The Empirical Link Between Financial Globalization and Macroeconomic Volatility in Sub-Saharan African Countries

1. Introduction

The Sub-Saharan Africa (SSA) region has been disproportionately bedevilled with chronic macroeconomic volatility despite the episodic structural reforms and intermittent macroeconomic stabilization programs (Kapingura et al., 2022). Macroeconomic volatility, which is the unpredictable fluctuations in the economic variables arising from shocks from the domestic and external markets, or the deviation of the short-term value of an economic variable from its long-term equilibrium value (David & Ampah, 2018), more prevalent and ensnaring in SSA countries than in developed countries (Kose et al., 2003; Yaya, 2024), Unfortunately, many SSA countries lack the policy instruments to address it compared with developed countries (Tolulope & Charles, 2020). Macroeconomic volatility is a serious sovereign problem that creates uncertainty in the business environment, leads to policy unpredictability, increases a country’s risk profile, and deteriorates citizens’ economic welfare (David & Ampah, 2021; Guresci, 2018; Xue, 2020). Therefore, the need for more empirical evidence on the wide range of factors contributing to macroeconomic volatility with a view to reining it in in SSA countries is quite an exigent one and this exigency motivates this study.

Interestingly, many SSA countries embarked on structural reforms, including external financial liberalization geared towards promoting the course of financial globalization, to achieve sustainable economic growth and macroeconomic stability (Blejer & Sagari, 1988). In the same vein, Khadraoui (2011) notes that SSA countries embraced financial globalization initiatives with the hopes that it would help ameliorate macroeconomic volatility. However, the chronic and persistent macroeconomic volatility in the SSA region calls for re-examining how financial globalization variables have influenced macroeconomic volatility in SSA countries. This is as the literature highlights that the prospects for financial globalization reducing economic volatility are brighter in developed countries than in developing countries (Dada et al., 2025).

Financial globalization, as a concept, relates to the integration of financial markets around the world and the unrestricted cross-border flow of capital, propelled by advancements in information and communication technology (Gudmundsson, 2016; Kılıçarslan & Dumrul, 2018; Naz et al., 2018). Notwithstanding, there are ongoing debates in policy and academic circles regarding the benefits and costs of financial globalization, as it affects macroeconomic volatility through many channels. Theoretically, financial globalization is expected to lower macroeconomic volatility through the promotion of financial sector development, enablement of technological diffusion, an increase in domestic capital stock, enhancement of production efficiency, entrenchment of global best practices, and imbibing of sound corporate and public governance systems (Balcilar et al., 2019; Flynn et al., 2019; Ghazouani et al., 2019; Kose et al., 2009; Presbitero et al., 2016; Tesega, 2022). Also, financial globalization should foster investment, income, and consumption smoothing during domestic real and financial sector shocks, facilitate global investment risk diversification (Awan et al., 2021), and provide the avenue for global capital allocation efficiency (Eozenou, 2008).

Conversely, financial globalization can precipitate macroeconomic volatility through several channels. First, financial globalization can increase the risk of financial crisis contagion from one region to another and expose the domestic financial markets to the fluctuations in the global financial cycles and global risk aversion levels, and the U.S. monetary policy rate adjustments (Miranda-Agrippino & Rey, 2021; Rey, 2015). For instance, many prior financial crises the world has experienced, such as the Latin America in the 1980s and 1990s (Hernández & Parro, (2008), the 1997 Asian financial crisis (Basu, 2003; Larose, 2003), and recently, the global financial crisis of 2007/8, and subsequently, the Europe debt crisis of 2010 (Aizenman et al., 2011; Akinyemi, 2025b) have been traced to financial globalization dynamics and vicissitudes. Literature has equally documented that fact that many SSA region has suffered serious macroeconomic problems after the implementation of external financial liberalization programs in the 1990s, and many of those challenges still exist till now (Gelbard & Leite, 1999; Moyo & Le Roux, 2020; Odhiambo, 2010). Second, financial globalization can cause macroeconomic volatility due to the pro-cyclicality of the external capital that flows into the SSA region, often leading to sharp reversals in capital flows or retrenchment. These volatile capital market movements negatively affect banking systems, drain available financial resources for capital investment, undermine the resilience of industrial production, and foment asset price volatility (Balcilar et al., 2019; Basu, 2003; Dada et al., 2025; Stiglitz, 2004b). Third, financial globalization can also impact macroeconomic volatility through the issues related to capital flight to safety and illicit financial flows from SSA countries to other developing and developed countries brought about the seamless international payment system and cryptocurrency operations, depleting domestic capital stock and making domestic capital formation difficult (Collin, 2020; Jume, 2021; Thiao, 2020). Finally, financial globalization can propagate macroeconomic volatility through excessive and unsustainable external debt accumulation enabled by easier access to international credit, thereby elevating fiscal sustainability risk, triggering sovereign credit ratings downgrades, and increasing the country risk premiums and the borrowing costs (African Union [AU], 2022; Kose et al., 2020). Therefore, these points indicate that financial globalization has multiple mechanisms through which it can influence macroeconomic volatility.

Empirically, the effects of financial globalization on macroeconomic volatility have been divergent and inconsistent. Some studies have reported that financial globalization increases macroeconomic volatility (Awan et al., 2021; Dada et al., 2025; Devereux & Yu, 2019; Kose et al., 2003; Neaime, 2005; Tolulope & Charles, 2020). Conversely, several others have concluded that financial globalization reduces macroeconomic volatility (Ahmed & Suardi, 2009; Aizenman et al., 2011; Buch et al., 2005; Cordell & Ospino, 2017; Meller, 2011; Yadav et al., 2018). Some other forms of relationships have been detected between financial globalization and macroeconomic volatility. For instance, a non-linear relationship has been found between financial globalization and macroeconomic volatility (Kose et al., 2003). Additionally, many studies have examined how financial globalization variables have affected other macroeconomic outcomes (Ahmed, 2021; Akinyemi, 2025; Batuo et al., 2018; Bhanumurthy & Kumawat, 2020; Dreher, 2006; Hermes & Lensink, 2003; Ikpesu, 2021; Sghaier, 2018; Tesega, 2022; Zahonogo, 2018). Apart from these issues, many existing studies have not focused on SSA countries. Equally, the indicators used to measure financial globalization vary widely across many existing studies; this may have accounted for the divergent results (Chinn & Ito, 2008). Many studies have used single indicators (Akinyemi, 2026; Ali et al., 2019; Kose et al., 2003), while others have employed a broad proxy, such as the financial openness index (Chinn & Ito, 2008; Meller, 2011; Tolulope & Charles, 2020), or the KOF Financial Globalization Index (Balcilar et al., 2019; Gozgor, 2018; Gulcemal, 2021; Hafezi et al., 2023). However, studies that have examined and compared the effects of the de facto, de jure, and overall financial globalization indices on macroeconomic volatility are rare. Hence, this study was instituted to bridge this empirical gap.

Against this background, this study examines and compares the effect of the overall financial globalization index on macroeconomic volatility with that of de facto and de jure financial globalization indices in SSA countries to determine whether there is convergence in their effects. Therefore, this study makes several contributions in several ways. First, this study uniquely examines the effects of different financial globalization indicators on macroeconomic volatility in SSA countries to assess the impact of financial and investment laws, regulations, and rules, as well as the actual capital flows, on macroeconomic volatility to enable SSA countries design and implement the appropriate global financial market integration strategy for effective macroeconomic volatility minimization and sovereign risk perception management. Also, it clarifies whether financial globalization measures matter for macroeconomic volatility modeling in SSA countries. Curiously, it has been observed that the different financial globalization measures respond divergently to macroeconomic outcomes (Tolulope & Charles, 2020) and that the distinction between the de facto and de jure measures is crucial in empirical analysis (Ahmed & Suardi, 2009; Chinn & Ito, 2008). Second, this study extends the work of Tolulope and Charles (2020). However, this study differs in that it focuses solely more on SSA countries and examines the macroeconomic volatility effects of the overall, de facto and de jure financial globalization measures, compared with the indication of capital account liberalization (de jure financial openness) and the ratio of total foreign investment to total investment (de facto financial openness) that Tolulope and Charles (2020) utilized in their study. Third, this study employs a robust econometric modelling strategy to enhance the validity and reliability of the empirical results by testing for cross-sectional dependence, autocorrelation, and heteroskedasticity before selecting the estimation techniques, methodological steps that many prior studies did not pay attention to (Abanikanda & Dada, 2023; Kose et al., 2003; Makoto, 2028; Tolulope & Charles, 2020). Consequently, PCSE and 2SGMM estimators were used to generate the panel regression model parameters.

Therefore, this study is beneficial to several groups of stakeholders, such as policymakers, central banks, regulators, market participants, and international financial institutions, in macroeconomic volatility modelling and in formulating policies to ensure sustained and sustainable macroeconomic stability in the SSA region. Also, this study offers valuable insights for academics, global investors, and financial analysts the financial globalization measures matter in the parameterization of systemic risk modelling in capital asset price formulation, thereby enhancing risk forecasting and capital allocation efficiency. Furthermore, the SSA region, though afflicted with a disproportionately high degree of macroeconomic volatility and several other economic development challenges, has a lot of opportunities as it continues to get integrated into the global financial markets and compete for a fair share of global FDI and portfolio investment flows (Adjei et al., 2025). Thus, this study will provide insights into how financial globalization dynamics have shaped the macroeconomic volatility to inform the reforms, policies, and programs that SSA countries need to implement to realize the potential benefits of financial globalization for economic development and macroeconomic stability.

This paper is divided into five main sections. Section One presents the introduction and background to the study. Section Two reviews the existing empirical and theoretical literature. Section Three discusses the methodology, including the econometric modelling strategy, while Section Four presents the empirical results and discussion. Finally, Section Five presents the conclusion, policy implications of the study, study’s limitations, and areas for future study.

2. Literature Review

2.1. Theoretical Foundation

This study is based on two distinct theories: the external capital flow theory and the real business cycle theory. The external capital flow theory, also known as the MacDougall-Kemp hypothesis, explains the impact of external capital flows from capital-rich countries to developing countries on macroeconomic outcomes. According to Agyapong and Bedjabeng (2020), the capital flow theory, developed by MacDougall in 1958 and Kemp in 1964, argues that free movement of capital from the capital-endowed country to the capital-poor country can foster a reduction in consumption and investment risk, while enabling the capital-poor countries to diversify their production bases. Chigbu et al. (2015) explain that the theory suggests that the quantum of FDI has a significant impact on economic development and macroeconomic stability. By the same token, when capital outflows for whatever reason, the financial market comes under severe strain, causing GDP to fall and national income to plummet (Aniekwe, 2022; Kurtishi-Kastrati, 2013; Patrick, 2023). However, this theoretical assertion has been challenged by the Lucas paradox, which highlights the contradictory behaviour of global capital flows. The Lucas paradox holds that external capital does not always flow from capital-rich to capital-poor countries, contrary to neoclassical economic theory. Instead, capital seems to flow in the opposite direction. This phenomenon, as hypothesised by the Lucas paradox (wherein capital does not flow to the capital-poor countries from the capital-rich countries despite the potential higher returns of capital), occurs political instability, corruption, policy uncertainty, and poor macroeconomic management, information asymmetry, high sovereign credit risk, differences in human capital development, poor production technology, decrepit physical infrastructure, and high business set-up and management costs (Gorniak, 2021; Janicka, 2016; Keskinsoy, 2017).

The real business cycle (RBC) theory is another relevant theory that explains factors that contribute to macroeconomic volatility. Mankiw (1989) explains that real business cycle theory posits that economic fluctuations result from changes in production technology that alter labor supply and consumption. Buttressing the postulations of the RBC, in open-economy models, financial openness increases an economy’s susceptibility to economic and financial crises through local production collapse, sudden retrenchment, and capital flow reversals, and loss of access to international capital (Awan et al., 2021; Dada et al., 2025). Mendoza (1991) further states that the RBC theory proposes that disturbances in production, technology, and financial structure are sources of economic fluctuations and procyclicality in consumption, investment, and employment. It is also argued that, due to the pro-cyclical and short-term sources of capital that developing countries have access to in international financial markets, they tend to experience macroeconomic instability. Financial globalization may not offer any tangible benefits to developing countries, such as those in SSA as Bolhuis et al. (2024) underscore how geoeconomic fragmentation and reversals in financial globalization efforts could create access to finance, trade, and the global supply chain, thwarting industrial production and deepening the multidimensional poverty in the SSA region. It is expected that financial globalization will boost the production capacity of developing countries by resolving the issues of underinvestment created by capital deficiencies.

2.2. Review of Empirical Literature

The relationship between financial globalization and macroeconomic volatility is nuanced and complex. Divergent empirical outcomes have been reported in the literature regarding the relationship between financial globalization and macroeconomic volatility, with some studies reporting that financial globalization reduces macroeconomic volatility. So many studies have concluded that financial globalization increases macroeconomic volatility. A non-linear relationship between financial globalization and macroeconomic volatility has also been reported. Many of the studies are explored in the ensuing paragraphs.

Regarding the reducing effect of financial globalization on macroeconomic volatility, Buch et al. (2005) examined the effect of financial openness on macroeconomic volatility. Based on 1960-2000 data for 24 OECD countries, analyzed using the GMM estimation technique. It was found that financial openness has a reducing rather than a magnifying effect on business cycle volatility, and the effect of financial openness on business cycle fluctuations largely depends on the type of underlying shock, with interest rate volatility affecting business cycle volatility more than government expenditure volatility. Aizenman et al. (2011) examined how the trilemma policy parameters (monetary autonomy, exchange rate stability, and financial integration) impact output volatility in 170 countries. Data were collected for the period 1972-2006, organized into 5-year panels, after which OLS and system GMM estimation techniques were used to estimate the regression coefficients. The study yielded several far-reaching conclusions. It was reported that greater monetary autonomy is positively associated with lower output volatility, whereas greater exchange rate stability is associated with greater output volatility. Still, the output volatility can be mitigated if a country holds foreign reserves equivalent to at least 20% of its GDP. Also, greater monetary policy autonomy could lead to higher inflation, while greater exchange rate stability and financial integration could foster lower inflation.

Using the panel threshold regression model, Meller (2011) examined the threshold level of financial openness on financial risk and output volatility, using 1980-2007 data for 26 developed and 36 developing countries. Financial openness increased output volatility in countries with high financial risk. At the same time, it decreased output volatility in countries with low financial risk (as measured by exposure to official, trade, and commercial debt). Makoto (2018) examined the effect of financial integration in Zimbabwe using the 2000-2016 data, analyzed with the panel ARDL estimator. It was revealed that financial integration reduced output volatility, while it had an insignificant effect on consumption volatility. Yadav et al. (2018) examined the effect of financial integration on macroeconomic volatility in Asian and developed countries. The study used data collected from 1980 to 2016 and analysed them using the GMM estimator. It was concluded that financial openness, trade openness, and broad money significantly and negatively affect macroeconomic volatility, as measured by four indicators: consumption, output, income, and consumption-income volatility.

Another strand of studies reported an increasing association between financial globalization indicators and macroeconomic volatility. For instance, Kose et al. (2003), a major and robust study from two decades ago. The study examined the relationship between global financial integration and macroeconomic volatility in industrialized and developing countries. The study was undertaken because previous studies had examined only the effect of financial globalization on economic growth variables, without considering its impact on macroeconomic volatility. The study data were obtained from 76 countries (21 industrial economies and 55 developing countries) over the period from 1960 to 1999. The collected data were analyzed using ordinary least squares (OLS) and the general method of moments (GMM) estimation techniques. The study found that financial openness, as measured by gross capital flows, increased the consumption-to-income volatility ratio, thereby contradicting the notion that financial integration contributes to international risk sharing. Furthermore, the study found that the relationship between financial integration measures and macroeconomic volatility is non-linear, suggesting that there is a threshold beyond which gross capital flows negatively affect the ratio of consumption volatility to income volatility. Neaime (2005) conducted a study on the implications of financial market integration for macroeconomic stability in the Middle East and North Africa (MENA) countries. The 1980-2002 data were obtained for the study, and the data were analyzed using the panel OLS regression model. The study found that financial openness, proxied by the ratio of gross capital flows to GDP, increased consumption volatility, and this unexpected empirical outcome may have been due to the weak monetary and fiscal policy coordination in the region. The study also found that greater financial sector development helps countries better manage external financial shocks, and that imposing capital account controls and restrictions may become counterproductive.

Devereux and Yu (2019) examined the relationship between international financial integration and crisis contagion in 140 countries following the Global Financial Crisis. The data used in the study cover the period from 1970 to 2012. The panel fixed effect model was estimated. The study found that financial integration led to global leverage escalation and increased the probability of crises. Additionally, it was concluded that financial integration heightens financial crisis contagion, but the effect of international financial integration on macroeconomic aggregates is very minute. Ashogbon et al. (2023), using the ARDL approach to analyse data collected from Nigeria from 1981 to 2020, found that public external debt has a long-run negative association with the exchange rate, while public domestic debt has a positive effect on the exchange rate, underscoring the need to reduce reliance on external debt. Similarly, Tolulope and Charles (2020) investigated how financial openness, trade openness, and financial sector development influence macroeconomic volatility in 51 African countries, using panel data from 1980 to 2017. The analysis employed estimation techniques such as OLS, fixed effects, and GMM. The study employed seven macroeconomic indicators, including per capita GDP growth rate volatility, GNP growth rate volatility, terms-of-trade-adjusted output growth rate volatility, GDP growth rate volatility, private consumption growth rate volatility, total consumption growth rate volatility, and gross fixed capital formation volatility. The results showed that de jure financial openness measures aggravated income volatility, whereas de facto measures reduced it. The test further revealed that financial openness is significantly positively associated with output volatility.

Similarly, Awan et al. (2021) examined the impact of financial globalization on output volatility in 22 Asian countries, using data collected for the period 1998-2015 and analysed using the system GMM estimator. The study concluded that financial globalization has a significant and positive influence on output volatility in the long run. However, the effect of financial globalization on output volatility in the three sub-samples—developed East Asia, South Asia, and West Asia—showed an insignificant yet positive relationship. Dada et al. (2025) examined the effect of financial globalisation on macroeconomic volatility in SSA countries, using data collected from 1991 to 2021 and analysing using the Driscoll-Kraay standard error (DKSE) and the panel spatial correction (PSCC) standard error. The study found that financial globalisation significantly increases macroeconomic volatility in the direct model. In the indirect model, financial globalisation, when interacted with economic and political institutional indicators, such as democratic accountability, political stability, and law and order, has a negative effect on macroeconomic volatility. Akinyemi (2026) examined the effects of financial globalization forces and financial sector development on country risk premiums in 35 emerging and developing economies (EDEs) using 2000-2021 data, analyzed using the seemingly unrelated regression technique and the two-step system GMM (2SGMM). The study found that external debt exacerbates country risk premiums, while FDI has an insignificant reducing effect. Also, Akinyemi and Owolabi (2026) investigated how financial globalization and financial sector development have contributed to the formation of equity risk premiums in 42 EDEs, using data from 2000 to 2021 and analyzing them with the PCSE and DKSE estimators. It was found that financial globalization is very weak in lowering equity risk premiums.

The other studies on the effect of financial globalization on macroeconomic volatility either present a non-significant effect or a non-linear/threshold effect. In this category, Rincón-Castro (2007) examined the effects of financial globalization on economic growth and macroeconomic volatility, data collected over the period 1984-2003 for a panel of 43 countries were used, employing the OLS estimation technique as the analytical model. It was found that financial globalization spurred economic growth, but its effect on macroeconomic volatility was neutral. Ali et al. (2019) examined the impact of globalization on unemployment and conducted the study using the data collected from 1980 to 2017 in Pakistan, adopting an ARDL model after conducting a Bounds test for long-run cointegration to analyze the data inferentially. The study concluded that diaspora remittances reduce unemployment, while FDI has a non-significant positive relationship with unemployment. Bhanumurthy and Kumawat (2020) examined the impact of financial globalization on the economic growth in South Asia. The data for the study were collected from 1990 to 2015, as there was no consensus on the effect of financial globalization on economic growth. The study concluded that financial globalization does not have a strong influence on economic growth; however, economic growth is strongly linked to financial globalization, suggesting that economic growth drives financial globalization. Additionally, the variable selection impact assessment revealed that the financial globalization indicator was significant, with the capital account openness indicator explaining economic growth more effectively than the KOF index of economic globalization, which itself measures financial and trade openness. In contrast, capital account liberalization measures capital openness. Equally, the study affirmed the importance of robust domestic fiscal and financial sectors in securing better growth outcomes from financial globalization.

Other studies investigated the effect of financial globalisation variables on economic growth and other macroeconomic performance indicators. Dabwor et al. (2020) investigated the effects of stock market returns and globalisation on economic growth in Nigeria using data collected from 1981 to 2018 and analysed with the fully modified least squares. The researchers adduced empirical evidence to show that globalisation has a positive effect on economic growth in Nigeria, while stock market returns showed a non-significant positive effect. Tesega (2022) examined the effect of financial globalisation, proxied by the de facto (quantity-based) measure, on financial development in 33 African countries, using data collected from 2005 to 2019. The collected data were analysed using the Driscoll-Kraay standard error (DKSE) and the panel-corrected standard error (PCSE) techniques, both of which can account for autocorrelation, heteroskedasticity, and cross-sectional dependence. The study found a U-shaped relationship between financial globalisation and financial sector development in Africa, indicating that lower levels of financial globalisation appear to have a negative effect on financial development, while higher levels have a positive effect. In investigating the impact of the overall globalization index, constructed using the three dimensions based on economic integration, social integration, and political integration, on economic growth, Dreher (2006), using panel data for the period 1970-2000, covering 123 countries, and analyzed using both OLS and GMM, concluded that the overall globalization index (comprising economic, political, and social globalization) has a positive effect on economic growth. However, the sub-dimension with the greatest effect on economic growth is economic globalization. Oshodin et al. (2024), using data collected from 1991 to 2022 and analysed with fixed- and random-effect GLS estimators, found that geoeconomic fragmentation deters economic development and increases macroeconomic vulnerability in West African countries. Recently, Adjei et al. (2025) studied the moderating role of governance indicators in the relationship between financial globalization and economic growth in 31 SSA countries, using data collected from 2002 to 2021 and analyzed using system GMM. The study revealed that government efficiency, reduced corruption, improved regulatory quality, rule of law, and consistent accountability enhance the positive outcome of financial globalization on economic growth.

The summary of existing studies in presented in Table 1. From the foregoing, studies in the SSA countries regarding the relationship between financial globalization indicators and macroeconomic volatility indicators are scarce. Hence, this study fills the empirical gap in this area of study. Apart from that, the connection between financial globalization and macroeconomic volatility is divergent, requiring more studies situated in the background or grounded in the situational context of the sample countries.

3. Methodology and Data

3.1. Data and Variables

The study’s population is the SSA countries, and the SSA region comprises 48 countries (Adjei et al., 2025). However, the countries without the required dataset for analysis were excluded. Therefore, the study focused on 39 SSA countries based on the availability of comprehensive data. The 48 SSA countries, as shown in Table 2, using the World Bank data, constituted 88.89% of 54 African countries, 82.13%, and 62.17% of the African average population and average GDP, from 2000 to 2023. However, according to 2024 World Bank data, SSA countries accounted for 85.31% and 68.08% of Africa’s population and GDP, respectively. The final sample of 39 SSA countries constituted 81.25% of the total number of SSA countries, accounting for 84.84% of the total SSA population, and controlling around 91.40% of the total SSA GDP, based on the 2024 World Bank data. Therefore, the sampled countries represent SSA well from socio-economic perspectives. Given this sample size and composition, generalising the study’s empirical findings to the entire SSA region is justified and can be extended, in a largely and prudent manner, to all African countries. The list of the sample countries is in Appendix 1.

The data used in the study were collected from 1996 to 2023. However, to calculate the volatility of the macroeconomic indicators used in the study, a 5-year rolling window technique was adopted. This 5-year rolling window calculation approach has been used in similar prior studies (Ihnatov & Capraru, 2014; Ma & Song, 2018; Yang & Liu, 2016; Yaya, 2024). With this approach, macroeconomic volatility indicators were calculated using the standard deviation over a 5-year rolling window. For the independent and control variables, a 5-year rolling average was computed to align with the dependent variables’ periodicity. Consequently, 5 years were lost, shrinking the period from 1996-2023 to 2000-2023. Hence, the actual study period is 2000 to 2023. This timeframe is justified because the SSA countries instituted structural reforms between 1980 and 1996 to address macroeconomic instability (Noorbakhsh & Paloni, 2001). Moreover, academic literature suggests that the process of financial globalisation, which rose steeply in advanced countries in the late 1970s, reached many developing and emerging countries from the 1990s to the 2000s (Aizenman, 2020). Also, WEF (2025) avers that global financial market integration intensified in the 2000s across developing countries.

Furthermore, the study used secondary data from databases of global economic and financial institutions, such as the Swiss Economic Institute, the World Bank, and the IMF, as done by many prior studies (Ahmed & Suardi, 2009; Akinyemi, 2025a, 2025b, 2026; Akinyemi & Owolabi, 2026; Cave et al., 2020; Ibrahim & Alagidede, 2017; Ikpesu, 2024). The use of data from these well-established, widely available sources of high-quality, relatively comprehensive data fosters reliability, replicability, and comparability of the empirical findings. The details of the data sources used in the study are presented in Table 3.

3.1.1. Variable Measurement

Dependent Variables

Macroeconomic volatility is the study’s dependent variable. Macroeconomic indicators provide information about the conditions of a general economy’s behavior, structure, and productivity. In this study, two indicators have been specified to proxy macroeconomic volatility: the Adjusted Hanke Misery Index Volatility (AHMIV) and the Okun Misery Index Volatility (OMIV).

Adjusted Hanke’s Misery Index Volatility as a Measure of Macroeconomic Volatility

The Adjusted Hanke’s misery index volatility (AHMIV) is derived from the Hanke’s misery index (HMI). The HMI is derived from the sum of the unemployment rate, inflation, and lending rate in an economy minus the real GDP growth rate (Das et al., 2023). In summary, HMI is calculated using equation (1):

$$HMI=Unemploymentrate+inflationrate+lendingrate-realGDPgrowthrate$$

(1)

The HMI was modified to arrive at the Adjusted Hanke’s misery index (AHMI) volatility. It is described as the adjusted HMI because it excludes the lending rate, which is influenced by factors such as a country’s financial sector development. Financial sector development is one of the independent variables. Hence, it has been excluded to reduce collinearity between the dependent and independent variables in the study. Hence, the AHMI is stated in equation 2 as:

$$AHMI=Unemploymentrate+inflationrate-realGDPgrowthrate$$

(2)

After calculating the AHMI, a 5-year rolling window standard deviation of the AHMI was used to derive the AHMIV. AHMIV is a broader index because it linearly combines three key macroeconomic variables: the unemployment rate, the inflation rate, and the real GDP growth rate.

Okun Misery Index Volatility as a Measure of Macroeconomic Volatility

The Okun misery index is the sum of the unemployment rate and the inflation rate. It measures the economic misery (economic discomfort) in a country and expresses the general macroeconomic condition (Anaele & Nyenke, 2021). The higher the index, the poorer the macroeconomic performance and the lower the quality of life of the people (Anaele & Nyenke, 2021; Das et al., 2023). After calculating the Okun Misery Index (OMI), a 5-year rolling window standard deviation of the OMI was used to derive the Okun Misery Index Volatility (OMIV). Just like AHMIV, OMIV is a broader index because it linearly combines two key macroeconomic variables: the unemployment rate and the inflation rate.

Independent Variables

The independent variable of the study is financial globalisation. Financial globalisation is a latent research variable. It is proxied by the overall KOF Financial Globalisation Index (KOFFIGI), which combines both the KOF Financial Globalisation De Facto Index (KOFFIGIDF) and the KOF Financial Globalisation De Jure Index (KOFFIGIDJ). This robust aggregate or composite index captures the actual flows occasioned by international integration or openness. The effect of the sub-components of the overall KOF Financial Globalization Index was also examined in this study. These sub-components are the KOF Financial Globalisation De Facto Index (KOFFIGIDF) and the KOF Financial Globalisation De Jure Index (KOFFIGIDJ). The KOF Financial Globalisation De Facto Index (KOFFIGIDF), a measure of capital flows and stock, is constructed from five variable indicators (Gygli et al., 2018), based on the initial work of Dreher (2006). These indicators are shown below: Foreign direct investment, foreign portfolio investment, international debt, international reserves, and international income payments. It uses five (5) cash flow variables that reflect the multidimensionality of globalisation. The KOF Financial Globalisation De Jure Index (KOFFIGIDJ) reflects policies, laws, and regulations that enable or restrict international capital flows, rather than the actual volume of capital flows or stock. The KOF financial globalisation indicators’ trend for SSA countries is depicted in Figure 1, which shows a moderate level across the region. The KOF Financial Globalization de facto Index has been used as a surrogate for financial globalization in similar prior studies (Akinyemi, 2025; Akinyemi & Owolabi, 2026; Balcilar et al., 2019; Gozgor, 2018; Gulcemal, 2021; Hafezi et al., 2023). Other studies have constructed a composite index from various indicators to represent financial globalization (Batuo et al., 2018; Chinn & Ito, 2008; Cordella & Ospino, 2017; Haelg, 2020; Ozkok, 2015).

Control Variables

Several control variables were included in the regression model to account for extraneous or confounding variables that have theoretical or empirical links to the dependent variable, thereby econometrically muting the variation related to these variables (Nielsen & Raswant, 2018). The control variables deployed in the study are fiscal balance (FBA), population growth rate (PGR), changes in the leading export commodities (CLECP), central government debt (CGD), and institutional quality index (IQI). Briefly explaining the empirical or theoretical connection of these control variables is expedient. The overall fiscal balance often serves as a measure of fiscal strength or vulnerability. The sounder the fiscal stance is, the lower the credit risk spread (Baldacci et al., 2011). Fiscal balance or government budget balance has also been used in the past as a control variable (Ahwireng-Obeng & Ahwireng-Obeng, 2019; Alexopoulou et al., 2009; Baldacci et al., 2011). Fiscal balance (FBA) should have a negative or positive coefficient. A fiscal surplus is a sign of fiscal sustainability and solvency and should have a negative coefficient. In contrast, a significantly persistent fiscal deficit indicates fiscal vulnerability and fragility and, thus, should have a positive coefficient. Population growth rate is another control variable used in the study. A larger population provides a platform for trade, labour supply, domestic consumption, and funds mobilization, and is a source of human capital, which is positively related to macroeconomic performance. A small population has been linked to higher volatility in GDP and GNI (Hnatkovska & Koehler-Geib, 2018). Nevertheless, it has been reported that small economies have recorded a high GDP and per capita income despite their vulnerability to exogenous shocks, a phenomenon tagged “Singapore Paradox”, which is a sign of strong economic resilience (Yaya, 2024).)

The population growth rate has been used as a control variable in similar studies (Ahmed, 2015; Gulcemal, 2021; Shahbaz et al., 2018). Another control variable used in the study is the central government debt stock-to-GDP ratio. Central government debt includes all financial liabilities of the central government, both short-term and long-term. The higher the debt level, the greater the fiscal sustainability risk. However, debt capital can be crucial to fiscal intervention programs and macroeconomic stabilization when there is a capital shortage or fiscal deficit. Prior studies have also used debt variables (Mpapalika & Malikane, 2019; Palic et al., 2017).

Furthermore, changes in the price of the leading export commodity are equally germane, especially for SSA countries, many of which are resource-dependent. Their resource dependency exposes them to significant external market shocks, including global trade challenges, global supply chain disruptions, geopolitical and geoeconomic conflicts, and international financial market perturbations. Resource-rich dummy variables, energy price index (Chinn & Ito, 2008; Presbitero et al., 2016), and oil rent, as a proxy for oil exports (Imeokparia et al., 2023), have been used in similar studies. The leading commodities of SSA countries are in Appendix 2. Finally, the institutional quality index was constructed using the World Bank Governance Indicators that consist of voice and accountability, political stability and absence of violence, government effectiveness, regulatory quality, rule of law, and control of corruption, as done in prior studies (Le et al., 2015; Ozkok, 2015; Shahbaz et al., 2018). The quality of institutions is important for macroeconomic volatility through control of corruption, policy effectiveness, political stability, and the creation of a conducive business environment, thereby giving investors confidence. Therefore, institutional quality should be negatively related to macroeconomic volatility indicators. Table 3 summarizes the measurement of the research variables, indicators, a priori expectations, data sources, and other relevant details.

3.2. Econometric Modelling Strategy

3.2.1. Model Specification

The theoretical model is derived from external capital flow theory, the real business cycle theory, and the fiscal theory of sovereign risk. The capital flow theory argues that external capital flows from capital-rich countries to capital-poor countries, facilitating socio-economic development and macroeconomic stability in countries with high capital productivity. The real business cycle theory explains that external and internal factors, such as technological efficiency, international trade, external capital flows, financial sector development, institutional effectiveness, and the degree of policy consistency, can influence macroeconomic volatility.

Therefore, the relevant econometric model for macroeconomic volatility used in this study is adapted from Haroon and Jehan (2022)as stated in equation (3.3), but with some modifications that include financial globalisation and financial-sector development indicators. Hence, the empirical test involves regressing macroeconomic volatility indicators on financial globalisation and financial sector development, while controlling for relevant variables, as stated in equation (3.3).

The functional or theoretical models for each of the macroeconomic volatility indicators: the Adjusted Hanke Misery Index Volatility (AHMIV) and the Okun Misery Index Volatility (OMIV).

$$AHMIV=f\left(KOFFiGI,FBA,PGR,CLECP,CGD,IQI\right)\left(3\right)$$

$$AHMIV=f\left(KOFFiGIdf,FBA,PGR,CLECP,CGD,IQI\right)\left(4\right)$$

$$AHMIV=f\left(KOFFiGIdf,FBA,PGR,CLECP,CGD,IQI\right)\left(5\right)$$

$$OMIV=f\left(KOFFiGI,FBA,PGR,CLECP,CGD,IQI\right)\left(6\right)$$

$$OMIV=f\left(KOFFiGIdf,FBA,PGR,CLECP,CGD,IQI\right)\left(7\right)$$

$$OMIV=f\left(KOFFiGIdf,FBA,PGR,CLECP,CGD,IQI\right)\left(8\right)$$

The econometric forms of these models are expressed below:

• Model 1

${AHMIV}_{it}={\beta }_0+{\beta }_1{KOFFiGI}_{it}+{\beta }_2{FBA}_{it}+{\beta }_3{PGR}_{it}+{\beta }_4{CLECP}_{it}+{\beta }_5{CGD}_{it}+{\beta }_6{IQI}_{it}+{\phi }_i+{\mu }_t+{\epsilon }_{it}\left(9\right)$

• Model 2

${AHMIV}_{it}={\beta }_0+{\beta }_1{KOFFiGIdf}_{it}+{\beta }_2{FBA}_{it}+{\beta }_3{PGR}_{it}+{\beta }_4{CLECP}_{it}+{\beta }_5{CGD}_{it}+{\beta }_6{IQI}_{it}+{\phi }_i+{\mu }_t+{\epsilon }_{it}\left(10\right)$

• Model 3

${AHMIV}_{it}={\beta }_0+{\beta }_1{KOFFiGIdj}_{it}+{\beta }_2{FBA}_{it}+{\beta }_3{PGR}_{it}+{\beta }_4{CLECP}_{it}+{\beta }_5{CGD}_{it}+{\beta }_6{IQI}_{it}+{{\phi }_i+{\mu }_t+\epsilon }_{it}\left(11\right)$

• Model 4

${OMIV}_{it}={\beta }_0+{\beta }_1{KOFFiGI}_{it}+{\beta }_2{FBA}_{it}+{\beta }_3{PGR}_{it}+{\beta }_4{CLECP}_{it}+{\beta }_5{CGD}_{it}+{\beta }_6{IQI}_{it}+{\phi }_i+{\mu }_t+{\epsilon }_{it}\left(12\right)$

• Model 5

${OMIV}_{it}={\beta }_0+{\beta }_1{KOFFiGIdf}_{it}+{\beta }_2{FBA}_{it}+{\beta }_3{PGR}_{it}+{\beta }_4{CLECP}_{it}+{\beta }_5{CGD}_{it}+{\beta }_6{IQI}_{it}+{\phi }_i+{\mu }_t+{\epsilon }_{it}\left(13\right)$

• Model 6

${OMIV}_{it}={\beta }_0+{\beta }_1{KOFFiGIdj}_{it}+{\beta }_2{FBA}_{it}+{\beta }_3{PGR}_{it}+{\beta }_4{CLECP}_{it}+{\beta }_5{CGD}_{it}+{\beta }_6{IQI}_{it}+{{\phi }_i+{\mu }_t+\epsilon }_{it}\left(14\right)$

where:

• FBA is the fiscal balance,
• PGR is the population growth rate,
• CLECP is the change in the leading export commodity price,
• CGD is the central government debt
• IQI is the institutional quality index.
• β₀ is the intercept or regression constant
• β₁ – β₆ are the regression coefficients/parameters to be estimated
• i is the cross-section units,
• t is the time series dimension
• ${\phi }_i$ is the unobserved country-specific effect
• ${\mu }_t$ is the unobserved time-specific effect
• ${ϵ}_{it}$ is the stochastic error term.

3.2.2. Model Estimation Techniques

The pre-estimation and model specification tests indicate the presence of slope heterogeneity, heteroskedasticity, autocorrelation, and cross-sectional dependence. Due to these econometric issues, the study employed the panel-corrected standard errors (PCSE) technique to estimate the main econometric models. The PCSE technique is renowned for addressing issues related to autocorrelation, heteroskedasticity, and cross-sectional dependence (Osinubi et al., 2025) and performs efficiently with dynamic heterogeneous panel data (Ikpesu et al., 2019). However, in their studies, it was evident that PCSE may not be effective in addressing endogeneity in regression models (Adeleye et al., 2022; Osinubi et al., 2025). Alternatively, a two-step GMM (2SGMM) estimator was deployed to ensure robustness and consistency. This is because GMM is efficient when the panel dataset is characterized by small T and large N, i.e., when there are few periods and many cross-sectional units, when there is a dynamic dependent variable that depends on its past values, when there is a presence of endogeneity, unobserved heterogeneity (fixed individual effects), and when there is a presence of heteroscedasticity and autocorrelation within individual units but not across cross-sectional units (Roodman, 2009). The first difference GMM (FD-GMM), introduced by Arellano and Bond (1991), works by transforming all explanatory variables into first differences and using the generalized method of moments to estimate parameter coefficients. FD-GMM can solve residual heterogeneity and endogeneity issues, but may produce inefficient results when the number of observations is small (Yang & Liu, 2016), and suffers from weak instruments in small samples (Adjei et al., 2025). The empirical work by Arellano and Bover (1995) and Blundell and Bond (1998) addressed small-sample bias issues by introducing system GMM (SGMM). The system GMM produces parameter estimates and achieves model efficiency by combining the level equations with the first-difference equation and adding lagged first-difference terms as instrumental variables. Therefore, the two-step system GMM (2SGMM) is particularly effective in addressing regressor-induced endogeneity, often caused by omitted variables, measurement error, or reverse causality (Adjei et al., 2025). To assess the instruments’ validity, two specification tests were conducted. The first is the Sargan/Hansen test for over-identification restrictions, while the second examines the hypothesis of no second-order serial correlation in the first-difference residuals (Yaya, 2024). The major challenge with GMM techniques is identifying valid instruments (Akinyemi, 2026; Hassan et al., 2026). The methodological rigor and thoroughness of the study have ensured that it will enhance the reliability and validity of its empirical outcomes.

4. Results and Discussion

4.1. Construction of Institutional Quality Indexa

The institutional quality index (IQI) was constructed based on the six World Bank governance indicators using the principal component analysis (PCA) technique, following the approach by Adeleye et al. (2023), PCA helps to combine strongly and positively correlated variables into an index, with the index constructed correlating with individual institutional quality indicators: Control of corruption (COC), government effectiveness (GEF), political stability (POS), regulatory quality (RQU), rule of law (ROL), and voice and accountability (VAA).

Panel A of Appendix 3 shows the Eigenvalues and eigenvectors. The first component (PC1) has an Eigenvalue of 4.817 and explains 80.3% of the series’ variation. The second component (PC2) has an Eigenvalue of 0.488, explaining only 8.84%% of the variation. Typically, an Eigenvalue greater than one signifies that the component captures more variance than its nominal proportion of the variables’ total variance. Therefore, the first component was used to generate the Institutional Quality Index (IQI). Panel B of Appendix 3 displays the correlation coefficients between the IQI and the six governance indicators, indicating a high positive correlation. This correlation structure suggests that the IQI can account for these variables simultaneously. For instance, IQI has a perfect and positive correlation with the rule of law.

4.2. Descriptive Statistics and Correlation Analysis

Descriptive statistics

Descriptive statistics are necessary to provide insights into the features and attributes of a distribution, typically including measures of central tendency, dispersion, skewness, and kurtosis. The mean, a measure of average; the standard deviation, a gauge of variance; the minimum and maximum values for each variable; and the total number of observations are presented in Table 4. The mean value of AHMIV is 9.0060 with a standard deviation of 58.258, minimum and maximum values of 0.466 and 1750.583. The standard deviation is greater than the mean, indicating a wide spread of the dataset around the mean. This suggests that predicting the future would be somewhat difficult. Also, this connotes that AHMIV is prone to volatility. The difference between the maximum and minimum values reflects the dataset’s range. OMIV has a mean of 6.929, a standard deviation of 50.243, a minimum of 0.262, and a maximum of 1754.914. A higher standard deviation than the mean value indicates wide dispersion of the dataset around the mean. Also, it implies that OMIV suffers from greater dataset inconsistency and unpredictability due to greater volatility. The dataset’s range is 11754.652, the difference between the maximum and minimum values.

The mean, standard deviation, minimum, and maximum values of KOFFIGI are 44.787, 11.622, 22.207, and 84.640, respectively. The standard deviation is lower than the mean value, implying minimal spread or dispersion of the dataset around the mean. Furthermore, it suggests a consistent, predictable dataset pattern for KOFFIGI. The dataset’s range is the difference between its maximum and minimum values. KOFFIGIDF has a mean value of 47.543, standard deviation of 15.537, minimum and maximum values of 16.447 and 99.068, respectively. A lower standard deviation indicates minimal spread of the dataset around the mean. It connotes a consistent, predictable dataset and is less susceptible to fluctuations. The difference between the maximum and minimum values is the dataset’s range. The mean value of KOFFIGIDJ is 41.846, with a standard deviation of 14.304, a minimum value of 17.972, and a maximum value of 77.131. The standard deviation is less than the mean, indicating minimal dispersion of the dataset around the mean. It suggests that the dataset is vulnerable to fluctuations, leading to an inconsistent, unpredictable structure. The dataset’s range is the difference between its maximum and minimum values. The descriptive statistics for the control variables are also presented in Table 4, indicating that the standard deviation values do not exceed the maximum values, suggesting that the dataset does not contain a high degree of extreme values.

Figure 1 further presents the trends in AHMIV, OMIV, KOFFIGI, KOFFIGIDF, and KOFFIGIDJ.

The pairwise correlation matrix results, in Table 5, show that the only higher collinearity among the explanatory variables was observed between overall financial globalisation and financial globalisation de facto index; between overall financial globalisation and financial globalisation de jure index, which are not used simultaneously in the same equation. Therefore, serious multicollinearity challenges leading to a singular matrix are not anticipated in this study. Additionally, a multicollinearity test was also conducted to assess concerns about the reliability of parameter estimates. Alem (2020) notes that two independent variables with a Pearson correlation coefficient exceeding 0.9 can cause multicollinearity issues. None of the independent variables has a high correlation coefficient. To check for multicollinearity, however, variance inflation factors were computed for each sub-model. The variance inflation factors (VIFs) are presented in Table 6, confirming the absence of multicollinearity. This conclusion is in line with the suggestion by Mehmood and Mustafa (2014) and Ikpesu (2024) that there is a case of severe multicollinearity only when the VIF exceeds 10, which usually occurs when $R^2$ exceeds 0.90.

4.3. Pre-Estimation and Model Specification Test Results

Testing for cross-sectional dependence amongst the cross-sectional units included in the study is necessary prior to examining the panel data unit root. The outcome will be helpful in determining the appropriateness or otherwise of the estimation technique. The null hypothesis of cross-sectional independence is rejected because the p-values of the CD-Test statistics are less than 0.05, thereby leading to acceptance of the alternative hypothesis of cross-sectional dependence. This suggests that the macroeconomic challenges encountered in these countries can be traced and linked to their similar economic conditions. Given the presence of cross-sectional dependence among the selected sub-Saharan African countries, it is imperative to conduct both first- and second-generation panel unit root tests. It should be noted that the second-generation unit roots tests of Cross-Sectional Augmented IPS (CIPS), which account for cross-sectional dependence, were conducted, and this is reported in Table 4 alongside the Levin, Lin, and Chu (LLC) and the Im, Pesaran, and Shin (IPS) first-generation panel unit root tests. The results of the cross-sectional dependence test in the fixed-effects regression model are reported in Table 8, providing irrefutable evidence of its presence.

Two categories of panel data unit root tests are employed: first-generation (panel data unit root tests without cross-sectional dependence) and second-generation (panel data unit root tests with cross-sectional dependence). Table 7 presents the unit root results. From the table, all the series become stationary in their first differences using the first-generation tests (LLC and IPS). The second-generation panel unit root test of (CIPS), which accounts for cross-sectional dependence, showed that all the series were stationary after 1st differences.

Panel data combine time-series and cross-sectional information, thereby improving parameter estimates. However, it is unlikely that the parameters of interest exhibit homogeneous characteristics (Blomquist & Westerlund, 2013). Therefore, testing for slope homogeneity is germane to determine the appropriateness or otherwise of heterogeneous panel techniques. Based on the Blomquist and Westerlund (2013) homogeneity test results in Table 8, the model parameters are heterogeneous. This is because the null hypothesis of homogeneous slope coefficients is strongly rejected in favour of the alternative hypothesis of heterogeneous slope coefficients for the AHMIV equation, whereas the reverse holds for the OMIV equation. This is because the Blomquist and Westerlund (2013) homogeneity test statistic is statistically significant at the 1 per cent level.

Besides, the cointegration test result specified by Westerlund (2007), which is more appropriate in the presence of cross-sectional dependence, as shown in Table 8, indicates a long-run relationship among the research variables. The group-wise heteroskedasticity test using the modified Wald test confirmed the presence of heteroskedasticity when the data were analyzed by ordinary least squares (OLS). Also, the Wooldridge serial correlation test result in Table 8 implies that autocorrelation is present in the models when an OLS model was estimated as part of pre-estimation protocols. All these issues informed the deployment of the panel-corrected standard error (PCSE) as the primary estimation technique, with the 2SGMM estimator for robustness checks. Four models are estimated using each of the two model estimation methods.

4.4. Panel Econometric Models

After examining the statistical and econometric properties of the variables, panel regression models were estimated to test the hypotheses and achieve the study’s objectives. Therefore, six (6) models were estimated, first using the baseline estimation technique, PCSE. Model 1 examines the effect of the overall financial globalization index on Adjusted Hanke Misery index volatility. Model 2 examines the effect of the financial globalization de facto index on the Adjusted Hanke Misery index volatility, while Model 3 tests the effect of the financial globalization de jure index on the Adjusted Hanke Misery index volatility. Similarly, Model 4 examines the effect of the overall financial globalization index on the Okun Misery index volatility. Model 5 examines the effect of the financial globalization de facto index on the Okun Misery index volatility, while Model 6 tests the effect of the financial globalization de jure index on the Okun Misery index volatility

The PCSE parameter estimates of Model 1 in Table 9 indicate that the overall financial globalisation index (KOFFIGI) is positively related to the Adjusted Hanke Misery Index Volatility (AHMIV). Thus, a unit increase in the overall financial globalisation index would cause a 2.945% significant increase in Adjusted Hanke Misery Index Volatility, other things being equal. As regards the control variables, the population growth rate (PGR) and changes in the leading export commodity price (CLECP) are positively and significantly associated with the Adjusted Hanke Misery Index Volatility. Conversely, fiscal balance (FBA), central government debt (CGD), and institutional quality (IQI) have significant negative coefficients, suggesting that they help reduce economic misery index volatility in SSA countries. Model 2 shows that the financial globalization de facto index has a positive and significant coefficient (2.5802), implying that a unit increase in the financial globalization de facto index would increase the Adjusted Hanke Misery Index Volatility by 2.580%. The signs and the significance of the control variables align with those in Model 1. However, Model 3 shows that the financial globalization de jure index has a negative and significant coefficient (0.8744), implying that a unit increase in the financial globalization de jure index would reduce the Adjusted Hanke Misery Index Volatility by 0.874%. The signs and significance of the control variables follow those in Model 1. The effects of financial globalization indicators on the Okun Misery Index Volatility (OMIV) are examined next. Model 4 indicates that the overall financial globalization index has a positive and significant coefficient (2.7541), implying that a unit increase in the overall financial globalization would increase the Okun Misery Index Volatility by 2.754%. The effects of the control variables are consistent with the results obtained in Models 1-3. Similarly, Model 5 shows that the financial globalization de facto index has a positive and significant coefficient (2.8151), indicating that a unit increase in the index would increase the Okun Misery Index Volatility by 2.815%. The signs and the significance of the control variables align with those in Model 1, except that fiscal balance has a negative but non-significant coefficient. However, Model 6 shows that the financial globalization de jure index has a negative and significant coefficient (0.9383), implying that a unit increase in the financial globalization de jure index would reduce the Okun Misery Index Volatility by 0.938%.

Figure 2 corroborates the relationships between the research variables.

Robustness and Sensitivity Analysis

The same number of models estimated using the PCSE was re-estimated using the two-step system generalized method of moments (2SGMM) to assess robustness and sensitivity. These models are shown in Table 10. The positive and significant effect of the overall financial globalization index on the adjusted Hanke misery index volatility, the Okun misery index volatility index, and the positive link that the financial globalization de facto index has with the Okun Misery index volatility are found to be robust when the 2SGMM technique was used, overwhelmingly indicating that overall and de facto financial globalization measures can aggravate macroeconomic volatility in the SSA countries. However, the financial globalization de jure index has a non-significant positive effect on adjusted Hanke misery index volatility in Model 3, while it has a significant positive effect on the Okun misery index volatility, as shown in Model 6 of Table 10. This result shows a reversal in signs when compared with the results obtained using the PCSE estimation technique. The positive effect could be due to geoeconomic and geopolitical fragmentation, which often leads countries to impose capital account restrictions on others. It could be a result of the indirect effect of hostility to foreign investment, including withdrawals, the rationalization of fiscal incentives, asset expropriation, and the erection of artificial barriers to business registration for foreign entities in the domestic economy. There is also a sign reversal in the effect of the population growth rate. This could be due to the advantage that population growth has in terms of the supply of cheap labour and domestic consumption, which could help minimize macroeconomic volatility during economic and financial market disturbances. Nevertheless, the significant and negative link between institutional quality and the adjusted Hanke misery index volatility, and the Okun misery index volatility index is robust and consistent. Just as the significant and positive association between changes in the leading export commodities and the adjusted Hanke misery index volatility is robust, as depicted in Models 2 and 3.

4.5. Discussion of Findings

The empirical results show that the overall financial globalization index and the financial globalization de facto index contribute to an increase in the adjusted Hanke misery index volatility and the Okun misery index volatility, using the PCSE parameter estimates. There are different channels through which financial globalization indicators increase macroeconomic volatility. First, financial globalization fosters contagion risk, whereby financial market crises in foreign countries are transmitted to domestic asset markets in SSA countries. Second, the existence of financial market imperfections, which make market arbitrage and speculative transactions thrive on a significant scale. Third, financial globalization facilitates easy access to international financial markets for borrowing, leading to excessive accumulation of external debt at high interest rates due to low sovereign credit ratings, which in turn increases the risk of high inflation, fiscal vulnerability, and lack of future ability to finance development programs. Fourth, the SSA countries have also been bedevilled by low levels of physical capital accumulation, which increases developing countries’ susceptibility to macroeconomic volatility. Fifth, SSA countries are equally susceptible to capital flow reversals, illegal capital flight, and the siphoning of funds overseas by the political class due to corruption, tax avoidance, and evasion schemes, thereby depleting domestic capital and ultimately raising the cost of capital in the domestic economy. Sixth and last, financial globalization can fail to reduce macroeconomic volatility due to the low level of FDI flowing to the SSA region. The low FDI flow is attributable to geoeconomic fragmentation and geopolitical polarization, as well as poor governance quality in many SSA countries. The increasing effect of financial globalization on macroeconomic volatility in this study squares with prior studies in this area of study (Kose et al., 2003; Neaime, 2005; Devereux & Yu, 2019; Tolulope & Charles, 2020; Dada et al., 2025).

Conversely, it has been shown that the de jure financial globalization index can lower macroeconomic volatility indicators. This is instructive and can be achieved through various mechanisms. First, the reducing effect could be due to an increase in the pursuit of financial liberalization by SSA countries, which enact laws and regulations that impede cross-border capital flows. Also, increased financial openness and regional cooperation on investment could have facilitated foreign countries’ participation in the domestic economy, thereby increasing jobs, growing the economy, and helping control inflation risk. Second, the ability to borrow in international financial markets to raise funds for infrastructure and other development projects. Finally, the reducing effect of financial globalization could be through reduced country risk premiums, which occur when countries perceive financial liberalization enabled by laws and regulations and investment pacts that allow an unconstrained flow of funds. Reduced country risk premiums translate into lower capital costs, which should boost economic growth, investment, and consumption. The key point here is that the de facto financial globalization measure dominates the de jure measure, since the empirical results using the overall financial globalization measure align with the de facto measure. Additionally, these results suggest that countries may implement laws and regulations that encourage cross-border capital flows and investments, but actual capital flows may differ and may not respond to the capital account liberalization strategy. This is intuitive, as many SSA countries still have access to only a disproportionately low share of global FDI flows and are chronically susceptible to capital reversals due to a highly capricious and pro-cyclical capital inflow structure. The lowering effect of financial globalization on macroeconomic volatility in this study conforms with prior studies in this area of study (Buch et al., 2005; Aizenman et al., 2011; Meller, 2011; Yadav et al., 2018).

These findings corroborate the real business cycle theory, which predicts that capital flow fluctuations can engender macroeconomic volatility. Conversely, the empirical findings fail to support the external capital theory that infers external capital flows from capital-rich countries to capital-rich countries should foster macroeconomic stability. This misalignment suggests the presence and existence of the Lucas Paradox, which argues that external capital flows, such as FDI, do not always achieve their macroeconomic objectives due to domestic structural issues and poor institutional quality in developing countries.

Coming to the control variables, the fiscal balance (FBA) has a negative and significant coefficient on the adjusted Hanke misery index volatility. A plausible explanation for this finding could be that the fiscal balance position is favourable to macroeconomic performance indicators through capital formation, or that the fiscal deficit, which averages -2.31% as shown in the descriptive statistics, poses no serious macroeconomic problems. This finding corroborates the work of Kose et al. (2003), Teixeira et al. (2008), Mendonça and Nunes (2010), Baldacci et al. (2011), Kennedy and Palerm (2014), and Ashogbon et al. (2023), who discovered that a fiscal balance position helps to taper macroeconomic volatility. The population growth rate is shown to exacerbate macroeconomic volatility indicators. This implies that the population growth rate corresponds with increasing macroeconomic volatility in the SSA region. This finding does not align with theoretical expectations regarding the relationship between population growth rate and macroeconomic volatility. The possible explanation for this aggravating effect of population growth rate on macroeconomic volatility could be issues related to informality and economic inequality that make fiscal consolidation policy ineffective. Another possible explanation could be the low productive capacities that generally exist in SSA countries due to poor human capital development and inefficient labour skills, which have made industrialization, technological innovation, and economic productivity difficult. This finding is consistent with the previous findings by Ali et al. (2019) that revealed population growth rate as a significant source of macroeconomic volatility.

Furthermore, changes in leading export commodity prices have been found to contribute to increases in macroeconomic volatility indicators, suggesting that fluctuations in these prices are a strong source of macroeconomic volatility in SSA countries. This finding supports the empirical results reported by Neaime (2005), Sanvicente and Carvalho (2020), and Abanikanda and Dada (2023), which conclude that export commodity price aggravates macroeconomic volatility indicators. Additionally, the central government debt is found to have a mitigating effect on the macroeconomic volatility. The economic rationale is that debt capital helps bridge financing gaps, thereby enabling the funding of capital projects and supporting social programs. This result confirms the previous findings by Aktug et al. (2013) and Yang and Liu (2016), which showed that debt levels reduce macroeconomic volatility indicators. Finally, the institutional quality index is found to be strongly lower than the macroeconomic volatility index. This indicates that robust institutional arrangements foster good governance, political stability, the rule of law, sound regulations, and a generally conducive environment, allowing businesses and investments to thrive. When businesses thrive, national economic productivity, global competitiveness, currency value stability, price stability, and job opportunities are enhanced. This result is consistent with empirical findings by Ahmed and Suardi (2009), Kennedy and Palerm (2014), Bezooijen and Bikker (2017), and Ashogbon et al. (2023).

5. Conclusions, Policy Implications, and Areas for Future Studies

5.1. Conclusions and Policy Implications

Macroeconomic volatility has remained chronic and persistent in SSA countries despite prior efforts to liberalize external financial markets, which have ensured unrestricted cross-border flows of capital and investment. This reform is expected to increase capital stock, promote economic growth, and foster macroeconomic stability. The extent to which financial globalization forces have shaped macroeconomic volatility in SSA countries is unclear. Therefore, this study investigates the empirical link between financial globalization variables and macroeconomic volatility in 39 SSA countries. The study obtained data from various reputable sources and used the panel-corrected standard error (PCSE) estimator, which is efficient in the presence of cross-sectional dependence, heteroskedasticity, and autocorrelation. Also, the study employed the two-step GMM (2SGMM) to obtain alternative parameter estimates for robustness and sensitivity analyses.

The empirical results indicate that financial globalization indicators, such as the overall financial globalization index (which covers both actual capital flows and laws and regulations on capital flows) and the financial globalization de facto index (which measures actual capital flows and stock), are significantly related to macroeconomic volatility. Whereas the financial globalization de jure index, which qualitatively assesses the ease or difficulty of capital mobility and investment cooperation, shows a reduction in macroeconomic volatility in SSA countries. This conclusion supports the proposition of real business cycle theory that fluctuations in external factors, such as FDI flows, can increase the levels of perturbations in macroeconomic variables. Additionally, the study indicates that fiscal balance, central government debt, and institutional quality are associated with decreasing macroeconomic volatility. Conversely, the population growth rate and changes in the prices of leading export commodities have an increasing effect on macroeconomic volatility in the SSA region.

Therefore, several policy implications have emerged from this study’s empirical findings. First, policymakers and regulators in SSA countries need to deepen reform efforts to ensure that the critical mass of external capital, especially FDI, is attracted to the SSA region, thereby increasing the capital available to fund investments for sustainable economic growth and macroeconomic stability. Two, the financial regulations and laws in SSA countries must remain clear and supportive of external capital flows and investments, such as cross-border mergers and acquisitions, asset trades in capital markets, and other real-sector investments. This also implies that tax laws and fiscal regulations must not be hostile to external capital and investments to build foreign investors’ confidence in the domestic economy. Three, SSA countries must devise a means of managing risks associated with geoeconomic fragmentation, international financial market imperfections, and geopolitical polarization that affect FDI allocation, access to capital, supply chain security, and trade flows. Fourth, sound fiscal consolidation and debt management practices are imperative for economic growth and macroeconomic stability through economic diversification, efficient public finance management, fiscal deficit minimization, avoidance of monetizing fiscal deficits, and improved revenue mobilization. Fifth, there is a need to reduce over-reliance on the export of commodities, which are often vulnerable to external price shocks, by promoting industrialization, structural transformation, and economic diversification in SSA economies. Natural resource rents could create some competitive advantage when efficiently managed through infrastructure financing, capital stock accumulation, and human capital development. Sixth, policymakers and governments in SSA countries should design policies that create economic opportunities for their populations and increase investment in skill development to improve economic productivity. The sheer population provides a veritable market for local consumption of goods and services if the populace earns a livable wage that sustains production and investment systems. Finally, the SSA governments should strengthen institutional effectiveness to ensure the rule of law, political stability, accountability, corruption control, predictable and fair regulations, respect for human rights, the absence of violence, the sanctity of property rights, and restraints on illegal asset seizures and expropriations. These factors are crucial to building confidence among investors and help reduce sovereign risk premiums ascribable to SSA countries, many of which have been afflicted by these challenges and have found it difficult to achieve sovereign credit rating upgrades and secure external capital at efficient borrowing costs.

5.2. Limitations of the Study

This study used secondary data from various sources, including the World Bank, the IMF, and other renowned multinational and multilateral institutions in governance, finance, and economics. This suggests that the researcher does not control the data collection process but relies solely on the data as published by these institutions. Therefore, any imprecision, inaccuracy, errors, or bias in the dataset may be difficult to ascertain, and their effect may be impossible to quantify. Notwithstanding these possibilities, the sources this study drew data from are widely used for empirical studies of this nature, which provides considerable comfort and confidence. Therefore, it is believed that these limitations do not detract from the findings’ wide acceptability, reliability, quality, replicability, efficiency, or policy relevance, or from the study’s unique contribution to empirical literature in this area of research endeavour.

5.3. Suggestions for Further Research

This study examined the effect of financial globalization on macroeconomic volatility in Sub-Saharan African (SSA) countries, while accounting for the relevant control variables. The amount of study in this area is very limited, and the existing results are very confounding and divergent. Hence, there remains a need to investigate the relationship between financial globalization and macroeconomic volatility in other world regions. Also, there is a need to examine whether financial globalization has a U-shaped, inverted U-shaped, or S-shaped relationship with macroeconomic volatility in different world economic regions.

Appendix A.1. List of 39 Sampled Sub-Saharan African Countries

Angola	Chad	Gambia	Mauritius	Sierra Leone
Benin	Comoros	Ghana	Mozambique	South Africa
Botswana	Cote d’Ivoire	Guinea	Namibia	Sudan
Burkina Faso	Equatorial Guinea	Guinea-Bissau	Niger	Tanzania
Burundi	Eritrea	Kenya	Nigeria	Togo
Cabo Verde	Eswatini	Lesotho	Rwanda	Uganda
Cameroon	Ethiopia	Madagascar	São Tomé and Príncipe	Zambia
Central African Republic	Gabon	Mali	Senegal

Appendix A.2. SSA Countries and Their Leading Export Commodities

S/n	Country	Region	Leading Export Commodities	Selected Item	Unit	In the Sample?
1	Angola	Southern Africa	Crude oil, diamonds	Crude oil	$/bbl	Yes
2	Benin	Western Africa	Cotton, gold	Cotton	$/kg	Yes
3	Botswana	Southern Africa	Pearls, precious metals	Copper	$/mt	Yes
4	Burkina Faso	Western Africa	Gold, cotton	Gold	$/troy oz	Yes
5	Burundi	Central Africa	Gold, coffee	Gold	$/troy oz	Yes
6	Cabo Verde	Western Africa	Fishery products, petroleum oils	Fishery products	$/mt	Yes
7	Cameroon	Central Africa	Petroleum oils, bitumen oils	Petroleum oils	$/bbl	Yes
8	Central African Republic	Central Africa	Cotton, pearls	Cotton	$/kg	Yes
9	Chad	Central Africa	Petroleum oils, gold	Petroleum oils	$/bbl	Yes
10	Comoros	Eastern Africa	Spices, iron ores	iron ores	$/dmtu	Yes
11	Congo, Dem. Rep.	Central Africa	Copper, non-ferrous metal	Copper	$/mt	No
12	Congo, Rep.	Central Africa	Petroleum oils, copper	Petroleum oils	$/bbl	No
13	Cote d’Ivoire	Western Africa	Cocoa, petroleum oils	Cocoa	$/kg	Yes
14	Equatorial Guinea	Central Africa	Petroleum oils, natural gas	Petroleum oils	$/bbl	Yes
15	Eritrea	Eastern Africa	Copper, gold	Copper	$/mt	Yes
16	Eswatini	Southern Africa	Forestry products, sugar	Sugar	$/kg	Yes
17	Ethiopia	Eastern Africa	Coffee, vegetables	Coffee	$/kg	Yes
18	Gabon	Central Africa	Petroleum oils, forestry products	Petroleum oils	$/bbl	Yes
19	Gambia, The	Western Africa	Forestry products, fruits and nuts	Forestry products	$/sheet	Yes
20	Ghana	Western Africa	Gold, petroleum oils	Gold	$/troy oz	Yes
21	Guinea	Western Africa	Gold, aluminium ores	Gold	$/troy oz	Yes
22	Guinea-Bissau	Western Africa	Fruits & nuts, forestry products	Forestry products	$/sheet	Yes
23	Kenya	Eastern Africa	Tea & mate, crude vegetable materials	Tea & mate	$/kg	Yes
24	Lesotho	Southern Africa	Pearls, non-alcoholic beverages	Pearls (Platinum)	($/troy oz)	Yes
25	Liberia	Western Africa	Iron ore, gold	Iron ore	$/dmtu	No
26	Madagascar	Eastern Africa	Spices, nickel	Nickel	$/mt	Yes
27	Malawi	Southern Africa	Tobacco, sugar	Tobacco	$/mt	No
28	Mali	Western Africa	Gold, cotton	Gold	$/troy oz	Yes
29	Mauritania	Eastern Africa	Iron ore, fishery products	Iron ore	$/dmtu	No
30	Mauritius	Eastern Africa	Fishery products, sugar	Fishery products	$/mt	Yes
31	Mozambique	Southern Africa	Aluminum, coal	Aluminium	$/mt	Yes
32	Namibia	Southern Africa	Pearls, fishery products	Fishery products	$/mt	Yes
33	Niger	Western Africa	Iron ores, petroleum oils	Iron ores	$/dmtu	Yes
34	Nigeria	Western Africa	Petroleum oils, natural gas	Petroleum oils	$/bbl	Yes
35	Rwanda	Eastern Africa	Iron ores, gold	Iron ores	$/dmtu	Yes
36	Sao Tome and Principe	Central Africa	Cocoa, petroleum oils	Cocoa	$/kg	Yes
37	Senegal	Western Africa	Petroleum oils, gold	Petroleum oils	$/bbl	Yes
38	Seychelles	Eastern Africa	Fishery products, petroleum oils	Fishery products	$/mt	No
39	Sierra Leone	Western Africa	Iron ore, pearls	Iron ore	$/dmtu	Yes
40	Somalia	Eastern Africa	Live animals, gold	Live animals	$/kg	No
41	South Africa	Southern Africa	Silver, platinum, gold	Silver	$/troy oz	Yes
42	South Sudan	Eastern Africa				No
43	Sudan	Eastern Africa	Petroleum oils, gold	Petroleum oils	$/bbl	Yes
44	Tanzania	Eastern Africa	Gold, fruits & nuts	Gold	$/troy oz	Yes
45	Togo	Western Africa	Gold, petroleum oils	Gold	$/troy oz	Yes
46	Uganda	Eastern Africa	Coffee, gold	Coffee	$/kg	Yes
47	Zambia	Southern Africa	Copper, tobacco	Copper	$/mt	Yes
48	Zimbabwe	Southern Africa	Tobacco, gold, nickel ores	Gold	$/troy oz	No
Data Source: World Bank Commodity Price Index (2024).

Appendix A.3. Principal Components of Governance Indicators for the IQI Construction

Panel A: Eigenvalues Eigenvalues: Sum =6; Average = 1)
Component				Eigenvalue		Difference		Proportion			Cumulative Eigenvalue		Cumulative Proportion
1				4.817		4.329		0.803			4.817		0.803
2				0.488		0.146		0.081			5.305		0.884
3				0.342		0.121		0.057			5.647		0.941
4				0.221		0.146		0.037			5.868		0.978
5				0.075		0.018		0.013			5.943		0.991
6				0.057		0.000		0.010			6.000		1.000
Eigenvectors (loadings)
Variable		PC 1		PC 2		PC 3		PC 4			PC 5		PC 6
COC		0.412		0.061		-0.376		-0.751			0.342		0.071
GEF		0.428		-0.294		-0.292		0.226			-0.420		0.647
POS		0.357		0.861		-0.011		0.331			0.108		0.105
RQU		0.418		-0.406		0.013		0.461			0.638		-0.201
ROL		0.443		-0.031		-0.134		-0.001			-0.529		-0.711
VAA		0.388		-0.060		0.869		-0.252			-0.083		0.143
Panel B: Corelation between IQI and governance indicators
Variables	Description		IQI		COC		GEF		POS	RQU		ROL		VAA
IQI	Institutional quality index		1.000
COC	Control of corruption		0.878		1.000
GEF	Government effectiveness		0.920		0.831		1.000
POS	Political stability		0.739		0.683		0.629		1.000
RQU	Regulatory quality		0.878		0.753		0.912		0.585	1.000
ROL	Rule of law		1.000		0.878		0.920		0.739	0.878		1.000
VAA	Voice and accountability		0.785		0.696		0.716		0.620	0.765		0.785		1.000