Nexus between financial integration, capital market development and economic performance: Does institutional structure matters?

Abstract

This study investigates the role of institutional structure on asymmetries dynamic impact of financial integration, capital market development on economic performance in Sub-Saharan Africa (SSA). The study classified economic performance into RGDPC, nominal gdp and human capital development, and employed (PNARDL) modeling framework, and a panel of 16 nations of SSA over the period 1996–2019. The finding of this research output can be summarize as thus: i.) In the long run, a rise in positive shock to the financial integration index leads to a rise in RGDPC, while a negative shock to FI leads to a fall in RGDPC. ii.) Both shocks (positive and negative) to MCAP reduce RGDPC. Institutional quality index (INSQI) is revealed to have a positive and significant impact on RGDPC in the long run and indeed intensify their asymmetries. iii.) Both shocks to FI exert inverse influence on nominal GDPC, while positive and negative shocks to MCAP exert a positive and negative influence on GDPC, respectively. INSQI also affects GDPC negatively and significantly and indeed reduces their asymmetries. iv.) Positive and negative shocks to FI reduce HCD as well as shocks to MCAP. INSQI, on the other hand, adds to HCD and intensifies their asymmetries. However, the lack of consistency in the results across the models suggest that the interplay between these variables are still undeveloped relative to other continents of the world, and the benefits are yet to be adequately harnessed.

1. Introduction

Financial integration has created a pathway for market co-movement, which is the rationale for the tremendous economic growth in the transformation emanting from globalization. This transformation could be attributed to market efficiency and productivity orchestrated by the flow of information [1]. Indeed, financial integration is considered the only platform for creating stronger ties with more established financial systems including the capital markets. From a theoretical perspective, financial integration entails removing obstacles to cross-country investment, treating domestic and foreign investors equally, harmonizing rules and regulations guiding the institution and its operation [2,3]. A good example of scholars supporting these views are Gurley and Shaw [4] and Goldsmith [5] from the Schumpeterian school of thought; both advocated for a free-regulated financial system, which includes expanding the domestic market to the global market.

Contrary to the above, followers of Keynesian researchers such as [6,7] suggested that government’s interest rate regulation and huge reserves could lead to decreased savings, less capital accumulation, and inefficient financial allocation. Vo, Vo, and Ho [8] and Arif-Ur-Rahman and Inaba [9] anchored on this study and developed a framework to examine China and African’s financial integration and economic growth, respectively, and found a linear association between financial integration, TFP, and growth. However, in Africa, the relationship between capital market development, financial integration, and economic development is still young due to the region's developmental constraints (path). There are two sides of the coin in this case; the first suggests that increased levels of financial integration could lead to an adverse effect on the economy alongside other potential costs of financial integration, including; countries with larger markets attract more funds, loss of macroeconomic stability and pertinent risk associated with high penetration of foreign financial institution. Some analysts and academia posit that Africa was immune to the global crisis due to its lowest ebb of integration across borders, especially financial markets. The second side of the coin points to the ameliorating effects of regional financial integration in accelerating economic growth and development across Africa through Pan-African banking, AfCFTA, and a Pan-African settlement system (see Refs. [2,3,10].

Driving the discussion further to continent-specific statistics, as of 2021, the Africa Regional Integration Index depicts a weak macroeconomic integration for the eight recognized regional economic communities (EAC, AMU, ECCAS, IGAD, ECOWAS, CEN-SAD, COMESA, and SADC) pegged at 0.399 from the 2019 score of 0.327 [10,11]. Despite the sublime efforts towards financial integration in Africa engineered by Pan-African banks and other financial institutions alongside agreements for regional financial cooperation (East African Community), financial market activities remain shallow, with particular reference to low capitalization, low liquidity, the short-term structure of instruments and a limited number of financial instruments (see Refs. [2,3,12].

The key question abound: why are the SSA nations failed to integrate financially over the year? To address this pertinent question, this paper makes an important contribution to repositories by identifying if the effect of institutional structure differs on the rising and falling of financial integration and capital market on economic performance. Interestingly, this is necessary for policy path in the region given the rising or falling in the capital flow, financial integration, and trade, especially in the aftermath of the COVID pandemic, given that financial integration shows the low-level capital market in a globalized village, depicting free-regulated financial system [8]. The differential effect of financial integration on economic growth and performance across the region is deemed a reflection of the role of governance structure towards the rising and falling effects of financial integration. This is particularly so because the recent trend has shown that theories -Schumpeterian views and Keynesian school do not support a clear direction of the influence of financial integration on economic growth and performance in Africa, and as such, attention have shifted to the contribution of governance or institution on growth and performance, financial integration nexus in Africa [13–14].

On the empirical front, studies have argued that if financial integration is properly aligned, it will attract a lower cost of capital and domestic savings via efficient resource allocation, risk diversification, and enhancing the local financial system [15]. Several empirical narrations have provided varying evidence on the dynamic influence of capital market and financial integration on economic growth. The first group argued that financial integration affect economic growth positively or directly (see Refs. [16,17]; Bonfiglioli, 2010; among others) but could not provide information regarding the co-movement of trade and as such, negate the international business cycle model which affirms the relationship between financial integration and economic development. Thus, our study provides another crucial contribution by demonstrating financial integration's rising and falling effects on economic performance. Obstfeld and Badri [18] and Sheshgelani (2016) argued that financial integration and economic growth are positively related only when the financial system, institutions, and macroeconomic policies are in good condition and well-designed. However, they could not provide any information on the unequal effect of financial integration on economic growth and development in Africa because of the region's varied characteristics. Meanwhile, recent repositories have shifted attention to examining the asymmetric relationship between a financial surge in capital inflows having strong currency appreciation pressures, asset price bubbles, or rapid credit growth that induce fragilities in the financial sector [19]. As a result, the impact of financial integration on economic development could generate differential effects.

Considering the above empirical shortcomings, it becomes critical to account for the empirical expositions on the relationship between the variation in the financial integration and capital market (rising and falling) and their respective effects on economic performance in SSA. Extant studies are submerged with a symmetric and linear model in estimation. Therefore, this study intends to contribute to the knowledge repository stated as thus; (i) ascertain the impacts of the asymmetric effects (rising or falling) of financial integration and capital market development on economic performance in the selected SSA. There is a dearth of empirical narration on the asymmetric condition relating to financial integration, capital development, and economic performance in SSA. Few studies in this area of scholarship only considered a symmetric relationship between financial integration and economic growth. This includes [8,13,15,16]. Undoubtedly, financial integration and the capital market in SSA have experienced a boom-and-bust cycle in the past four decades under the aegis of financial reformation targeting trade, economic liberalization, and structural adjustment programs [20,21]. Existing studies may have neglected the differential effect of the boom-and-bust cycle in the case of financial integration and capital market on economic performance. (ii) establish the role effect of institutional structure on the interaction of financial integration, capital development, and economic performance in SSA. Furthermore, existing studies may have imposed many restrictions on the choice of variable, especially on the economic performance, while analyzing the relationship between financial integration, capital development, institutional structure, and economic performance, as these changes observed in the policy consideration may have impacted asymmetrically on the causal relationship. Therefore, this study decomposed financial performance into real gross domestic capita, nominal GDP, and human capital development to explain long-aged empirical debate in the area of study. iii.) On the empirical front, this study intends to prove that the traditional linear and symmetric assumption in economic relationships is prone to bias and unrealistic estimates, and such could lead to spurious conclusions. Therefore, the concept of asymmetric upholds when the positive and negative changes in the independent variables on dependent variables are significantly at variance.

In this study, if the effects of positive and negative changes in the financial integration and capital market on economic development (real GDP and human capital development) are significantly at variance, the association is called asymmetric relationship. On the other side of the coin, if the effect of a unit change (rise/fall) in the financial integration and capital market on economic development are similar, we conclude on the presence of symmetric assumption; as rightly pointed out by Liddle and Huntington [22] that such equal response from the dependent variable to the independent variable is far from realities. It implies that, empirically, such upward or downward may generate varying effects on the dependent variable and can only be identified through an asymmetric method.

We hypothesize that positive and negative financial integration and capital market developments have a significant impact on economic outcomes, and thus we use the Shin and Greenwood-Nimmo [23] Panel nonlinear autoregressive distributed lag (PNARDL) technique and incorporate the key independent variables. It is an extension of PARDL, It captures simultaneously the effects of positive and negative partial changes of the independent variables on the dependent variable; its framework is dynamic and can account for both long- and short-run partial changes. It is superior to other techniques and has been carefully implemented (see, [24,25]). The above narration opens the study. Section 2 is a literature review. Section 3 examines estimation and data sources, followed by Section 4’s empirical analysis. Section 5 is a summary and conclusion.

2. Literature review

The theoretical orientation of this study is anchored on financial restrictions under the umbrella of two school of thought: Schumpeterian views, Keynesian school thought and McKinnon-Shaw school of thought. The Schumpeterian school of thought believed that an advanced financial state do stimulate growth and advancement through the concept of disruptive creation of financial systems and its services; driving saving–investment flow to work more effectively , see McKinnon [7] and Shaw [6] and among others. The implication is that financial integration through cross-border investments are indeed an avenue for return on investment, which is largely a factor of domestic institutional structure of the host country via the capital market. Another school of thought, Keynesian school which is the proponent of Schumpeterian views explains that efficient financial system can minimize governance when it comes to handling the economy through financial repression such as regulatory reserves, regulated rate of interest, and control of forex [26,27] could serve as plausible amplifier for economic growth by maintaining economic performance. However, McKinnon-Shaw school of thought believed that high cost of capital regulation and retained earnings could bring about low savings and fund collection and sub-optimality in resources and financial allocation [28,29], and proffers an interest rate free adjustment mechanism and market driven system.

2.1. Financial integration and economic growth

Research have shown that financial integration or openness can influence economic growth in both ways, but there is no consensus on how these relationships are influenced. Among the indirect effects of financial integration on growth, Obstfeld [30] cited productivity enhancement, specialization and efficient capital allocation as examples. Furthermore, financial openness has been suggested by Klein and Olivei [31] and Levine [32] as a way to increase economic growth through the transfer of financial systems from advanced nations. It has shown to exhibit a direct influence on GDP, while indirect impact studies have revealed that a direct linkage between financial integration and growth can only be realized if the financial system, institutions, and macroeconomic policies are all in good health [33]. Others who oppose financial integration posit that its benefits from indirect channel influence are insignificant and not traceable [34]. Bonfiglioli [17] studied financial integration and total productivity growth from 1975 to 1999 and found a positive association. In a similar study, Friedrich et al. [35] employed trade, governance indicators, credit to private sectors, and financial openness to evaluate integration-growth linkages in Europe. The study document that positive effects of financial integration are particularly evident in countries that are close to the EU nations, implying that political integration might improve financial integration benefits. Adopting the VECM technique on data from 1981 to 2011, Mahajan and Verma [16] found support for the favorable effect of financial openness on GDP. On the other hand, Badri and Sheshgelani [36] did a panel study on the association between CPS, financial integration, and GDP in 24 OIC states from 2005 to 2013. They observed that financial globalization and economic growth have inverse linkage. Rodrick and Subramanian [37] explored the financial globalization-economic growth nexus in emerging markets. They found that financial openness neither boosts nor reduces growth and, as such, assert that financial globalization's role in growth is mere speculation and not realistic.

Recent studies have emerged considering various proxies of financial integration and economic growth; for instance, Phutkaradze et al. [15] explored the association between financial globalization and GDP for the Republic of Georgia with data covering 1995 to 2016. Employing the OLS technique, they found an insignificant relationship in the financial-growth nexus. They thus conclude that positive exertion of financial integration depends mainly on the currency stability but will influence a nation negatively in times of currency fluctuations. Ezzeddine and Hammami [38] applied ECM to examine the impact of international financial integration on economic growth in Tunisia covering the period 1970–2012. In the short run, financial integration has an insignificant impact on economic growth, whereas, in the long run, it has a positive and significant effect on economic growth in Tunisia. Demystifying financial integration into its broad components, trade integration index and financial integration index, Orji et al. [39] deployed the instrumental variable (IV) regression econometric technique to decipher if financial integration influenced economic growth in ECOWAS to achieve sustainable economic performance. In their study, Nguyen et al. [40] employed three measures of financial integration: economic integration, including overall integration, financial integration, and trade integration, and to probe into their linkages with economic growth in Vietnam from 1986 to 2015. Based on the application of Autoregressive Distributed Lag (ARDL) and Granger causality test. A positive relationship was confirmed between financial integration and economic growth; further analysis revealed a mutual causal association between financial integration and economic growth. Tekin [41] argued that financial integration could have a heterogeneous effect on economic growth in the selected 52 economies from 2000 to 2019. Using the generalized method of moments and quantile regression, the study found that financial integration has a positive and significant impact on economic growth, and the effects vary between high- and low-income countries. In a similar study, Kurantin and Osei-Hwedie [42] employed digital (e-economy) to analyze the relationship between financial integration and economic performance. Using ordinary least squares (OLS), the result provides evidence of a positive relationship between financial integration and economic development, among others. Khalid and Marasco [43] probed the moderating role of financial integration on the FDI growth nexus in 134 developing countries from 1989 to 2017. Based on the application of the GMM technique, the result validated the moderating roles of financial integration on the relationship between the FDI-growth nexus. Further, it disclosed a positive and complementary role of financial integration on economic growth. Ahmad et al. [44] have highlighted the dynamic relationship between financial globalization, among other variables, and the economic growth of the G7 countries. The discussion revolves around the emergence of a sustainable environment and rising economic growth. Their analysis shows a positive relationship between financial globalization, among other variables, and economic growth.

2.2. Market capitalization and economic performance

Research on the link between market capitalization and growth is extensive. According to Flaviabarna and Mura [45]; asset market development has a beneficial influence on GDP by using the regression method to evaluate the linkage between market capitalization and the economic growth of Romania and found support for direct connectivity between them. In a similar study, Oprea and Stoica [46] utilized a different econometric method of ARDL to examine the influence of the capital markets' integration on economic growth in the EU nations from 2004 to 2016. They observed that asset market integration directly and significantly links GDP. Khetsi and Mongale [47] explored the influence of capital markets on growth with data covering the period 1971 to 2013 in South Africa. Employing the OLS technique, they found support for a direct association between economic growth and capital market measures, including market capitalization and the value of transactions. In recent studies, Alam and Hussein [48] utilized multiple regression to analyze the association between capital market GDP of OMAN with data spanning 1993– 2015 and realized a direct association between the dependent and explanatory variables. In like manner, Algaeed [49] examine the relationship between asset development and economic growth in Saudi Arabia with data spanning 1985–2018. Employing FMOLS and ARDL techniques, the study found that Capitalization and liquidity exert an inverse relationship with economic growth. At the same time, stock market variables add to economic growth.

The second group of scholars revealed a mixed association between market capitalization and growth which depends greatly on the measure of the capital market development used. In this line, Lenee and Oki [50] explored the effect of asset market growth on GDP for Mexico, Indonesia, Nigeria, and Turkey with data covering 2000 and 2012. Adopting different measures of economic growth such as gross domestic product, GDS and capital formation as the dependent variables while market capitalization and value of listed securities represent the capital market development, they used the panel least square method and established that number of listed securities exert inverse influence on GDP, while it exhibits direct relationship with both domestic savings and capital formation. In a similar study, Ailemen et al. [51] used the ASI (all-share index), MCAP, value of transactions and number of deals to measure capital market development and examine the linkage between security and growth; they found that all share index exert a negative influence on GDP whereas the other three measures exhibit positive impact on economic growth. Onuora [52] utilized data spanning 2001–2017 in the analysis of the association between capital market development and economic growth of Nigeria, and by adopting the OLS technique, the findings showed an insignificant linkage between capital market revenue and GDP while adequate security and capital market revenue were positively and significantly associated.

Furthermore, Acha and Akpan [53] used VAR and causality to investigate the association between stock market performance and economic growth in Nigeria from 1987 to 2014. The data used by Abina and Maria [54] covered the years 1985–2017 and were analyzed using Error correction and Granger Causality techniques. They observed a long run positive correlation between the two variables. One-way linkage was seen between GDP and market capitalization and the value of new issues. Okere and Ndubuisi [55] use the ARDL technique to examine the link between oil price, security market, and economic performance in Nigeria as an OPEC nation from 1981 to 2014. They find that crude oil price determines growth whereas stocks have a little effect. In a similar study, Bista [56] used the same technique in Nepal to explore the link between security market development and GDP from 1993 to 2014. Using the ARDL approach to measure stock market development by market capitalization, the study found that capital market drives economic growth in the short and long term, whereas inflation slows growth. More causal tests demonstrate a two-way causal linkage between the security development and GDP in Nepal.

2.3. Financial integration, capital market development, institutional quality and economic growth nexus

Examining the role institutions quality plays in the financial integration - economic growth nexus for SSA, Egbetunde and Akinlo [13] adopted a panel GMM technique on a data covering the period of 1980– 2010. Their findings show that financial openness exhibit negative and significant influence on economic growth while the two institutional quality indicators (rule of law and government effectiveness) have inverse effect on economic growth. It was concluded that the SSA did not benefit from financial globalization due to poor institutional structure. Extending the data to from 1980 to 2015 and using same econometric method, Egbetunde and Akinlo [57] still confirm an inverse linkage between financial openness and economic growth. However, when institutional quality is incorporated, they found that government effectiveness reduces the negative effect of financial integration on growth. Klein [14] tested for non-linear association between the responsiveness of economic growth to capital account integration and institutional quality using quadratic estimation for the case of 71 nations. The result support a non-linear (that is an inverted U-shape) linkage between the responsiveness of growth to capital account openness and institutional quality.

In a recent study, Assas et al. [58] investigated whether financial integration is a veritable tool for economic growth in GCC Countries using data from 1981 to 2019. The authors show that financial integration increases economic growth under stable institutional conditions and conclude that a structural shift to institutional quality, among others, can help achieve economic prosperity. Haruna and Bakar [59] in their study, tried to examine the impact of financial liberalization and institutional quality proxied by corruption on economic growth in 5 sub-Saharan African countries. Using Driscoll and Kraay standard errors under the pooled ordinary least squares framework, some interesting results were uncovered pertinent to policy direction. In the first instance, financial liberalization positively impacts economic growth in SSA. Again, the total effect of financial liberalization and corruption negatively influences economic growth. Kim et al. [60] investigated the heterogeneous effect of financial liberalization on the alleviation of income inequality for a cpanel distribution of developing and developed economies covering 1989–201. They focused on democratization (institution), the stock market, banking development, and their influence on human capital and growth. After undertaking a series of econometric analyses, they established that financial liberalization, institution, stock, and banking development exert a positive and significant impact on income inequality and human capital development.

3. Methodology

3.1. Measurement of variables

The study covered 16 nations of SSA.1 In this study, we adopt Gross Domestic Product at constant price (RGDP), gross domestic product at current price (GDPC), the justification shows that GDPC measures the economy's demand for financial intermediary services. It is anticipated that economic expansion will increase demand for financial intermediary services, Human capital development as the dependent variables to capture another aspect of economic performance. Also, the index of financial integration index, market capitalization and were used as the main explanatory variables. However, index of financial integration (comprising financial, economic and political globalization) and index of institutional quality (including Voice and Accountability, Political Stability and Absence of Violence/Terrorism, Government Effectiveness, Regulatory Quality, Rule of Law and Control of Corruption) were generated via principal component analysis. This is justified because in order to deal with multicollinearity, principal component analysis (PCA) is often used to collapse a large number of correlated variables into a smaller number of uncorrelated ones. While credit to private sector (CPS); the justification for incorporating this variable in the model is based on the fact that it is a barometer for the health of the financial system, evidence of banks' ability to convert mobilized deposits into household and firms credits, and, in keeping with the discussion of economic performance, a sign that firms and household can gain access to the capital they need to make investments that will boost economic output. Hence, a positive coefficient is expected. Trade openness (OPN) was adopted as control variable for the three models. The justification is that it indicates the degree to which an economy is open, it is commonly regarded as a key factor in the growth of economic performance in many countries. Thus, a positive coefficient is expected. Data employed covers the period of 1996–2019 due to availability. The variables for the study are measured in Table 1, below:

Table 1.

Measurement and description variable used in the study.

Symbols	Variable	Measurement	Data source	Previous recent studies that utilized the variable
RGDPC	Real Gross Domestic Product Per Capita	Measured in constant 2010 US$	WDI (2020) [61]	Onuora [52]: Acha & Akpan [53]: Algaeed [49]
GDPC	Nominal GDP	Measured current 2010 US$	WDI (2020) [61]
HCD	Human Capital Development	Measured (0– 1). A combination of 4 items.	http://www.theglobaleconomy.com/
FI	Financial Integration Index	Financial integration index computed from three components of financial index such as (financial globalization, political globalization and economic globalization)	WDI (2020) [61]	Egbetunde and Akinlo [13]; Egbetunde and Akinlo [57]; Badri and Sheshgelani [36]’ Phutkaradze Tsintsadze and Phutkaradze [15]
MCAP	Market capitalization	Market capitalization % GDP. It captures the size of the selected African stock market	WDI (2020) [61]	Lenee and Oki [50]; Ailemen et al. [51]; Alam and Hussein [48]
INSQI	Institutional quality index	Perceptions of governance measured by computing the index of six dimensions of governance via principal component analysis.	WGI (2021) [61]	Egbetunde and Akinlo [13]; Egbetunde and Akinlo [57]; Klein [14]
CPS	Domestic credit to private sector	Private credit issued by banks to private sectors presented as percentage of GDP.	WDI (2020) [61]	Phutkaradze, Tsintsadze and Phutkaradze [15]
OPN	Degree of trade openness	Export plus import/GDP	WDI (2020) [61]

Note: WDI denote the World Bank website, WGI stands for Worldwide Governance Indicators.

Source: Authors Computation

3.2. Descriptive properties of variables

Table 2 displayed the descriptive statistics and the mean values of LRGDPC, LGDPC, LOPN and LCPS are 6.896, 6.660, 4.138 and 3.545 respectively in descending order. This indicates that LRGDPC contains the largest mean value followed by LGDPC, LOPN and LCPS respectively. Also, LGDPC has the highest maximum value of 8.23, followed by LRGDPC (8.16), LHCD (6.01), LCPS (5.76), LOPN (5.74) LMCAP_P (5.69) and FI_P (5.25) respectively in descending order.

Table 2.

Descriptive statistics.

Variables	MEAN	STD.DEV	MIN	MAX	OBS
LRGDPC	6.895624	0.5378016	5.8972	8.1555	375
LGDPC	6.659994	0.6959206	4.9323	8.2269	375
LHCD	−0.794124	0.4614869	−1.380312	6.011267	375
FI_P	1.677603	1.243227	−0.8741489	5.244871	375
FI_N	−1.257111	0.9714222	−3.621455	0.8117124	375
LMCAP_P	1.262025	1.33553	−1.143883	5.699912	375
LMCAP_N	−1.315807	1.326371	−5.833946	0.8831288	375
INSQI	0.0229465	1.027555	−2.287492	3.028002	375
LCPS	3.545318	1.343828	0.1492817	5.75786	375
LOPN	4.138418	0.38557	3.031099	5.740918	375

Source: Authors computation

The correlation analysis in Table 3 was employed to determine if the regressors have a perfect or precise linear representation with each other. The result reveals an inverse correlations between positive and negative shocks to stock market capitalization (LMCAP)) and LRGDPC. The rest of the explanatory variables exert positive correlations with RGDPC. On the second specification which is LGDPC model, negative shocks to financial integration index and MCAP have negative correlation with LGDPC while the rest have positive correlations with it. Similar result was indicated by the HCD model as both negative shocks to FI and MCAP exhibit negative correlation with HCD while the rest are positively correlated with it. As rule of thumb, if the correlation statistic is greater than 80% is observed between the independent variables, it indicates the presence of multicollonearity as such the variables not a candidate of linearly.

Table 3.

Correlation analysis.

	Spec 1- LRGDPC MODEL
	LRGDPC	FI_P	FI_N	LMCAP_P	LMCAP_N	INSQI	LCPS	LOPN
LRGDPC	1.0000
FI_P	0.116	1.0000
FI_N	0.0312	−0.8078	1.0000
LMCAP_P	−0.0183	0.501	−0.4766	1.0000
LMCAP_N	−0.0392	−0.5076	0.435	−0.8649	1.0000
INSQI	0.408	0.0099	0.1217	−0.2869	0.2326	1.0000
LCPS	0.385	−0.0748	0.1005	−0.1309	0.0895	0.0936	1.0000
LOPN	0.0836	0.1749	−0.1679	−0.2112	0.2611	0.1842	−0.0727	1.0000
	Spec 2- LGDPC MODEL
LGDPC	1.0000
FI_P	0.3665	1.0000
FI_N	−0.262	−0.8078	1.0000
LMCAP_P	0.1326	0.501	−0.4766	1.0000
LMCAP_N	−0.1127	−0.5076	0.435	−0.8649	1.0000
INSQI	0.4052	0.0099	0.1217	−0.2869	0.2326	1.0000
LCPS	0.366	−0.0748	0.1005	−0.1309	0.0895	0.0936	1.0000
LOPN	0.0992	0.1749	−0.1679	−0.2112	0.2611	0.1842	−0.0727	1.0000
	Spec 3- LHCD MODEL
LHCD	1.0000
FI_P	0.1746	1.0000
FI_N	−0.1628	−0.8078	1.0000
LMCAP_P	0.0434	0.501	−0.4766	1.0000
LMCAP_N	−0.039	−0.5076	0.435	−0.8649	1.0000
INSQI	0.1594	0.0099	0.1217	−0.2869	0.2326	1.0000
LCPS	0.1372	−0.0748	0.1005	−0.1309	0.0895	0.0936	1.0000
LOPN	0.1431	0.1749	−0.1679	−0.2112	0.2611	0.1842	−0.0727	1.0000

Source: Authors compilation

3.3. Model specification

Building on the previous studies, Bonfiglioli [17] Badri and Sheshgelani [36]; Egbetunde and Akinlo [13] Egbetunde and Akinlo [57]; we extend these studies by investigating jointly how financial integration (FI) and capital market development (MCAP) influence economic performance (EP) captured by (RGDP, GDPC and HCD) by accounting for the role of institutional quality (INSQI). Variables such as credit to private sector and degree of trade openness are used as control variables. The justification for incorporating domestic credit to private sector and trade openness are i.) domestic credit to private sector is that it is an ingredient for economic growth and performance input due to the ability to spur investment and productivity in the economy. The apriori expectation is positive (see Ref. [15]. ii.) Trade openness demonstrates the level of integration between African countries and the rest of the globe, and increasing trade is predicted to boost growth. The functional forms of the models are specified thus:

$${EP}^{RGDP,GDPC,HCD}=f\left(FI,MCAP,INSQI,CPS,OPN\right)$$ (1)

where: RGDPC, GDPC and HCD are the economic performance indicators denoting per capita gross domestic product, nominal GDP per capita and human capital development, each serving as a robust check another in their respective model, FI is index of financial integration, MCAP is market capitalization representing capital market development, INSQI is index of institutional quality already explained in Table 1, CPS stands for credit to private sector while OPN is the degree of trade openness.

The econometric form and log form of equations (1a), (1b) and (1c) are as follows:

$${LEP}_{it}^{RGDP,GDPC,HCD}={\beta }_0+{{\beta }_{1}LFI}_{it}+{\beta }_2{LMCAP}_{it}+{\beta }_3{INSQI}_{it}+{\beta }_4{LCPS}_{it}+{\beta }_5{LOPN}_{it}+{\mu }_{it}$$ (2)

From equation (2), L stands for natural log of the variables 1a to 1c, while the error term are μ_it , ε_it and $\omega$_it, the subfix ‘i’ and ‘t’ stand for the country (i = 1 … 16) and time (1996–2020). The coefficients, ${\beta }_1$–${\beta }_5$, ${\alpha }_1–{\alpha }_5$, and b₁–b₅ denote the long-run estimates of the independent variables.

3.3.1. Panel non-linear ARDL

The framework for the empirical narration is built on the study of Shin and Greenwood-Nimmo's [23] NARDL approach capable of handling panel of long T with heterogeneous features. They econometric reason for employing this method is highlighted as thus: (a) the duration of the study is large T with small cross-sectional dimension N, thus T > N, where N = 16 nations while T = 25 years. (b) The technique provides asymmetries nonlinearly which linear regression fails to account for. (c) It has the capacity to offer a combination of long, short-run and the error correction mechanism in a single model specification as long as all variables are integrated at order one or mixed and none ascends to order two. (d) Finally, it accounts for possible long and short-run negative and positive asymmetric impact of the exogenous variables on the endogenous variable, and capable of handling cross-sectional dependency in a panel data series. As a rule of thumb, our medaling strategy begin with the log-linear equation shown below as thus:

$${LEP}_{it}^{RGDP,GDPC,HCD}={\alpha }_{0i}+{\alpha }_{1i}{LEP}_{t-1}+{\alpha }_{2i}{FI}_{t-1}+{\alpha }_{3i}{LMCAP}_{t-1}+{\alpha }_{4i}{INSQI}_{t-1}+{\alpha }_{5i}{LCPS}_{t-1}+{\alpha }_{6i}{LOPN}_{t-1}+{\epsilon }_{it}$$ (3)

i = 1,2 3 … … ….N; , T = 1, 2, 3, … ….T.

Building on the works of Oprea and Stoica [46]; Algaeed, A. H. [49]; and Bista [56] who employed PARDL, and secondly, we incorporate the non- models to help us compute the long and short-term asymmetric expected equations as shown as thus :

$${LEP}_{it}^{RGDP,GDPC,HCD}={\delta }_{0i}+{\delta }_{1i}{LEP}_{t-1}+{\delta }_{2i}^{+}{FI}_{t-1}^{+}+{\delta }_{2i}^{-}{FI}_{t-1}^{-}+{\delta }_{3i}^{+}{LMCAP}_{t-1}^{+}+{\delta }_{3i}^{-}{LMCAP}_{t-1}^{-}+{\delta }_{4i}{INSQI}_{t-1}+{\delta }_{5i}{LCPS}_{t-1}+{\delta }_{6i}{LOPN}_{t-1}{\epsilon }_{it}$$ (4)

${where\delta }_{it}$ and${\partial }_{it}$ indicate coefficients for long-term parameters to be computed and ${FI}_{t-1}^{+}$ , ${FI}_{t-1}^{-}$ , ${LMCAP}_{t-1}^{+}$ and ${LMCAP}_{t-1}^{-}$ denote the positive and negative partial sum process variation in FI, and LMCAP respectively. The values of ${FI}_{t-1}^{+}$ , ${FI}_{t-1}^{-}$ , ${LMCAP}_{t-1}^{+}$ and ${LMCAP}_{t-1}^{-}$ can be obtained2

As stated by Shin et al. [23] and Pesaran et al. [62], we substitute Eq. (4) into Eq. (3) to obtain the asymmetric PARDL model by distinguishing the long-run and short-run asymmetric associations as follows:

$${LEP}_{it}^{RGDP,GDPC,HCD}={\vartheta }_{0i}+{\vartheta }_{1i}{LRGDPC}_{t-1}+{\vartheta }_{2i}^{+}{FI}_{t-1}^{+}+{\vartheta }_{2i}^{-}{FI}_{t-1}^{-}+{\vartheta }_{3i}^{+}{LMCAP}_{t-1}^{+}+{\vartheta }_{3i}^{-}+{\vartheta }_{4i}INSQI+{\vartheta }_{5i}{LCPS}_{t-1}+{\vartheta }_{6i}{LOPN}_{t-1}+\sum _{j=1}^{\rho 1}{\gamma }_{ij}\Delta {lRGDPC}_{i,i-j}+{\sum _{j=0}^{\rho 2}{\lambda }_{ij}\Delta {FI}_{t-j}^{+}+\sum _{j=0}^{\rho 3}{\lambda }_{ij}\Delta {FI}_{t-j}^{-}+\sum _{j=0}^{\rho 4}{\lambda }_{,ij}\Delta {LMCAP}_{t-j}^{+}+\sum _{j=0}^{\rho 5}{\lambda }_{ij}\Delta {LMCAP}_{t-j}^{-}+\sum _{j=0}^{\rho 6}{\lambda }_{ij}\Delta {lCPS}_{t-j}+\sum _{j=0}^{\rho 7}{\lambda }_{,ij}\Delta {lOPN}_{i-j}+\mu i7+\epsilon t}_{it}$$ (5)

In equation (5), Δ denote differenced variables, $\rho 1$ to $\rho 7$ stand for corresponding lag orders, $\vartheta$ = (${\vartheta }_{1i}$, ${\vartheta }_{2i}^{+}$, ${\vartheta }_{2i}^{-}$, ${\vartheta }_{3i}^{+}$, ${and\vartheta }_{3i}^{-}$) represent the coefficients of the long term positive and negative shocks of FI and MCAP, on RGDPC while $\sum _{j=0}^{\rho 2}{\lambda }_{ij}\Delta {FI}_{t-j}^{+}+\sum _{j=0}^{\rho 3}{\lambda }_{ij}\Delta {FI}_{t-j}^{-}$ and $\sum _{j=0}^{\rho 2}{\gamma }_{,ij}\Delta {LMCAP}_{i-j}^{+}+\sum _{j=0}^{\rho 3}{\gamma }_{3,ij}\Delta {LMCAP}_{i-j}^{-}$ are the short-run positive and negative influence of financial integration index and stock market development on RGDPC.

4. Discussion of results

4.1. CSD (cross sectional dependence) and statioarity test

The first estimation protocol begin with the cross sectional dependence as shown in Table 4. Accordingly, we checked for the common features of the series by leveraging on cross sectional dependence tests and Pesaran CD test under the null assumption that there exist no cross-sectional dependence within the panel while the H_A asserts that there is exist a cross-sectional dependency within the panel. Therefore, the result reveals a p-values of all variables being statistically significant at the 1% level, which implies the rejection of H₀ and accepting that there is cross-sectional dependence within the panel.

Table 4.

Cross-sectional dependence test.

Variables	Breusch-Pagan LM	Pesaran scaled LM	Bias-corrected scaled LM	Pesaran CD
LRGDPC	1090.6***(0.000)	68.01***(0.000)	67.70***(0.000)	26.36***(0.000)
LGDPC	2139.63***(0.000)	140.40***(0.000)	140.09***(0.000)	46.16***(0.000)
LHCD	1572.98***(0.000)	101.3***(0.000)	100.98***(0.000)	36.49***(0.000)
FI_P	1939.20***(0.000)	126.57***(0.000)	126.26***(0.000)	43.71***(0.000)
FI_N	1919.81***(0.000)	125.23***(0.000)	124.92***(0.000)	43.51***(0.000)
LMCAP_P	1390.98***(0.000)	88.74***(0.000)	88.43***(0.000)	33.87**(0.032)
LMCAP_N	1250.81***(0.000)	79.07***(0.000)	78.76***(0.000)	27.68***(0.000)
INSQI	521.77***(0.000)	28.76***(0.000)	28.45***(0.000)	11.27***(0.000)
LCPS	1204.24***(0.000)	75.85***(0.000)	75.54***(0.000)	2.66***(0.000)
LOPN	422.56***(0.000)	21.91***(0.000)	21.60***(0.000)	3.88***(0.000)

Hint: *, ** and *** stands for 10, 5 and 1% levels of significance respectively.

To determine the stationary states of our variables, we employed a 2nd generation unit stationary test because of the presence of cross-sectional dependency within our panel as proposed by Pesaran [63]. Therefore we estimated both the cross-sectional ADF (CADF) and cross-sectional augmented IPS (CIPS) tests that most appropriate in the presence of cross-sectional dependence. Table 5 represents the results of the 2nd generation unit root tests. The CIPS result shows that all the variables except LMCAP_P and LMCAP_N are not stationary but become stationary after differencing them once. Also, the CADF reveals that all the variables except LGDPC are stationary only at first difference. Hence the results show mixed order of integration of the variables irrespective of the test applied. That is, they are of I(1) and I(0).

Table 5.

Panel unit root test.

Variable	level		first Difference		Order of Integration
	intercept	intercept and trend	intercept	intercept and trend
Pesaran CIPS test
LRGDPC	−1.583	−1.832	−3.764***	−4.102***	I(1)
LGDPC	−1.936	−2.161	−4.069***	−4.314***	I(1)
LHCD	−1.745	−2.385	−4.242***	−4.801***	I(1)
FI_P	−1.943	−1.943	−4.325***	−4.489***	I(1)
FI_N	−2.372	−2.569	−4.95***	−5.042***	I(1)
LMCAP_P	−2.918***	−3.394***	−4.825***	−4.851***	I(0)
LMCAP_N	−3.972***	−3.849***	−5.304***	−5.384***	I(0)
INSQI	−2.016	−2.307	−4.407***	−4.533***	I(1)
LCPS	−1.389	−2.26	−4.601***	−5.073***	I(1)
LOPN	−2.444	−2.74	−4.777***	−4.791***	I(1)
Pesaran CADF test
LRGDPC	−2.208	−2.314	−2.795***	−2.973***	I(1)
LGDPC	−2.549**	−2.994**	−3.206***	−3.568***	I(0)
LHCD	−1.697	−1.828	−2.777***	−3.106***	I(1)
FI_P	−1.876	−2.012	−3.05***	−3.21***	I(1)
FI_N	−1.818	−1.894	−3.168***	−3.332***	I(1)
LMCAP_P	−1.908	−2.637	−3.171***	−3.176***	I(1)
LMCAP_N	−2.676	−2.487	−3.535***	−3.73***	I(1)
INSQI	−1.905	−2.432	−3.357	−3.5***	I(1)
LCPS	−1.277	−2.321	−3.549***	−4.213***	I(1)
LOPN	−2.039	−2.218	−3.267***	−3.207***	I(1)

Hint: ** and *** represent statistical significance at 5% and 1% levels respectively.

4.2. PNARDL outcomes

Table 6 represents the panel nonlinear autoregressive distributed lag result for the three specifications employed in this work. The result contains both long and short term outcomes of the PNARDL model. From the real GDP model (LRGDPC), the findings reveal that in the long term period, a positive shock to the financial integration index brings about a positive and significant impact on real GDP. That is, a unit rise in positive shock to the financial integration index leads to a 0.122 rise in RGDPC. Also, a unit rise in the negative shock to FI brings about a 0.042 reduction in RGDPC. This implies that both positive and negative shocks give rise to a different effect on RGDPC. Concerning the linkage between the stock market (LMCAP) and RGDPC, the outcome shows that a unit rise in the positive shock to MCAP will lead to a 0.088 reduction in RGDPC. In contrast, a unit increase in the negative shock to MCAP brings about a 0.074 decrease in RGDPC. This means that, unlike the FI, both positive and negative shocks to MCAP have the same inverse influence on RGDPC. The institutional quality index (INSQI) is revealed to have a positive and significant impact on RGDPC in the long run, which could lead to the asymmetric or differential effects of financial integration and market capitalization on economic performance. A unit increase in INSQI leads to a 0.230 unit rise in RGDPC. In the same manner, both credits to private sectors (CPS representing financial development) is not statistically significant and trade openness (OPN) exert positive and significant impact on the RGDPC in the long run, which in parri-passu with the theoretical expectations and indeed validates the roles of trade openness in stimulating economic performance.

Table 6.

Result of panel NARDL.

Variables	LRGDPC MODEL	LGDPC MODEL	LHCD MODEL
Long-run Equation
FI_P	0.122***(0.026)	−0.022***(0.058)	−0.106***(0.0003)
FI_N	−0.042**(0.038)	−0.111*(0.063)	−0.129***(0.0004)
LMCAP_P	−0.088**(0.036)	0.326***(0.050)	−0.055***(0.0003)
LMCAP_N	−0.074(0.047)	−0.098(0.091)	−0.061***(7.182)
INSQI	0.230***(0.025)	−0.164***(0.053)	0.2237***(0.0006)
LCPS	0.132(0.055)	−0.262(0.025)	0.0181(0.0004)
LOPN	0.154***(0.051)	0.085***(0.091)	0.0843***(0.0004)
Short-run equation
ECT(-1)	−0.166***(0.041)	−0.187**(0.043)	−0.118**(0.016)
FI_P	−0.016(0.011)	0.029(0.022)	−0.220(0.289)
FI_N	−0.019(0.016)	−0.0322(0.028)	−0.515(0.588)
LMCAP_P	−0.038(0.063)	0.00034(0.086)	0.0867(1.084)
LMCAP_N	0.051(0.026)	−0.0578(0.047)	−0.059(0.946)
INSQI	−0.024**(0.012)	−0.022(0.038)	−0.086(0.083)
LCPS	−0.091(0.052)	−0.3478**(0.088)	−0.459(0.403)
LOPN	0.010(0.032)	- 0.3119***(0.099)	0.987(0.731)
Cons	0.932(0.226)	0.237(0.061)	−0.2007(0.152)
Long Run Asymmetry
FI	5.0282	12.6029	8.0231
	[0.0201]	[0.0000]	[0.0001]
LMCAP	9.848523	15.1121	[18.0503]
	[0.0000]	[0.0000]	[0.0000]
Short Run Asymmetry
FI	1.407445	1.766952	0.022119
	[0.4351]	[0.1724]	[0.9781]
LMCAP	0.715692	1.853911	0.090925
	[0.4896]	[0.1582]	[0.9131]

Note: *, ** and *** stands for statistical significance at 10%, 5% and 1% levels respectively. The values in brackets denote the standard errors. Values in [] represents the probability values under the asymmetry.

The result of the nominal GDP model (GDPC) indicates that a unit rise in both positive and negative shocks to FI will bring about 0.022 and 0.111 reductions in GDPC, respectively. This also implies that both shocks to FI exert the same influence on nominal GDP in the long run. Also, a positive shock to MCAP is seen to exert a positive influence on GDPC, while a negative shock to MCAP has a negative influence on GDPC. This shows that different shocks to MCAP exhibit a different effect on GDPC in the long run. Institutional quality index and trade openness exhibit inverse and significant influence on GDPC, while CPS does not influence GDPC, as indicated by its insignificant probability value. The variance in the findings could be a result of choices of variables and indeed convey a picture of the complex nature of economic development and performance in Africa.

Concerning the Human capital development model, the findings show that both positive and negative shocks to FI and MCAP negatively impact HCD in the long term. That is, a unit rise in both positive and negative shocks to the financial integration index and stock market (MCAP) brings about 0.106, 0.129 (positive and negative of FI), 0.055 and 0.061 (positive and negatives of MCAP) reductions in HCD respectively. This entails that both shocks to FI and MCAP have a similar effect on HCD. On the other hand, a unit increase in the institutional quality index and CPS will lead to 0.224 and 0.018 increases in HCD in the long run, while a unit rise in OPN leads to a 0.084 decrease in HCD.

In the short run, however, the ECT for all the models are negative and significant, with error terms of 0.166, 0.187 and 0.118, respectively. The signs and significant outcomes of the ECM met the economic expectation, which implies restoring the long-run equilibrium after an exogenous shock irrespective of the model adopted. As for the explanatory variables, INSQI exerts negative and significant impact on the RGDPC model, while CPS and OPN exhibit negative and significant influence on the GDPC model. The rest of the variables do not have any short-term impact on the rest of the dependent variables. Overall, these findings depict the underdeveloped nature of INSQI, CPS and OPN in Africa.

The long-run asymmetric effects indicate that for all the specifications (RGDPC, LGDPC, and HCD models), we reject the null hypothesis of no asymmetric linkage between dependent variables and financial integration index and between dependent variables and stock market development (MCAP) since the Wald test probability values are all less than 5% level of significance. This means that there is a long-run asymmetric influence of FI and MCAP on RGDP, GDPC and HCD models. As for the short-run asymmetric influence, the Wald test probability values for the three models indicate acceptance of the null hypothesis since the probability values are all greater than the 5% significant level. Hence, no short-run asymmetric linkages exist between dependent variables and financial integration index and between dependent variables and stock market development.

4.3. Discussion of findings

We tested the stationarity status of our variables using both CIPS and CADF. The CIPS result reveals that all the variables except LMCAP_P and LMCAP_N are not stationary but become stationary after differencing them once. CADF, on the other hand, indicates that all the variables except LGDPC are stationary only at first difference. Hence, the results show mixed order integration of the variables irrespective of the test applied. They are of I(1) and I(0). This test is informed by the findings of the CSD test as can be seen in Table 4. Using ARDL/PMG method in testing the PNARDL, the long-run outcome indicate that an increase in positive shock to financial integration index leads to a rise in RGDPC. In contrast, an increase in the negative shock to FI leads to a fall in RGDPC. This implies that both positive and negative shocks give rise to and lead to a different effect on RGDPC. These findings corroborate the findings of Klein [14]; who found similar interplay between financial integration and economic growth. Also, the positive outcome is in tandem with the findings of Friedrich et al. [35]; Mahajan and Verma [16]; and Bonfiglioli [17]. In comparison, the negative association is in collaboration with the findings of Badri and Sheshgelani [36] and Egbetunde and Akinlo [57]. On the linkage between the stock market capitalization (LMCAP) and RGDPC, the outcome shows that both shocks to MCAP reduce RGDPC. This means that, unlike the FI, both positive and negative shocks to MCAP have the same negative influence on RGDPC. These result tallies with the outcome of Algaeed [49] but against the findings of Flaviabama and Mura [45]; Oprea and Stoica [46]; Khetsi and Mongale [47]: and Alam and Hussain [48]. The institutional quality index (INSQI) is revealed to have a significant positive impact on RGDPC in the long run. This outcome contradicts the findings of Egbetunde and Akinlo [13]. However, using nominal GDP to represent economic performance reveals that both shocks to FI exert negative influence on GDPC while positive and negative shocks to MCAP exert a positive and negative influence on GDPC, respectively. INSQI also affects GDPC negatively and significantly. With HCD as economic performance, positive and negative shocks to FI reduce HCD as well as shocks to MCAP. INSQI, on the other hand, adds to HCD. This means that the influence of explanatory variables on dependent (economic performance) is purely dependent on the indicator used to measure economic performance. The short-run estimates indicate that ECTs for economic performance, irrespective of the measure or the indicator adopted, are negative and significant. The sign and the coefficients of the ECM meet the apriori expectation, which also suggests restoring the long-run equilibrium after an exogenous shock.

The results of the controlled variables also justify their inclusion in the models. Credit to the private sector exerts a positive and insignificant impact on the three-model specification. It implies that the financial system has inherent underdeveloped structure and cannot drive economic performance in selected SSA countries. It implies that channeling the financial inflows from the international economies and meagre resources of the economy to productive and growth-enhancing activities of the financial system is jeopardized. The trade openness coefficient is positive and statistically significant across the model. This means that trade openness is an essential driver of economic performance in selected SSA countries. This result is in conformity with the spillover effects hypothesis as documented in the previous studies.

The findings of this research output can be summarized as thus: i.) In the long-run, a rise in positive shock to financial integration index leads to a rise in RGDPC while a negative shock to FI leads to fall in RGDPC. ii.) Both shocks (positive and negative) to MCAP reduce RGDPC. Institutional quality index (INSQI) is revealed to have a direct and significant impact on RGDPC in the long run and indeed intensify their asymmetries. iii.) Both shocks to FI exert inverse influence on nominal GDPC while positive and negative shocks to MCAP exert positive and negative influence on GDPC respectively. INSQI also affects GDPC negatively and significantly and indeed reduces their asymmetries. iv.) Positive and negative shocks to FI reduce HCD as well as shocks to MCAP. INSQI on the other hand, adds to HCD and indeed intensify their asymmetries.

4.4. Conclusion and policy recommendation

The key goal of this study is to determine the role of institutional quality in the in the interplay between financial integration, market capitalization and economic development in the 16 SSA nations. The study used three different indicators to measure economic performance, such as, real gross domestic product, nominal domestic product and human capital development. Results of the PNARDL reveals different outcomes based on the measure of economic growth used. For instance, when economic performance is measured by real GDP, it was observed that the differential and asymmetric linkage between financial openness index and economic growth exist. But using nominal GDP to represent economic performance, financial integration index irrespective of the shock exerts negative linkage with economic growth. And with human capital development measuring economic performance, both positive and negative shocks to FI dampens HCD in the region. For the association between stock market (MCAP) and economic performance, both shocks to MCAP significantly reduce economic growth when performance is measured by human capital development and real GDP. But when nominal GDP is used, only positive shock to MCAP significantly adds to economic performance. Institutional quality index (INSQI) on the other hand exerts a long run positive influence on economic growth when economic performance is measured by real GDP and human capital development while it reduces economic performance when it is proxied by nominal GDP in the region. In all, the nature of the interplay is purely dependent on the indicator used to measure economic performance. The error correction mechanism (ECM) have a negative signs and at same time statistically significant which is in tune to the theoretical expectations and this further suggests restoration of the long-run equilibrium after an exogenous shocks. The lack of consistency of our research output in the selected SSA shows that financial integration is still underdeveloped and the benefits of international trade have not been fully harnessed. By implication, the negative influence of six institutional quality index on RGDP and HCD implies that the institutional qualities were not properly entrenched in the SSA region and as such resulted to lower or poor economic growth. Also, this shows a weak judicial system which may equally deter foreign investor coming into the region. A careful policy option is for SSA to i.) To change and strengthen the quality of the region's institutional structure and dispel the myth that SSA is a high-risk investment zone. This is crucial because SSA economic success is sensitive to financial integration (rising and falling). (ii) Trade regulations and barriers must be relaxed, and (iii) the Africa Continental Free Trade Agreement (AfCFTA) will help synergize trade in SSA.

Our research also suggests that the financial integration index can increase RGDPC while the impact of positive and negative shocks to market capitalization can diminish it. Therefore, governments in SSA need to take advantage of financial integration's positive effects on growth by formulating a clear policy framework to shield the financial sector from the negative effects of institutional shock. Therefore, there is need to strengthen the SSA's regulatory, supervisory, institutional, and legal frameworks to boost economic confidence. By facilitating the flow of most economic activity through the financial system, this policy framework will foster the growth of the financial industry.

4.5. Limitations

The main limitation of this study is the lack of data on financial integration in most of the SSA counties, as it did not permit us to extend our sample size beyond 2019 and below 1996. Our sources stopped data update in 2019 while data were missing in some African countries. Therefore, we recommend that future research should examine the issues in details as data become available. Another limitation is the statistical deficiency from the additive assumption in the financial development index and institutional structure sourced from World Bank Development Website. In this study, it is assumed that the improvement in the quality of institutional structure, financial integration, and market capitalization will have a corresponding effect on economic performance regardless of the extent of the asymmetric behavior of the variables. However, in reality, the interaction of these variables and their effect on economic growth (real and nominal gdp) and human capital development are more complex. Hence, this study would benefit more by using more advance econometric technique such method moment quantile regression (MMQR) that capture heterogeneous behaviors of countries in the SSA. Also, further studies should consider unbundling institutional structure since it comprises Voice and Accountability, Political Stability and Absence of Violence/Terrorism, Government Effectiveness, Regulatory Quality, Rule of Law and Control of Corruption) and examine their respective moderating on financial integration, capital market development, and economic performance nexus in SSA. Good knowledge of this institution will demand different kinds of analysis other than the econometric method and huge data analysis.

Author contributions

Eugene Iheanacho: Conceived and designed the analysis.

Kingsley I. Okere: Analyzed and interpreted the data; Wrote the paper.

John Okey Onoh: Contributed analysis tools or data; wrote the paper.