Trivariate Modelling of the Nexus Between Financial Development, Globalization and Economic Growth: Insight from African Countries

: Financial development has been identified as main drivers of economic growth. However, empirical probe of this nexus remains inconclusiveness due use of an inappropriate proxy by previous studies, and the inability of previous studies to consider globalization in this nexus. To this end, we probe the finance-growth nexus in the presence of globalization by applying the Pooled Mean Group (PMG) estimator to a sample of 21 countries spanning 1990–2017. The empirical results affirm the supply-leading hypothesis which indicates that financial development spur economic growth. In addition, our estimate provides evidence of a positive linear relationship between globalization and economic growth. Further, results indicate that physical capital investment plays an important role in accelerating economic performance of African economies. Based on these findings, it is important for African countries to promote globalization-financial development policies in order to have access to alternative sources of external financing and attract foreign investment that can spur growth of African countries.


Introduction
The objective of every economy is to achieve sustainable economic development because economic progress is essential for poverty reduction and infrastructure development in the long run (Shahbaz, Nasir, Hille and Mahalik, 2020). Financial development increases a country's resilience and boosts economic growth. It mobilizes savings, promotes information sharing, improves resource allocation, and facilitates diversification and management of risk. It also promotes financial stability to the extent that deep and liquid financial systems with diverse instruments help dampen the impact of shocks (Sahay et al., 2015). Thus, a well-developed financial system attracts foreign direct investment, reducing the costs of acquiring information, enforcing contracts, and augments growth (Levine, 2005 andAng, 2008a). Theoretically, relationship between finance and economic growth has been a controversial issue as it has led to the establishment of three positions, which are the supply-leading hypothesis, the demand-following hypothesis, and the neutrality hypothesis. The first view is the supplyleading hypothesis also known as finance-led hypothesis suggests that financial development causes industrial productivity; hence, economic growth. The view supported by Bagehot (1873), Schumpeter (1911), Gurley and Shaw (1955), Goldsmith (1969), McKinnon (1973), King and Levine (1993), Levine et al. (2000), and Christopoulos and Tsionas (2004), Shan and Jianhong, (2006), Odhiambo (2008) argues that the supply of financial services creates the impetus for enterprises to demand for them which ultimately causes growth. The implication of this view point is that policies that move toward the development of financial systems facilitate economic growth. On the other hand, the demandfollowing hypothesis assert enterprise leads, finance follows suggesting that it is growth which creates the demand for financial services and not vice versa. The third position, the neutrality hypothesis implies that there is no causation between financial development and economic growth and finance is seen as an over-stressed determinant of economic growth. Developing countries across the globe face a myriad of problems ranging from poor governance, poor programme implementation, and corruption, just to mention a few (Adelowokan, 2012). African countries exhibited the weakest economic performance relative to other regions of the world. For instance, in 2010 the average per capita GDP in Africa is US$ 1,669, which is far below the lower middle income groups' average of US$ 2,530.5 (Abdulsalam, Salina and Mohammed, 2015).
Globalization is the process through which corporations, governments, and other organizations around the world increasingly interact (Zaidi, Zafar, Shahbaz & Hou, 2019). It represents a set of economic, political, and cultural processes that manifest in increased interdependence among nations (Goldberg& Pavcnik, 2007;Mills, 2009;and Sethi, Chakrabarti and Bhattacharjee, 2020). Globalization plays an important role on economic development by providing competitive environment via opening access to foreign investors in local financial markets, strengthens the quality of financial institutions and enhanced integration of economies by virtue of trade openness and financial flows (Mishkin, 2009;Kandil et al. 2015 andAtil et al., 2020). Rousseau and Sylla (2003) also reported that nations should remain open to rest of the world, as innovative products, investment, and new ideas will come to host countries, generating wealth and raising per-capita income.
A plethora of studies with mixed findings on the nexus between financial development and economic growth are available in the existing literature (Adeniyi, Oyinlola, Omisakin and Egwaikhide, 2015;Samargandi, Fidrmuc, and Ghosh, 2015;Kenza & Eddine, 2016;Muazu and Ibrahim, 2018;Botev and Jawadi, 2019;Alimi and Adediran, 2020;Wu et al 2020;and Redmond and Nasir, 2020). The present study contends that the divergent findings recorded could be because the authors did not account for globalization in their regressions. In addition, existing empirical studies have used a single indicator to examine the finance-growth nexus. Given the complexity of services provided by the financial system, capturing financial development with a single indicator could lead to potential bias and mislead the findings. To this end, this study employs financial development index of World Bank Global Financial Development index and KOF index that considers economic, social and political aspects of globalization to probe the link among financial development, globalization and economic growth in African economies. Given the motivation around the attainment of economic development, there is need to evaluate the distinctive role of financial development and globalization in African economies to help project the trend of economic performance. This empirical exercise is important to help reveal the source of economic development in African economies and guide finance and growth policies appropriately.
The rest of the paper is structured as follows. The next section takes a look at the empirical facts on globalization in finance-growth nexus. Section 3 present the methodology while section 4 focuses on result and discussion of finding and section 5 concludes the paper.

Empirical Facts on globalization, financial development and economic growth
The link between financial development, globalization and economic growth has received great interest by scholars and policy makers during the past few years. Among these is the study of Ahmed, Zhang and Cary (2021) Wu et al (2020) validates the supply-leading and demand-following hypotheses for Japan and India but support the supply-leading theory in China. Samargandi, Fidrmuc, and Ghosh (2015) analyzed the link between financial development and economic growth in a sample of 52 middle-income countries over the 1980 -2008 period. They find an inverted U-shaped relationship between finance and growth in the long run while the short run relationship was insignificant. For 11 MENA countries over the period of 1980 and 2012, Kenza & Eddine (2016) established that financial intermediary has a negative effect on the growth rate in the MENA countries both in the short and long run.
In Nigeria, Adeniyi, Oyinlola, Omisakin and Egwaikhide (2015) employed the threshold modelling to examine the relationship between financial development and economic growth in Nigeria using data covering the period 1960 -2010. They discovered that financial development impedes growth. Similarly, in investigating the finance-growth nexus for a sample of 100 developed and developing countries, Botev and Jawadi (2019) shows that finance stimulate economic growth in developed countries but exact an insignificant influence in close economies. For SSA countries, Brueckner and Lederman (2015) probed the linkage between trade openness and growth. They report that greater openness to international trade augment economic growth in both short-run and long run. In the same region, Muazu and Ibrahim (2018) finds that financial development hinders SSA growth when financial sector development is not complemented by real sector growth. Examining 30 countries spanning 1990 and 2016, Redmond and Nasir (2020) incorporated natural resources, trade openness, institutional quality and international trade into finance-growth nexus. Their empirical finding reveals that natural resource abundance and trade openness promotes growth while international trade and financial development inhibits growth.
Recently, Atil et al (2020) considered the link between natural resources and financial development through globalization, oil price and economic growth in Pakistan spanning 1972 and 2017. The outcome of their study suggests that natural resources, economic growth and oil price aid financial development while globalization dampen financial development. By employing ARDL, FMOLS and DOLS estimation technique, Guan, Kirikkaleli, Bibi and Zhang (2020) also investigated the same nexus in the presence of human capital in China. Their empirical result validates the demand-following and resources curse hypotheses which suggests that natural resources impedes financial development whereas globalization and human capital foster financial development in China. Osei and Kim (2020) probed the contribution of foreign direct investment in finance-growth nexus for 62 middle-and high-income countries between 1987 and 2016. Their panel threshold methodology results suggested foreign direct investment improves economic growth but retards economic growth whenever financial development exceeds the threshold of 95%. For 40 African countries, Ibrahim and Sare (2018) explored the determinants of financial development and established that trade openness and human capital stimulates Africa's financial development.
Using a sample of 23 European countries spanning 1989 and 2016, Nasreen, Mahalik, Shahbaz and Abbas (2020) employed GMM estimation to analyze the linkage between Financial Globalization, Institutions, Economic Growth and financial sector development. Their empirical finding indicates that financial globalization inhibits financial development while institutions support financial inflows and augment financial development in European countries. In a related study, Nasreen et al. (2015) explore the linkages between financial development, institutions, globalization and economic growth. The finding of the PVAR technique reveals that globalization promote financial development through financial reforms and thereby spur economic growth. With the aid of Westerlund panel cointegration test and annual data from 1984 to 2008, Law et al. (2015) assessed the nexus between globalization, institutional reforms and financial development in East Asian economies. The study discovered that globalization strengthen financial development via institutional reforms. However, in India, Shahbaz et al. (2018a) estimated the link between globalization, institutional quality, economic growth and financial development. The result of the study reveals that globalization and institutional quality have detrimental effects on financial development whereas economic growth accelerates financial development.
By applying panel cointegration and causality methodologies for 32 developing and developed countries, Kandil et al. (2015) examined the role of globalization (proxied by globalization index) on financial development. The empirical result of the study suggests that globalization hinders financial development while financial development spurs globalization by relaxing the of constraints of external financing. Further, Kandil et al. (2017) probe the influence of globalization in the finance-growth nexus in China and India. The outcome of their study shows that globalization spurs economic growth in India but hampers growth in China. Concentrating on 3 economic blocs (BRCIS, MINTS and ECOWAS), Muye and Muye (2017) investigated the relationship between financial development, institutional quality and globalization. Their empirical result indicates that globalization promotes financial and quality institution boosts globalization-financial development nexus. The outcome of the study of Balcilar et al. (2019) indicates that economic globalization stimulates by improving the effectiveness of financial institutions in 36 developed and developing countries. In investigating the association between environmental degradation, globalization, financial development, economic growth and energy consumption in India over the period of 1980 and 2015, Sethi, Chakrabarti and Bhattacharjee (2020) employed ARDL technique and finds that globalization, economic growth increased energy consumption and degrades the environment. Likewise, Shahbaz, et al. (2020) employs Bootstrapping ARDL methodology to examine the link among economic growth, R&D expenditures, financial development, and energy consumption and carbon dioxide emissions in UK. According to their empirical finding, energy consumption and financial development degrades the environment while R&D expenditures reduce CO2 emissions. By incorporating corruption into finance-growth linkage in 142 countries spanning 2002 and 2016, Song, Chang and Gong (2020) established that economic growth strengthens financial development in both developed and developing countries while corruption worsen financial development in developing countries.
In sum, the above literature suggests a diverse finding on the nexus between financial development and growth and its determinants across countries. To the best of our knowledge, no research work has been carried out on African countries using comprehensive globalization and financial development index, making it a motivation.

Data and Methodology
For this study, we analyze annual data from 1990 to 2018 for 25 African economies to probe the finance-growth nexus in the presence of globalization. The justification for the selected countries and time frame is based on availability of data. In order to simplify coefficient interpretations and overcome the problem of heteroscedasticity, all the dataset is transformed to natural logarithm (Faisal, Sulaiman and Tursoy, 2019). The data employed in this study are economic growth (proxied by GDP per capita measured at constant A2010 US Dollars), Following Dreher (2006) and Gygli et al. (2019), globalization is measure as economic, social and political globalization. In line with Sahay et al (2015) and Nasreen et al., (2020), financial development is measured financial access (proxied by commercial bank branches per 100,000 adults and ATMS per 100,000 adults), financial depth (proxied by private-sector credit to GDP, stock market capitalization to GDP, stock traded to GDP and liquid liability to GDP) and efficiency (proxied by stock market turnover ratio). Data on economic growth is sourced from World Bank Indicator 2020 edition while data on globalization is collected from the KOF index of globalization of Dreher (2006) and Gygli et al. (2019). Lastly, data on financial development is collected from World Bank Global Financial Development Database (GFDD). The selected countries for this study are: Angola, Algeria, Benin, Botswana, Cameroon, Cote d'lvoire, Egypt, Ethiopia, Gabon, Ghana, Kenya, Morocco, Mali, Mauritania, Mauritius, Namibia, Nigeria, Senegal, Seychelles, Tunisia and South Africa.

Econometric Strategy
Building on the previous empirical studies, the general form of the finance demand function is modelled as: ( Where Y is economic growth, FIN denotes financial development and GLOB is globalization. Expressing equation (3.1) in econometric form becomes: where Y denote economic growth, � denotes country -specific intercept, FIN is financial development and GLOB is globalization while i denotes the country, t is the time period, error term is symbolized by µ it and Z it denote the control variable (macroeconomic uncertainty (proxied by CPI-based inflation) and domestic investment (proxied by gross capital formation as a ratio of GDP)) because they have been empirically proven to have vital impact on economic growth (Osei and Kim (2020). In order to probe the presence of globalization in finance-growth nexus, panel autoregressive distributive lag model (PARDL) approach developed by Pesaran et al. (1999) using the autoregressive distributed lag ARDL(p,q) method, where p is the lag of the dependent variable, and q is the lag of the independent variables. This estimation technique has several merits over static models such as OLS, fixed and random effects and GMM models. The PARDL framework considered the heterogeneity of the dynamic panel setting, the short-run dynamic and the long run equilibrium of the model (Rafindadi, Muye & Kaita, 2018;Samargandi, et al. 2015). Second, this method is consistent in the face of I (0) and/or I(1) variables or the mixture (Rafindadi, Muye & Kaita, 2018); and also, it yields consistent estimates in the presence of endogeneity (Chen & Vujic, 2016;Chudik et al., 2013;Pesaran et al., 1999;Pesaran & Smith, 1995). Lastly, the short-run and long-run models are estimated simultaneously (Pesaran et al., 1999). Thus, this technique gets rid of problem of slope heterogeneity across countries, order of integration in variables and cross-sectional dependence (Rafindadi, Muye & Kaita, 2018;Chen & Vujic, 2016).
The mean group (MG) estimator introduced by Pesaran and Smith (1995) estimates the long-run parameters by taking an average of the long-run coefficients of each cross-section. The MG assumes heterogeneity in all coefficients (both short-run and long-run and the intercepts) across units. The dynamic fixed effect (DFE) estimator evolves from the fixed effects estimator, with the lagged term of the dependent variable incorporated as one of the independent variables. The DFE estimator assumes homogeneity in all coefficients (both short-run and long-run) across units except the intercepts. The pooled mean group (PMG) estimator proposed by Pesaran, Shin and Smith (1999) is an intermediate estimator between DFE and MG. The PMG allows only the long-run slope coefficients to be homogeneous. The appropriateness of these estimators is determined using the Hausman test.
The general form of panel autoregressive distributed lag (ARDL) (p, q) is model as follows: Where t symbolize time element (annual), country index is represented by i and � i denotes country fixed effects. In addition, k time lags. In equation (3.4), the dependent variable is denoted by Y it ; the vector X it represents financial development, globalization and control variables (investment and macroeconomic uncertainty). Equation (3.3) can be re-parameterized into: Equation [3.7] in Error correction form is specified as: is the adjustment in Y it to the deviation from its long-run relationship and short-run coefficients linking Y it with its lag values and other independent variables are * ij � and * ij � . In addition, i � is the error-correction coefficient estimates which measures the speed of adjustment of Y it toward its long-run equilibrium in case of a change in any of the independent variables X it . In order to ensure that long run relationship exist among the variables, the speed of adjustment � i must be negative and significant.

Result and Discussion
This section focuses on result presentation and discussed in stages, starting with the preliminary analysis, after which the PARDL result is discussed. We carried out descriptive statistics and correlation matrix in order to provide firsthand descriptions of the relevant data employed in this study. Volatility in economic growth is higher than volatility stems in financial development and globalization index. The correlation matrix shows a positive association between growth and urbanization, which is not surprising for a heavily income per capita and financial development and globalization index. Similarly, we observe that growth and inflation are negatively related. It is important to check for stationary and nonstationary properties of variables because nonstationary variables invalidated the assumptions of a regression analysis and could result in a spurious regression (Roquez-Diaz and Escot, 2018). In order to check the unit root property of the variables employed in this study, first generation panel unit root tests such as Levin, Lin and Chu (LLC), Im, Pesaran and Shin (IPS) and Breitung panel unit root tests are utilized. The null hypothesis for these tests is the panel data has unit root while the alternative hypothesis is that the panel data has no unit root. The empirical result of the panel unit root test is presented in Table 2. We observed that gross domestic product per capita (LY), globalization index (LGLOB) and investment (LINV) contain unit root while financial development index and macroeconomic uncertainty are found to be stationary at level. After the first difference, all the variables are stationary. This shows that the variables have a mixed order of integration. According to Pesaran et al. 1995Pesaran et al. , 1999, panel ARDL is applicable if variables contain a mixed order of integration, based on this, we proceed to evaluate the nexus between financial development and economic growth in the presence of globalization. We employ cross dependence test developed by Pesaran (2004) test to check for cross-sectional dependence among the variables employed for this study. We apply CD test because African countries tend to exhibit similar traits and are interconnected via trade, globalization due economic union and other networks. The result of the Pesaran CD (2004) cross dependency test is displayed in the first column of Table 3 and the result indicates that all variables considered in this study have cross-sectional bias problem. Since the performance of traditional unit root tests such as LLC, IPS, Breitung tests can be influence by the presence of cross dependence (Nathaniel et al. 2020;and Ahmed and Le, 2020). Thus, in order to overcome this problem, we employed Cross-Sectional Augmented Panel Unit Root (CIPS) and Cross-Sectional Augmented Dickey-Fuller (CADF) panel Unit root test of Pesaran (2007) which gives reliable and consistent result. The result of the CIPS and PESCADF panel unit root presented in Table 4 suggests that all the variables are stationary at level except economic growth (LY). After first difference, economic growth becomes stationary.  Table 4 report the result of the panel ARDL (pooled mean groug, mean group and dynamic fixed effect) on the nexus between financial development, globalization and economic growth among 21 African economies spanning 1990 and 2017. In order to select the best estimator among PMG, MG and DFE, we subject the results from these estimators to Hausman test. A nonrejection of the null hypothesis implies the adoption of the PMG estimator while the rejection indicates the adoption of the MG estimator. In other words, the PMG estimator is the efficient estimator under the null while the MG estimator is the efficient estimator under the alternative hypothesis. The result of the Hausman test support the PMG estimator as the efficient estimator, then the use of a MG or DFE estimator is unsuitable. Thus, the interpretation of this study will be based on PMG estimator. The result of the PMG estimator indicates that financial development exerts a positive and significant impact on economic growth in both short and long run, that is, a 1% increase in financial development foster African economies by 2.451 and 0.302 percent in both long run and short run respectively. This finding validates the result of Botev and Jawadi (2019) that finance stimulate economic growth in developed countries but refutes the result of Muazu and Ibrahim (2018) and Redmond and Nasir (2020) whose study finds that financial development hinders SSA growth. The result supports the supply-leading hypothesis which suggests that well-developed financial sector that minimizes transaction and monitoring costs and asymmetric information is a prerequisite to the process of economic growth. Furthermore, globalization has a positive and significant influence on economic growth in the long run though insignificant in the short run. Our results show that 1% of globalization resulted in an increase of 0.986% of economic growth in the long term and 0.0975% in the short-term, and this finding is aligned with Brueckner and Lederman (2015). This finding suggests that interacting African economies with the rest of the world provides competitive environment by opening access to foreign investors in local financial markets and helps firms to attract more external financing, competition, and investment opportunities; promotes economies of scale; and improves production processes which paves the way for economic development. The coefficient of physical capital investment is positive and significant in both short and long run, thus confirming the important role played by physical capital in African's growth. However, the coefficient of macroeconomic uncertainty (proxied by consumer price index) is negative and significant in the long run but insignificant in the short run. This implies that macroeconomic uncertainty impedes African economies. Lastly, a significant negative coefficient of the error correction is necessary to substantiate a statistically significant relationship over the long run. The results of empirical testing affirm a short-and long-run relationship between financial development, globalization and economic growth among African countries in OPEC countries. Table 4 suggests that error correction coefficient (ECM) is -0.0985, which is significant at the 1% level, implying that a 9.85% adjustment takes place after temporal shocks in the long-run equilibrium relationships between economic growth, financial development and globalization.

Conclusion
This study aims to investigate the link between growth, financial development and globalization using a multivariate framework on panel data sets of 21 African countries from 1990 to 2017. The findings of this study are vital for understanding the relationship among financial development, globalization and economic growth in African countries. Particularly, our findings suggest that financial development and globalization plays a significant role in promoting economic growth of African countries. Further, results indicate that physical capital investment has an important role in accelerating economic performance of African economies. Based on these findings, it is important for African countries to promote globalization-financial development policies in order to have access to alternative sources of external financing and attract foreign investment that can spur growth of African countries.