The role of financial globalization in the long-run volatility between forex and stock markets during COVID-19: Evidence from Africa

Insaidoo, Michael; Brafu-Insaidoo, William Gabriel; Peprah, James Atta; Cantah, William Godfred Article The role of financial globalization in the long-run volatility between forex and stock markets during COVID-19: Evidence from Africa Research in Globalization Provided in Cooperation with: Elsevier Suggested Citation: Insaidoo, Michael; Brafu-Insaidoo, William Gabriel; Peprah, James Atta; Cantah, William Godfred (2024) : The role of financial globalization in the long-run volatility between forex and stock markets during COVID-19: Evidence from Africa, Research in Globalization, ISSN 2590-051X, Elsevier, Amsterdam, Vol. 9, pp. 1-10, https://doi.org/10.1016/j.resglo.2024.100242 This Version is available at: https://hdl.handle.net/10419/331170 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc/4.0/ The role of financial globalization in the long-run volatility between forex and stock markets during COVID-19: Evidence from Africa Michael Insaidoo a , * , William Gabriel Brafu-Insaidoo b , James Atta Peprah c , William Godfred Cantah b a Department of Economics and Actuarial Science, Faculty of Accounting and Finance, University of Professional Studies, Accra, Ghana b Department of Data Science and Economic Policy, School of Economics, University of Cape Coast, Ghana c Department of Applied Economics, School of Economics, University of Cape Coast, Ghana ARTICLE INFO Keywords: Stock market volatility Exchange rate volatility Financial globalization COVID-19 Long-run volatility ABSTRACT This study examines the long-run volatility between forex and stock markets and the role of financial globalization in this relationship in Africa, during the COVID-19 pandemic period, using panel Fully Modified Ordinary Least Squares (FMOLS) and panel Dynamic Ordinary Least Squares (DOLS) approaches. Our empirical outcomes revealed bi-directional long-run volatility between the two financial markets in the COVID-19 pandemic period. The results further established that, financial globalization reduces forex markets’volatility effects on stock markets’volatility, whilst it heightens stock markets’volatility effects on forex markets’volatility in Africa, during the COVID-19 pandemic period. The implications of this study, include the need to harness the stabilising potential of financial globalization in the long-run volatility between forex and stock markets, primarily through asset diversification, enhanced information flow, and market efficiency in African financial markets. Introduction Increased international equity flows heighten the demand and supply of currencies in which these equities are denominated, resulting in some level of interdependence between the forex and stock markets. This integration of financial markets exposes these markets to fluctuations in global and national economies (Lakshmanasamy, 2021). Similarly, exchange rates are vulnerable to fluctuations in global financial market which tends to affect economies around the globe. In market economies, the mechanisms of free-floating and managed floating exchange rates, to some degree, douses the exchange rate volatility (Karoui, 2006). Domestic stock prices and its returns are vulnerable to fluctuations in global markets and exchange rate volatility, which causes volatility in domestic stock market. Extended stock market volatility disrupts the price mechanism of the capital markets, which can result in holders of foreign equities to face greater risks of exchange rate volatility. Globalization over the past five decades have led to increased interest in international equity investments (United States’National Research Council [NRC], 1995). The upsurge in investments in international equity has occasioned a hike in activities in foreign currencies’ market. As a result of the increased interdependency, volatility between stock and forex markets is heightened resulting in riskier investments in international portfolio, which further leads to poor performance of these investments (Kanas, 2000). Stock and forex markets, which are subsets of financial markets are vulnerable to different crisis, which occasions their increased volatility (Zhao et al., 2023). With COVID-19 pandemic being seen as one of the hardest hit global crisis (Barai &Dhar, 2021; Naseer et al., 2023), does the recent pandemic heightens the comovement of forex and stock markets in Africa. Moreover, Alagidede (2008) shows, that African markets are not well integrated with each other, and revealed weak linkage between these markets and the rest of the world, indicating that Africa’s market respond to local rather than global information. Financial globalization according to Arestis and Basu (2004) is the phenomenon in which financial markets of different countries become interconnected into a unified system or an unrestricted flow of financial resources across international borders. Financial globalization is connected to the stock market in a number of ways. Financial globalization allows investors in one country to invest in stocks of companies in another, leading to increased capital flows, which has the potential to impact stock prices. It also provides investors with greater opportunities * Corresponding author. E-mail address: [email protected] (M. Insaidoo). Contents lists available at ScienceDirect Research in Globalization journal homepage: www.sciencedirect.com/journal/research-in-globalization https://doi.org/10.1016/j.resglo.2024.100242 Received 6 January 2024; Received in revised form 29 July 2024; Accepted 2 August 2024 Research in Globalization 9 (2024) 100242 Available online 6 August 2024 2590-051X/© 2024 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ ). to diversify their portfolios (stocks) internationally, reducing risk and potentially increasing returns. In addition, local companies can raise capital (through sale of stocks) by attracting foreign investment through Foreign Direct Investment (FDI). Further, since financial globalization relies on the forex market, changes in currency values can affect the competitiveness of multinational companies, influencing their stock prices. With respect to financial globalization connection to forex markets, these are the key linkages. As businesses and investors engage in crossborder transactions, the demand for different currencies increases, driving activity in the forex market. Moreover, due to the exposure of investors to various currencies as they engage in international investment, exchange rate fluctuations directly impact the returns on these investments. Additionally, since global trade depends on forex market, exchange rate fluctuations can affect the competitiveness of exports and imports, influencing trade balances. Furthermore, as multinational companies make investments globally, they participate in forex market to manage currency risk and optimize returns on their investments. Financial globalization, according to Cordella and Ospino Rojas (2017), play either a stabilizing or destabilizing role in financial market volatility. The former role provides an avenue for more efficient risk sharing, whilst the latter role provides a vehicle for contagion. Pursuant to this discussion, does financial globalization stabilizes or destabilizes the long-run volatility between stock and forex markets in Africa, during the COVID-19 pandemic period. This study, in addition to examining the long-run volatility between stock and forex markets in eight African countries, during the COVID-19 pandemic period, assesses the moderating role of financial globalization in this relationship. Whilst empirical literature is inundated with studies on the co-movement of stock and forex markets, empirical studies on the moderating role of financial globalization in this relationship is to the best of our knowledge absent. Thus, this study proposes to examine the moderating role of financial globalization in the long-run volatility between stock and forex markets, using eight African countries data, during the COVID-19 pandemic period. The aim is to examine whether these economies integration in the global financial architecture increases or reduces the co-movement of the two financial markets in the African setting. The rest of the study evolves as follows. The second section presents the literature review, whilst the third section provides the data and methodology used for the paper. Discussions on the results are presented in the fourth section, with the fifth section capturing the conclusion and policy implications of the results of the study. Literature review Theoretical review This study is premised on the flow-oriented model propounded by Dornbusch and Fischer (1980), and the stock-oriented model attributed to Branson and Henderson (1985), and Frankel (1992). The floworiented model asserts a positive co-movement of exchange rates and stock prices. The idea underlying this model is that, exchange rate determines the trade balance or current account balance of a country. The model explains that, depreciation in local currency makes local firms competitive with their cheaper products in the international markets, translating into cheaper exports. These competitive exports drive up demand, leading to an increased exports volume, which translate into an appreciation of wealth of local firms, with the potential of an upsurge in the value of domestic stock prices of these local firms. In this regard, the stock market is the recipient of volatility, whilst the foreign exchange market becomes the originator. On the contrary, the stock-oriented model, categorized into portfolio balance model and monetary model, posit that market activities in financial assets (equity and bonds) determines the exchange rate. The portfolio balance model asserts that, when domestic stock prices rise, investors who are holders of foreign and domestic assets, including foreign and domestic currencies would be incentivized to increase their investments in domestic assets. To this end, foreign assets are sacrificed to raise more domestic currency for local investment purposes, resulting in appreciation of the local currency. The monetary model on the other hand, postulates a weak or no co-movement of stock prices and exchange rate. Empirical review Volatility between forex and stock markets in COVID-19 pandemic period The literature on examination of the volatility between the stock and forex markets in the COVID-19 pandemic period is limited. Van Der Westhuizen et al. (2022) using dataset spanning January 1979 to August 2021, investigates the inter-dependence and volatility spillovers between stock and forex markets for South Africa, incorporating the effect of COVID-19 pandemic in these relationships, using the EGARCH methodology. The study shows bidirectional transmission of volatility between forex and stock markets, confirming the presence of contagion between these markets. The study further confirmed that, volatility spillovers were more pronounced during the COVID-19 pandemic period, confirming the hike in contagion during periods of turbulence. In a similar study, Rai and Garg (2022) among other objectives, used the BEKK-GARCH approach, to examine the volatility spillover between stock prices and exchange rates in BRIICS economies. The findings show a general transmission of volatility between the forex and stock markets in most of these economies. Specifically, the study found that, during the COVID-19 period, domestic stock returns plummeted which led to capital outflows resulting in an increase in exchange rates. Financial globalization and financial market volatility Similarly, the studies on the assessment of the impact of financial globalization on financial market volatility is limited. For instance, using the generalized method of moments (GMM) approach, Gaies et al. (2020) explored the effects of financial globalization on growth in developing countries, particularly considering exchange rate. Amongst other findings, the study revealed a negative impact of financial globalization on exchange rate volatility. In a similar study, Esqueda et al. (2012) using a dynamic panel data framework and data over the 1995–2007 period, examined the impact of financial globalization on stock market volatility. The study revealed that, financial globalization reduces stock market volatility in 22 emerging markets, whilst it did not have any impact on stock market volatility in 22 industrial economies. Similarly, Cordella and Ospino Rojas (2017) using 84 countries, assessed the impact of financial globalization on stock market volatility for the 1992 –2016 period. The results established a reduction of stock market volatility by financial globalization in steady periods, whilst it increased the volatility in turbulent periods. The study also indicated that, comparatively, the volatility reduction properties of financial globalization dominate, and is more pronounced in frontier markets. As evidenced in the review of previous studies, none of them has conducted a panel study to examine the moderating role of financial globalization in the long-run volatility between forex and stock markets in Africa, during COVID-19 period. This therefore provides grounds for the present study. Hypothesis To ascertain the moderating role of financial globalization in the long-run volatility between forex and stock markets in Africa, during COVID-19 period, the following hypotheses are stated: Ho: Financial globalization does not moderate the long-run volatility between forex and stock markets during COVID-19. Ha: Financial globalization moderates the long-run volatility between forex and stock markets during COVID-19. M. Insaidoo et al. Research in Globalization 9 (2024) 100242 2 Data, model and the empirical methodology Data To assess the long-run volatility between forex and stock markets, during COVID-19 pandemic period, the study has compiled dataset on daily prices and returns of stock indices and exchange rates of eight African countries, namely, Botswana, Egypt, Ghana, Kenya, Mauritius, Morocco, South Africa, and Tunisia. Eleven stock markets were used for this study, with South Africa and Kenya providing three and two stock markets respectively. These African countries/stock markets were selected based on the availability of data. In the forex market, national local currencies are quoted against the US dollar, which serves as the base currency. The stock prices used for this study are quoted in local currencies. The stock indices and exchange rates data are sourced from investing.com, whilst the financial globalization index is sourced from KOF Swiss Economic Institute. The stock index of Ghana was sourced from the Ghana Stock Exchange. The sample period for the COVID-19 period spans from March 11, 2020, to December 30, 2020. The dataset for this study starts from March 11, 2020 because this is the date the WHO declared COVID-19 as a global pandemic. Whilst the health crisis, the magnitude of the COVID19, has tendency to adversely impact financial markets, in the same vein, the discovery and distribution of COVID-19 vaccines constitute a positive news, which can translate in a positive influence on financial markets. To decouple the possible positive influence of the COVID-19 vaccine in the co-movement of forex and stock markets in Africa, the dataset for this study ends at December 30, 2020. This end date is because on December 31, 2020, WHO issued its first emergency use validation for a COVID-19 vaccine, which was aimed at ensuring equitable global access to COVID-19 vaccines (see World Health Organization, 2020). The sources of data, as well as the description of variables used for the study are reported in Table 1. Justification of variables For this study, the variables of interest are stock market volatility, exchange rate volatility, and financial globalization, whilst the control variables are volatility index, stock returns, and exchange rate returns. To examine the long-run volatility between stock and forex markets, the two dependent variables used for this study are stock market volatility and exchange rate volatility. This is in pursuant to the studies of Hung (2022), Jebran (2018), and Leung et al. (2017). These volatilities were measured using the EGARCH methodology. To achieve the objectives of this study, the independent variable of interest is exchange rate volatility in a model where stock market volatility is the dependent variable, and when exchange rate volatility is the dependent variable, stock market volatility becomes the independent variable of interest. Exchange rate returns have been found to have an influence on exchange rate volatility. The local currency depreciation causes a fall in the rate of returns, which results in an increase in exchange rate volatility (Mohammed et al., 2021). This suggests an inverse relationship between exchange rate returns and its volatility. A company’s profitability induces demand for its stocks, with higher profit levels causing an upsurge in the interest in the company’s stocks (Hermuningsih, 2013). Consequently, a company’s stocks with improved returns (profitability) would likely attract high demand for the company’s stocks. Based on market mechanism, the surge in demand would likely cause a hike in stock price, leading to its increased volatility. This implies a positive association of stock returns to its volatility. This study attempts to investigate the moderating role of financial globalization in the relationship between stock market volatility and exchange rate volatility. To this end, financial globalization has been shown to have an influence on these volatilities (see Esqueda et al., 2012; Gaies et al., 2020). The KOF Swiss Economic Institute provides the financial globalization, de facto index used in this study, which uses five indicators in its measurement. These indicators are foreign direct investment (measured as sum of stocks of assets and liabilities of foreign direct investment as a percentage of Gross Domestic Product (GDP)), portfolio investment (measured as sum of stocks of assets and liabilities of international equity portfolio investments as a percentage of GDP), international debt (measured as sum of inward and outward stocks of international portfolio debt securities and international bank loans and deposits as a percentage of GDP), international reserves (which includes foreign exchange (excluding gold), special drawing rights holdings and reserve position in the International Monetary Fund (IMF) as a percentage of GDP), and international income payments (measured as sum of capital and labour income to foreign nationals and from abroad as a percentage of GDP). The index scores each country out of 100, with a high score indicating how more globalised a country is. Ashraf (2021) asserts that, in a cross-country Table 2 Descriptive statistics during COVID-19 pandemic period. Variable Mean Max Min SD OBS SV 0.0001 0.0019 0.0000 0.0002 2255 XV 0.0000 0.0003 0.0000 0.0000 2255 VIX 27.0613 45.4100 19.9700 5.0491 2255 SR 0.0004 0.1027 −0.1723 0.0108 2255 XR −0.0005 0.0396 −0.0284 0.0062 2255 FG 57.1737 99.000 29.9260 19.8762 2255 Notes: Standard Deviation is denoted by SD, observations are depicted by OBS, SV is Stock market volatility, XV is Exchange rate volatility, VIX is Volatility index, SR is Stock returns, XR is Exchange rate returns, FG is Financial globalization, Min is Minimum, and Max is Maximum. Source: Authors’Construction (2023). Table 1 Summary of Variable Description. Variable Measurement Source Stock volatility EGARCH generated Authors’construction Exchange rate volatility EGARCH generated Authors’construction investing.com Volatility Index CBOE Volatility Index Stock returns Log (stock price in period t/stock price in period t−1) investing.com/Authors’construction Exchange rate returns Log (exchange rate in period t/exchange rate in period t−1) investing.com/Authors’construction Financial globalization Financial globalization, de facto index 2022 KOF Globalization Index –KOF Swiss Economic Institute SV*FG The interactive term of stock volatility and financial globalization Authors’construction XV*FG The interactive term of exchange rate volatility and financial globalization Authors’construction Notes: EGARCH is Exponential Generalized Autoregressive Conditional Heteroscedastic, SV is Stock Market Volatility, XV is Exchange Rate Volatility, and FG is Financial Globalization M. Insaidoo et al. Research in Globalization 9 (2024) 100242 3 setting, due to differences in institutional and cultural environments, investors react differently to similar events. Resultantly, this study controls for these differential reactions by using Volatility index as country fixed-effects variable. Volatility index have been shown to have influence on these volatilities. Feng et al. (2021) for instance, established that, volatility index which depicts fear, increases exchange rate volatility, whilst Zhu et al. (2019) provide evidence of the US stock market volatility being impacted by volatility index. Pre-diagnostic tests Correlation matrix and Variance Inflation Factor (VIF) tests The pairwise correlation matrix is conducted among the variables in the model. The objective of the correlation matrix is to ensure that there is no multicollinearity among the exogenous variables in the model. In addition, the direction and strength between any two variables in the model is shown by this matrix. Similarly, a VIF test, which is another tool for the detection of multicollinearity in a panel regression model is conducted. Cross-sectional dependence (CD) tests The dependence on cross-sections may arise as a result of factors such as financial globalization and the movement of international capital between countries and regions. Consequently, failing to consider the potential cross-sectionality in a panel dataset could result in unreliable and biased estimates. This study employs CD tests introduced by Pesaran et al. (2008) and Pesaran (2021) and the Langrange multiplier (LM) test attributed to Breusch and Pagan (1980) to identify potential collinearity in the dataset. Second-generation unit root tests When CD is detected within a series, it is advisable to perform second-generation unit root tests like Cross-section Im-Pesaran (CIPS) and Cross-section Augmented Dickey Fuller (CADF) unit root tests, to assess the stationarity of the series. The tests are effective in handling cross-sectional dependence and heterogeneity. Panel cointegration tests The panel cointegration tests introduced by Pedroni (1999, 2004) are conducted to test for cointegration among the variables in the model. The Pedroni (1999, 2004) panel cointegration test is a statistical test used to determine whether two or more time series are cointegrated, which means they share a long-run equilibrium relationship despite short-term deviations from this relationship. The test is used to identify cointegration among multiple time series in a panel dataset. The Pedroni (1999, 2004) test is a robust alternative to the standard cointegration test, which can be sensitive to outliers and heteroscedasticity in the data. Model specification To generate the volatility of stock and forex markets, the Exponential Generalized Autoregressive Conditional Heteroscedastic (EGARCH) model attributed to Nelson (1991), which have been widely used in volatility related studies in financial literature is used (see Bal et al., 2018; Jebran, 2018; Ngo Thai, 2019; Insaidoo et al., 2021). The improved stability of optimization routines and the absence of parameter restrictions places the EGARCH model above other models. The EGARCH model is specified as: In σ 2 j,t= ω t+βjln( σ j,t−1)2γ ε t−1  σ 2 t−1 √+θ[| ε t−1|  σ 2 t−1 √− 2 π √](1) where the conditional variance which captures a one-period ahead variance estimated on any prior important event is depicted by σ 2 j,t, as shown in equation (1). The conditional density function is denoted by ω t, whilst the GARCH effect which captures the model’s symmetric effect is depicted by θ.βcaptures the perseverance in conditional volatility, whilst the leverage effect is denoted by γ. In pursuit of robustness of results, two empirical approaches are used in this study. The Panel Fully Modified Ordinary Least Squares (Panel FMOLS) and the Panel Dynamic Ordinary Least Squares (Panel DOLS). Panel FMOLS was used due to its ability to handle time-invariant unobserved heterogeneity, improved estimation of regression coefficients, robustness to omitted variable bias, ability to handle panel data with multiple observations per unit, and flexibility in modelling the error term. Additionally, the Panel DOLS ability to estimate both short-term and long-term relationship, makes it a versatile tool for analysing dynamic relationships. In addition, Panel DOLS ability to handle missing data and non-stationarity, makes it a useful approach for analysing such dataset. The functional form of Panel FMOLS and Panel DOLS are presented as: yit =γi+jitφ+ ∪iti=1,⋯.11,t=11March2020−30December2020 (2) Table 3 Pairwise Correlation Matrix. Variables SV XV VIX SR XR FG SV 1.0000 XV 0.2426*** 1.0000 VIX 0.3280*** 0.1216*** 1.0000 SR 0.0009 0.0499** −0.0802*** 1.0000 XR 0.0091 −0.0913*** 0.0874*** −0.0791*** 1.0000 FG 0.2055*** 0.4861*** −0.0000 0.0252 −0.0364* 1.0000 Notes: *** p <0.01, ** p <0.05, * p <0.10, SV represents Stock volatility, XV denotes Exchange Rate volatility, VIX refers to Volatility Index, SR depicts Stock returns, XR represents Exchange Rate returns, and FG is Financial Globalization. Source: Authors’Construction (2023). Table 4 Variance Inflation Factor. Variable Model 6 Variable Model 8 Exch. Rate Volatility 1.34 Stock volatility 1.18 Fin. Globalization 1.32 Volatility Index 1.14 Volatility Index 1.03 Fin. Globalization 1.05 Stock returns 1.01 Exch. Rate Returns 1.01 Mean VIF 1.17 Mean VIF 1.09 Notes: Model 6 uses Stock volatility as the dependent variable, and Model 8 uses Exchange Rate volatility as the dependent variable. Source: Authors’Construction (2023). M. Insaidoo et al. Research in Globalization 9 (2024) 100242 4 here φrepresents the slope symbol with a dimension of (m, 1), which denotes a matrix of size (1, 1), and also unveils stationary disturbance terms, whilst γirepresents individual-fixed effects presented as a vector jit(m, 1) within integrated schemes of level one, denoted as I(1) for all I, where jit =jit−1+ ∪it. Equation (2) indicates cointegration regression which implies yit is co-integrated with jit. An asymptotically normal behaviour of FMOLS and DOLS estimators is reported. The equation of FMOLS rectifies the issue of serial correlation and endogeneity of OLS regression. The mathematical form of the Panel FMOLS and the Panel DOLS are presented below: ζFMOLS −[∑ N i=1∑ T t=1 ( μ it − μ i)]1[∑ N i=1{∑ T t=1 ( μ it − μ i)yit +TΔu}] (3) where y.it is the transformed form of the yit to rectify endogeneity issue and Δ∊Uindicates a serial correlation term. Likewise, DOLS estimator by default takes care of autocorrelation and endogeneity issue in panel data regression as follows: yit =γi+φlrit +∑ d h=d fihΔrit+k+ υ itt=1,⋯,Ti =1,⋯,N(4) where υ it shows the deviations, φindicates firm-related impact and fih shows values of lag or lead of first-difference independent variables. DOLS estimator is shown as follows: ζDOLS =∑ N t=1{SitSʹ it}1∑ t=1 T {Sitz.it}(5) where Sit = { lit −liΔli,t−v,⋯,Δli,t+v}is 2{v +1} ×1 regressor’s vector. Empirical method Pursuant to Al-Awadhi et al. (2020) and Ashraf (2020), this study employs panel-data regression methodology due to its suitability in empirical analysis in comparison to the conventional event-study approaches. Wooldridge (2010), and Hsiao (2014) posits that the dependent and independent variables’time-variant associations are identified, and the problems of individual heterogeneity, multicollinearity, and estimation biases are minimized with the use of panel-data regression approach. To examine how the stock market volatility has been impacted by exchange rate volatility, the panel-data regression models used are as follows: SVj,t= α 01 + α 02XVj,t+ α 03VIXt+ α 04SRj,t+ α 05FGj,t+ ε 0j,t(6) where SVj,tin equation (6) is stock market volatility for country jat time t,XVj,tdepicts exchange rate volatility for country jat time t,VIXtdepicts Volatility Index at time t,SRj,tis stock returns for country jat time t, financial globalization for country jat time tis depicted by FGj,t, α 01, …, α 05 are the regression coefficients to be estimated, and ε 0j,tdenotes Table 5 Cross-sectional dependence results. Variables Breusch-Pagan LM Pesaran Scaled LM Pesaran CD SVit 405.5294*** 33.42167*** 1.445496 XVit 1152.636*** 104.6555*** 0.620734 VIXt353.0414*** 28.41713*** 7.966238*** SRit 63.82323 0.841262 0.371058 XRit 58.72863 0.355511 −0.631068 Note: *** p <0.01, SV is Stock market volatility, XV is Exchange rate volatility, VIX is Volatility index, SR is Stock returns, and XR is Exchange rate returns. Source: Authors’Construction (2023). Table 6 CADF and CIPS unit root test results. CADF CIPS Variables Level 1st diff. Level 1st diff. SVit −3.534*** −6.090*** −2.942*** −6.190*** XVit −4.002*** −6.190*** −4.364*** −6.190*** VIXt−6.190*** −6.190*** −6.190*** −6.190*** SRit −6.190*** −6.190*** −6.190*** −6.190*** XRit −6.190*** −6.190*** −6.190*** −6.190*** FGit 0.514 −4.590*** 0.511 −4.590*** Note: *** p <0.01, SV is Stock market volatility, XV is Exchange rate volatility, VIX is Volatility index, SR is Stock returns, XR is Exchange rate returns, and FG is Financial Globalization. Source: Authors’Construction (2023). Table 7 Pedroni panel cointegration test results (Stock market volatility as dependent variable). Within Dimension (Panel) Statistics Weighted Statistics v-statistic 2.554*** 0.282 rho-statistic −42.912*** −44.629*** PP-statistic −28.503*** −26.652*** ADF-statistic −27.562*** −19.631*** Between Dimension (Group) rho-statistic −42.138*** PP-statistic −27.495*** ADF-statistic −15.436*** Note: *** p <0.01. Source: Authors’Construction (2023). M. Insaidoo et al. Research in Globalization 9 (2024) 100242 5 the error term. To evaluate the moderating role of financial globalization in the co-movement of exchange rate volatility and stock market volatility, equation (6) is modified as follows: SVj,t= α 11 + α 12XVj,t+ α 13VIXt+ α 14SRj,t+ α 15FGj,t+ α 16XVj,t*FGj,t+ ε 1j,t (7) where α 11,…, α 16 are the regression coefficients to be estimated, XVj,t* FGj,tis the interactive term of exchange rate volatility and financial globalization, with all the other variables in equation (7) remaining same as the ones in Eq. (6). Further, to examine the impact of stock market volatility on exchange rate volatility, reversal panel-data regression models are used as follows: XVj,t= α 21 + α 22SVj,t+ α 23VIXt+ α 24XRj,t+ α 25FGj,t+ ε 2j,t(8) where XRj,tis exchange rate returns for country jat time t, α 21,…, α 25 are the regression coefficients to be estimated, with all the other variables in Eq. (8) remaining same as the ones in Eq. (6). Finally, to evaluate the moderating role of financial globalization in the effect of stock market volatility on exchange rate volatility, Eq. (8) is modified as follows: XVj,t= α 31 + α 32SVj,t+ α 33VIXt+ α 34XRj,t+ α 35FGj,t+ α 36SVj,t*FGj,t+ ε 3j,t (9) where α 31,…, α 36 are the regression coefficients to be estimated, SVj,t* FGj,tis the interactive term of stock market volatility and financial globalization, with all the other variables in Eq. (9) remaining same as the ones in Eq. (8). Eq. (7) is partially differentiated with respect to Exchange rate volatility which gave rise to Eq. (10). This helped us to identify the net effect of Exchange rate volatility on Stock market volatility. ∂ SVj,t ∂ XVj,t =a12 +a16FGj,t(10) here ∂ is the difference operator; SVj,tis Stock market volatility; XVj,tis Exchange rate volatility and FGj,tis the average value of financial globalization. Similarly, Eq. (9) is partially differentiated with respect to Stock market volatility, which gave rise to equation (11). This helped us to identify the net effect of Stock market volatility on Exchange rate volatility. ∂ XVj,t ∂ SVj,t =a32 +a36FGj,t(11) The definition of variables in Eq. (11) are the same as in Eq. (10). Empirical results This section begin by analyzing the descriptive statistics of stock market volatility, exchange rate volatility, volatility index, stock returns, exchange rate returns, and financial globalization, during the COVID-19 pandemic period. This is then followed by an analysis of the results of the pre-diagnostic tests. Following this, the results of the estimates of the effect of exchange rate volatility on stock market volatility are presented and discussed. Further, the results of the estimates of the influence of stock market volatility on exchange rate volatility are presented and discussed. Finally, the results of the estimates of the moderating role of financial globalization in the co-movement of stock market volatility and exchange rate volatility are presented and discussed. Analysis of descriptive statistics Table 2 reports a set of descriptive statistics on stock market volatility, exchange rate volatility, volatility index, stock returns, exchange rate returns, and financial globalization, during the COVID-19 period. The average volatility recorded for the stock and forex markets is 0.0001 and 0.0000 respectively, with minimum volatility of 0.0000 for both markets. The standard deviations of stock and forex markets are given as 0.0002 and 0.0000 respectively. The maximum volatility for stock and forex markets are given as 0.0019 and 0.0003 respectively. The average, standard deviation, minimum, and maximum volatility index is given as 27.0613, 5.0491, 19.9700 and 45.4100 respectively. The average return for the stock markets is given as 0.0004, with minimum, maximum, and standard deviation of −0.1723, 0.1027, and 0.0108 respectively. The average return for the forex markets is −0.0005, with standard deviation of 0.0062, minimum returns of −0.0284 and maximum returns of 0.0396. The minimum and maximum index of financial globalization achieved is 29.9260 and 99.0000 respectively, with a mean index of 57.1737 and standard deviation of 19.8762. The average index for financial globalization implies that, whilst the African region stands to moderately enjoy its positives, such as increased access to capital, improvement in financial market efficiency, increased competition, and risk diversification, the region also has the potential to be temperately vulnerable to external shocks. Results of pre-diagnostic tests In Table 3, the study reports on the results of the pairwise correlation matrix for the COVID-19 pandemic period. The objective of the correlation matrix is to ensure that, there is no multicollinearity among the exogenous variables in the model. In addition, the direction and strength between any two variables in the model is shown by this matrix. The results in Table 3 clearly suggest, that the exogenous variables generally show strong correlation and no problem of multicollinearity. According to Krammer (2010), the absence of multicollinearity is affirmed when the correlation coefficient among independent variables remains below 0.85. Examination of the correlation matrix in Table 3 for this study reveals that, all coefficients are below 0.85, thereby indicating an absence of multicollinearity concerns. Table 8 Pedroni panel cointegration test results (Exchange rate volatility as dependent variable). Within Dimension (Panel) Statistics Weighted Statistics v-statistic 1.123 −0.628 rho-statistic −38.041*** −40.150*** PP-statistic −25.286*** −25.242*** ADF-statistic −21.777*** −12.506*** Between Dimension (Group) rho-statistic −40.212*** PP-statistic −27.841*** ADF-statistic −12.710*** Note: *** p <0.01. Source: Authors’Construction (2023). M. Insaidoo et al. Research in Globalization 9 (2024) 100242 6 Similarly, Table 4 reports on the results of the variance inflation factor (VIF), which is another tool for the detection of multicollinearity in a regression model. A VIF value equal or less than 10 indicate an absence of multicollinearity, which implies consistency in the standard errors of the estimated model. From Table 4, the VIF for the variables for the COVID-19 pandemic period are close to 1. This indicates an absence of multicollinearity and thus, suggests that the standard errors of the estimated model are consistent. The estimated model regressed stock market volatility on all variables in model 6, and model 8 regressed exchange rate volatility on all variables. The results of the panel cross-sectional dependence (CD) test presented in Table 5, indicate that, even at the 1 % level of significance, the null hypothesis of CD is rejected for Stock market volatility, Exchange rate volatility, and Volatility index. This rejection implies the presence of CD and suggests that the first-generation panel unit root framework is not suitable for this study. However, the results indicate that Stock market returns and Exchange rate returns does not suffer from CD. As highlighted by Breitung and Pesaran (2008), CD can emerge due to spatial spillover effects and unobservable factors among nations and regions. The second-generation unit root tests, specifically (Cross-section Augmented Dickey-Fuller (CADF) and Cross-section Im-Pesaran (CIPS) were conducted. The findings indicate that all the variables, except financial globalization, exhibits stationarity at level. Financial globalization became stationary at first difference. The results of the CADF and CIPS tests are presented in Table 6. After performing unit root tests, we utilized the Pedroni (1999, 2004) panel cointegration tests to determine the presence of cointegration relationship among the variables in the model. Table 7 and Table 8 report the findings of the panel cointegration tests using stock market volatility and exchange rate volatility respectively as dependent variable. As evidenced in Table 7, ten of the eleven statistics demonstrate statistically significant values at a 1 % significance level, whilst in Table 8, nine of the eleven statistics show statistically significant values at 1 % significance level. These statistics provide evidence of long-term relationship among these variables. Consequently, the null hypothesis of no cointegration among the variables in the two models are rejected. Drivers of stock market volatility Table 9 reports the main empirical results of the Panel FMOLS and Panel DOLS, using stock market volatility as the dependent variable. Model 6 is the baseline specification. For model 6 and model 7 for both methodologies, volatility index has positive and significant relationship with stock market volatility. This is in tandem with the findings of Zhu et al. (2019), that provides evidence of the US stock market volatility being impacted by volatility index. Further, model 6 and model 7 for both Panel FMOLS and Panel DOLS results show a positive and significant association of stock market returns with stock market volatility. This corroborates the findings of Al-Rjoub and Azzam 2012, that found a positive association of stock returns with stock market volatility in Jordan. In addition, the results of model 6 and model 7 for both methodologies, show significant and positive influence of financial globalization on stock market volatility. This contradicts the findings of Cordella and Ospino Rojas (2017), which using 84 countries revealed that financial globalization reduces stock market volatility. Long-run volatility from forex to stock market As shown in the results of model 6 for both Panel FMOLS and Panel DOLS from Table 9, the exchange rate volatility enters positive and significant, showing a positive co-movement of the two volatilities. This mirrors the results of Sui and Sun (2016), that established influence of forex volatility on stock market volatility for BRICS countries during the 2007/08 global financial crisis period. This result is also in tandem with the findings of Mikhaylov, 2018,Fasanya and Akinde, 2019, and Baranidharan and Alex, 2020, that showed positive effect of forex volatility on stock market volatility for Brazil and Russia, Nigeria, and South Africa respectively, in a non-crisis period. This result also confirms the findings of Van Der Westhuizen et al. (2022) and Rai and Garg (2022) that found positive effects of forex volatility on stock market volatility for the South Africa and BRIICS countries respectively, in the COVID-19 pandemic period. On the contrary, the results of model 7 for both methodologies indicate negative and significant association of exchange rate volatility with stock market volatility as evidenced in Table 9. Whilst exchange rate volatility initially increases stock market volatility due to risk perceptions occasioned by COVID-19, financial globalization introduces mechanisms, such as hedging, diversification, and market integration that can mitigate these effects over time. This explains the change of relationship from positive to negative when the interaction term (exchange rate volatility*financial globalization) was incorporated in model 7. This incorporation is aimed at examining the moderating role of financial globalization in the long-run relationship between exchange rate volatility and stock market volatility. The results of model 7 in both the Panel FMOLS and Panel DOLS show that, financial globalization moderates a significant and positive influence of exchange rate Table 9 Panel FMOLS and Panel DOLS Results (Stock volatility as dependent variable). Panel FMOLS Panel DOLS Variables Model 6 Model 7 Model 6 Model 7 Exchange Rate vol. 0.6531*** −4.9027*** 0.6885*** −5.2559*** (0.1427) (0.6483) (0.1430) (0.8776) Volatility Index 0.0000*** 0.0000*** 0.0000*** 0.0000*** (0.0000) (0.0000) (0.0000) (0.0000) Stock returns 0.0006* 0.0006* 0.0013** 0.0011* (0.0004) (0.0000) (0.0006) (0.0006) Fin. Globalization 0.0000*** 0.0000*** 0.0000*** 0.0000* (0.0000) (0.0000) (0.0000) (0.0000) XV*FG 0.0835*** 0.0902*** (0.0099) (0.0129) Observations 2244 2244 2222 2222 R 2 0.3482 0.3763 0.5699 0.6411 Adjusted R 2 0.3441 0.3721 0.5397 0.6095 SER 0.0001 0.0001 0.0001 0.0000 LRV 0.0000 0.0000 0.0000 0.0000 Net Effect − − 0.1287 − − 0.0988 Notes: Values in the parenthesis are the robust standard error; *** p <0.01, ** p <0.05, * p <0.10, XV*FG represents an interaction term of Exchange Rate volatility and Financial Globalization Source: Authors’Construction (2023). M. Insaidoo et al. Research in Globalization 9 (2024) 100242 7 volatility on stock market volatility. Drivers of exchange rate volatility Table 10 reports the Panel FMOLS and Panel DOLS empirical results, using exchange rate volatility as the dependent variable. Model 8 is the baseline specification. The results of both models and methodologies indicate positive and significant relationship between volatility index and exchange rate volatility. This is in tandem with the findings of Feng et al. (2021), which established significant and positive relationship between volatility index and exchange rate volatility. Further, the results of both models and methodologies, reveal significant and negative influence of exchange rate returns on exchange rate volatility. This mirrors the findings of Mohammed et al. (2021), which established a negative relationship in Ghana. Moreover, the results of both models and methodologies, show the existence of significant and positive relationship between financial globalization and exchange rate volatility. This is in contrast with the findings of Gaies et al. (2020), which established a negative association in selected emerging and developing countries. Long-run volatility from stock to forex market The results of both models and methodologies as presented in Table 10, show significant and positive co-movement of stock market volatility and exchange rate volatility. This contradicts the findings of Sui and Sun (2016), that found no evidence of effect of stock market volatility on forex market volatility in BRICS countries, during the 2007/ 08 GFC period. This results also confirm the findings of Qin et al. (2018) and Singh et al. (2021), that showed positive influence of stock market volatility on forex market volatility in China and Japan, and in India, China and South Africa respectively, in a tranquil period, and the findings of Van Der Westhuizen et al. (2022), which established positive effect of stock market volatility on forex market volatility in South Africa, during the COVID-19 pandemic period. This study includes an interaction term (stock market volatility*financial globalization) in model 9 to examine the moderating role of financial globalization in the influence of stock market volatility on exchange rate volatility. The results of model 9 in both methodologies show that, financial globalization has significant and positive moderating role in the co-movement of stock market volatility and exchange rate volatility. Long-run volatility between forex and stock markets Gleaning from the results of Table 9 and Table 10, the long-run volatility between stock and forex markets is bidirectional, during the COVID-19 pandemic period. This result is in tandem with the findings of Bal et al. (2018), that revealed significant association of stock and forex market volatility, during the 2007/08 GFC period, and the findings of Qin et al. (2018), and Singh et al. (2021), that established significant relationship between forex and stock markets, during tranquil periods. Theoretically, the bidirectional volatility between forex and stock markets in Africa as established by this study, backs the theories of both “stock-oriented”and “flow-oriented”models. Whilst the former model posits that stock prices persuade the exchange rate, the persuasion of the stock prices by the exchange rate as stipulated by the latter model is supported by this study. An economic rationale underpinning this result may be sourced from the informational transmission workings of the “flow-oriented”model, which posits that, the real income and output of an economy are affected, through the effect of exchange rate fluctuations on the competitiveness of local firms in the international markets. In addition, as the present value of future cash flows determines stock prices, the stock market eventually responds to the fluctuations in the exchange rate. The stock markets reactions to changes in the exchange rate, during crisis periods are not surprising, particularly in the COVID19 pandemic period. For the former, Do et al. (2015) asserts that, the continuous and significant real growth rate in the US between 2002 to middle of 2009, strengthened the US dollar, which is the world base currency. Similarly, Miller (2020) posits that, the safe-haven property of the US dollar, was highly exhibited due to its high demand during the COVID-19 crisis. These phenomena underpin the noticeable informational flows from forex to stock markets during crisis periods as expected. Financial globalization and its moderating role For the direct effect, financial globalization is revealed to have significant and positive influence on stock market volatility. This implies that, in turbulent periods, such as COVID-19 crisis, in the African setting, financial globalization serves as a vehicle of contagion in increasing stock market volatility. Further, financial globalization indirect effect via the forex market is positive in the COVID-19 pandemic period. This implies that, the positive direct influence of financial globalization on stock market volatility is reinforced by the positive indirect influence of financial globalization on stock market volatility via exchange rate volatility. This finding is in conformity with the study of Cordella and Table 10 Panel FMOLS and Panel DOLS Results (Exchange Rate volatility as dependent variable). Panel FMOLS Panel DOLS Variables Model 8 Model 9 Model 8 Model 9 Stock volatility 0.0402*** 0.1043*** 0.0685*** 0.1882*** (0.0088) (0.0265) (0.0125) (0.0399) Volatility Index 0.0000*** 0.0000*** 0.0000** 0.0000* (0.0000) (0.0000) (0.0000) (0.0000) Exch. Rate returns −0.0005*** −0.0005*** −0.0007** −0.0005 (0.0002) (0.0002) (0.0003) (0.0003) Fin. Globalization 0.0000*** 0.0000*** 0.0000*** 0.0000*** (0.0000) (0.0000) (0.0003) (0.0000) SV*FG −0.0009** −0.0013** (0.0003) (0.0005) Observations 2244 2244 2222 2222 R 2 0.6179 0.6203 0.7024 0.7417 Adjusted R 2 0.6156 0.6177 0.6815 0.7189 SER 0.0000 0.0000 0.0000 0.0000 LRV 0.0000 0.0000 0.0000 0.0000 Net Effect 0.0528 0.1139 Notes: Values in the parenthesis are the robust standard error; *** p <0.01, ** p <0.05, * p <0.10, SV*FG represents an interaction term of Stock volatility and Financial Globalization Source: Authors’Construction (2023). M. Insaidoo et al. Research in Globalization 9 (2024) 100242 8