = j.

Pesaran and Yamagata (2008) developed standardised dispersion statistics that cover a broader spectrum of analysis. In contrast to Swamy's (1970) model, which is only limited to models where N is relatively smaller than T, it takes the Pesaran and Yamagata model into account and extends it to wider panels. The model is represented as:

$$
\tilde{\mathbf{A}} = \sqrt{\mathbf{N}} \left( \frac{\mathbf{N}^{-1} \tilde{\mathbf{s}} - \mathbf{k}}{\sqrt{2\mathbf{k}}} \right) \tag{3}
$$

with s being a modified version of Swamy's (1970) slope homogeneity test.

$$\mathbf{\tilde{s}} = \sum\_{i=1}^{N} \left( \ddot{\boldsymbol{\beta}}\_{i} - \hat{\boldsymbol{\beta}}\_{\text{WFE}} \right) \frac{\mathbf{x}\_{i}^{\prime} \mathbf{M}\_{\text{Y}} \mathbf{x}\_{i}}{\tilde{\sigma}\_{\text{i}}^{2}} \left( \ddot{\boldsymbol{\beta}}\_{i} - \boldsymbol{\beta}\_{\text{WFE}} \right) \tag{4}$$

where .. β<sup>i</sup> represents the pooled OLS estimator, βˆ WFE is the pooled estimator of the weighted fixed effect, M<sup>γ</sup> is an identity matrix, and <sup>σ</sup><sup>2</sup> <sup>i</sup> is the estimator of σ<sup>2</sup> i .

In addition, the Δ test's small sample properties can be improved with normally distributed errors by using the following variance and mean bias-adjusted version:

$$
\tilde{\Lambda}\_{\rm adj} = \sqrt{\mathbf{N}} \left( \frac{\mathbf{N}^{-1} \tilde{\mathbf{s}} - \mathbf{E} \left( \tilde{\mathbf{z}}\_{\rm it} \right)}{\sqrt{\mathbf{var} \left( \tilde{\mathbf{z}}\_{\rm it} \right)}} \right) \tag{5}
$$

with E- z . it <sup>=</sup> k, and var- z . it = 2k(T − k − 1)/(T + 1).

The analysis result (Table A2) did not reject the null hypothesis of homogeneous coefficients. Therefore, in the WAEMU area, there will be replication in other countries of every significant economic relationship or change in one country.

From the previous analysis, we could have gone straight to a bootstrap panel Granger causality according to Kónya (2006); however, this requires cross-sectional dependency and cross-border heterogeneity, which was not the case for us.

Following the analysis, four tests—Hadri Lagrange, Levin Lin Chu, Im-Pesaran-Shin, and Fisher—(Hadri 2000; Levin et al. 2002; Im et al. 2003) were carried out to check the stationarity; this provided information about the degree of integration for each variable.

As shown in Table A3, only Growth, SMTR, FD, OP, FDI, and INF were stationary. By computing the first difference, the stationarity for SMC, STTV, FMA, FD, CC, INV, and PSE was restored.

The non-stationarity of some variables suggests a long-run relationship between them that can affect growth. Therefore, a little digression into cointegration analysis is necessary. Since there are multiple cointegration tests and remembering that our panels were homogeneous (Table A2), the best option was that of Kao (1999), which allows five sets of tests specifically designed for homogeneous panels to be performed. Kao's cointegration test can be carried out together with those of Pedroni (1999) and Westerlund (2005). However, since the Westerlund test assumes heterogeneous coefficients, it did not fit into our context. Pedroni's test assumes both homogeneous and heterogeneous coefficients. Therefore, besides the Kao test, we also carried out the Pedroni cointegration test. Both tests' results confirmed the cointegration (Table A4).

For further analysis and to take robustness into account, we performed a dynamic ordinary least squares (DOLS) regression.

The fully modified ordinary least squares (FMOLS) regression was developed by Phillips and Hansen (1990), and later Stock and Watson (1993) developed the DOLS. These models help generate coefficients that are asymptotically efficient because they consider both serial autocorrelation and endogeneity. They are only used if the variables are stationary at the first difference. The ordinary least squares (OLS) regression is often skewed when the variables are non-stationary and cointegrated, while DOLS and FMOLS are not. However, FMOLS is designed and works well with heterogeneous data, while DOLS offers more flexibility. Because of our data, we only had a DOLS regression.

As can be seen in Table A5, there was a significant and positive long-run relationship between stock and financial market variables and growth, except for STTV, which hinders growth over the long run. The control variables (except for OP and FDI) also showed significant long-run relationships with growth. CC, INV, and PSE contribute to economic growth, while INF does not. These results reflect reality since the reduction of corruption reduces the loss of money and allows the perfect implementation of projects and policies, thus improving the economy. The more investment, the more employment and increased economic activity. As for education, it is the very basis for any nation; and it starts from primary school. Based on these first observations, one might be tempted to say that stock exchange developments are contributing to economic growth in WAEMU countries; however, such a conclusion would be far too easy.

We proceeded with a causality analysis by running a Granger non-causality test. The most widely used Granger causality tests are designed for heterogeneous panels. Juodis et al. (2021) proposed a novel approach that helps perform a Granger non-causality test in both heterogeneous and homogeneous panels that best fit our data. Table A6 shows the outcome of the Granger non-causality test.

A first test was performed with Growth as the dependent variable; then reverse causality was implemented to assess the effect of Growth on each of the dependent and control variables. There was a one-way causality between FMA and FMD to Growth and from Growth to INV. For the other independent variables and control variables, a bidirectional causality was observed in Growth.

According to the results so far, the financial market (the BRVM) is linked to economic growth and contributes to this in WAEMU countries. This conclusion is supported by the results of the DOLS regression, which showed significant long-run relationships between Growth and SMC, STTV, SMTR, FMA, FMD, and FD. The only negative relationship observed over the long run was with STTV, which showed significance at the 1% level (Table A5). The results also confirmed the supply-leading hypothesis for the stock and financial markets (all six variables cause growth); in other words, the development of the stock and financial market contributes to economic growth. The demand-driven hypothesis was confirmed for the stock exchange (the three stock exchange variables were caused by Growth versus one in three for the financial market variables); that is to say, the increase of the economy significantly participates to the improvement of the stock exchange, and somewhat to the development of the financial market.

Regarding the control variables, although INV had a long-run relationship with Growth, it had no causality. The remaining control variables (CC, OP, FDI, PSE, INF) Granger caused growth (Table A6); meaning that they have an impact on the economy. The positive impact of CC is not surprising, as better anti-corruption policies will lead to a healthier economy in the long run.

After the causality analysis, an ARDL regression was implemented. Since the number of the ARDL's regressive variables was limited, we split our original model into two equations for its application. We outline our ARDL representation in Equation (6) for stock and financial market variables and in Equation (7) for control variables.

$$\text{Groupth} = \beta\_0 + \Sigma \beta\_l \text{SMC}\_{t-1} + \Sigma \beta\_l \text{STTV}\_{t-j} + \Sigma \beta\_k \text{SMTR}\_{t-k} + \Sigma \beta\_l \text{FMA}\_{t-l} + \Sigma \beta\_{\text{m}} \text{FMD}\_{t-\text{m}} + \Sigma \beta\_{\text{n}} \text{FD}\_{t-\text{n}} + \varepsilon\_l \tag{6}$$

Growtht = β<sup>0</sup> + ΣβiCCt−<sup>i</sup> + ΣβjOPt−<sup>j</sup> + ΣβkINVt−<sup>k</sup> + ΣβlFDIt−<sup>l</sup> + ΣβmPSEt−<sup>m</sup> + Σ βnINFt−<sup>n</sup> + et (7)

where i, j, k, l, m, and n are the number of lags for the independent variables in the models. Akaike Information Criterion (AIC) and Schwartz Bayesian Criterion (SBC) were used to determine the optimal lag number and the ARDL (1,0,0,0,0,4) was chosen for both equations. We then formulated our unrestricted error correction models:

$$\begin{array}{rcl} \mathsf{A}\mathsf{G}\mathsf{c}\mathsf{w}\mathsf{w}\mathsf{h}\_{\mathsf{t}} = \beta\_{0} & + \Sigma\beta\_{\mathsf{t}}\mathsf{A}\mathsf{c}\mathsf{c}\mathsf{w}\mathsf{w}\mathsf{h}\_{\mathsf{t}-\mathsf{i}}\,\Sigma\beta\_{\mathsf{t}}\mathsf{A}\mathsf{S}\mathsf{M}\mathsf{C}\_{\mathsf{t}-\mathsf{j}} + \Sigma\beta\_{\mathsf{k}}\mathsf{A}\mathsf{S}\mathsf{T}\mathsf{T}\mathsf{V}\_{\mathsf{t}-\mathsf{k}} + \,\Sigma\beta\_{\mathsf{t}}\mathsf{A}\mathsf{S}\mathsf{M}\mathsf{T}\mathsf{R}\_{\mathsf{t}-\mathsf{i}} + \,\Sigma\beta\_{\mathsf{m}}\mathsf{A}\mathsf{P}\mathsf{M}\mathsf{A}\_{\mathsf{t}-\mathsf{m}} \\ & + \Sigma\beta\_{\mathsf{n}}\mathsf{A}\mathsf{F}\mathsf{M}\mathsf{D}\_{\mathsf{t}-\mathsf{n}} + \Sigma\beta\_{\mathsf{o}}\mathsf{A}\mathsf{F}\mathsf{D}\_{\mathsf{t}-\mathsf{o}} + \gamma\_{\mathsf{1}}\mathsf{S}\mathsf{M}\mathsf{C}\_{\mathsf{t}-\mathsf{i}} + \gamma\_{\mathsf{2}}\mathsf{S}\mathsf{T}\mathsf{T}\mathsf{V}\_{\mathsf{t}-\mathsf{2}} + \gamma\_{\mathsf{3}}\mathsf{S}\mathsf{M}\mathsf{T}\mathsf{R}\_{\mathsf{t}-\mathsf{i}} + \gamma\_{\mathsf{4}}\mathsf{F}\mathsf{M}\mathsf{A}\_{\mathsf{t}-\mathsf{i}} \\ & + \gamma\_{\mathsf{3}}\mathsf{F}\mathsf{M}\mathsf{D}\_{\mathsf{t}-\mathsf{k}} + \gamma\_{\mathsf{6}}\mathsf{F}\mathsf{D}\_{\mathsf{t}-\mathsf{k}}\,\varepsilon\_{\mathsf{t}} \end{array} \tag{8}$$

$$\begin{split} \Delta \text{Growth}\_{t} &= \mu\_{0} \quad + \Sigma \mu\_{t} \Lambda \text{Growth}\_{t-1} \Sigma \mu\_{t} \Lambda \text{CC}\_{t-j} + \Sigma \mu\_{t} \Lambda \text{OP}\_{t-k} + \Sigma \mu\_{t} \Lambda \text{INV}\_{t-1} + \Sigma \mu\_{\text{m}} \Lambda \text{FDI}\_{t-\text{m}} + \Sigma \mu\_{\text{n}} \Lambda \text{PSE}\_{t-\text{n}} \\ &+ \Sigma \mu\_{\text{o}} \Lambda \text{INF}\_{t-\text{o}} + \Lambda\_{1} \text{CC}\_{t-1} + \Lambda\_{2} \text{OP}\_{t-2} + \Lambda\_{3} \text{INV}\_{t-3} + \Lambda \gamma\_{4} \text{FDI}\_{t-4} + \Lambda\_{5} \text{PSE}\_{t-5} + \Delta\_{6} \text{INF}\_{t-6} \text{e}\_{t} \end{split} \tag{9}$$

We estimated long- and short-run coefficients. The long-run relationship results of the DOLS analysis were robust with the ARDL results. Therefore, only the ARDL results for the short-run relationship were reported (Table A7).

From the observation of Table A7, we see that only STTV and SMTR had a significant positive relationship to Growth in the short run at the 1% and 5% levels, respectively; so a one point increase in the SSTV (SMTR) increases growth by 11.4781 (0.7832) points in the short run. So while STTV promotes growth in both the short and long run, SSTV hinders growth in the short run and promotes it in the long run. Other stocks and financial market variables do not appear to impact economic growth in the short run. As for the control variables, CC, OP, and FDI had a significant negative relationship to growth in the short run; any increase in point in this variable will reduce economic growth. The short-run negative effect of CC is understandable because the West African states have been under the influence of Western countries for decades and still leave the doors open to mismanagement of the countries' economic resources. A sudden breakdown of corruption in these countries will undoubtedly shake the system and affect the economy in the short run. However, the equilibrium is restored over time for the benefit of citizens and the economy, as shown by the positive long-run results (Table A5). This long-run horizon may be difficult to estimate at present since WAEMU countries are still under the yoke of France, which steers policies from outside. It implies that corruption control is unlikely to be effective at the high level of government decision-making. The short-run effects will be those observed outside of the government system's core: the part which is directly connected to the population, and which explains the immediate negative impact on the economy. Only INV, PSE, and INF had no short-run impact on economic growth. One comment on OP is that most WAEMU countries are oil price takers, and the few countries that produce cannot set prices at will because they are post-colonial states. Post-colonial states are de jure independent but are constrained by their economic systems so that policy is steered from outside (Ian 2019).

It seems understandable that if a negative effect on the economy is observed in the short run, no recognisable effects are to be expected in the long run. The negative impact of FDI in both the short and long run can be explained by the fact that foreign investment is mainly aimed at capitalising on the weaknesses of WAEMU countries and serving Western countries. Concerning PSE, it is widely used in assessing economic growth in underdeveloped or developing countries. Primary education is the foundation for any government to upgrade and improve its educational level, and our results confirmed its importance to the economy. The real benefits of primary education can only be seen in the long run. Our inflation results were contrary to those of Barcola and Kebalo (2018). They found that inflation does not affect economic growth in West Africa and has a positive effect on the economy within a certain threshold. While some economically independent West African countries, such as Ghana, Nigeria, etc., could benefit from INF, our results demonstrate that in WAEMU countries, inflation significantly hinders economic growth in the long run. This can be explained by the fact that WAEMU member countries, as mentioned earlier, are using a colonial currency. Since that currency limits them, they do not receive direct shocks from the international price movements and are literally in a state of "stability." Thus, the impact of inflation on the economy will take longer to be observed, compared to economically independent countries.

The diagnostic checks showed that there was no serial correlation and heteroscedasticity. The three tests carried out to control structural breaks also yielded negative results; namely, they confirmed the null hypothesis of no structural breaks over the study period (Table A7).

We complemented our static analysis with a dynamic analysis using a Markovswitching dynamic regression. Although the WAEMU's regional stock exchange shows no structural break, there may be times of regime changes (international or regional) that will affect the market and the economy. We assume two regimes in the economy: high and low volatility. According to Bautista (2003), the state of volatility comes from an unobserved first-order Kth state Markov process. It can be described by transition probabilities.

$$\mathbf{p}\left(\mathbf{s}\_{\mathbf{t}}=\frac{\mathbf{k}}{\mathbf{s}\_{\mathbf{t}-1}}\mathbf{i}\right)=\mathbf{p}\_{\mathbf{i}\parallel}\tag{10}$$

With Pij, the probability is that state i will be followed by state j. Ertugrul and Ozturk (2013) later specified that for the first-order Markov assumption it is necessary that the probability of being in a state depends entirely on the previous state. The simplified transition probability matrix, according to Coskun Yener et al. (2017), is shown in Equation (11) below:

$$\mathbf{p} = \left| \frac{\mathbf{p}\_{11}}{\mathbf{p}\_{12}} \frac{\mathbf{p}\_{21}}{\mathbf{p}\_{22}} \right| \text{ where } \sum\_{\mathbf{j}=1}^{2} \mathbf{p}\_{\mathbf{j}} = 1 \tag{11}$$

As can be seen from Table A8, the dynamic regression confirmed the findings of the static analysis. SMC promotes economic growth in the long run but can hinder it in a system with high volatility. STTV has a short-term positive influence on economic growth with high volatility and in the long run a negative impact on growth with low volatility. SMTR contributes to economic growth in the short and long run with high and low volatility. FMA, FMD, and FD all promote long-run economic growth with low and high volatility regimes, respectively. Details of other variables can be found in Table A7.

From the above analysis, the study's results show that WAEMU member countries are homogeneously interconnected. Similarly, their financial and real sectors are interdependent. The individually addressed regional stock exchanges and the globally considered financial markets are closely connected to WAEMU's economic growth. All the stock and financial market proxies cause (contribute to) economic growth, and a bidirectional causality has been observed between economic growth and the stock and financial market proxies, except for financial market access and depth.

The study has revealed that the stock exchange is driving economic growth in the region, which is very important. Only STTV hinders economic growth in the long run with low volatility regimes. Financial markets are driving economic growth globally. Increased FMA, increased FMD, and FD will contribute to WAEMU's economic growth in the long run. These observations suggest that the development of BRVM and its stock exchange play an important role in boosting the economies of WAEMU countries.

In the long run, measures must also be taken to prevent or undo the negative effects of the STTV.

In the remaining variables, we found that INV does not cause (has no impact on) economic growth in WAEMU countries, despite having long-run positive relationships with WAEMU countries. FDI and OP cause (lead to) economic downturns in the short run, and INF impedes economic growth in the long run.

The impact of CC in WAEMU is negative in the short run but positive in the long run. Primary school education (PSE) has no impact on growth in the short run but has a significant positive impact in the long run.

#### **5. Conclusions**

The relationship between economic growth and financial development remains a hot topic for extensive research and debate. The existence of a causal relationship between financial development and economic growth is still unknown. The available literature that proves the existence of causality remains uncertain regarding the direction of the causality.

This paper has examined the direction of the causal relationship between stock exchange development and economic growth in eight WAEMU countries that have shared a unique stock exchange (BRVM) for 27 years. The applicable methodology was Granger's non-causal relationship, completed by ARDL and Markov-switching analysis.

The study reveals that WAEMU member countries are not simple members of the Union but are all interconnected. Any significant economic situation in one country will reverberate on other countries. Further, the regional stock exchange is confirmed to be a major participant the economic development of WAEMU countries. The level of market capitalization, the value of share traded and the market turnover contribute to boosting the regional economy. In return, the growth of the economy contributes to the development of BRVM. The index of financial development further confirms this mutual relationship between BRVM development and WAEMU economic growth. Thus, it is a call for the eight countries' governments to work hand in hand with institutional and individual investors to guarantee a healthier development of the regional stock exchange for a better economy. Increased investment on the stock exchange is a key economic booster for the WAEMU member countries.

As access to the financial market is also important to boost the stock exchange activities and further promote the economy, governments should put more efforts into sensitizing the populations regarding financial investment and the use of the regional stock exchange. A sad reality of frontier markets is that, due to the low-income, the larger part of the population is focus on its day-to-day survival and is not educated for financial investment. The result is that, unlike developing or developed countries, most of the investors of BRVM are institutional, and very few are individuals.

It is well known that the WAEMU government uses financial markets to raise funds for public projects and investments, but these funds are often raised through the bond market. For example, WAEMU member countries mobilised US \$17 billion in the regional money markets in 2020. The Republic of Benin alone completed one and set a record as of early January 2021 with a Eurobond issuance of 1 billion euros in the international bond market (Ecofinagency 2021a, 2021b). Given the importance of the stock exchange, the WAEMU government needs to rethink the activities of the regional stock exchanges and pay close attention to the optimal use of the stock exchange for the economic development of the country. This should start by raising awareness and educating the populations for investment in the regional stock exchange. Following this, measures should be taken to increase and facilitate market access to both individuals and institutional investors. More efforts should also be put into increasing the market depth for BRVM.

In the same perspective, WAEMU governments should gradually aim at giving priority to BRVM as financial partner and consider leaving the grip of the colonial currency. As observed, investments in WAEMU member countries do not promote economic growth, because most of the funds raised for projects or local development come from foreign investors, partners, or foreign companies. This also explains why when foreign direct investment is added, it does not take long for the economy to receive a blow. In addition, the control of monetary policies by France, as well as the use of the colonial currency, do not leave enough room for an efficient use of BRVM by WAEMU.

This study sheds more light on BRVM, which has not been the subject of extensive research. Diouf and Boutin-Dufresne (2012) focused on WAEMU's growth financing from the securities market and reviewed its achievements and prospects. Ouedraogo and Drabo (2019) have investigated WAEMU's regional integration and economic growth, and Zoungrana et al. (2021) have studied the effect of the occurrence of COVID-19 on BRVM stock returns. Other researchers have also conducted some research on WAEMU or BRVM, but none have focused on the relationship between the two. The relationship between financial market development and national economic growth has been widely studied around the world, and this study is the first of its kind. It provides a survey of the BRVM and economic growth in WAEMU countries under the pressure of a colonial currency.

Another contribution of this paper is that the financial markets considered as a whole, and some of them specifically considered (the stock exchange in this study), show a positive impact on the WAEMU. It also supports the stock and financial market supply leading hypothesis and, primarily, the stock exchange demand-driven hypothesis in WAEMU. Current research results are also consistent with most of the results of previous studies on the causal relationship between stock exchange development and the real economy in lowand middle-income countries (Kolapo and Adaramola 2012; Demirguc-Kunt et al. 2012; Ikikii and Nzomoi 2013; Mittal 2014; Milka 2021; Maku 2020; etc.).

A limitation may be that empirical analysis is not country specific. Proven country interconnections suggest that the economic events of one country are shared with others, but the findings relate to WAEMU as a whole. Therefore, further research may cause similar assessments at the WAEMU-country level. A future study can also focus on how financial debts, especially bonds, affect the economic growth of WAEMU member countries. Moreover, given the growing complaint that the FCFA as a currency slows the development of many African countries, future studies could analyse its role as a vector reducing the performance of the BRVM and the extent to which it hinders the economic prosperity of WAEMU countries. From a governance point of view, later research can examine to which extent policy makers can influence the development of both BRVM and WAEMU countries.

**Funding:** This research has been self-sponsored.

**Data Availability Statement:** The datasets used and analyzed in this study are available from the corresponding author on justified request.

**Acknowledgments:** Sincere gratitude to my dear supervisor, advisor, and teachers who contributed to this manuscript. Their moral support, expert advice, and suggestions will always be remembered.

**Conflicts of Interest:** The author declares no conflict of interest.

#### **Appendix A**



Note: Table of the descriptive statistics for the variables. The sample contains 216-year observations from 1989 to 2015.

**Table A2.** Cross-sectional independence and slope homogeneity test.


Note: The null hypothesis (Ho) for <sup>a</sup> is that there is cross-section independence. For b, the coefficients are homogeneous.



Note: For Hadri Lagrange (Hadri), Ho: all the panels are stationary; the alternative hypothesis (Ha) is that some panels contain unit roots. For Levin Lin Chu (LLC), Ho: there are unit roots in the panels, Ha: the panels are stationary. For Im-Pesaran-Shin (IPS), Ho: there are unit roots in all panels, Ha: some panels are stationary. For Fisher, Ho: all the panels contain unit roots, Ha: at least one panel is stationary.


**Table A4.** Cointegration test.

Note: The tests presented in this table have Ho of no cointegration. Ha is that all panels are cointegrated.



Note: This table reports the results of the DOLS regression for the cointegrated variables. The z-values are shown in parentheses below the coefficients. \*, \*\*, and \*\*\* indicate significance at the 10%, 5%, and 1% level, respectively.


Note: The test's null hypothesis is that the independent variable does not cause the dependent variable. "C" means causes, and "NC" means does not cause. In test 1, the dependent variable is Growth. In test 2, the direction of causality is reversed, with Growth being the independent variable. \*, \*\*, and \*\*\* indicate significance at the 10%, 5%, and 1% level, respectively.



Note: The dependent variable is Growth. <sup>a</sup> Breusch-Godfrey Lagrange Multiplier (LM) test for autocorrelation; the *p*-values is in parentheses. <sup>b</sup> Durbin's alternative test for autocorrelation; the *p*-values is in parentheses. <sup>c</sup> Breusch-Pagan/Cook-Weisberg test for heteroscedasticity; the *p*-values is in parentheses. <sup>d</sup> White test for homoscedasticity; the *p*-values is in parentheses. Cusum, Cusum Square, and Single tests are all for the detection of structural breaks. e,f the test statistics are lower than the critical value at 1, 5, and 10%, so we confirm the null hypothesis of no structural break. <sup>g</sup> *p*-values are all > 0.050; thus, we confirm the null hypothesis of no structural break. \*, \*\*, and \*\*\* indicate significance at the 10%, 5%, and 1% level, respectively.


**Table A8.** Markov-switching dynamic regression.

Note: The dependent variable is Growth. \*, \*\* and \*\*\* indicate significance at the 10%, 5% and 1% level, respectively.

**Figure A1.** Economic growth and selected stock and financial market variables' evolution from 1989 to 2015 in Benin, Burkina Faso, Cote d'Ivoire, and Guinea Bissau. Source: Prepared by authors, based on collected data.

**Figure A2.** Economic growth and selected stock and financial market variables' evolution from 1989 to 2015 in Mali, Niger, Senegal, and Togo. Source: Prepared by authors, based on collected data.

#### **References**


Francis, Bill B., and Eric Ofori. 2015. Political Regimes and Stock market Development. *Eurasian Economic Review* 5: 111–37. [CrossRef] Frees, Edward W(Jed). 1995. Assessing Cross-Sectional Correlations in Panel Data. *Journal of Econometrics* 64: 393–414. [CrossRef]


Mittal, Rishab. 2014. The Effects of Market Capitalization Ratio on GDP Growth and Capital Market Robustness in Newly Industrialized Countries. *UChicago Undergraduate Business Journal* 1: 1–18.

Mohtadi, Hamid, and Sumit Agarwal. 2007. *Stock exchange Development and Economic Growth: Evidence from Developing Countries*. Working Paper. Milwaukee: Department of Economics, University of Wisconsin-Milwaukee.


### *Article* **What Does Vietnam Gain When Its Currency Depreciates?**

**Nguyen Thi Thanh Binh**

Department of Accounting, Chaoyang University of Technology, Taichung City 41349, Taiwan; tbnguyen@cyut.edu.tw

**Abstract:** The study investigates how the depreciation of the Vietnam dong (VND) against the US dollar (USD) affected export turnover and the stock market in Vietnam during the period from 2000 to 2020. A Markov triple regime-switching model is developed for time-series data involving multistructural breaks. Empirical results reveal that the impact of exchange rates on export turnover and stock price existed both in the long and short run. In the short run, the depreciation of VND led to (i) an increase in export turnover after 12 months; (ii) a decrease in export turnover of the high-growing regime in the short term; (iii) a reduction in stock returns in most cases. In addition, the common cycle from order receipt, preparation, production, and export is about 12 months for all states. The high volatility of export turnover was associated with high export growth. The commonly used phrase of "high risk, high return" seems to not be true for Vietnam's stock market. The results of this study suggest the feasibility of a slight appreciation of VND against USD, which is the key to escape from being labeled a currency manipulator by the US Treasury.

**Keywords:** currency; export; stock returns; triple regime-switching model; Vietnam

**JEL Classification:** C22; L85; P44

**Citation:** Thi Thanh Binh, Nguyen. 2021. What Does Vietnam Gain When Its Currency Depreciates? *Economies* 9: 185. https://doi.org/10.3390/ economies9040185

Academic Editors: Robert Czudaj and Andreia Dionísio

Received: 12 October 2021 Accepted: 18 November 2021 Published: 19 November 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

#### **1. Introduction**

Exchange rates are a hot topic for academic debate and speculative market forces. There are two macroeconomic variables of emerging economies such as Vietnam that play an important role with foreign investors, namely, inflation and exchange rate. The literature *de facto* has not yet paid much attention to studying the advantages and disadvantages of currency devaluation, and its connection to exports and the stock market, which has seen different changes in the implementation of the monetary policy and exchange rates over the past two decades. In particular, the US Department of the Treasury officially labeled Vietnam a currency manipulator at the end of 2020. Being labeled a currency manipulator puts Vietnam at risk of being restricted by the US under US law from accessing procurement contracts, and government and development financing, under US law.

A depreciating currency theoretically supports exports. A weak currency means that domestic goods are cheaper abroad. Therefore, it increases both exports and stock prices, as more businesses raise their capital through the securities market, pushing up the stock price. There is a growing economic literature dealing with the possible effects of exchange rates on exports, such as Sercu and Hulle (1992); Arslan and Wijnbergen (1993); Aristotelous (2001); Hall et al. (2010); Berman et al. (2012); Choudhri and Hakura (2015); Paudel and Burke (2015); Nguyen and Do (2020); Chen et al. (2021). However, there is no clear consensus in the empirical literature on the direction of the relationship, and positive, negative, mixed, or no effects. One of the first studies investigating the interaction between exchange rates and stock prices was conducted by Franck and Young (1972). Numerous articles then successively reported the short- and long-run relationship between them, such as Fang and Loo (1996); Ajayi et al. (1998); Kanas (2000); Homma and Benzion (2005); Phylaktis and Ravazzolo (2005); Hau and Rey (2006); Pan et al. (2007); Caporale et al. (2014); Bahmani-Oskooee and Saha (2015); Reboredo et al. (2016); Sui and Sun (2016); Dahir et al. (2018);

Andriansyah and Messinis (2019); Lee and Brahmasrene (2019); Nguyen et al. (2020); Ding (2021). However, the results and direction of the relationship between exchange rate and stock price are contradictory across studies.

The triple regime-switching model was developed for examining whether the depreciation of the Vietnam dong (VND) to the US dollar (USD) could lead to an increase in future export turnover and stock price in Vietnam. Over the period from 31 July 2000 to 31 December 2020, empirical results show that the depreciation of VND most effectively promotes export growth in a moderate-growth state to which the group of agricultural products, raw materials, and unprocessed goods belong. The high-growth state where the depreciation of VND negatively affects exports should belong to the FDI sector, where export prices do not depend much on the exchange rate in the short term. Meanwhile, the effect on the stock market is generally negative. Empirical results contradict the usual wisdom of "high risk, high return" in the stock market of Vietnam.

This study makes three important contributions to the existing literature. First, as far as it could be ascertained, this is the first study investigating the gain and loss of VND depreciation on export turnover and stock price in Vietnam within asymmetric frameworks. Second, this is one of few studies that take the triple regime model into account for more appropriate results. Third, empirical results suggest that the effect of exchange rates on export and stock price exists both in the long and short run.

These results contribute to the discussion of Vietnam's monetary policy when the US raised the issue of currency manipulation and put Vietnam on the observation list. The remainder of this paper is organized as follows. Section 2 presents a data description. Section 3 outlines the empirical methodology and analysis of the results. Section 4 offers conclusions.

#### **2. Data and Descriptive Statistics**

#### *2.1. Data*

Data for this study were obtained from two sources. Monthly data of the export value (EXP) and VND/USD exchange rates (EXR) of Vietnam were obtained from the database of the International Monetary Fund (IMF) from 31 July 2000 to 31 December 2020. The monthly data of VN-INDEX (VNI) were taken from Vietstock in the same period.

#### *2.2. Descriptive Statistics*

Let EXR, EXP, and VNI denote the vector of exchange rates (EXR), export value (EXP), and VN-INDEX (VNI), respectively. The trends of EXR, EXP, and VNI during the period of 31 July 2000 to 31 December 2020 are plotted in Figure 1. EXP was on an upward trend during the sample period. A significant decrease occurred every February. The VNI was rather stable within the range from 101 to 500 points in the first 5 years since trading commenced on 28 July 2000. In the period from June 2006 to June 2008, Vietnam's index had an unexpectedly strong growth and nearly peaked at 1138 points in March 2007. The 2008 global economic crisis caused the downturn of Vietnam's stock index, which suddenly fell sharply below 300 points in early 2009. Although it was so badly affected by the crisis, VNI quickly recovered after the bleak period, as evidenced by the second half of 2009 to the end of 2020. Vietnam's index maintained continuity and increased, while the exchange rates of VND to USD remained stable except for the drastic increase in the period of April 2008 to February 2011. Table 1 reports the descriptive statistics of EXP, EXR, and VNI, showing that Vietnam's HOSE index reached its maximum at 1174 points in March 2018, export value at USD 27,702.47 million in August 2020, and an exchange rate at VND23,261/USD in May 2020. In addition, skewness and kurtosis show the right-skewed and leptokurtosis of stock return distribution. Jarque–Bera statistics significantly reject the normality of three variables.

**Figure 1.** Trend of exchange rate (EXR), export (EXP) and stock index (VNI).


**Table 1.** Descriptive statistics.

Note: Sample period spans from July to December 2020. J–B, statistic of Jarque–Bera normal distribution test. Ljung–Box test used for testing for variable autocorrelation. *LB(12)* that uses the lag length of 12 months is the statistics of the Ljung–Box test. *LB*2(12) that applies the lag length of 12 months is the Ljung–Box statistics for squared residuals. \*\*\* indicates 1% significance level.

Figure 2 shows that, beginning in 2007, the central bank began to loosen the range of effective rates from 0.25% to 0.5%; at the end of 2007, it was further enlarged to 0.75. During this period, the exchange rate regime remained with a small oscillation amplitude. The year 2008 was eventful with many unexpected occurrences for the exchange rate of VND/USD. In the first half of 2008, the central banks applied tightening monetary measures; the interest base rate was adjusted from 8.25% to 8.75%, and inflation was pushed. In the second half of 2008, the rate of VND/USD suddenly increased from 16,600 to 16,998. Then, 2009 was when monetary policy had to face unpredictable challenges arising from the inadequacies of the economy, and the adverse impact of the financial crisis and economic recession. To increase supply and stabilize the foreign exchange market, banks deployed more drastic measures, such as widened exchange rates from +/−3% to +/−5%. Generally, the gap between the average monthly return of stock (0.97%) and the growth of VND/USD (0.20%) was narrow (0.70%). Since the outbreak of the COVID-19 pandemic at the end of January 2020, the US Federal Reserve (FED) has lowered its operating interest rates to 0–0.25% (lower–upper range) and relaunched the quantitative easing (QE) in an effort to rescue the US economy from the pandemic-induced recession. The unprecedented easing policy along with the negative economic growth outlook of the US in 2020 caused the dollar to decline. On 2 November, the dollar strength index fell 2.4% since the beginning of the year. However, as of 30 October, the VND/USD exchange rate has increased slightly by 0.2% compared to the beginning of the year. If investors sell the US dollar (USD) for the Vietnamese dong (VND) to invest in the stock market, their returns adjusted for inflation were negative from July to December 2020, and the gap between the average growth rate of EXP and EXR was about 1%.

**Figure 2.** Trend of exchange rate volatility, export growth, and stock returns.

#### **3. Empirical Method and Results**

#### *3.1. Empirical Method*

The Markov switching dynamic regression (MSDR) of Hamilton (1989), namely, the regime-switching model, is one of the most popular nonlinear time-series models. It involves multiple structures that can characterize the dynamic behaviors of data under different regimes. The basic model with switching intercept is as follows:

$$Y\_{l} = \mu(s\_{l}) + \beta(s\_{l})X\_{l} + \varepsilon\_{l} \tag{1}$$

where

$$\mu = \begin{cases} \quad \mu\_1 \text{ if } s\_t = 1 \text{ (Region 1)}\\ \quad \mu\_2 \text{ if } s\_t = 2 \text{ (Region 2)}\\ \quad \quad \dots \\ \quad \mu\_n \text{ if } s\_t = n \text{ (Region n)} \end{cases}$$

*Yt* is a dependent variable that follows a process depending on the value of unobserved state *st*. *st* is assumed to have n possible regimes. *Xt* is a vector of exogenous variables, *μ*(*st*) is the conditional mean of *Yt* in each specified regime. *<sup>t</sup>* is an independent and identically distributed (i.i.d.) normal error. The regression model was assumed to be linear in regime *n*.

The Markov switching regression model extends the basic exogenous probability framework by specifying a first-order Markov process for regime probabilities, where *st* ∈ {0, 1}, and regime transitions are calculated according to

$$P[s\_t = a | s\_{t-1} = b] = p\_{ab}(t) \tag{2}$$

These probabilities are presented in a transition matrix of an ergodic n regimes Markov process as follows:

$$p(t) = \begin{bmatrix} p\_{11}(t) & \dots & p\_{1n}(t) \\ \dots & \dots & \dots \\ p\_{n1}(t) & \dots & p\_{nn}(t) \end{bmatrix} \tag{3}$$

where element *ab* represents the probability of transitioning from regime a in period *t* − 1 to regime b in period *t*.

Following Hamilton's (1989), probabilities can be parameterized in terms of a multinomial logit. As each row of the transition matrix specifies a complete set of conditional probabilities, a separate multinomial specification for each row *a* of the matrix is as follows:

$$p\_{ab}(G\_{t-1}, \vartheta\_a) = \frac{\exp\left(G\_{t-1}' \vartheta\_{ab}\right)}{\sum\_{s=1}^{n} \exp\left(G\_{t-1}' \vartheta\_{ab}\right)}\tag{4}$$

where *a* = 1, . . . , *n*, *b* = 1, . . . , *n*, *ϑan* = 0, and *Gt*−<sup>1</sup> contains a constant.

The basic switching model can be extended to the Markov switching dynamic regression to allow for dynamics in the form of lagged exogenous variables:

$$Y\_t = \mu(s\_t) + \sum\_{i=1}^{q} \beta\_i(s\_t)(X\_{t-i}) + \sigma(s\_t)\varepsilon\_t \tag{5}$$

where *<sup>t</sup>* is iid standard normally distributed, and standard deviation σ(*st*) is regimedependent.

For testing the relationship between variables, the autoregressive AR(p) process was added to the model for capturing the remained autocorrelation of residuals:

$$\mathcal{Y}\_t = \mu(\mathbf{s}\_t) + \sum\_{i=1}^q \beta\_i(\mathbf{s}\_t)(X\_{t-i}) + \sigma(\mathbf{s}\_t)\mathbf{e}\_t + \sum\_{i=1}^p q\_i \mathcal{Y}\_{t-i} \tag{6}$$

#### *3.2. Empirical Results*

Let *Yt* denote the vector of EXP, EXR, and VNI. Then, *Yt* yields the annual growth rates after taking the differences of its logs:

$$\log\_l \, = \log(\mathcal{Y}\_l \,) - \log(\mathcal{Y}\_{l-1} \,) \tag{7}$$

The Markov switching dynamic regression discussed in the preceding section is only suitable for stationary data. The growth rates of EXP, EXR, and VNI are plotted in Figure 2. These three series were stationary after taking the differences in their log. Results of ADF and DF-GLS unit root tests are reported in Table 2. The equations of unit root tests include both a constant and a time trend. The optimal lag length was selected according to the minimal SIC. Testing results showed that three variables were stationary after taking differences in their log.


Note: Constant and time trend included in all test equations. Maximal lag length applied for the test is 15 periods. Numbers in parentheses are the adequate lag order of the ADF test and DF-GLS test, determined by the minimal SIC. \*\*\*, \*\* and \* indicate 1%, 5%, and 10% significance levels, respectively. Critical values of ADF derived from Mackinnon (1996). Critical values of DF-GLS derived from Elliott–Rothenberg–Stock (1996).

The residual-based cointegration test of Engle and Granger (1987) was applied for testing the linear and nonlinear long-run effect of EXR on EXP and VNI. Table 3 summarizes the results of the residual-based cointegration test. Both the Engle–Granger tau statistic (t-statistic) and the normalized autocorrelation coefficient (z-statistic) are uniformly failing to reject the null of no cointegration at conventional levels. These test statistics suggest that it is unable to reject the null hypothesis of no cointegration between variables in both linear and nonlinear models. In other words, the long-term influences of the exchange rate on exports and stock index clearly exist.


**Table 3.** Engle–Granger test for cointegration.

Note: Long-run influence of Ln\_EXR (exchange rate) on Ln\_EXP (export) and Ln\_VNI (stock index) examined on the basis of residual-based cointegration test of Engle and Granger (1987). Maximal applied lag was 15 periods. Optimal lag selected according to SIC. Numbers in parentheses are P values (see MacKinnon (1996) for reference).

The Markov switching dynamic regression shows different dynamics across unobserved regimes using regime-related parameters to adapt to structural breaks or other multistate phenomena. To determine the number of regimes for the MSDR model, the structural break test of Bai–Perron (1998) was applied. This method detects the breakpoints of the relationship between Ln\_EXP vs. Ln\_EXR and Ln\_VNI vs. Ln\_EXR. As Table 4 shows, two breakpoints were detected for both equations. In other words, the MSDR model with three states was appropriate for examining the nonlinear dynamic effect of EXR on EXP and VNI. The three states represent low-, moderate-, and high-growth states.

**Table 4.** Structural breakpoint test.


Note: Bai–Perron test (1998) applied for detecting breakpoints of the relationship between Ln\_EXP vs. Ln\_EXR, and Ln\_VNI vs. Ln\_EXR. F values reject nulls of 0, 1, and 2 breakpoints, but test of 3rd breakpoint did not reject the null. Timepoints of breaks are the first date of the subsequent regime; \*\*\*, 1% significance level.

Markov switching dynamic regression was applied to Model 6 using the observations of the whole sample. Estimation results summarized in Tables 5 and 6 were obtained by applying multivariate Markov switching dynamic regression to the after-change sample. *p*<sup>11</sup> is the estimated probability of staying in Regime 1 in the next period, and *p*<sup>22</sup> is the probability of staying in Regime 2. The estimated standard deviations for the entire process are represented by Log(σ), which shows periods of high and low volatility.

Parameters in Table 5 and smoothing probabilities in Figure 3 characterize the influence of exchange rate on export growth in a given month earlier. Table 5 presents the estimation for five selected periods of exchange rate: same period, 3 months, six months, one year, and two years earlier. As reported in Table 5, μ denotes the mean of three regimes: Regime 1 is the high growth state of export (mean of 4.3%), Regime 2 is the moderate growth state (mean of 0.5%), and Regime 3 is the low growth state (mean of −1.6%). The transition probabilities for Regime 1 to 1 and Regime 2 to 2 are *p*<sup>11</sup> and *p*22, respectively. Both *p*<sup>11</sup> and *p*<sup>22</sup> showed that the three regimes were highly persistent. The implied standard deviations of Log(σ) are 0.163, 0.059, and 0.003, respectively, indicating that Regime 1 corresponds to the high-volatility period, Regime 2 corresponds to the medium-volatility period, and Regime 3 corresponds to the low-volatility period. Exchange rate volatility had positive and statistically significant impact on export in all three regimes.


**Table 5.** Effect of exchange rate on export growth.

Note: Q10uh<sup>−</sup> <sup>1</sup> 2 and Q10u2h−<sup>1</sup> represent 10th standardized residual and squared standardized residual of Ljung–Box statistic, respectively. p11 and p22 stand for Markov transition probabilities. Log L is the value of maximum likelihood function. Values inside square brackes are standard errors. Values inside parentheses are p statistics. \*, \*\* and \*\*\* represent 10%, 5% and 1% significant level, respectively.

Results show that the significant signs of coefficients on the lagged exchange rate were not consistent in the three regimes. The depreciation of VND against the USD one year earlier positively affected export growth in all three regimes. These positive effects existed in all cases of Regime 2, where the influence of VND depreciation in three months and two years earlier on export growth was positive and significant. Regime 2 might be the group of agricultural products, raw materials, and unprocessed goods. The depreciation of VND most effectively promotes export growth in a moderate growth state and in one year earlier for all regimes. However, it has the opposite effect in most of the remaining cases of Regimes 1 and 3. Regime 1 where the devaluation of VND negatively affects export should belong to the FDI sector as their export prices do not depend much on the devaluation of VND in the short term. This sector greatly contributed in increasing Vietnam's export capacity and accounts for over 70% of the total export turnover of the country.

Table 6 shows that Regime 1 belongs to the low-return state and has a mean of −2.30%; Regime 2 is the moderate-return state that has a mean of 0.04%; Regime 3 is the high-return state that has a mean of 1.50%. The estimates of *p*<sup>11</sup> and *p*<sup>22</sup> imply that the three regimes are highly significant.

Results indicate that the depreciation of VND against the USD three months earlier positively and significantly affected stock returns in Regime 2, but it did not in Regimes 1 and 3. Results also indicate that the depreciation of VND negatively and significantly affected stock returns in Regime 3 when stock returns were high.


**Table 6.** Effect of exchange rate growth on stock returns.

Note: Q10- uh<sup>−</sup> <sup>1</sup> 2 and Q10- u2h−<sup>1</sup> represent the 10th standardized residual and the squared standardized residual of Ljung–Box statistic, respectively. P11 and p22 stand for Markov transition probabilities. Log L is the value of maximum likelihood function. Values inside square brackets are standard errors. Values inside parentheses are p statistics. \*, \*\*, and \*\*\* represent 10%, 5%, and 1% significance levels, respectively.

Results show estimates of Log(σ) in the high-, low-, and medium-volatility regimes. Implied standard deviations were 0.136, 0.023, and 0.059, respectively. The commonly used phrase in investment of "high risk, high return" does not hold in the case of Vietnam's stock market, as taking a high risk does not guarantee that relative high returns can be achieved. Regime 1, with high risk and low returns, lasted the longest. The Ljung–Box diagnostic test statistics *<sup>Q</sup>*10- *uh*<sup>−</sup> <sup>1</sup> 2 for the residuals and *Q*10 *u*2*h*−<sup>1</sup> for the squared residuals indicate that models in Tables 5 and 6 can describe salient features in export growth rates and stock returns.

Figure 3 is the plot of the smoothing probabilities P(*st* = *a*|ℵ*T*) of the triple regime model that explores how exchange rate growth affects export growth and stock returns. The plot of exchange rate effect on export growth (ΔLn\_EXR →ΔLn\_EXP) shows that both the red vertical dashed lines at June 2005 and May 2012 are in Regime 2. The highest volatility period was 2008–2009, which coincided with the global financial crisis. This high volatility carries high export growth caused by the depreciation of VND one year earlier.

The plot of exchange rate effect on stock returns (ΔLn\_EXR →ΔLn\_VNI) shows that most of the high variance regime is located in the first half of the sample. The probabilities of one appeared the most in Regime 1 during February 2006 to May 2009, where the red vertical dashed lines signify the first breakpoint identified by the Bai–Perron test (1998) of the stock index in February 2006. That is, the stock is in a state of low return and high volatility. The second breakpoint occurred at January 2013 where Regime 2 is switched to Regime 3, and the stock is in a state of high return.

**Figure 3.** Markov switching smoothed probabilities of triple regime model.

Figure 4 plots the impact of the risk generated by the exchange rates on the export and stock market. It shows that the shock of exchange rates on export is negative, then becomes positive, and slowly diminishes. This implies that the exchange rate first makes the export decrease then slowly rise with a small magnitude, and disappears after 12 months. From the impact on the stock market, it can be found that the effect is not significant and then slowly disappears after 6 months.

To summarize, this study examined the behaviors of exports and stock prices to changes in exchange rates in Vietnam. The VND/USD exchange rate is carefully monitored and controlled by the State Bank of Vietnam (SBV). Generally, the SBV manages the exchange rates to vary within a given range. When exchange rates rise sharply in response to political or economic shocks, the SBV usually intervenes in the market to ensure the stability of the exchange rates. As such, Vietnam's export and stock market may behave differently during different regimes.

The long-term effect of exchanges rates' volatility on the export and stock markets do exist in Vietnam. The signs of these effects are examined with the triple regime-switching model. VND devaluation only significantly increases the export value after 12 months. This shows that the cycle from order receipt, preparation, production, and export is usually 12 months. It also works for most of the periods during the regime of moderate export growth, and it has the opposite effect on the regime of high growth. When considering the combined effect of all 3 regimes, the VND depreciated causes the export value to decrease in the first 2 months; then, it gradually increases but with a narrow range and disappears after 12 months. This implies that, when export products are priced in USD, their prices are lower, causing the export turnover to be lower in the first 2 months, then gradually increasing in the following months on the basis of the devaluation of VND. The devaluation

of VND increases stock returns only effectively in the previous 3 months, and this effect only significantly appears in the medium-return period and does not last long. In most remaining cases, the opposite is true. Its combined effect is not significant. Empirical results contradict the usual wisdom of "high risk, high return".

**Figure 4.** Impulse response of export growth and stock returns to volatility of exchange rate.

#### **4. Conclusions**

This study developed a triple regime-switching model in which low-, medium-, and high-growth regimes play roles in explaining a substantially detailed relationship between exchange rates, and export turnover and stock returns. The proposed model allows for multistate phenomena to better capture the time-varying aspect of the effect of exchange rates. Applications of the proposed model on the effect of exchange rates suggest that the depreciation of VND most effectively promotes export growth in a moderate-growth state to which the group of agricultural products, raw materials, and unprocessed goods belong. The high-growth state where the depreciation of VND negatively affects exports should belong to the FDI sector where export prices do not depend much on the exchange rate in the short term. Meanwhile, the effect on the stock market is generally negative. Empirical results contradict the usual wisdom of "high risk, high return" in the stock market of Vietnam.

In the context that the USD is depreciating sharply compared to other currencies in the world, the slight appreciation of VND against USD not only does not harm the competitiveness of export businesses, but could also stimulate investment capital flow into Vietnam, reduce the burden of foreign debt payment, lower the trade imbalance between the US and Vietnam, and be the key to escaping from being labeled a currency manipulator by the US Treasury. This policy adjustment of the State Bank is necessary.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Conflicts of Interest:** The author declares no conflict of interest.

#### **References**


### *Article* **Dynamic Linkages among Saudi Market Sectors Indices**

**Farouq Altahtamouni \*, Hajar Masfer and Shikhah Alyousef**

Financial Sciences Department, Applied College, Imam Abdulrahman Bin Faisal University, P.O. Box 1982, Dammam 31441, Saudi Arabia; hmmasfer@iau.edu.sa (H.M.); salyousef@iau.edu.sa (S.A.) **\*** Correspondence: fraltahtamouni@iau.edu.sa

**Abstract:** This study aims to test the causal relationship between Saudi stock market index (TASI) and sectoral indices throughout the period from 2016–2020. The study data were extracted through the main index of the Saudi market and the indices of the available data of 19 sectors out of 21 sectors. The unit root test was used along with the Granger causality test, in addition to multiple regression tests in order to analyze the study hypotheses. The study shows that all index series were stationary at the zero level I (0), and the results also show that there were bidirectional and unidirectional causal relationships between TASI and sectoral indices, and that TASI effectively mirrors all the changes that occur in the Saudi stock market.

**Keywords:** TASI; unit root; granger causality; sectoral indices

#### **1. Introduction**

In developed and emerging economies, a stock market is one of the fundamental pillars playing active roles in that development, for it acts as an intermediary between borrowers and lenders. An efficient stock market refers to how successful a market is in incorporating stock prices, which reflect the stock value (Keane 1983). This market may accelerate the development process in an economy through raising savings and efficient allocation of resources.

Throughout recent years, stock market integration has become a major topic in finance literature and has gained wide currency due to its high significance for many parties involved (Youcef and Adewale 2017). This integration could generate considerable economic growth across the economy, enhance the allocation of capital, decrease costs of capital, and raise the risk sharing efficiency.

On the contrary, Rehman and Hazazi (2014) explain that this co-movement among economies may increase homogeneity in market performance in reaction to international financial impacts. This might cause a peril from some events such as a credit crunch, capital flight, a contagion effect, and so on.

Accordingly, the transmission of volatility and crises among different sectors of a stock market is an important issue for policy makers, investors and researchers (Mohammed et al. 2020; Ahmed 2016).

There is a significant concern with studies considering the relationship between the general stock market index and the sector indices. For example, Sharabati et al. (2013) and Aravind (2017) illustrate that there is a strong tendency for stock prices to move in accordance with the overall stock market and in parallel with the direction of other stocks in the same sector. Besides, the movement of stock prices in one sector may affect stock prices in another sector. Whenever a company has a large market capitalization, many changes might occur in other stock prices in the same sector (Sharabati et al. 2013). Thus, understanding and examining the relationship between various sectors' indices on one hand and the general index on the other would greatly benefit all parties involved.

Over the last three decades, emerging markets have raised profits enormously, for they provide many investment opportunities in financial sectors (Rehman and Hazazi 2014).

**Citation:** Altahtamouni, Farouq, Hajar Masfer, and Shikhah Alyousef. 2022. Dynamic Linkages among Saudi Market Sectors Indices. *Economies* 10: 16. https://doi.org/ 10.3390/economies10010016

Academic Editor: Robert Czudaj

Received: 28 October 2021 Accepted: 28 December 2021 Published: 4 January 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

Saudi Arabia, which is considered as the biggest oil exporter country in the world, has one of the strongest stock markets among the Gulf Council Countries (GCC), the Middle East, and North Africa (MENA) (Mustafa 2012). The Saudi stock market (Tadawul) was formally established in 1984, and it is now one of the leading emerging markets in the Arab world. Then, in 2001, and as a result of the rising of the trading volume in Saudi stock, a new official Saudi stock market index called TASI was established (Tadawul All Share Index) (Tadawul 2019). Additionally, the extant statistic reports of Tadawul (2019) mentioned that the number of companies listed on the stock exchange increased from 163 in 2013 to 199 at the end of 2019, with total market capitalization of US\$ 2,406.78 billion.

In this context, through conducting the Granger causality between this market's indices, the investigation on the Saudi Arabia's stock market (Tadawul) would contribute to clarifying the volatility of Tadawul.

#### **2. Study Objectives**

This study aims at:


#### **3. The Importance of the Study**

This study adds new information about a developing market such as the Saudi market, which has proven its worth during the previous years as one of the most important markets globally and in the Middle East, especially as it includes the Aramco company which is considered the largest company in the world in terms of market value. Since the Kingdom of Saudi Arabia is considered one of the Group of Twenty countries (G 20), meaning that it is one of the largest economies in the world, and this makes this study of great importance. Finally, this study is considered important because it is the first study that aims to examine the relationships between the Saudi market index (TASI) and sectoral indices, and through its results, this study attempts to provide a service to investors that helps them make their investment decisions.

#### **4. Literature Review**

The co-movement among sectors of the equity exchange and the extent of its impact on the general index of the market have long been the subject of researcher attention. Scholars use different research methods and statistics to analyze and understand this relationship. Many use Pearson's correlation method to investigate the relationship between one sector and others in the market (Rajamohan and Muthukamu 2014) or to investigate the linkage among all industries within a market (Mohanty et al. 2019; Cao et al. 2013). Other researchers employ Granger causality to test the effect of industry indices (Mustafa 2012; Ahmed 2016; Aravind 2017). In addition, some studies use ANOVA to analyze the variation among sectoral indices (Sharabati et al. 2013), while others use Vector Error Correction Model (Arbrlaez et al. 2001;Vardhan et al. 2015). Furthermore, Sharabati et al. (2013) employed Johansen's multivariate cointegration analysis to study the abovementioned correlation.

The findings of these studies share some points and differ in others. For instance, Rajamohan and Muthukamu (2014) examined the relationship between the banking industry index and other sectors in the National Stock Exchange of India and found a direct link between the banking industry and other sectors in the market. This result is consistent with Vardhan et al. (2015), who concluded that the banking industry is a leading industry in the Indian Stock Market. Moreover, some studies dealt with the long/short-term relationship like Mustafa (2012), which investigates the co-movement of the Saudi stock market sub-sectors and their connection with the whole market. Considering this to be a long-term relationship, unlike a short-term relationship, there seems to be a positive

correlation among Saudi sub-sectors indices. There is also a long-term causality between the sub-sectors and the market portfolio's movement. Similarity, Ahmed (2016) comes to the same conclusion on the Egyptian Stock Exchange and Arbrlaez et al. (2001) did so in their analysis of the Colombian Stock Exchange. Furthermore, there are several research findings about the significance of this connection that vary by sector. By way of illustration, Mohanty et al. (2019) found that all industries in the Bombay stock Market have a positive influence on the market index, except for health care and telecommunications. The health index has an inverse correlation with all other sectors. Yet, the telecommunications industry is directly correlated with the automotive and health sectoral indices, and it is inversely related to the indices for information technology, banks, fast-moving consumer goods, and oil. Likewise, Aravind (2017) demonstrated a unidirectional linkage among the Indian Stock Exchange index, FMCG and IT. However, the banking industry shows a highly significant correlation to the market index. On the other hand, Sharabati et al. (2013) analyzed the data of the Jordanian Stock Exchange to determine the correlations among sectoral indices and the market overall. They found a remarkable positive correlation among the sectors in the long run, especially in the sectors of finance, industry and services respectively. Another study examined the sectoral behavior of the Chinese Stock Market index and divides it into two periods based on economic conditions (Cao et al. 2013). The first occurs when the market experiences a sharp rise and fall in stock prices and sectoral indices show consistent behavior together. The other takes place when the market is in a normal condition, and the findings vary depending on the industry. Some industries indices show strong linkage to the market portfolio, including financial services, energy, and industrials. Nevertheless, some show a weak linkage to the market, such as telecom, IT, and utilities. In addition, some industries show no linkage to the market, such as health, consumer discretionary, and consumer staples (Cao et al. 2013).

Arbrlaez et al. (2001) found that the co-movement among sectoral indices is becoming more significant over time and suggest that this conclusion could be applicable in other emerging markets. There are some studies that examine not only the relationship between market indicators but also between different market indices. The most recent study on this subject is a study of Joshi et al. (2021), where researchers aimed to examine the degree of interdependence between 22 indices in the American and European regions from 2005 to 2018 by using ADF and a causality test. The results indicate that there is a significant amount of interdependence between stock markets. It was also observed that there is an association between markets. Stoupos and Kiohos (2021) tested, through their studies, the integration between the European area stock markets after the end of 2010 debt crisis. The results revealed that the stock market integration was strong between Germany and EA core member-states, but disparate for the EA periphery. In contrast, there are only indications regarding the EA, Eastern Mediterranean and Baltic stock markets' integration with the DAX-30. And there is the Kapar et al. (2020) study, which is one of the studies that examined the integration of financial markets using data from the Dubai Financial Market Stock Exchange, the Abu Dhabi Stock Exchange and the FTSE Nasdaq Dubai UAE 20 index, by applying a vector error correction model and a permanent–transitory decomposition to the series of prices. The researchers found a long-run equilibrium relationship between the three financial indices, suggesting that UAE stock markets are integrated, and they found that shocks to any of these markets affect the other markets in the long and the short run through the equilibrium condition. Furthermore, Nasser and Hajilee (2016) examined stock market integration among five selected emerging stock markets (Brazil, China, Mexico, Russia and Turkey) and the developed markets of the US, UK and Germany by using monthly data from 2001 to 2014. The results show evidence of the existence of short-run integration among stock markets in emerging countries and developed markets, and the long-run coefficients for stock market returns in all emerging countries showed a significant relationship only with Germany's stock market return. Therefore, our study enriches the preceding literature with an up-to-date time series analysis for the Saudi Stock Exchange

(Tadawul). Moreover, it will appeal to investors and portfolio managers by presenting the effectiveness of portfolio diversification in the Saudi equity market.

#### **5. Research Methodology**

*5.1. Sample and Data*

The study sample consists of daily data for the Saudi Stock Market General Index (TASI: Tadawul All Share Index) and 21 sectoral indices from 2016–2020. The study relies on 19 sectors in order to provide data for the indicators of these sectors, as shown in Table 1.

**Table 1.** Indices and index symbols.


\* All the symbols were developed by researchers except for the general index.

#### *5.2. The Study Hypotheses*

The hypotheses of the study were formulated according to the questions of the study and its goals in order to test the relations between the study variables.

5.2.1. Hypotheses of the Time-Series Non-Stationarity Test

**Hypothesis 1 (H1).** *TASI returns are nonstationary or have a unit root.*

**Hypothesis 2 (H2).** *The sectoral indices returns are nonstationary or have a unit root.*

5.2.2. Hypothesis Testing Causality between TASI and Sectoral Indices

**Hypothesis 3 (H3).** *TASI returns do not cause sectoral returns.*

**Hypothesis 4 (H4).** *Sectoral indices returns do not cause TASI returns.*

5.2.3. Hypothesis of the Multiple Regression

**Hypothesis 5 (H5).** *The sectoral indices have no statistically significant effect on the TASI.*

*5.3. Statistical Methods Used in the Study*

The following tests were used to examine the hypotheses of the study:

#### 5.3.1. Tests of Nonstationarity (Unit Root Test)

Most of the time series of macroeconomic variables are affected by an instability procedure and the presence of a unit root, and regression on a model that contains nonstationary time series will lead to a spurious skew among the data and cause problems in the analysis (Granger and Newbold 1974). Thus, it is not possible to use t-ratios to determine the effect of one variable on another variable in the event of nonstationarity. This is due to the presence of a trend factor which reflects certain conditions that affect all variables, making them change in the same direction despite the absence of a real relationship among them.

Therefore, most studies assume that the time series used in the analysis are stationary series. The stationarity of the time series means that the mean and variance of the series are stationary over time and that the covariance between two time periods depends on the time differences between those two periods (Lags) and not on the real value of time.

If the time series is stationary in its original form, then it is said to be integrated of the zero order I (0), and if it is nonstationary, the time differences must be taken for it until it becomes stationary. One important condition of the cointegration technique is to prove that the time series used in the cointegration model are integrated in the same order.

The nonstationary hypothesis in time series is usually tested by the Autocorrelation Function (ACF) test and the Unit Root Test, a test that examines the long-term statistical properties of variables.

One of the most important tests examining the unit root is the Augmented Dickey-Fuller (ADF) test (Engle and Granger 1987), which suggests the following equations:

With constant and without trend Δ*Yt* = *α* + *δYt* − 1 + *μ<sup>t</sup>* With constant and with trend Δ*Yt* = *α* + *α*1*T* + *δYt* − 1 + *μ<sup>t</sup>* Where *Yt* = Index Time Series and = Linear Trend.

The null hypothesis of both ADF tests is that a series is nonstationary; hence, rejection of the unit root hypothesis is necessary to support stationarity.

#### 5.3.2. Granger's Causality Test

Before discussing the causal relationship test for time series, it is worth mentioning that this test shows the short-term relationship, unlike the cointegration relationship test, which shows the long-term relationship. It is also noteworthy that these series are stationary and integrated at the same level. (Granger 1986).

The mathematical equation to measure Granger causality depends on the linear regression model used in the prediction method. Granger's study (Granger 1969) is one of the most important studies demonstrating the concept of causality among variables.

To measure the causal relationship in the short term among the indicators, the following equations were used:

$$\begin{aligned} Y\_t &= \sum\_{i=1}^n \alpha\_i Y\_{t-i} + \sum\_{j=1}^n \beta\_j X\_{t-j} + \mu\_{1t} \\ X\_t &= \sum\_{i=1}^n \lambda\_i Y\_{t-i} + \sum\_{j=1}^n \sigma\_j X\_{t-j} + \mu\_{2t} \end{aligned}$$

The two equations postulate that the current values, *Yt* and *Xt*, are related to their past values, and vice versa. Unidirectional causality from *Xt* and *Yt* is indicated if the estimated coefficients on the lagged *Xt* are statistically different from zero as a group (i.e., Σ*β* = 0) and if the set of estimated coefficients on the lagged *Yt* are not statistically different from zero if Σ*λ* = 0. The converse is also the case for unidirectional causality from *Yt* to *Xt*. Feedback or bilateral causality exists when the sets of *Xt* and *Yt* coefficients are statistically different from zero in the above two regressions (Gujarati and Porter 2009).

#### 5.3.3. Multiple Regression

To test the efficiency of the market index in reflecting the changes that occur in the market, a multiple linear regression (MLR) test was employed, using the ordinary least squares (OLS) method.

#### **6. Results and Discussion**

#### *6.1. Descriptive Statistics*

Table 2 reveals the descriptive analysis of the indicators, which shows average mean returns, median, standard deviations, skewness, kurtosis, etc. The normal distribution of index returns was also tested through Jarque–Bera test. The table also shows that the highest average return is the sector index return for Media and Entertainment. Moreover, it can be seen through the table that the least-risk index, as measured by standard deviation, was TASI (0.009981), which means that the market's portfolio is well diversified compared to other sectors. When measuring normality of the indices in question, it was found that all indicators had normal distributions, as indicated by the test results. It is clear, too, that the probability values obtained for all return series in Jarque–Bera test were statistically significant at 5 percentage level of significance (*p*-values 0.00 < 0.05).


**Table 2.** Descriptive statistics.

#### *6.2. Results of Stationarity Test (Unit Root Test)*

Table 3 shows the results of the stationarity test of the indices returns after applying the Augmented Dickey-Fuller test. The results indicate that all indices were stationary with the presence of the constant in the equation and with the presence of the constant and linear trends. Consequently, it can be said that the indices are individually integrated to order zero (I (0)), which means that any of sectoral indices can be used to predict another. Thus, it is possible to reject the null Hypothesis H1, which states that the TASI returns are nonstationary.


**Table 3.** Result of stationarity test.

\* Significant at 1%.

#### *6.3. Results of Causality Test*

Table 4 indicates that, when applying the Granger causality test, it can be concluded that the general Saudi market index (TASI) affects 10 sectoral indices out of the 19 indices of the study sample. This indicates that the TASI return can be used to predict the returns of the sectoral indices that move along with the movements of the TASI, and those indices are: energy, materials, commercial and Professional svc, consumer durables and apparel, retailing, food and staples retailing, food and beverages, pharma, biotech and life science, insurance, and banks. It can also be noted that there are only two indices whose movements affect TASI, transportation and pharma, biotech and life science. The only index that both affects and is affected by the general index is pharma, biotech and life science. Those results indicate that the null Hypothesis H3 can be rejected, which states that TASI returns do not cause sectoral returns. These results also indicate that the null Hypothesis H4 can be rejected, which states that sectoral index returns do not cause TASI returns.

Table 5 indicates that the most influential index among the rest of the sectoral indices was the transportation sector index, which affects 17 sectors, while the least influential index was the media and entertainment sector index, which affects one sector. Moreover, Table 5 shows that the most affected index is the insurance sector index, which is affected by 16 sectors, and the least affected index is the services sector index, which is affected by only one sector.

These results indicate that we can use some series of index returns to predict the return values of other indices, which supports the assumption that there are investment risks in the financial portfolios and reduces the importance of diversifying financial portfolios.


**Table 4.** Granger causality test between TASI and sectoral indices.

causal relationship.


Means that it is significant at 1%, \*\* means it is significant at 5%, \*\*\* means it is significant at 10%.

**Table 5.** Granger causality test between sectoral indices.

#### *6.4. Results of Regression Analysis*

Table 6 shows the results of multiple linear regression test of TASI on sectoral indices. The results indicate the existence of a statistically significant effect of sectoral indices on TASI, and the model explains 99.75% of the changes in the market value. It was found through the results that all sectoral indices have a positive and statistically significant impact on TASI. This enhances the strength of the model for its adoption as an illustrative model for changes in TASI and indicates that TASI is an index that reflects all changes in TADAWUL.

**Table 6.** Results of multiple linear regression.


#### **7. Conclusions**

This study was based on testing the causal relationship between the TASI index of the Saudi stock market and determining whether the TASI is an effective tool for expressing the changes taking place in the market and in market sectors. This study attempted to draw conclusions that may help investors in the Saudi stock market (Tadawul) make efficient decisions about diversifying their investments and support their decisions about diversification efficiency. The Augmented Dickey- Fuller test (ADF) and the Granger causality test were used to test these relationships. The results showed that there is stationarity in these chains. The results also showed that there are bidirectional and unidirectional causal relationships between TASI and the sectoral indices. These results indicate the existence of short-term relationships among the indices comprising the Saudi stock market, and these results also show that there is no benefit of diversification. Moreover, the results of the regression analysis revealed that it is possible to rely on TASI to show the movement of Tadawul. These results confirm that the general index of the Saudi market (TASI) is a clear mirror of the changes taking place in sectors in the market. The results of the study were consistent with past findings (Rajamohan and Muthukamu 2014; Vardhan et al. 2015; Mustafa 2012; Ahmed 2016; Arbrlaez et al. 2001; Aravind 2017; Sharabati et al. 2013). We hope that researchers in the future will apply these techniques to studies of similar markets and compare the results of this study with the results from markets that operate in similar environment and conditions.

**Author Contributions:** Conceptualization, F.A.; methodology, F.A.; software, F.A.; validation, F.A.; formal analysis, F.A.; investigation, F.A., H.M. and S.A.; resources, F.A., H.M. and S.A.; data curation, S.A.; writing—original draft preparation, F.A., H.M.; writing—review and editing, F.A.; visualization, F.A.; supervision, F.A.; project administration, F.A.; funding acquisition, self funding from F.A., H.M. and S.A. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Data Availability Statement:** The data are available from the authors upon request.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

