Interconnectedness of Stock Indices in African Economies Under Financial, Health, and Political Crises

Abstract

This study examines the interconnectedness of African stock markets during three major global crises: the 2008 Global Financial Crisis (GFC), the COVID-19 pandemic, and the Russia–Ukraine conflict. We use daily stock index data from 2007 to 2023 for ten African countries and apply a Time-Varying Parameter Vector Autoregressive (TVP-VAR) model. The results reveal that volatility connectedness among African markets intensified during all three crises, peaking during the COVID-19 pandemic followed by the 2008 GFC and the Russia–Ukraine conflict. Short-term connectedness consistently exceeded long-term connectedness across all crises. South Africa and Egypt acted as dominant transmitters of volatility, highlighting their systemic importance, while Morocco showed increased influence during the COVID-19 pandemic. These findings suggest that African markets are more globally integrated than previously assumed, making them vulnerable to external shocks. Policy implications include the need for stronger regional financial cooperation, the development of early warning systems, and enhanced intra-African investment to improve market resilience and reduce contagion risk.

1. Introduction

Within the field of international finance, the interconnectedness of stock markets plays a pivotal role in shaping economic landscapes, especially during periods of crisis (Kenourgios et al., 2011; Sarkar & Patel, 1998; Yang & Bessler, 2008). The last two decades have witnessed unprecedented challenges, with events such as the Global Financial Crisis (GFC), COVID-19 pandemic, and geopolitical tensions such as the ongoing conflict between Ukraine and Russia each leaving an indelible mark on the world economy. This study delves into the impacts of these crises on the connectedness of stock indices in African economies, seeking to unravel the complex dynamics that underlie the interactions within and across these markets. First, the 2008 GFC was characterized by the collapse of major financial institutions and subsequent global recession, marking a turning point in the modern financial landscape. The global financial crisis was started by subprime mortgages, and academic researchers noted that this was reflected in a rise in correlation and interactions in global financial markets. Second, the COVID-19 pandemic emerge as an unparalleled crisis, not only affecting public health but also sending shockwaves through the global economy. Third, the persistent conflict between Ukraine and Russia has introduced geopolitical uncertainties that resonate across financial markets.

The body of stock exchanges in Africa has some 28 member stock exchanges covering 38 countries, with over 2400 listed companies. African stock exchanges had an overall market capitalization of about USD 1.6 trillion in 2023, compared to USD 57 billion in 2020, making up about 2% of global stock market capitalization. This study utilizes stock index data from ten African stock market indices: Botswana, Côte d’Ivoire, Egypt, Kenya, Morocco, Tunisia, South Africa, Nigeria, Mauritius, and Zambia. The selected countries and indices are representative of diverse African regions and have been chosen to capture a comprehensive view of interconnectedness of African stock markets. According to the classification of African markets published by MSCI (2024), Egypt and South Africa are classified as emerging markets, while Côte d’Ivoire, Kenya, Mauritius, Morocco, and Tunisia are classified as frontier markets. In 2023, the stock market of Nigeria was reclassified to Standalone Market status because of FX liquidity issues that have continued to impact the accessibility of the Nigerian equity market; likewise, Botswana has been reclassified as a frontier market.1

The study of interconnectedness among stock markets has a longstanding tradition, which has been intensified by recent global crises such as the COVID-19 pandemic and the Russia–Ukraine conflict. A review of this literature indicates a notable gap in research regarding the impact of global crises on African stock markets. While there has been extensive research on developed markets, further investigation is needed in order to understand the effects of events such as the GFC, COVID-19 pandemic, and Russia–Ukraine conflict on African stock markets. Youssef et al. (2021) have shown that the COVID-19 pandemic significantly increased the dynamic connectedness between stock markets, with economic policy uncertainty playing a crucial role in this interdependence. The pandemic also amplified the probability of negative market states, highlighting the importance of understanding stock market behavior during periods of extreme uncertainty. The Russian invasion of Ukraine has similarly affected global financial markets, as firms with strong ties to Russia have experienced significant stock price impacts. However, studies on the efficiency of African stock markets suggest that they may not be as far behind their global counterparts as previously thought, indicating potential for growth and development in these markets. It is clear that while developed markets have been thoroughly examined, further research is needed to understand the effects of global crises on African stock markets and their potential for efficiency and growth. The literature review in the next section situates our study within the broader debate on financial contagion, emphasizing the unique vulnerabilities and opportunities for resilience in African stock markets.

With its many diverse economies, the African continent faces unique challenges that warrant closer examination of the interconnectedness of its stock indices. Our study contributes to this literature by investigating the impact and nature of three major crises on individual African stock markets: the 2008 GFC, the COVID-19 pandemic, and the Russia–Ukraine conflict. We use daily stock index data from ten African countries over a recent 17-year period to analyze short-term and long-term volatility connectedness through a Time-Varying Parameter Vector Autoregressive (TVP-VAR) model. This approach allows us to capture dynamic interactions and avoid the biases associated with static models.

Through empirical analysis, we seek to assess the interconnectedness of stock indices within African economies during the three identified crisis periods. This involves examining the degree of correlation, network structures, and other relevant metrics. By comparing the outcomes across different crises, we aim to discern patterns and variations in the response of African stock markets, providing valuable insights for policymakers, investors, and researchers alike.

1.1. Contribution of the Paper

The novelty of this research lies mainly in its systematic comparison of African stock market integration across three major exogenous shocks of different magnitudes and natures, covering a recent period that remains underdocumented for certain events, notably the Russia–Ukraine conflict. The literature on African countries has largely focused on analyzing a single event at a time, particularly the 2008 financial crisis, with very few studies offering a comparative and longitudinal empirical mapping of regional stock market integration in the face of multiple successive and distinct crises. Our approach reveals differentiated integration dynamics according to the type of shock (financial, health-related, geopolitical), providing an original perspective that has rarely been explored in the existing literature.

The findings of this study carry important policy implications for financial stability and regional integration in Africa. The heightened interconnectedness of African stock markets during global crises underscores the need for stronger regional financial cooperation and the development of joint monitoring and crisis response mechanisms. The systemic roles of South Africa and Egypt as primary transmitters of volatility highlight the importance of closer regulatory oversight and their involvement in regional stabilization efforts. Policymakers should also promote intra-African investment and financial market integration in order to enhance risk-sharing and diversification. Finally, the use of high-frequency data and advanced models suggests the value of establishing early warning systems for timely detection and mitigation of financial contagion.

1.2. Organization of the Paper

The remainder of this paper is organized as follows: Section 2 presents the literature review; Section 3 contains the methodology; Section 4 provides data and a summary of the statistics; empirical results and discussions are contained in Section 5; finally, Section 6 concludes the paper.

2. Literature Review

2.1. Theoretical Underpinnings

Regional economic integration in Africa has gained momentum in response to globalization, the need for larger markets, and the desire to enhance competitiveness on the global stage. By fostering cooperation among countries, integration initiatives such as the African Union, ECOWAS, COMESA, and SADC aim to create unified markets, diversify economies, and boost intra-African trade and investment. In particular, intra-African trade as a share of total African trade has increased from 9.8% in 2000 to 15.9% in 2023 according to the UNCTAD database, reflecting the gradual progress of the continent towards deeper economic ties. Financial and trade liberalization are central to these efforts, with many African countries reducing tariffs, harmonizing regulations, and promoting cross-border financial flows to support increased economic activity and stability. The New Partnership for Africa’s Development (NEPAD) underscores the importance of regional integration as a pathway to economic growth, poverty reduction, and improved global competitiveness, emphasizing joint action in infrastructure, governance, and macroeconomic stability.

The leading theoretical framework for understanding African stock market integration and diversification benefits is rooted in modern portfolio theory, which posits that international diversification reduces risk through exposure to imperfectly correlated markets. Empirical studies on African markets reveal low correlations both within the continent and with global indices, offering diversification potential. For instance, Alagidede (2008) demonstrated that African stock markets exhibit weak stochastic trends with global markets, responding predominantly to local rather than international factors. This finding was corroborated by Alagidede et al. (2011), who noted average monthly correlations of 14% with developed markets and 13% with emerging markets. The International Capital Asset Pricing Model (ICAPM) extends the traditional CAPM by incorporating exchange rate risk and country-specific factors, providing a framework for understanding how global investors price assets in emerging markets Endri et al. (2024). It has been tested in African contexts, particularly in South Africa, where multifactor ICAPM models incorporating global indices and currency risks outperformed domestic CAPM variants as financial integration deepened post-1995. Peerbhai et al. (2018) found that South Africa’s equity market was increasingly aligned with global risk premiums, with adjusted R² values for ICAPM models increasing from 7.79% in 1995 to 17.39% in 2015, reflecting increasing exposure to international systematic risks. However, persistent segmentation in other African markets driven by liquidity constraints and institutional underdevelopment limits ICAPM’s universal applicability.

2.2. Empirical Literature

In this section, we review the most relevant studies pertaining to the topic of this paper. We categorize the existing literature into three main groups. First, we examine studies that explore connectedness within developed financial economies. Second, we discuss research focused on the connectedness of African stock markets. Third, we review studies that address methodologies for measuring connectedness.

2.2.1. Review of Connectedness in Developed Stock Markets

Many empirical works have compared the integration and connectedness of developed stocks during the GFC, COVID-19 pandemic, and Russia–Ukraine conflict. For instance, Agarwal et al. (2024) investigated the dynamic interconnectedness and volatility spillover effects among the stock markets of the United States, China, Germany, Japan, and India from January 2020 to March 2024, a period marked by significant global shocks including the COVID-19 pandemic. Utilizing the DCC-GARCH model and the Diebold–Yilmaz method, their study quantified both the time-varying correlations and the direction and magnitude of volatility transmission between these major economies. The findings reveal that the US and Germany acted as net transmitters of volatility, while China, Japan, and India were primarily net receivers. Notably, the total spillover among these five markets was estimated at 39.37%, with the exception of the Germany–China pair, where long-term information transmission was limited.

Jurkowska et al. (2024) explored the dynamic interconnectedness among the stock market sectors of the ‘Fragile Five’ countries (Brazil, India, Indonesia, South Africa, and Turkey) and examined how volatility shocks as measured by the VIX propagated across these markets. Utilizing a Time-Varying Parameter Vector Autoregression (TVP-VAR) model, their study provided evidence of significant and fluctuating spillover effects both within and between sectors in these emerging economies. The findings highlight that VIX shocks representing global risk aversion play a crucial role in intensifying market connectedness, especially during periods of heightened uncertainty. Their analysis revealed that certain sectors are more susceptible to external volatility shocks, underscoring the importance of sectoral diversification and risk management for investors and policymakers in the ‘Fragile Five’ markets.

Siddiqui et al. (2022) examined the transmission of financial shocks from developed economies (the United States, the United Kingdom, and Japan) to emerging markets during the COVID-19 pandemic. Utilizing the Markov regime-switching model to identify crisis periods and the DCC-GARCH model to analyze time-varying correlations, their research confirmed that certain emerging markets experienced contagion from developed markets, including those in Asia, Africa, and the Middle East. These findings highlight the varying impact of contagion across regions, suggesting that investors in these regions have similar diversification opportunities. This research is pivotal for policymakers and international agencies in formulating post-crisis strategies.

2.2.2. Review of Connectedness in African Stock Markets

While previous studies typically focused on a single crisis period or limited market interactions, our paper provides a comprehensive comparative analysis across three major crises, offering unique insights into how different types of crises impact the interconnectedness of African markets.

A study by Yaya et al. (2024) entitled “African stock markets connectedness: Quantile VAR approach” investigated the interconnectedness of stock markets in Africa using a Quantile Vector Autoregressive (QVAR) framework. The study revealed that stock markets in South Africa, Nigeria, Morocco, and Kenya played significant roles in transmitting shocks across the continent, especially during market extremes. Matsuki et al. (2016) examined the evolution of regional financial integration in African stock markets. Despite increased regional economic cooperation, stock market integration has been slow. The authors suggested promoting further regional integration in order to foster economic growth through developed financial markets while minimizing financial stability risk.

Nyakurukwa and Seetharam (2023) investigated the integration of African stock markets by using an innovative information-theoretic framework to quantify the flow of information between markets. Their study utilized the non-parametric transfer entropy method to measure the direction and magnitude of information transfer among fifteen African stock markets over a 20-year period. This approach allowed them to capture both linear and nonlinear dependencies, providing a robust analysis of market interconnectedness compared to traditional econometric models. The findings revealed limited integration among African stock markets, with significant heterogeneity in the degree of information flow across countries. South Africa emerged as a dominant transmitter of information, while smaller markets such as Botswana and Zimbabwe were primarily receivers. The study also highlighted that periods of global financial crises amplify information flows, suggesting increased co-movement during times of market stress. These results underscore the segmented nature of African stock markets and their potential for diversification benefits while also emphasizing the need for policies aimed at fostering deeper financial integration across the continent.

Recently, Peterson et al. (2024) examined the impact of the COVID-19 pandemic on co-movement and information transmission across twelve stock markets in Sub-Saharan Africa using daily data from 2012 to 2023. Methodologically, their study employed multiple wavelet techniques to analyze time–frequency dynamics and Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) to isolate short-, medium-, and long-term market behaviors. They also applied Rényi’s and Shannon’s transfer entropy to quantify information spillovers alongside nonlinear causality tests for robustness. The results revealed heightened short-term co-movements during the pandemic, particularly among Botswana, Mauritius, and South Africa, with Shannon’s entropy showing non-negative information flows across horizons. While most African markets avoided full contagion, these findings challenge assumptions around the immunity of African stock markets to global shocks, demonstrating that diversification benefits diminish during crises due to increased synchronization. It is worth noting that the wavelet and entropy frameworks provide novel insights into asymmetric nonlinear relationships, aligning with the adaptive and heterogeneous market hypotheses.

2.2.3. Review of Connectedness Methodology

This comprehensive literature review reveals that the dynamic relationships among global stock markets have been extensively studied using a range of econometric models and approaches, each with distinct advantages and limitations. Traditional methods such as rolling-window VAR models and event study frameworks have been widely employed to capture time-varying interdependencies and spillover effects. However, these approaches often suffer from drawbacks such as sensitivity to window size selection, loss of valuable observations, and vulnerability to outliers (Korobilis & Yilmaz, 2018). In recent years, the Time-Varying Parameter Vector Autoregressive (TVP-VAR) model has emerged as a robust alternative, offering several methodological improvements. The TVP-VAR model estimated via the Kalman filter in a state-space framework allows model coefficients and covariances to evolve over time, providing a more flexible and accurate depiction of dynamic connectedness across markets (Hurduzeu et al., 2024). Recent empirical studies have demonstrated the superiority of the TVP-VAR model in tracking systemic risk and measuring volatility spillovers in both developed and emerging markets.

In this study, we employ a sophisticated Time-Varying Parameter Vector Autoregressive (TVP-VAR) model to address methodological limitations in prior research such as outlier sensitivity and parameter estimation biases. By introducing a novel application of the TVP-VAR framework to high-frequency data, our analysis achieves unprecedented precision in measuring real-time market dynamics. Leveraging data up to 2023, this paper delivers the most current evaluation of African stock market interconnectedness, including a focused examination of the understudied Russia–Ukraine conflict, which is a pivotal geopolitical event that has rarely analyzed in regional contexts.

It is worth mentioning that there is a wide literature addressing time-varying market integration, which is a concept similar to market interconnectedness in some aspects. Indeed, market interconnectedness refers to the short-term transmission of shocks or volatility across financial markets, highlighting dynamic interdependence, especially during crises. In contrast, market integration reflects the medium-term and long-term alignment of market fundamentals, where prices and returns move together due to shared risk factors and capital mobility. While interconnectedness captures how markets influence each other in the short run, integration assesses whether they function as unified system in the long run. For further details about the integration concept, the reader is referred to Bekaert and Harvey (1995), Errunza and Losq (1985), and Harvey (1991).

3. Methodology

In prior research, the event study estimation approach has been employed to conduct similar analyses (Umar et al., 2022). However, this method is limited to capturing the impact of an event within the predefined event window. In this study, we extend the analysis by examining the response of each African stock market to the crisis in question.

The Time-Varying Parameter Vector Autoregression (TVP-VAR) model offers several advantages over alternative approaches. First, it is less sensitive to outliers, leading to more robust estimations. Second, it mitigates parameter estimation biases that may result from arbitrarily selecting a window size. Additionally, this estimation method prevents data loss by utilizing a Kalman filter to determine the variance and covariance matrices. Furthermore, the TVP-VAR model introduces an innovative approach that facilitates analysis of high-frequency data such as intraday and daily observations (Antonakakis et al., 2020).

The TVP-VAR model has gained increasing popularity in macroeconomic time series analysis, and has been employed to investigate such varied economic phenomena as monetary policy, fiscal policy, and financial markets (Nakajima et al., 2010). While the event study approach has been widely used in finance, economics, and accounting research since the 1970s, it has notable limitations, including susceptibility to outliers and parameter estimation biases arising from arbitrary window size selection. The TVP-VAR model overcomes these limitations by incorporating time-varying parameters and stochastic volatility, thereby enabling more robust estimations and preserving data integrity (Lubik & Matthes, 2015 for more details).

3.1. Connectedness Approach Based on TVP-VAR

3.1.1. Connectedness Approach Based on TVP-VAR

This paper employs the novel TVP-VAR frequency connectedness approach proposed by Chatziantoniou et al. (2023), which effectively builds upon the foundational work of Baruník and Křehlík (2018) and Antonakakis et al. (2020).

In this section, we first provide a brief introduction to the TVP-VAR-based connectedness approach developed by Antonakakis et al. (2020), which seamlessly integrates the connectedness index of Diebold and Yılmaz (2012) with the TVP-VAR model of Koop and Korobilis (2014).

The TVP-VAR$\left(p\right)$ model can be expressed as

$${\mathit{z}}_t={\mathbf{\Phi }}_{1t}{\mathit{z}}_{t-1}+{\mathbf{\Phi }}_{2t}{\mathit{z}}_{t-2}+\dots +{\mathbf{\Phi }}_{pt}{\mathit{z}}_{t-p}+{\mathit{\epsilon }}_t,\mathrm{with}{\mathit{\epsilon }}_t\sim N\left(0,{\mathbf{\Sigma }}_t\right),$$

(1)

where ${\mathit{z}}_t$ and ${\mathit{\epsilon }}_t$ are $N\times 1$ vectors, ${\mathbf{\Sigma }}_t$ represents the $N\times N$ time-varying variance–covariance matrix, and ${\mathbf{\Phi }}_{it},i=1,\dots ,p$ denotes the $N\times N$ time-varying VAR coefficient. Using the matrix lag-polynomial $\mathbf{\Phi }\left(L\right)=\left[{\mathit{I}}_N-{\mathbf{\Phi }}_{1t}L-\dots -{\mathbf{\Phi }}_p{L}^p\right]$ and applying the Wold representation theorem, the stationary TVP-VAR process can be expressed as a time-varying parameter moving average process of infinite order, denoted TVP-VMA$(\infty )$:

$${\mathit{x}}_t=\mathbf{\Psi }\left(L\right){\epsilon }_t,$$

where $\mathbf{\Phi }\left(L\right)={\left[\mathbf{\Psi }\left(L\right)\right]}^{-1}$. Because $\mathbf{\Psi }\left(L\right)$ contains an infinite of lags, it is approximated by computing ${\mathbf{\Psi }}_h$ for horizons $h=1,\dots ,H$ (see Chatziantoniou et al., 2023 for further details). In this paper, we adopt a TVP-VAR$\left(1\right)$ model, as suggested by the Bayesian Information Criterion (BIC).

Using the TVP-VMA coefficients ${\mathbf{\Psi }}_h$, we can compute the Generalized Forecast Error Variance Decomposition (GFEVD), which quantifies the impact of a shock in variable j on the forecast error variance of variable i. The GFEVD is provided by

$$C_{ijt}\left(H\right)={\displaystyle \frac{{\left({\mathbf{\Sigma }}_t\right)}_{ij}^{-1}{\sum }_{h=0}^H{\left({\left({\mathbf{\Psi }}_h{\mathbf{\Sigma }}_t\right)}_{ijt}\right)}^2}{{\sum }_{h=0}^H{\left({\mathbf{\Psi }}_h{\mathbf{\Sigma }}_t{\mathbf{\Psi }}_h^{\prime }\right)}_{ii}}},$$

(2)

and

$${\tilde{C}}_{ijt}\left(H\right)={\displaystyle \frac{C_{ijt}\left(H\right)}{{\sum }_{k=1}^N{C}_{ijt}\left(H\right)}},$$

(3)

where ${\tilde{C}}_{ijt}\left(H\right)$ represents the proportion of the forecast error variance of variable i attributable to shocks in variable j at horizon H. By applying row normalization of ${\tilde{C}}_{ijt}\left(H\right)$, we ensure that ${\sum }_{i=1}^N{\tilde{C}}_{ijt}\left(H\right)=1$ and ${\sum }_{j=1}^N{\sum }_{i=1}^N{\tilde{C}}_{ijt}\left(H\right)=N$.

Using Equations (2) and (3), we can compute several connectedness measures, including:

• Net pairwise directional connectedness, defined as

  $${\mathrm{NPDC}}_{ijt}\left(H\right)={\tilde{C}}_{ijt}\left(H\right)-{\tilde{C}}_{jit}\left(H\right),$$

  which indicates whether variable j has a greater (or lesser) influence on variable i compared to the reverse. A positive value (${\mathrm{NPDC}}_{ijt}\left(H\right)>0$) suggests that variable j exerts a stronger influence on variable i than vice versa, while a negative value (${\mathrm{NPDC}}_{ijt}\left(H\right)<0$) suggests the opposite.
• Total directional connectedness TO others, defined as

  $${\mathrm{TO}}_{it}\left(H\right)=\sum _{i=1,i\ne j}^N{\tilde{C}}_{jit}\left(H\right),$$

  quantifies the extent to which shocks originating from variable i are transmitted to all other variables j.
• Total directional connectedness FROM others, defined as

  $${\mathrm{FROM}}_{it}\left(H\right)=\sum _{j=1,i\ne j}^N{\tilde{C}}_{ijt}\left(H\right),$$

  captures the degree to which variable i is affected by shocks originating from all other variables $j.$
• Net total directional connectedness, defined as

  $${\mathrm{NET}}_{it}\left(H\right)={\mathrm{TO}}_{it}\left(H\right)-{\mathrm{FROM}}_{it}\left(H\right),$$

  represents the difference between the total directional connectedness TO others (${\mathrm{TO}}_{it}\left(H\right)$) and the total directional connectedness FROM others ${\mathrm{FROM}}_{it}\left(H\right)$. A positive value (${\mathrm{NET}}_{it}\left(H\right)>0$) indicates that variable i is a net transmitter of shocks, exerting more influence on the system than it receives; conversely, a negative value (${\mathrm{NET}}_{it}\left(H\right)<0$) signifies that variable i is a net receiver of shocks.
• Total averaged connectedness index, defined as

  $${\mathrm{TACI}}_t\left(H\right)=N^{-1}\sum _{i=1}^N{\mathrm{TO}}_{it}\left(H\right)=N^{-1}\sum _{i=1}^N{\mathrm{FROM}}_{it}\left(H\right),$$

  measures the average impact that a shock in one variable has on all others, providing an indication of overall network interconnectedness and systemic market risk (Chatziantoniou et al., 2023).

3.1.2. Connectedness in the Frequency Domain

By integrating the TVP-VAR connectedness framework with the spectral representation of variance decompositions introduced in the BK-18 model, we can analyze volatility connectedness among variables of interest in the frequency domain. Using the frequency response function $\mathbf{\Psi }\left(e^{-i\omega }\right)={\sum }_{h=0}^{\infty }{e}^{-i\omega h}{\mathbf{\Psi }}_h$, where $i=\sqrt{-1}$ and $\omega$ represents the frequency, we proceed with the spectral density representation of ${\mathit{z}}_t$ at frequency $\omega$. The spectral density of ${\mathit{z}}_t$ over $\omega$ is defined as the Fourier transformation of the TVP-VMA$(\infty )$ process:

$${\mathit{S}}_z\left(\omega \right)=\sum _{h=-\infty }^{\infty }E\left({\mathit{z}}_t{\mathit{z}}_{t-h}^{\prime }\right)e^{-i\omega h}=\mathbf{\Psi }\left(e^{-i\omega h}\right){\mathbf{\Sigma }}_t{\mathbf{\Psi }}^{\prime }\left(e^{+i\omega h}\right).$$

The frequency-domain GFEVD derived from the spectral density and the GFEVD is provided by

$$\begin{array}{c}\hfill C_{ijt}\left(\omega \right)={\displaystyle \frac{{\left({\mathbf{\Sigma }}_t\right)}_{jj}^{-1}{\left|{\sum }_{h=0}^{\infty }{\left(\mathbf{\Psi }\left(e^{-i\omega h}\right){\mathbf{\Sigma }}_t\right)}_{ijt}\right|}^2}{{\sum }_{h=0}^{\infty }{\left(\mathbf{\Psi }\left(e^{-i\omega h}\right){\mathbf{\Sigma }}_t\mathbf{\Psi }\left(e^{i\omega h}\right)\right)}_{ii}}}{\tilde{C}}_{ijt}\left(\omega \right)={\displaystyle \frac{C_{ijt}\left(\omega \right)}{{\sum }_{k=1}^N{C}_{ijt}\left(\omega \right)}}.\end{array}$$

To capture connectedness within specific frequency ranges, we aggregate over a frequency interval of interest: ${\tilde{\theta }}_{ijt}\left(d\right)={\int }_a^b{\tilde{\theta }}_{ijt}\left(\omega \right)d\omega$, where $d=(a,b):a,b\in (-\pi ,\pi ),a<b.$ This allows us to compute frequency-based connectedness measures that provide insights into spillover effects within a specific frequency range d:

$$\begin{array}{ccc}\hfill {\mathrm{NPDC}}_{ijt}\left(d\right)& =& {\tilde{C}}_{ijt}\left(d\right)-{\tilde{C}}_{jit}\left(d\right)\hfill \\ \hfill {\mathrm{TO}}_{it}\left(d\right)& =& \sum _{i=1,i\ne j}^N{\tilde{C}}_{jit}\left(d\right)\hfill \\ \hfill {\mathrm{FROM}}_{it}\left(d\right)& =& \sum _{j=1,i\ne j}^N{\tilde{C}}_{ijt}\left(d\right)\hfill \end{array}$$

$$\begin{array}{ccc}\hfill {\mathrm{NET}}_{it}\left(d\right)& =& {\mathrm{TO}}_{it}\left(d\right)-{\mathrm{FROM}}_{it}\left(d\right)\hfill \\ \hfill {\mathrm{TACI}}_t\left(d\right)& =& N^{-1}\sum _{i=1}^N{\mathrm{TO}}_{it}\left(d\right)\hfill \\ & =& N^{-1}\sum _{i=1}^N{\mathrm{FROM}}_{it}\left(d\right).\hfill \end{array}$$

Moreover, we have

$$\mathrm{CN}\left(H\right)=\sum _d\mathrm{CN}\left(d\right),$$

where $\mathrm{CN}(·)=[\mathrm{NPDC},\mathrm{TO},\mathrm{FROM},\mathrm{NET},\mathrm{TACI}]$ represents the set of connectedness measures described above. This formulation ensures that the aggregation of frequency-specific connectedness measures corresponds to the connectedness obtained in the time domain.

4. Data and Empirical Analysis

Table 1. Sample of African stock markets.

Country	Region	Index	Market Capitalization (End of Year 2023)
			Billions USD
Botswana	Southern	Botswana Gaborone Index	3.849
Côte d’Ivoire	West	BRVM Composite Share Index	10.871
Egypt	North	Egyptian Exchange EGX 30 Price Index	36.801
Kenya	East	Nairobi Securities Exchange Ltd 20 Index	7.830
Mauritius	East	Mauritius Stock Exchange SEMDEX Index	5.815
Morocco	North	MASI Free Float All Shares Index	62.797
Nigeria	West	NGX All Share Index	45.073
South Africa	Southern	FTSE/JSE Africa All Share Index	990.845
Tunisia	North	Tunisia Stock Exchange TUNINDEX	6.674
Lusaka	Southern	Lusaka Stock Exchange All Share Index	4.738

offers an overview of the size and scope of several major stock exchanges across Africa, highlighting the continent’s diverse financial landscape spanning both large and small economies. Among the listed exchanges, South Africa’s FTSE/JSE Africa All-Share Index reported the highest market capitalization, reaching an impressive USD 990.845 billion by the end of 2023. In contrast, Botswana’s Gaborone Index registered the lowest market capitalization at USD 3.849 billion, illustrating the relatively modest scale of financial market activity in that country.

Daily stock market data were collected from January 2007 to December 2023, covering a period that encompassed several major financial crises over the past two decades. The dataset was carefully cleaned to remove non-trading days (e.g., holidays) and any instances of date mismatches. This time span captures critical global events, including the Global Financial Crisis (August 2007 to March 2009), COVID-19 pandemic (January 2020 to January 2022), and Russia–Ukraine war (24 February 2022 to 31 December 2023). The analysis begins by computing the daily log returns for each stock market index over the full sample period, defined as $R_t=log(P_t/P_{t-1})$, where $R_t$ denotes the return of a stock market index at time t, while $P_t$ and $P_{t-1}$ represent the daily closing prices at times t and $t-1$, respectively.

Figure 1 depicts the price trajectories and performance of selected African stock market indices from 2007 to 2023. The figure highlights periods of pronounced volatility, particularly during the early part of the sample (2007–2009), around the onset of the COVID-19 pandemic in March 2020, and in February 2022. These periods of heightened market turbulence align with major global events: the Global Financial Crisis (GFC), the World Health Organization’s declaration of COVID-19 as a global pandemic, and the escalation of the Russia–Ukraine conflict. Although some markets, particularly in North Africa, experienced additional episodes of volatility such as those triggered by the Arab Spring between January and April 2011 (Abdelbaki, 2013), the present study concentrates on the three aforementioned crises. Accordingly, our analysis focuses on these pivotal events in order to examine their impact on African stock markets.

Table 2. Summary statistics of selected African stock markets returns.

	Mean	Median	Std. Dev.	Max	Min	Skewn.	Kurtosis	$\mathbf{J}-\mathbf{B}$
Full Sample (3538 observations)
Botswana	0.000103	0.000000	0.003543	0.053067	−0.047746	0.137278	44.749515	295,572.20
Côte d’Ivoire	0.000183	0.000000	0.008457	0.066363	−0.059997	0.285167	6.865978	7008.76
Egypt	0.000352	0.001120	0.017063	0.111799	−0.179916	−0.867309	9.096413	12,660.25
Kenya	−0.000376	−0.000279	0.009494	0.113183	−0.100908	0.723686	22.464227	74,797.06
Mauritius	0.000146	0.000000	0.008175	0.102660	−0.123917	−1.204711	54.133421	433,366.19
Morocco	0.000063	0.000144	0.008547	0.068859	−0.097596	−1.246406	21.106394	66,672.86
Nigeria	0.000229	0.000000	0.011912	0.117584	−0.098059	0.309291	9.582480	13,612.94
South Africa	0.000321	0.000804	0.013390	0.072615	−0.102268	−0.500790	5.472921	4571.30
Tunisia	0.000372	0.000231	0.005952	0.059626	−0.069906	−0.774553	17.276830	44,414 .73
Lusaka	0.000502	0.000000	0.009015	0.091993	−0.091137	0.104777	14.958871	33,038.39
During Global Financial Crisis (347 observations)
Botswana	−0.001244	−0.000311	0.005507	0.038637	−0.032126	−0.258368	12.775382	2397.50
Côte d’ivoire	−0.000424	−0.000334	0.010402	0.062301	−0.036436	0.776441	5.023931	406.87
Egypt	−0.001786	0.000993	0.025041	0.083224	−0.179916	−1.578228	8.038722	1094.53
Kenya	−0.001841	−0.002641	0.016819	0.113183	−0.050176	1.914915	10.452991	1817.33
Mauritius	−0.000825	−0.000285	0.015713	0.080051	−0.123917	−0.908193	13.404981	2683.13
Morocco	−0.000311	0.000088	0.012816	0.068859	−0.084442	−0.894471	9.088329	1259.36
Nigeria	−0.002867	−0.001711	0.016036	0.081651	−0.098059	−0.311470	6.636753	653.25
South Africa	−0.000756	0.000263	0.022901	0.072079	−0.077031	−0.102979	1.125613	19.72
Tunisia	0.000695	0.000350	0.007712	0.037507	−0.053158	−0.366567	10.405257	1596.67
Lusaka	−0.000990	0.000000	0.010616	0.038356	−0.050217	−0.689241	3.790812	239.81
During COVID-19 (435 observations)
Botswana	−0.000137	0.000000	0.001884	0.013793	−0.016421	−1.331018	23.526555	10,266.14
Côte d’lvoire	0.000640	0.000722	0.007827	0.038845	−0.042588	−0.139039	4.115472	313.31
Egypt	−0.000292	0.000748	0.015170	0.087508	−0.123768	−2.037495	20.439248	7955.35
Kenya	−0.000811	−0.000190	0.008456	0.020644	−0.083336	−2.949528	22.928776	10,263.62
Mauritius	−0.000056	0.000152	0.012735	0.102660	−0.108237	−2.109710	34.750563	22,432.65
Morocco	0.000314	0.000567	0.011142	0.068808	−0.097596	−2.702092	30.698063	17,786.93
Nigeria	0.001262	0.000410	0.010403	0.060478	−0.050444	0.394764	6.417757	768.04
South Africa	0.000624	0,001418	0.015995	0.072615	−0.102268	−1.418143	9.124524	1674.41
Tunisia	0.000042	0.000303	0,006226	0.018974	−0.069906	−4.100744	40.060436	30,604.53
Lusaka	0.000919	0.000000	0.007339	0.055748	−0.031540	1.878595	12.844979	3282.28
During Russia-Ukraine conflict (393 observations)
Botswana	0.000556	0.000175	0.001476	0.009474	−0.003762	2.493791	11.048729	2429.37
Côte d’Ivoire	0.000023	−0.000196	0.005594	0.030701	−0.019044	0.647517	3.790406	266.59
Egypt	0.002088	0.001258	0.015887	0.070604	−0.063723	−0.038230	2.634854	115.99
Kenya	−0.000578	−0.000660	0.006596	0.022053	−0.025702	−0.058260	1.365497	31.68
During Russia-Ukraine conflict (393 observations)
Mauritius	−0.000151	−0.000171	0.003309	0.014233	−0.013711	0.019745	2.940524	144.19
Morocco	−0.000185	0.000247	0.007930	0.031194	−0.038939	−0.573102	3.616044	239.22
Nigeria	0.001160	0.000389	0.008461	0.036158	−0.032799	−0.003680	3.290474	180.33
South Africa	0.000095	0.000129	0.013238	0.050092	−0.054364	0.134649	2.030294	70.24
Tunisia	0.000537	0.000378	0.003776	0.014861	−0.014465	0.335719	1.369017	39.04
Lusaka	0.001334	0.000000	0.006337	0.052813	−0.021257	3.442511	21.260682	8249.70

provides a comprehensive statistical summary of financial market returns for ten African countries across four distinct periods: the full sample (3538 observations), the Global Financial Crisis (347 observations), the COVID-19 pandemic (435 observations), and the Russia–Ukraine conflict (393 observations).

For the full sample period, average daily returns (Mean) were positive for most countries, with the exception of Kenya, which recorded a slight negative return (−0.000376). Zambia posted the highest average return (0.000502), followed by Tunisia, Egypt, and South Africa, all of which exceeded 0.0003. The standard deviation (Std. Dev.), which measures the volatility or dispersion of returns, varied significantly across countries. Egypt, South Africa, and Nigeria exhibited the highest volatility, indicating greater fluctuations in returns, while Tunisia demonstrated the lowest volatility, suggesting relative stability. The return distributions deviated notably from normality, as indicated by elevated skewness and kurtosis values, particularly in Mauritius, Morocco, and Kenya. This non-normal behavior, including the presence of asymmetric distributions and heavy tails, was further confirmed by the Jarque–Bera test, which rejected the null hypothesis of normality across all markets.

During the Global Financial Crisis, the majority of countries experienced negative average returns, reflecting widespread financial losses. Nigeria (−0.2867%) and Kenya (−0.1841%) were the most adversely affected. Egypt and South Africa exhibited the highest volatility during this period, indicating increased uncertainty and risk. Notably, Egypt showed significant negative skewness (−1.5782), implying a higher likelihood of extreme negative returns. Most countries displayed high kurtosis, highlighting the presence of fat tails in the return distributions and a higher probability of extreme events. The Jarque–Bera statistics reaffirmed the departure from normality during this crisis period.

During the COVID-19 pandemic, Kenya (−0.000811) and Egypt (−0.000292) again reported negative average returns, suggesting continued vulnerability. In contrast, most markets, including Nigeria (0.001262) and Zambia (0.000919), demonstrated resilience with positive returns. Mauritius exhibited the highest volatility, while Botswana was the most stable, indicating varied levels of return predictability across countries. Kurtosis values were notably high during this period, ranging from 4.115 in Ivory Coast to 40.060 in Egypt, confirming the persistence of fat-tailed distributions and deviation from normality.

In the period corresponding to the Russia–Ukraine conflict, most countries recorded positive average returns, with the exception of Kenya, Mauritius, and Morocco. Egypt posted the highest average return (0.002088) but also the highest volatility (0.015887), reflecting substantial fluctuations in market performance. Compared to earlier periods, kurtosis values across most countries were lower, suggesting that return distributions were relatively closer to normality during this time.

In summary, this analysis underscores the substantial influence of major economic events on the behavior of African stock indices, particularly highlighting heightened volatility, deviations from normal return distributions, and directional fluctuations during periods of crisis. These findings are consistent with prior research that has documented similar patterns of increased volatility and non-normal return behavior during global economic downturns, including the Global Financial Crisis and the COVID-19 pandemic (Bello et al., 2022). For example, during the COVID-19 pandemic, African stock markets demonstrated heterogeneous responses, with some experiencing negative returns driven by heightened investor uncertainty and others revealing potential diversification benefits (Isiaka & Terver, 2023). These patterns suggest that African financial markets may exhibit distinctive characteristics in their responses to global shocks, pointing to the need for further research into their resilience, structural dynamics, and potential role in global portfolio diversification.

5. Results and Discussion

5.1. Analysis of the Total Dynamic Volatility Connectedness Index

The dynamic behavior of volatility connectedness among African stock markets is analyzed using time-varying measures. Figure 2 illustrates the evolution of the total average connectedness index (depicted by the black-shaded area) and its decomposition into short-term (red-shaded area) and long-term (green-shaded area) components.

Total Connectedness (Black Area): This represents the overall level of volatility connectedness across the selected markets. It is computed as the sum of short- and long-term components, and reflects the aggregate degree of interdependence in market volatility. Over the sample period, the total connectedness index exhibits pronounced fluctuations, with several peaks and troughs corresponding to major global economic events.

Short-Term Connectedness (1–5 Days, Red Area): This component captures volatility spillovers occurring over a short-term horizon, typically between 1 and 5 days. It constitutes the largest share of the total connectedness index, indicating that short-term linkages dominate volatility transmission among African markets. The short-term connectedness index displays sharp spikes during periods of heightened uncertainty, such as the 2008 GFC and the COVID-19 pandemic. These spikes suggest rapid and widespread reactions to shocks in which markets exhibit synchronous movements and elevated co-movements in volatility. For instance, many African markets experienced significant negative abnormal returns during the GFC, contributing to a surge in short-term connectedness. A similar pattern is observed during the COVID-19 period, highlighting the sensitivity of regional markets to global financial disruptions.

Long-Term Connectedness (Above 5 Days, Green Area): This component reflects persistent volatility spillovers over longer horizons (greater than 5 days). Compared to the short-term component, long-term connectedness accounts for a smaller share of the total connectedness index, suggesting that long-lasting volatility transmission is less prevalent across African stock markets. Nevertheless, long-term connectedness increases notably during major crises such as the GFC and the COVID-19 pandemic, indicating that the impact of these events extended beyond short-lived reactions and had more enduring effects on market interdependence.

Figure 3 demonstrates that the degree of total connectedness among African stock markets is highly time-varying, fluctuating between approximately 10% and 90%. Several notable spikes in the Total Averaged Connectedness Index (TACI) are observed throughout the sample period, each corresponding to major global economic events.

The first significant increase in connectedness occurred between 2007 and 2009, coinciding with the Global Financial Crisis (GFC). This period saw a sharp rise in market interdependence as global financial stress transmitted across regions. Another pronounced spike is observed in late February 2020, marking the onset of the COVID-19 pandemic. Although the virus reached Africa later than other regions, with the first confirmed case reported in Egypt on 14 February 2020, the TACI surged as global uncertainty intensified and financial markets across the continent responded to worldwide shocks.

A further wave of heightened connectedness is evident in February 2022, aligning with the outbreak of the Russia–Ukraine conflict. In addition to these major events, the graph also reveals elevated connectedness during other global episodes, such as the 2011–2012 European debt crisis and the Chinese stock market crash of 2015–2016.

These patterns suggest that African stock markets are not insulated from global financial developments. Indeed, they exhibit significant sensitivity to external shocks, particularly those originating in developed economies, underscoring the interconnected nature of modern financial systems.

5.2. Analysis of Total Averaged Volatility Connectedness

Table 3, Table 4, Table 5, Table 6, Table 7, Table 8, Table 9, Table 10, Table 11, Table 12, Table 13 and Table 14 present the results of the total average volatility connectedness among African stock markets across four key periods, where the tables labeled (b) and (c) provide the decomposition of connectedness into short-term (1–5 trading days) and long-term (more than 5 trading days) components. These results are obtained using a Time-Varying Parameter Vector Autoregression (TVP-VAR) model combined with 10-steps-ahead Generalized Forecast Error Variance Decomposition (GFEVD) to estimate the average volatility connectedness across markets.

During the full sample period, the Total Connectedness Index (TCI) is estimated at 14.69%, indicating that on average 14.69% of the forecast error variance in this network of African stock markets is attributable to cross-market volatility spillovers, with the remaining 85.31% explained by idiosyncratic factors specific to each market.

The level of connectedness increases markedly during crisis periods. During the COVID-19 health crisis, the TCI rises to 33.47%, consisting of 27.94% from short-term connectedness and 5.54% from long-term connectedness. This is followed by the Global Financial Crisis, during which the TCI reaches 27.41% (22.15% short-term; 5.26% long-term), and the Russia–Ukraine war period, with a TCI of 20.35% (17.58% short-term; 2.78% long-term).

The comparative analysis across periods reveals that short-term volatility connectedness is consistently higher than long-term connectedness, suggesting that African stock markets respond more immediately to shocks, with the intensity of spillovers diminishing over time. This pattern reflects the rapid transmission of market sentiment and information in the short term, while longer-term effects tend to be more muted or absorbed over time.

Volatility connectedness captures the extent to which shocks in one market affect the volatility in others. This phenomenon can be analyzed directionally through outward and inward connectedness. Outward connectedness measures the degree to which a country’s stock market transmits volatility to others, whereas inward connectedness quantifies the extent to which it is influenced by external shocks.

Net spillover values are derived by subtracting inward connectedness from outward connectedness. A positive net spillover indicates that a country is a net transmitter of volatility, exerting more influence on other markets than it receives. Conversely, a negative net spillover implies that a country is a net receiver, and is more influenced by external volatility shocks than it contributes to the broader network.

5.2.1. Full Sample Period (2007–2023)

During the full sample period (see Table 3, Table 4 and Table 5), African stock markets displayed varying degrees of volatility connectedness. Egypt exhibited the highest outward connectedness at 20.74%, followed by South Africa at 16.98%. Similarly, Egypt and South Africa recorded the highest inward connectedness at 21.77% and 20.62%, respectively, underscoring their central roles in the transmission of volatility within the region. The net spillover values indicate that both Egypt and South Africa acted as net transmitters of volatility, while countries such as Kenya and Tunisia were net receivers, absorbing more volatility from external sources than they transmitted.

The diagonal elements in the connectedness tables represent self-connectedness, that is, the proportion of shocks each market absorbs internally. These values exceed 80% for all markets, suggesting a predominance of market-specific volatility. For example, Botswana recorded a particularly high self-connectedness of 90.11%, indicating that the vast majority of its volatility is driven by domestic factors rather than spillovers from other markets.

Table 3. Averaged volatility connectedness (full sample).2

	Botswana	Côte d’Ivoire	Egypt	Kenya	Mauritius	Morocco	Nigeria	South Africa	Tunisia	Lusaka	FROM
Botswana	90.11	0.98	0.8	1.03	1.08	1.2	1.59	0.97	0.86	1.37	9.89
Côte d’Ivoire	0.99	89.55	1.48	1.1	0.97	1.4	1.29	0.95	1.17	1.11	10.45
Egypt	1.54	1.22	79.26	1.99	2.11	3.18	1.39	6.86	1.12	1.34	20.74
Kenya	1.79	1.06	3.01	83.08	2.01	1.81	2.44	2.28	1.18	1.36	16.92
Mauritius	1.26	0.93	2.73	1.37	83.25	2.78	1.7	2.66	1.82	1.5	16.75
Morocco	2.09	0.8	3.25	1.54	2.68	82.99	1.74	1.96	1.68	1.26	17.01
Nigeria	1.13	0.97	1.68	1.8	2.05	1.71	86.8	2.03	0.9	0.95	13.2
South Africa	0.97	0.88	5.69	1.55	1.93	2.1	1.51	83.02	1.21	1.13	16.98
Tunisia	1.73	1.01	1.82	1.25	2.25	2.24	1.15	1.85	85.99	0.72	14.01
Lusaka	1.43	1.29	1.31	1.19	1.46	1.36	1.42	1.08	0.43	89.04	10.96
TO	12.91	9.13	21.77	12.81	16.53	17.78	14.23	20.62	10.37	10.74	146.9
Inc.Own	103.03	98.68	101.04	95.89	99.78	100.77	101.03	103.65	96.35	99.78	$\mathrm{cTCl}/\mathrm{TCl}$
NET	3.03	−1.32	1.04	−4.11	−0.22	0.77	1.03	3.65	−3.65	−0.22	$16.32/14.69$
NPT	8	1	6	2	4	6	6	6	1	5

Table 3. Averaged volatility connectedness (full sample).2

	Botswana	Côte d’Ivoire	Egypt	Kenya	Mauritius	Morocco	Nigeria	South Africa	Tunisia	Lusaka	FROM
Botswana	90.11	0.98	0.8	1.03	1.08	1.2	1.59	0.97	0.86	1.37	9.89
Côte d’Ivoire	0.99	89.55	1.48	1.1	0.97	1.4	1.29	0.95	1.17	1.11	10.45
Egypt	1.54	1.22	79.26	1.99	2.11	3.18	1.39	6.86	1.12	1.34	20.74
Kenya	1.79	1.06	3.01	83.08	2.01	1.81	2.44	2.28	1.18	1.36	16.92
Mauritius	1.26	0.93	2.73	1.37	83.25	2.78	1.7	2.66	1.82	1.5	16.75
Morocco	2.09	0.8	3.25	1.54	2.68	82.99	1.74	1.96	1.68	1.26	17.01
Nigeria	1.13	0.97	1.68	1.8	2.05	1.71	86.8	2.03	0.9	0.95	13.2
South Africa	0.97	0.88	5.69	1.55	1.93	2.1	1.51	83.02	1.21	1.13	16.98
Tunisia	1.73	1.01	1.82	1.25	2.25	2.24	1.15	1.85	85.99	0.72	14.01
Lusaka	1.43	1.29	1.31	1.19	1.46	1.36	1.42	1.08	0.43	89.04	10.96
TO	12.91	9.13	21.77	12.81	16.53	17.78	14.23	20.62	10.37	10.74	146.9
Inc.Own	103.03	98.68	101.04	95.89	99.78	100.77	101.03	103.65	96.35	99.78	$\mathrm{cTCl}/\mathrm{TCl}$
NET	3.03	−1.32	1.04	−4.11	−0.22	0.77	1.03	3.65	−3.65	−0.22	$16.32/14.69$
NPT	8	1	6	2	4	6	6	6	1	5

Table 4. Averaged volatility connectedness in the short term (1–5 trading days) (full sample).3

	Botswana	Côte d’Ivoire	Egypt	Kenya	Mauritius	Morocco	Nigeria	South Africa	Tunisia	Lusaka	FROM
Botswana	76.51	0.77	0.67	0.77	0.89	1.03	0.88	0.83	0.64	1.16	7.64
Côte d’Ivoire	0.77	80.19	1.27	0.95	0.8	1.16	1.05	0.81	1.03	0.92	8.77
Egypt	1.02	0.96	69.04	1.74	1.86	2.73	1.01	5.64	0.9	1.22	17.08
Kenya	1.03	0.83	2.21	68.94	1.44	1.42	1.69	1.74	0.91	1.07	12.35
Mauritius	0.9	0.8	2.25	1.15	71.31	2.3	1.3	2.2	1.56	1.27	13.74
Morocco	1.44	0.69	2.78	1.26	2.4	72.13	1.25	1.68	1.36	1.14	14
Nigeria	0.65	0.75	1.23	1.4	1.52	1.29	71.51	1.53	0.66	0.75	9.77
South Africa	0.74	0.79	5.21	1.41	1.82	1.9	1.32	75.97	1.05	1.07	15.33
Tunisia	1.16	0.85	1.42	1.02	1.84	1.81	0.7	1.52	74.2	0.6	10.92
Lusaka	1.15	1.07	1.13	1.05	1.3	1.21	1.07	0.98	0.36	80.3	9.32
TO	8.88	7.52	18.17	10.76	13.87	14.86	10.26	16.93	8.48	9.2	118.92
Inc.Own	85.39	87.71	87.21	79.7	85.18	86.99	81.77	92.91	82.69	89.49	$\mathrm{cTCl}/\mathrm{TCl}$
NET	1.24	−1.25	1.09	−1.59	0.13	0.86	0.49	1.61	−2.44	−0.12	$13.21/11.89$
NPT	6	1	6	2	6	6	5	7	1	5

Table 4. Averaged volatility connectedness in the short term (1–5 trading days) (full sample).3

	Botswana	Côte d’Ivoire	Egypt	Kenya	Mauritius	Morocco	Nigeria	South Africa	Tunisia	Lusaka	FROM
Botswana	76.51	0.77	0.67	0.77	0.89	1.03	0.88	0.83	0.64	1.16	7.64
Côte d’Ivoire	0.77	80.19	1.27	0.95	0.8	1.16	1.05	0.81	1.03	0.92	8.77
Egypt	1.02	0.96	69.04	1.74	1.86	2.73	1.01	5.64	0.9	1.22	17.08
Kenya	1.03	0.83	2.21	68.94	1.44	1.42	1.69	1.74	0.91	1.07	12.35
Mauritius	0.9	0.8	2.25	1.15	71.31	2.3	1.3	2.2	1.56	1.27	13.74
Morocco	1.44	0.69	2.78	1.26	2.4	72.13	1.25	1.68	1.36	1.14	14
Nigeria	0.65	0.75	1.23	1.4	1.52	1.29	71.51	1.53	0.66	0.75	9.77
South Africa	0.74	0.79	5.21	1.41	1.82	1.9	1.32	75.97	1.05	1.07	15.33
Tunisia	1.16	0.85	1.42	1.02	1.84	1.81	0.7	1.52	74.2	0.6	10.92
Lusaka	1.15	1.07	1.13	1.05	1.3	1.21	1.07	0.98	0.36	80.3	9.32
TO	8.88	7.52	18.17	10.76	13.87	14.86	10.26	16.93	8.48	9.2	118.92
Inc.Own	85.39	87.71	87.21	79.7	85.18	86.99	81.77	92.91	82.69	89.49	$\mathrm{cTCl}/\mathrm{TCl}$
NET	1.24	−1.25	1.09	−1.59	0.13	0.86	0.49	1.61	−2.44	−0.12	$13.21/11.89$
NPT	6	1	6	2	6	6	5	7	1	5

Table 5. Averaged volatility connectedness in the long tern (5–infinity trading days) (full sample).

	Botswana	Côte d’Ivoire	Egypt	Kenya	Mauritius	Morocco	Nigeria	South Africa	Tunisia	Lusaka	FROM
Botswana	13.6	0.22	0.13	0.26	0.19	0.17	0.72	0.14	0.21	0.21	2.25
Côte d’Ivoire	0.21	9.36	0.21	0.15	0.17	0.23	0.24	0.14	0.14	0.19	1.69
Egypt	0.52	0.26	10.22	0.24	0.25	0.45	0.38	1.22	0.22	0.12	3.65
Kenya	0.75	0.23	0.8	14.14	0.57	0.39	0.75	0.54	0.26	0.29	4.57
Mauritius	0.36	0.13	0.48	0.22	11.94	0.48	0.4	0.46	0.26	0.23	3.01
Morocco	0.65	0.11	0.47	0.29	0.27	10.86	0.49	0.29	0.31	0.12	3.01
Nigeria	0.47	0.22	0.45	0.41	0.53	0.42	15.28	0.49	0.25	0.2	3.43
South Africa	0.23	0.09	0.48	0.13	0.1	0.2	0.19	7.05	0.16	0.07	1.65
Tunisia	0.57	0.16	0.4	0.22	0.41	0.44	0.46	0.32	11.78	0.13	3.09
Lusaka	0.28	0.21	0.18	0.13	0.17	0.15	0.36	0.09	0.07	8.75	1.64
TO	4.04	1.62	3.6	2.05	2.66	2.93	3.97	3.69	1.88	1.54	27.98
Inc.Own	17.64	10.98	13.82	16.19	14.6	13.79	19.26	10.74	13.67	10.29	$\mathrm{cTCl}/\mathrm{TCl}$
NET	1.79	−0.07	−0.05	−2.52	−0.35	−0.08	0.54	2.04	−1.21	−0.1	$3.11/2.80$
NPT	7	5	6	0	4	5	6	8	1	3

Table 5. Averaged volatility connectedness in the long tern (5–infinity trading days) (full sample).

	Botswana	Côte d’Ivoire	Egypt	Kenya	Mauritius	Morocco	Nigeria	South Africa	Tunisia	Lusaka	FROM
Botswana	13.6	0.22	0.13	0.26	0.19	0.17	0.72	0.14	0.21	0.21	2.25
Côte d’Ivoire	0.21	9.36	0.21	0.15	0.17	0.23	0.24	0.14	0.14	0.19	1.69
Egypt	0.52	0.26	10.22	0.24	0.25	0.45	0.38	1.22	0.22	0.12	3.65
Kenya	0.75	0.23	0.8	14.14	0.57	0.39	0.75	0.54	0.26	0.29	4.57
Mauritius	0.36	0.13	0.48	0.22	11.94	0.48	0.4	0.46	0.26	0.23	3.01
Morocco	0.65	0.11	0.47	0.29	0.27	10.86	0.49	0.29	0.31	0.12	3.01
Nigeria	0.47	0.22	0.45	0.41	0.53	0.42	15.28	0.49	0.25	0.2	3.43
South Africa	0.23	0.09	0.48	0.13	0.1	0.2	0.19	7.05	0.16	0.07	1.65
Tunisia	0.57	0.16	0.4	0.22	0.41	0.44	0.46	0.32	11.78	0.13	3.09
Lusaka	0.28	0.21	0.18	0.13	0.17	0.15	0.36	0.09	0.07	8.75	1.64
TO	4.04	1.62	3.6	2.05	2.66	2.93	3.97	3.69	1.88	1.54	27.98
Inc.Own	17.64	10.98	13.82	16.19	14.6	13.79	19.26	10.74	13.67	10.29	$\mathrm{cTCl}/\mathrm{TCl}$
NET	1.79	−0.07	−0.05	−2.52	−0.35	−0.08	0.54	2.04	−1.21	−0.1	$3.11/2.80$
NPT	7	5	6	0	4	5	6	8	1	3

5.2.2. Global Financial Crisis (2007–2009)

The Global Financial Crisis (GFC) period (Table 6, Table 7 and Table 8) reveals a notable increase in volatility connectedness across all African stock markets. South Africa again leads in outward connectedness with a value of 30.71%, reaffirming its position as a key transmitter of volatility in the region during times of financial stress. This result is consistent with earlier studies that have identified South Africa as a central player in volatility transmission within emerging markets (Andrew & Alain, 2011). Additionally, the three North African stock markets—Egypt, Morocco, and Tunisia—exhibited significant outward connectedness during this period (Xiaoyang et al., 2024), highlighting their active roles in regional spillovers during the crisis.

Notably, Côte d’Ivoire experienced a substantial increase in outward connectedness relative to the full sample period, indicating a heightened influence on other markets during the GFC. South Africa’s net connectedness was particularly elevated, reaching 14.57%, further emphasizing its status as a major source of volatility spillovers during the crisis.

During the GFC, self-connectedness (the share of volatility attributable to domestic shocks) declined significantly compared to the full sample period, ranging between 67% and 80%. Tunisia recorded the lowest level of self-connectedness, while Kenya exhibited the highest. This decline in self-connectedness across markets suggests that external shocks played a more prominent role in driving volatility during the crisis. The variation in self-connectedness across countries highlights the heterogeneous nature of volatility transmission in African markets, which contrasts with the relatively more stable and predictable patterns observed in developed economies (Onwumere et al., 2018).

Table 6. Averaged volatility connectedness during the Global Financial Crisis.

	Botswana	Côte d’Ivoire	Egypt	Kenya	Mauritius	Morocco	Nigeria	South Africa	Tunisia	Lusaka	FROM
Botswana	75	1.92	1.37	2.5	4.51	2.26	1.22	2.94	3.55	4.73	25.00
Côte d’Ivoire	1.93	76.16	1.19	2.34	2.85	1.37	2.35	4.9	2.46	4.45	23.84
Egypt	0.78	2.09	69.29	4.1	2.18	3.45	3.39	8.37	2.25	4.1	30.71
Kenya	1.25	1.31	4.04	80.77	1.17	3.51	1.18	3.63	1.88	1.26	19.23
Mauritius	2.82	2.82	3.63	2.5	71.49	3.83	1.34	5.49	3.19	2.89	28.51
Morocco	1.68	2.29	5.04	3.37	3.61	69.74	1.87	4.89	5.04	2.47	30.26
Nigeria	1.74	2.45	4.42	2.45	2.12	1.52	73.74	4.95	2.3	4.32	26.26
South Africa	2.71	4.51	4.18	3.01	3.69	3.23	1.39	69.5	5.05	2.71	30.50
Tunisia	2.48	2.7	3.48	1.47	7.92	3	2.12	6.54	67.26	3.03	32.74
Lusaka	4.34	4.19	1.6	1.93	2.72	1.39	5.94	3.36	1.59	72.92	27.08
TO	19.74	24.29	28.95	23.68	30.76	23.56	20.8	45.07	27.31	29.96	274.12
Inc.Own	94.74	100.45	98.24	104.45	102.25	93.31	94.54	114.57	94.57	102.88	cTCl/TCl
NET	−5.26	0.45	−1.76	4.45	2.25	−6.69	−5.46	14.57	−5.43	2.88	$30.46/27.41$
NPT	2	4	5	6	4	3	2	9	4	6

Table 6. Averaged volatility connectedness during the Global Financial Crisis.

	Botswana	Côte d’Ivoire	Egypt	Kenya	Mauritius	Morocco	Nigeria	South Africa	Tunisia	Lusaka	FROM
Botswana	75	1.92	1.37	2.5	4.51	2.26	1.22	2.94	3.55	4.73	25.00
Côte d’Ivoire	1.93	76.16	1.19	2.34	2.85	1.37	2.35	4.9	2.46	4.45	23.84
Egypt	0.78	2.09	69.29	4.1	2.18	3.45	3.39	8.37	2.25	4.1	30.71
Kenya	1.25	1.31	4.04	80.77	1.17	3.51	1.18	3.63	1.88	1.26	19.23
Mauritius	2.82	2.82	3.63	2.5	71.49	3.83	1.34	5.49	3.19	2.89	28.51
Morocco	1.68	2.29	5.04	3.37	3.61	69.74	1.87	4.89	5.04	2.47	30.26
Nigeria	1.74	2.45	4.42	2.45	2.12	1.52	73.74	4.95	2.3	4.32	26.26
South Africa	2.71	4.51	4.18	3.01	3.69	3.23	1.39	69.5	5.05	2.71	30.50
Tunisia	2.48	2.7	3.48	1.47	7.92	3	2.12	6.54	67.26	3.03	32.74
Lusaka	4.34	4.19	1.6	1.93	2.72	1.39	5.94	3.36	1.59	72.92	27.08
TO	19.74	24.29	28.95	23.68	30.76	23.56	20.8	45.07	27.31	29.96	274.12
Inc.Own	94.74	100.45	98.24	104.45	102.25	93.31	94.54	114.57	94.57	102.88	cTCl/TCl
NET	−5.26	0.45	−1.76	4.45	2.25	−6.69	−5.46	14.57	−5.43	2.88	$30.46/27.41$
NPT	2	4	5	6	4	3	2	9	4	6

Table 7. Averaged volatility connectedness in the short term (1–5 trading days) during the Global Financial Crisis.

	Botswana	Côte d’lvoire	Egypt	Kenya	Mauritius	Morocco	Nigeria	South Africa	Tunisia	Lusaka	FROM
Botswana	62.51	1.67	0.97	1.83	3.52	1.54	0.73	2.76	2.78	3.98	19.78
Côte d’Ivoire	1.66	69.15	1.07	1.66	2.59	1.28	2	4.31	2.19	3.69	20.45
Egypt	0.62	1.71	59.49	2.94	1.92	3.02	2.16	6.25	1.96	3.18	23.76
Kenya	0.97	1.16	2.71	63.71	0.75	2.54	0.7	2.63	1.56	0.95	13.98
Mauritius	2.64	2.12	2.95	1.76	59.66	3.15	0.72	4.81	2.8	1.93	22.87
Morocco	1.38	2.11	3.83	2.73	3.35	61.05	1.28	4.51	4.83	2.2	26.24
Nigeria	1.07	1.62	2.69	1.73	1.37	0.81	50.98	3.08	1.57	2.48	16.43
South Africa	2.43	4.25	3.92	2.55	3.46	2.78	1.25	64.47	4.75	2.55	27.95
Tunisia	2.32	2.63	2.82	1.32	6.49	2.74	2	6.23	60.49	2.49	29.04
Lusaka	3.62	3.25	1.24	1.29	2.54	1.24	3.92	2.51	1.42	64.11	21.04
TO	16.7	20.53	22.22	17.82	25.99	19.1	14.76	37.1	23.87	23.45	221.54
Inc.Own	79.21	89.68	81.71	81.52	85.65	80.15	65.74	101.57	84.36	87.56	$\mathrm{cTCl}/\mathrm{TCl}$
NET	−3.08	0.07	−1.54	3.84	3.12	−7.14	−1.66	9.15	−5.17	2.41	$24.62/22.15$
NPT	1	4	5	7	6	1	4	8	3	6

Table 7. Averaged volatility connectedness in the short term (1–5 trading days) during the Global Financial Crisis.

	Botswana	Côte d’lvoire	Egypt	Kenya	Mauritius	Morocco	Nigeria	South Africa	Tunisia	Lusaka	FROM
Botswana	62.51	1.67	0.97	1.83	3.52	1.54	0.73	2.76	2.78	3.98	19.78
Côte d’Ivoire	1.66	69.15	1.07	1.66	2.59	1.28	2	4.31	2.19	3.69	20.45
Egypt	0.62	1.71	59.49	2.94	1.92	3.02	2.16	6.25	1.96	3.18	23.76
Kenya	0.97	1.16	2.71	63.71	0.75	2.54	0.7	2.63	1.56	0.95	13.98
Mauritius	2.64	2.12	2.95	1.76	59.66	3.15	0.72	4.81	2.8	1.93	22.87
Morocco	1.38	2.11	3.83	2.73	3.35	61.05	1.28	4.51	4.83	2.2	26.24
Nigeria	1.07	1.62	2.69	1.73	1.37	0.81	50.98	3.08	1.57	2.48	16.43
South Africa	2.43	4.25	3.92	2.55	3.46	2.78	1.25	64.47	4.75	2.55	27.95
Tunisia	2.32	2.63	2.82	1.32	6.49	2.74	2	6.23	60.49	2.49	29.04
Lusaka	3.62	3.25	1.24	1.29	2.54	1.24	3.92	2.51	1.42	64.11	21.04
TO	16.7	20.53	22.22	17.82	25.99	19.1	14.76	37.1	23.87	23.45	221.54
Inc.Own	79.21	89.68	81.71	81.52	85.65	80.15	65.74	101.57	84.36	87.56	$\mathrm{cTCl}/\mathrm{TCl}$
NET	−3.08	0.07	−1.54	3.84	3.12	−7.14	−1.66	9.15	−5.17	2.41	$24.62/22.15$
NPT	1	4	5	7	6	1	4	8	3	6

Table 8. Averaged volatility connectedness in the long term (5–infinity trading days) during the Global Financial Crisis.

	Botswana	Côte d’Ivoire	Egypt	Kenya	Mauritius	Morocco	Nigeria	South Africa	Tunisia	Lusaka	FROM
Botswana	12.49	0.25	0.4	0.67	0.99	0.72	0.49	0.18	0.76	0.75	5.22
Côte d’Ivoire	0.26	7.01	0.11	0.68	0.26	0.09	0.35	0.59	0.28	0.76	3.39
Egypt	0.17	0.38	9.8	1.16	0.26	0.43	1.23	2.11	0.29	0.91	6.94
Kenya	0.28	0.15	1.32	17.06	0.41	0.97	0.48	1	0.31	0.31	5.25
Mauritius	0.18	0.7	0.68	0.74	11.83	0.68	0.62	0.68	0.39	0.96	5.64
Morocco	0.3	0.18	1.2	0.64	0.26	8.69	0.58	0.37	0.21	0.27	4.02
Nigeria	0.68	0.83	1.73	0.72	0.75	0.7	22.76	1.87	0.73	1.84	9.83
South Africa	0.28	0.26	0.26	0.45	0.23	0.46	0.14	5.03	0.29	0.16	2.55
Tunisia	0.17	0.07	0.66	0.16	1.43	0.26	0.12	0.3	6.77	0.54	3.7
Lusaka	0.73	0.94	0.36	0.65	0.18	0.15	2.02	0.85	0.16	8.81	6.04
TO	3.04	3.77	6.73	5.87	4.77	4.46	6.03	7.96	3.43	6.51	52.58
Inc.Own	15.53	10.78	16.53	22.93	16.6	13.16	28.8	13	10.21	15.33	$\mathrm{cTCl}/\mathrm{TCl}$
NET	−2.18	0.38	−0.22	0.62	−0.87	0.44	−3.8	5.42	−0.27	0.48	$5.84/5.26$
NPT	3	5	6	5	3	6	1	7	4	5

Table 8. Averaged volatility connectedness in the long term (5–infinity trading days) during the Global Financial Crisis.

	Botswana	Côte d’Ivoire	Egypt	Kenya	Mauritius	Morocco	Nigeria	South Africa	Tunisia	Lusaka	FROM
Botswana	12.49	0.25	0.4	0.67	0.99	0.72	0.49	0.18	0.76	0.75	5.22
Côte d’Ivoire	0.26	7.01	0.11	0.68	0.26	0.09	0.35	0.59	0.28	0.76	3.39
Egypt	0.17	0.38	9.8	1.16	0.26	0.43	1.23	2.11	0.29	0.91	6.94
Kenya	0.28	0.15	1.32	17.06	0.41	0.97	0.48	1	0.31	0.31	5.25
Mauritius	0.18	0.7	0.68	0.74	11.83	0.68	0.62	0.68	0.39	0.96	5.64
Morocco	0.3	0.18	1.2	0.64	0.26	8.69	0.58	0.37	0.21	0.27	4.02
Nigeria	0.68	0.83	1.73	0.72	0.75	0.7	22.76	1.87	0.73	1.84	9.83
South Africa	0.28	0.26	0.26	0.45	0.23	0.46	0.14	5.03	0.29	0.16	2.55
Tunisia	0.17	0.07	0.66	0.16	1.43	0.26	0.12	0.3	6.77	0.54	3.7
Lusaka	0.73	0.94	0.36	0.65	0.18	0.15	2.02	0.85	0.16	8.81	6.04
TO	3.04	3.77	6.73	5.87	4.77	4.46	6.03	7.96	3.43	6.51	52.58
Inc.Own	15.53	10.78	16.53	22.93	16.6	13.16	28.8	13	10.21	15.33	$\mathrm{cTCl}/\mathrm{TCl}$
NET	−2.18	0.38	−0.22	0.62	−0.87	0.44	−3.8	5.42	−0.27	0.48	$5.84/5.26$
NPT	3	5	6	5	3	6	1	7	4	5

5.2.3. COVID-19 Pandemic (2020–2021)

The COVID-19 pandemic period (Table 9, Table 10 and Table 11) witnessed a significant surge in volatility connectedness among African stock markets. This period marked the highest level of interconnection observed in the study, with the Total Connectedness Index (TCI) reaching 37.19%. Morocco experienced a particularly notable increase in both outward and inward connectedness, recording values of 49.05% and 53.93%, respectively. Egypt and South Africa also demonstrated elevated inward connectedness, at 50.72% and 45.40%, respectively, reaffirming their central roles in regional volatility transmission.

Net connectedness values identify Mauritius as a major net transmitter of volatility during this period, with a net value of 11.44%, indicating that its market exerted substantial influence on others. In contrast, Tunisia, Kenya, and Botswana were net receivers of volatility, suggesting they were more affected by external shocks than they contributed to the regional volatility network.

Self-connectedness also declined during the COVID-19 period, most notably in Morocco, where it fell to 50.95%. This reduction reflects the heightened interconnectedness among African markets and underscores the pervasive impact of the pandemic on regional financial dynamics.

Table 9. Averaged volatility connectedness during the COVID-19 pandemic.

	Botswana	Côte d’Ivoire	Egypt	Kenya	Mauritius	Morocco	Nigeria	South Africa	Tunisia	Lusaka	FROM
Botswana	80.87	1.93	1.77	1.01	1.98	1.14	1.38	4.32	1.18	4.41	19.13
Côte d’Ivoire	0.74	71.79	1.42	2.5	5.57	2.16	9.74	1.5	2.77	1.79	28.21
Egypt	0.62	1.71	55.53	3.96	8.13	12.97	1.32	11.9	2.71	1.14	44.47
Kenya	0.68	1.19	4.56	65.73	6.27	4.31	4.31	6.84	2.65	3.46	34.27
Mauritius	1.03	1.9	8.71	3.27	58.87	12.72	1.16	5.72	5.19	1.43	41.13
Morocco	1.07	2.36	12.34	3.01	12.91	50.95	3.26	6.35	6	1.73	49.05
Nigeria	1.05	10.01	1.48	1.9	3.28	1.18	76.12	1.54	0.9	2.54	23.88
South Africa	2.43	1.24	12.52	5.19	5.37	7.37	1.08	59.76	2.78	2.26	40.24
Tunisia	1.06	2.21	5.81	3.5	6.5	10.26	1.22	5.41	63.07	0.95	36.93
Lusaka	2.57	1.92	2.1	1.22	2.54	1.81	2.59	1.81	0.89	82.54	17.46
TO	11.25	24.47	50.72	25.58	52.56	53.93	26.05	45.4	25.06	19.73	334.75
Inc.Own	92.13	96.26	106.25	91.31	111.44	104.88	102.17	105.16	88.13	102.27	$\mathrm{cTCl}/\mathrm{TCl}$
NET	−7.87	−3.74	6.25	−8.69	11.44	4.88	2.17	5.16	−11.87	2.27	$37.19/33.47$
NPT	0	5	7	3	7	6	5	6	2	4

Table 9. Averaged volatility connectedness during the COVID-19 pandemic.

	Botswana	Côte d’Ivoire	Egypt	Kenya	Mauritius	Morocco	Nigeria	South Africa	Tunisia	Lusaka	FROM
Botswana	80.87	1.93	1.77	1.01	1.98	1.14	1.38	4.32	1.18	4.41	19.13
Côte d’Ivoire	0.74	71.79	1.42	2.5	5.57	2.16	9.74	1.5	2.77	1.79	28.21
Egypt	0.62	1.71	55.53	3.96	8.13	12.97	1.32	11.9	2.71	1.14	44.47
Kenya	0.68	1.19	4.56	65.73	6.27	4.31	4.31	6.84	2.65	3.46	34.27
Mauritius	1.03	1.9	8.71	3.27	58.87	12.72	1.16	5.72	5.19	1.43	41.13
Morocco	1.07	2.36	12.34	3.01	12.91	50.95	3.26	6.35	6	1.73	49.05
Nigeria	1.05	10.01	1.48	1.9	3.28	1.18	76.12	1.54	0.9	2.54	23.88
South Africa	2.43	1.24	12.52	5.19	5.37	7.37	1.08	59.76	2.78	2.26	40.24
Tunisia	1.06	2.21	5.81	3.5	6.5	10.26	1.22	5.41	63.07	0.95	36.93
Lusaka	2.57	1.92	2.1	1.22	2.54	1.81	2.59	1.81	0.89	82.54	17.46
TO	11.25	24.47	50.72	25.58	52.56	53.93	26.05	45.4	25.06	19.73	334.75
Inc.Own	92.13	96.26	106.25	91.31	111.44	104.88	102.17	105.16	88.13	102.27	$\mathrm{cTCl}/\mathrm{TCl}$
NET	−7.87	−3.74	6.25	−8.69	11.44	4.88	2.17	5.16	−11.87	2.27	$37.19/33.47$
NPT	0	5	7	3	7	6	5	6	2	4

Table 10. Averaged volatility connectedness in the short term (1–5 trading days) during the COVID-19 pandemic.

	Botswana	Côte d’lvoire	Egypt	Kenya	Mauritius	Morocco	Nigeria	South Africa	Tunisia	Lusaka	FROM
Botswana	72.43	1.68	0.72	0.74	1.55	0.98	1.29	3.63	0.7	2.88	14.17
Côte d’lvoire	0.67	63.94	1	2.12	4.73	1.83	9.52	1.23	2.37	1.55	25.01
Egypt	0.43	1.44	48.43	3.37	7.66	11.68	1.01	10.1	2.18	0.81	38.67
Kenya	0.62	0.94	3.96	56.17	5.09	3.78	3.04	5.47	2.41	1.89	27.19
Mauritius	0.83	1.68	7.22	2.9	51.1	11.09	1.05	4.79	4.45	1.1	35.1
Morocco	0.79	1.92	10.41	2.48	12.2	45.97	2.53	5.59	5.31	1.34	42.57
Nigeria	0.79	8.15	1.2	1.55	2.99	1.07	64.58	1.32	0.79	2.06	19.92
South Africa	2.26	1.12	11.16	4.96	5.13	6.51	0.76	54.5	2.45	1.6	35.94
Tunisia	0.81	1.99	3.87	2.52	5.56	7.94	1.04	3.96	52.49	0.53	28.22
Lusaka	2.24	1.62	0.77	0.78	1.86	1.38	2.36	1.18	0.4	69.99	12.58
TO	9.43	20.53	40.3	21.42	46.76	46.27	22.59	37.27	21.05	13.76	279.37
Inc.Own	81.86	84.47	88.72	77.59	97.86	92.24	87.17	91.76	73.54	83.75	$\mathrm{cTCl}/\mathrm{TCl}$
NET	−4.74	−4.48	1.62	−5.77	11.66	3.7	2.67	1.33	−7.17	1.18	$31.04/27.94$
NPT	1	4	5	3	9	6	6	5	1	5

Table 10. Averaged volatility connectedness in the short term (1–5 trading days) during the COVID-19 pandemic.

	Botswana	Côte d’lvoire	Egypt	Kenya	Mauritius	Morocco	Nigeria	South Africa	Tunisia	Lusaka	FROM
Botswana	72.43	1.68	0.72	0.74	1.55	0.98	1.29	3.63	0.7	2.88	14.17
Côte d’lvoire	0.67	63.94	1	2.12	4.73	1.83	9.52	1.23	2.37	1.55	25.01
Egypt	0.43	1.44	48.43	3.37	7.66	11.68	1.01	10.1	2.18	0.81	38.67
Kenya	0.62	0.94	3.96	56.17	5.09	3.78	3.04	5.47	2.41	1.89	27.19
Mauritius	0.83	1.68	7.22	2.9	51.1	11.09	1.05	4.79	4.45	1.1	35.1
Morocco	0.79	1.92	10.41	2.48	12.2	45.97	2.53	5.59	5.31	1.34	42.57
Nigeria	0.79	8.15	1.2	1.55	2.99	1.07	64.58	1.32	0.79	2.06	19.92
South Africa	2.26	1.12	11.16	4.96	5.13	6.51	0.76	54.5	2.45	1.6	35.94
Tunisia	0.81	1.99	3.87	2.52	5.56	7.94	1.04	3.96	52.49	0.53	28.22
Lusaka	2.24	1.62	0.77	0.78	1.86	1.38	2.36	1.18	0.4	69.99	12.58
TO	9.43	20.53	40.3	21.42	46.76	46.27	22.59	37.27	21.05	13.76	279.37
Inc.Own	81.86	84.47	88.72	77.59	97.86	92.24	87.17	91.76	73.54	83.75	$\mathrm{cTCl}/\mathrm{TCl}$
NET	−4.74	−4.48	1.62	−5.77	11.66	3.7	2.67	1.33	−7.17	1.18	$31.04/27.94$
NPT	1	4	5	3	9	6	6	5	1	5

Table 11. Averaged volatility connectedness in the long term (5–infinity trading days) during the COVID-19 pandemic.

	Botswana	Côte d’Ivoire	Egypt	Kenya	Mauritius	Morocco	Nigeria	South Africa	Tunisia	Lusaka	FROM
Botswana	8.45	0.25	1.05	0.27	0.44	0.16	0.09	0.69	0.47	1.53	4.95
Côte d’Ivoire	0.08	7.84	0.42	0.38	0.84	0.33	0.23	0.27	0.41	0.24	3.2
Egypt	0.19	0.28	7.11	0.59	0.47	1.29	0.31	1.8	0.53	0.33	5.79
Kenya	0.06	0.25	0.6	9.57	1.19	0.53	1.27	1.37	0.24	1.57	7.08
Mauritius	0.2	0.22	1.5	0.37	7.77	1.63	0.11	0.93	0.74	0.34	6.03
Morocco	0.28	0.44	1.93	0.53	0.71	4.98	0.73	0.77	0.7	0.39	6.48
Nigeria	0.26	1.86	0.28	0.34	0.29	0.11	11.54	0.22	0.11	0.48	3.95
South Africa	0.17	0.11	1.37	0.24	0.24	0.86	0.31	5.27	0.33	0.67	4.3
Tunisia	0.25	0.22	1.95	0.98	0.95	2.32	0.17	1.45	10.58	0.42	8.71
Lusaka	0.33	0.31	1.33	0.45	0.68	0.43	0.24	0.63	0.48	12.55	4.88
TO	1.82	3.94	10.42	4.16	5.8	7.66	3.46	8.13	4.02	5.97	55.38
Inc.Own	10.27	11.78	17.53	13.72	13.57	12.64	15	13.4	14.6	18.52	$\mathrm{cTCl}/\mathrm{TCl}$
NET	−3.13	0.74	4.63	−2.92	−0.23	1.18	−0.49	3.83	−4.7	1.09	$6.15/5.54$
NPT	2	4	7	4	6	4	5	6	3	4

Table 11. Averaged volatility connectedness in the long term (5–infinity trading days) during the COVID-19 pandemic.

	Botswana	Côte d’Ivoire	Egypt	Kenya	Mauritius	Morocco	Nigeria	South Africa	Tunisia	Lusaka	FROM
Botswana	8.45	0.25	1.05	0.27	0.44	0.16	0.09	0.69	0.47	1.53	4.95
Côte d’Ivoire	0.08	7.84	0.42	0.38	0.84	0.33	0.23	0.27	0.41	0.24	3.2
Egypt	0.19	0.28	7.11	0.59	0.47	1.29	0.31	1.8	0.53	0.33	5.79
Kenya	0.06	0.25	0.6	9.57	1.19	0.53	1.27	1.37	0.24	1.57	7.08
Mauritius	0.2	0.22	1.5	0.37	7.77	1.63	0.11	0.93	0.74	0.34	6.03
Morocco	0.28	0.44	1.93	0.53	0.71	4.98	0.73	0.77	0.7	0.39	6.48
Nigeria	0.26	1.86	0.28	0.34	0.29	0.11	11.54	0.22	0.11	0.48	3.95
South Africa	0.17	0.11	1.37	0.24	0.24	0.86	0.31	5.27	0.33	0.67	4.3
Tunisia	0.25	0.22	1.95	0.98	0.95	2.32	0.17	1.45	10.58	0.42	8.71
Lusaka	0.33	0.31	1.33	0.45	0.68	0.43	0.24	0.63	0.48	12.55	4.88
TO	1.82	3.94	10.42	4.16	5.8	7.66	3.46	8.13	4.02	5.97	55.38
Inc.Own	10.27	11.78	17.53	13.72	13.57	12.64	15	13.4	14.6	18.52	$\mathrm{cTCl}/\mathrm{TCl}$
NET	−3.13	0.74	4.63	−2.92	−0.23	1.18	−0.49	3.83	−4.7	1.09	$6.15/5.54$
NPT	2	4	7	4	6	4	5	6	3	4

5.2.4. Russia–Ukraine Conflict (2022–2023)

In the most recent period (Table 12, Table 13 and Table 14) corresponding to the Russia–Ukraine conflict, the patterns of connectedness among African stock markets exhibited notable shifts. Egypt and Côte d’Ivoire recorded the highest levels of outward connectedness, at 27.21% and 25.97% respectively, while Botswana and Nigeria registered the lowest (13.38% and 17.04%). Egypt continued to exhibit elevated inward connectedness (36.06%), although it was less dominant compared to its levels during the Global Financial Crisis and the COVID-19 pandemic.

Net spillover values indicate that Egypt and Nigeria acted as significant net transmitters of volatility during this period, whereas Kenya and Tunisia remained net receivers. Notably, self-connectedness increased relative to other crisis periods, although it remained below the levels observed during the full sample period. This suggests a partial reversion to more domestically driven volatility, albeit with continued regional spillover.

In summary, the analysis of volatility connectedness across distinct crisis periods reveals that South Africa has consistently played a central role as a key transmitter of volatility to other African markets, particularly during episodes of financial stress such as the Global Financial Crisis. This is consistent with prior research identifying South Africa as a dominant player in African financial markets, often serving as a conduit for shock transmission during bearish market phases (Yaya et al., 2024). Similarly, Egypt has consistently shown high levels of outward connectedness, underscoring its importance as a volatility source within the region. This finding aligns with studies on MENA stock markets that highlight Egypt’s influence on volatility transmission, particularly during periods marked by political or economic instability (Kirkulak Uludag & Ezzat, 2017).

The variation in net spillover values across different periods illustrates the dynamic nature of volatility connectedness in African markets, driven by both global and regional economic events. For instance, evidence suggests that African markets tend to exhibit increased co-movement during crises such as the COVID-19 pandemic, with common shocks rather than direct market-to-market transmission often explaining the observed interlinkages (Peterson et al., 2024). Morocco also emerged as a notable transmitter of volatility during the COVID-19 period, reflecting changes in regional market dynamics under global stress.

Overall, this analysis highlights the importance of considering temporal dynamics and external shocks when assessing volatility connectedness in African stock markets. Markets such as South Africa and Egypt consistently play pivotal roles in regional volatility transmission, underscoring their influence within the broader African financial network.

Table 12. Averaged volatility connectedness during the Russia–Ukraine conflict.

	Botswana	Côte d’Ivoire	Egypt	Kenya	Mauritius	Morocco	Nigeria	South Africa	Tunisia	Lusaka	FROM
Botswana	86.62	0.98	1.63	0.68	1.23	0.72	2.6	1.15	1.39	3	13.38
Côte d’Ivoire	1.39	74.03	7	1.57	2.49	2.49	3.39	1.4	1.25	5	25.97
Egypt	0.59	6.31	72.79	1.17	5.45	1.17	1.3	4.82	1.89	4.5	27.21
Kenya	1.23	2.44	2.56	79.96	1.55	1.8	3.3	2.42	1.73	3	20.04
Mauritius	1.35	2.88	7.19	1.12	77.58	2.05	2.17	1.2	0.66	3.81	22.42
Morocco	2.58	0.94	1.53	1.81	2.71	81.73	3.35	0.94	2.29	2.12	18.27
Nigeria	1.2	2.62	1.52	1.66	0.91	1.91	84.37	1.05	1.03	3.73	15.63
South Africa	1.26	1.59	5.18	1.34	1.42	1.81	1.92	81.44	1.87	2.16	18.56
Tunisia	2.84	1.24	3.26	2	1.22	3.17	2.07	2.51	80.19	1.51	19.81
Lusaka	2.56	4.8	6.19	0.91	1.6	0.4	3.22	1.54	1.03	77.75	22.25
TO	15.01	23.82	36.06	12.25	18.58	15.52	23.32	17.02	13.13	28.83	203.54
Inc.Own	101.63	97.85	108.85	92.21	96.16	97.26	107.68	98.47	93.32	106.58	cTCI/TCI
NET	1.63	−2.15	8.85	−7.79	−3.84	−2.74	7.68	−1.53	−6.68	6.58	$22.62/20.35$
NPT	6	3	9	2	4	3	7	2	1	8

Table 12. Averaged volatility connectedness during the Russia–Ukraine conflict.

	Botswana	Côte d’Ivoire	Egypt	Kenya	Mauritius	Morocco	Nigeria	South Africa	Tunisia	Lusaka	FROM
Botswana	86.62	0.98	1.63	0.68	1.23	0.72	2.6	1.15	1.39	3	13.38
Côte d’Ivoire	1.39	74.03	7	1.57	2.49	2.49	3.39	1.4	1.25	5	25.97
Egypt	0.59	6.31	72.79	1.17	5.45	1.17	1.3	4.82	1.89	4.5	27.21
Kenya	1.23	2.44	2.56	79.96	1.55	1.8	3.3	2.42	1.73	3	20.04
Mauritius	1.35	2.88	7.19	1.12	77.58	2.05	2.17	1.2	0.66	3.81	22.42
Morocco	2.58	0.94	1.53	1.81	2.71	81.73	3.35	0.94	2.29	2.12	18.27
Nigeria	1.2	2.62	1.52	1.66	0.91	1.91	84.37	1.05	1.03	3.73	15.63
South Africa	1.26	1.59	5.18	1.34	1.42	1.81	1.92	81.44	1.87	2.16	18.56
Tunisia	2.84	1.24	3.26	2	1.22	3.17	2.07	2.51	80.19	1.51	19.81
Lusaka	2.56	4.8	6.19	0.91	1.6	0.4	3.22	1.54	1.03	77.75	22.25
TO	15.01	23.82	36.06	12.25	18.58	15.52	23.32	17.02	13.13	28.83	203.54
Inc.Own	101.63	97.85	108.85	92.21	96.16	97.26	107.68	98.47	93.32	106.58	cTCI/TCI
NET	1.63	−2.15	8.85	−7.79	−3.84	−2.74	7.68	−1.53	−6.68	6.58	$22.62/20.35$
NPT	6	3	9	2	4	3	7	2	1	8

Table 13. Averaged volatility connectedness in the short term (1–5 trading days) during the Russia–Ukraine conflict.

	Botswana	Côte d’lvoire	Egypt	Kenya	Mauritius	Morocco	Nigeria	South Africa	Tunisia	Lusaka	FROM
Botswana	77.39	0.85	1.43	0.56	1.08	0.64	2.29	1.03	1.23	2.63	11.74
Côte d’Ivoire	1.2	68.4	6.79	1.31	2.42	2.16	3.23	1.3	1.11	4.95	24.46
Egypt	0.54	5.47	63.99	1.09	4.33	1.08	1.14	4	1.66	3.85	23.14
Kenya	1.14	2.25	2.05	67.21	1.33	1.53	2.81	1.92	1.55	2.43	17
Mauritius	1.26	2.28	5.32	1.02	63.04	1.64	1.61	0.96	0.54	2.95	17.58
Morocco	2.32	0.81	1.33	1.6	2.51	70.09	2.85	0.84	1.97	1.76	15.98
Nigeria	1.07	2.22	1.18	1.39	0.72	1.41	72.25	0.95	0.83	3.12	12.89
South Africa	1.18	1.37	4.58	1.22	1.35	1.59	1.76	73.75	1.64	1.96	16.66
Tunisia	2.44	1.11	3.14	1.59	1.12	2.87	1.95	2.34	70.84	1.28	17.85
Lusaka	2.31	3.85	5.05	0.73	1.39	0.35	2.62	1.28	0.89	68.06	18.47
TO	13.47	20.19	30.87	10.51	16.25	13.27	20.26	14.62	11.42	24.92	175.76
Inc.Own	90.85	88.59	94.86	77.72	79.29	83.36	92.51	88.37	82.26	92.98	$\mathrm{CTCl}/\mathrm{TCI}$
NET	1.72	−4.28	7.73	−6.49	−1.33	−2.71	7.37	−2.04	−6.43	6.45	$19.53/17.58$
NPT	6	3	9	2	5	3	7	2	0	8

Table 13. Averaged volatility connectedness in the short term (1–5 trading days) during the Russia–Ukraine conflict.

	Botswana	Côte d’lvoire	Egypt	Kenya	Mauritius	Morocco	Nigeria	South Africa	Tunisia	Lusaka	FROM
Botswana	77.39	0.85	1.43	0.56	1.08	0.64	2.29	1.03	1.23	2.63	11.74
Côte d’Ivoire	1.2	68.4	6.79	1.31	2.42	2.16	3.23	1.3	1.11	4.95	24.46
Egypt	0.54	5.47	63.99	1.09	4.33	1.08	1.14	4	1.66	3.85	23.14
Kenya	1.14	2.25	2.05	67.21	1.33	1.53	2.81	1.92	1.55	2.43	17
Mauritius	1.26	2.28	5.32	1.02	63.04	1.64	1.61	0.96	0.54	2.95	17.58
Morocco	2.32	0.81	1.33	1.6	2.51	70.09	2.85	0.84	1.97	1.76	15.98
Nigeria	1.07	2.22	1.18	1.39	0.72	1.41	72.25	0.95	0.83	3.12	12.89
South Africa	1.18	1.37	4.58	1.22	1.35	1.59	1.76	73.75	1.64	1.96	16.66
Tunisia	2.44	1.11	3.14	1.59	1.12	2.87	1.95	2.34	70.84	1.28	17.85
Lusaka	2.31	3.85	5.05	0.73	1.39	0.35	2.62	1.28	0.89	68.06	18.47
TO	13.47	20.19	30.87	10.51	16.25	13.27	20.26	14.62	11.42	24.92	175.76
Inc.Own	90.85	88.59	94.86	77.72	79.29	83.36	92.51	88.37	82.26	92.98	$\mathrm{CTCl}/\mathrm{TCI}$
NET	1.72	−4.28	7.73	−6.49	−1.33	−2.71	7.37	−2.04	−6.43	6.45	$19.53/17.58$
NPT	6	3	9	2	5	3	7	2	0	8

Table 14. Averaged volatility connectedness in the long term (5–infinity trading days) during the Russia–Ukraine conflict.

	Botswana	Côte d’Ivoire	Egypt	Kenya	Mauritius	Morocco	Nigeria	South Africa	Tunisia	Lusaka	FROM
Botswana	9.24	0.14	0.2	0.12	0.14	0.08	0.31	0.12	0.16	0.37	1.63
Côte d’Ivoire	0.18	5.63	0.21	0.26	0.07	0.34	0.16	0.1	0.14	0.06	1.51
Egypt	0.06	0.85	8.8	0.09	1.12	0.09	0.16	0.82	0.23	0.66	4.07
Kenya	0.09	0.2	0.51	12.75	0.22	0.28	0.49	0.5	0.18	0.57	3.04
Mauritius	0.08	0.6	1.87	0.1	14.54	0.41	0.56	0.24	0.12	0.86	4.84
Morocco	0.27	0.13	0.21	0.21	0.2	11.64	0.5	0.09	0.32	0.36	2.29
Nigeria	0.13	0.41	0.34	0.27	0.19	0.5	12.12	0.09	0.2	0.61	2.74
South Africa	0.08	0.23	0.6	0.12	0.07	0.22	0.16	7.69	0.22	0.2	1.9
Tunisia	0.4	0.13	0.11	0.4	0.11	0.3	0.12	0.17	9.35	0.23	1.96
Lusaka	0.25	0.95	1.14	0.18	0.21	0.05	0.6	0.26	0.14	9.68	3.79
TO	1.54	3.63	5.19	1.74	2.33	2.26	3.06	2.4	1.71	3.91	27.78
Inc.Own	10.78	9.26	13.99	14.5	16.87	13.9	15.18	10.09	11.06	13.6	$\mathrm{CTCl}/\mathrm{TCl}$
NET	−0.09	2.12	1.12	−1.3	−2.51	−0.04	0.31	0.5	−0.25	0.13	$3.09/2.78$
NPT	3	5	6	3	2	4	5	5	6	6

Table 14. Averaged volatility connectedness in the long term (5–infinity trading days) during the Russia–Ukraine conflict.

	Botswana	Côte d’Ivoire	Egypt	Kenya	Mauritius	Morocco	Nigeria	South Africa	Tunisia	Lusaka	FROM
Botswana	9.24	0.14	0.2	0.12	0.14	0.08	0.31	0.12	0.16	0.37	1.63
Côte d’Ivoire	0.18	5.63	0.21	0.26	0.07	0.34	0.16	0.1	0.14	0.06	1.51
Egypt	0.06	0.85	8.8	0.09	1.12	0.09	0.16	0.82	0.23	0.66	4.07
Kenya	0.09	0.2	0.51	12.75	0.22	0.28	0.49	0.5	0.18	0.57	3.04
Mauritius	0.08	0.6	1.87	0.1	14.54	0.41	0.56	0.24	0.12	0.86	4.84
Morocco	0.27	0.13	0.21	0.21	0.2	11.64	0.5	0.09	0.32	0.36	2.29
Nigeria	0.13	0.41	0.34	0.27	0.19	0.5	12.12	0.09	0.2	0.61	2.74
South Africa	0.08	0.23	0.6	0.12	0.07	0.22	0.16	7.69	0.22	0.2	1.9
Tunisia	0.4	0.13	0.11	0.4	0.11	0.3	0.12	0.17	9.35	0.23	1.96
Lusaka	0.25	0.95	1.14	0.18	0.21	0.05	0.6	0.26	0.14	9.68	3.79
TO	1.54	3.63	5.19	1.74	2.33	2.26	3.06	2.4	1.71	3.91	27.78
Inc.Own	10.78	9.26	13.99	14.5	16.87	13.9	15.18	10.09	11.06	13.6	$\mathrm{CTCl}/\mathrm{TCl}$
NET	−0.09	2.12	1.12	−1.3	−2.51	−0.04	0.31	0.5	−0.25	0.13	$3.09/2.78$
NPT	3	5	6	3	2	4	5	5	6	6

5.3. Comparison of Net Total Direction Volatility Connectedness During the Three Crises

This section presents a comparative analysis of volatility connectedness among African stock markets during the three selected crisis periods. The figures labeled (a), (b), and (c) respectively illustrate the dynamic evolution of directional volatility spillovers to others, from others, and their difference (referred to as the net total directional connectedness). Additionally, the figures labeled (d) display the directional volatility connectedness and the network structure of volatility transmission among African stock markets, providing a visual representation of the interdependencies and the intensity of spillovers across markets during each crisis period.

5.3.1. GFC Period

An analysis of the charts reveals that for most indices, the total volatility (represented by the black line) exhibited both pronounced and moderate peaks throughout the GFC period, reflecting heightened uncertainty and increased price fluctuations in these markets. This pattern highlights the broader impact of the Global Financial Crisis, which was felt across African financial markets.

Figure 4a shows that South Africa experienced the highest level of total volatility, marked by sharp peaks, indicating a strong market response to the crisis. The “1–5 days” category, representing short-term volatility, contributed significantly to overall volatility, suggesting a high degree of short-term interconnectedness with other markets during this period. Figure 4b demonstrates that Tunisia exhibited the greatest sensitivity to external shocks among the countries analyzed, followed by Mauritius and Morocco, which showed moderate levels of responsiveness. These findings are consistent with those of Guyot et al. (2014), who observed increased integration of North African markets with global financial systems during the crisis, including Tunisia and Morocco. In contrast, Nigeria displayed relatively low levels of volatility connectedness and sensitivity to external shocks. This could suggest either a lower degree of integration into global financial markets or greater resilience to external disturbances during the GFC period.

Nigeria’s case offers a compelling empirical example of how oil price fluctuations can influence market volatility and connectedness. Between July and December 2008, oil prices experienced substantial volatility, falling from a peak of nearly USD 150 per barrel to approximately USD 40 per barrel. Despite this dramatic change, Nigeria’s stock market maintained relatively low connectivity, which may be attributed to two key factors:

• Oil Price Recovery: By mid-2009, oil prices had rebounded to approximately USD 70 per barrel, providing a stabilizing effect for Nigeria’s oil-dependent economy. This recovery acted as a buffer, helping to mitigate the impact of external shocks and contributing to the relative stability of the Nigerian stock market during the latter part of the crisis period. Empirical evidence supports this observation. For instance, Alimi and Ekpenyong (2021) demonstrated that fluctuations in crude oil prices have a direct and significant influence on the Nigerian stock market, with rising oil prices often associated with improved market performance. However, as noted by Adenekan et al. (2020), the positive impact of oil price changes on stock market performance may be short-lived, suggesting that the relationship is subject to temporal variation and broader macroeconomic dynamics.
• Limited Financial Integration: Compared to South Africa, Nigeria’s financial markets were less integrated with the global financial system. This limited integration served as a buffer against the full effects of the Global Financial Crisis. While the Nigerian financial system was not directly exposed to the complex financial instruments at the core of the crisis, such as mortgage-backed securities, it was nonetheless affected through indirect channels. These included reduced global demand for crude oil, a key export commodity, along with capital outflows resulting from heightened risk aversion among foreign investors.

Figure 4c shows that South Africa exhibited the highest levels of connectedness among the African economies, functioning as both a major transmitter and receiver of volatility shocks. This pattern reflects South Africa’s position as the most developed financial market on the continent and aligns with findings from prior studies. South Africa is widely recognized as a key financial hub in Africa, playing a central role in regional economic activity and financial integration (Cenfri, 2020). Moreover, recent research confirms South Africa’s dominant role in both transmitting and absorbing financial shocks across African markets (Yaya et al., 2024).

5.3.2. COVID-19 Pandemic Period

The COVID-19 pandemic triggered significant declines in global equity markets, including African frontier stock markets. However, these markets also experienced a recovery, reaching record-high levels between September 2021 and January 2022. This was particularly the case in South Africa, Egypt, Nigeria, Zambia, Tunisia, and Morocco (Figure 1). This pattern suggests that African equities are more integrated into the global financial system than previously assumed, making them vulnerable to international spillovers during periods of global financial uncertainty. The pandemic period witnessed a marked increase in connectedness across all markets, reflecting the global and systemic nature of the shock.

In Figure 5a, South Africa consistently exhibited the highest “TO others” connectedness, indicating its dominant role as a transmitter of volatility to other African markets. Morocco, Nigeria, and Egypt also demonstrated relatively high outward connectedness, underscoring their importance in the regional transmission of volatility shocks.

Figure 5b shows that South Africa, Egypt, Mauritius, Morocco, and Tunisia experienced the highest levels of “FROM others” connectedness, suggesting that these markets were particularly susceptible to volatility spillovers originating from other countries.

As shown in Figure 5c, South Africa generally acted as a net transmitter of volatility, exhibiting positive net connectedness values, especially during the early part of the pandemic; meanwhile, most other African stock markets alternated between being net transmitters and net receivers of volatility, particularly between January and May 2020. These findings are in line with those of Emenike (2022), who reported evidence of bidirectional volatility spillovers among Southern African Customs Union stock markets during the COVID-19 pandemic period (Emenike, 2022).

During this period, significant bilateral connectedness was observed between several pairs of stock markets, including Morocco and Tunisia, South Africa and Tunisia, Egypt and Tunisia, Mauritius and Côte d’Ivoire, and Nigeria and Morocco. Notably, long-term connectedness exceeded short-term connectedness, indicating more persistent volatility spillovers in the aftermath of the initial shock.

Heightened and prolonged volatility was particularly evident in countries that rely heavily on tourism, such as Egypt, Mauritius, Tunisia, Kenya, Morocco, and South Africa. These economies were severely impacted by international border closures and travel restrictions following the declaration of COVID-19 as a global pandemic in March 2020. The tourism sector in these countries experienced sharp contractions, with international arrivals declining by 80–90% in 2020. Given tourism’s substantial contribution to GDP and employment in these economies, the resulting economic disruption was considerable (UNCTAD, 2021).

In summary, the COVID-19 pandemic period was characterized by increased volatility transmission and heightened susceptibility to external shocks across African stock markets. While African markets are generally considered to be weakly integrated with global financial systems, the COVID-19 pandemic underscored their exposure to global events and their evolving roles in regional volatility dynamics. Some markets acted as net transmitters of volatility, while others served as net receivers. The literature further suggests that despite this susceptibility, certain African markets may offer potential diversification benefits during periods of global uncertainty (Bernard & Sin-Yu, 2020). This view is supported by Yaya et al. (2024), who found that investor attention related to the COVID-19 pandemic influenced stock returns differently across African countries.

5.3.3. Russia–Ukraine War Period

The Russia–Ukraine war has had profound global economic repercussions, with African markets being no exception. Figure 6a shows that during this period, countries with economies that are heavily dependent on commodity exports, such as Nigeria (oil), Botswana (diamonds), and Zambia (copper and cobalt), have experienced heightened volatility, particularly at the onset of the conflict. This increased volatility can be attributed to sharp fluctuations in global commodity prices, which are often sensitive to geopolitical tensions and shifts in global demand. The imposition of economic sanctions and widespread disruptions to supply chains have further exacerbated these price fluctuations, amplifying the economic shocks experienced by resource-dependent economies.

The effects of the conflict are also evident in smaller trade-dependent economies such as Mauritius. Given Mauritius’s strong integration into global trade networks, external shocks originating from major geopolitical events tend to reverberate through its financial system. As such, the charts in Figure 6b–d are likely to reflect increased market volatility and connectedness during this period, highlighting the economy’s vulnerability to external disturbances.

Figure 6b shows that Egypt’s stock market also demonstrates increased connectedness during the war period, particularly due to its reliance on wheat imports from Ukraine and Russia. Disruptions in wheat supply chains resulting from the conflict have led to a surge in global wheat prices, significantly impacting Egypt’s economy (Tanchum, 2022). As one of the world’s largest importers of wheat, Egypt is highly sensitive to global price shocks in this sector, making its financial markets particularly reactive to disruptions in the global grain trade during periods of geopolitical instability.

As the continent’s most advanced and diversified economy, South Africa continued to exhibit high levels of connectedness throughout the war period (see Figure 6d). Its position as a regional economic hub ensures that both regional developments and global economic trends are reflected in its financial indicators. Given the country’s active participation in global markets and its reliance on both exports and imports of key commodities, the war’s impact on international commodity markets and global trade flows is likely to be mirrored in South Africa’s market dynamics.

6. Conclusions

The interconnectedness of stock indices in African economies has received growing scholarly attention, particularly in light of the financial, health, and geopolitical crises that have shaken the global economy in recent decades. This study offers a comprehensive analysis of the impact of three major global crises—the Global Financial Crisis (GFC), COVID-19 pandemic, and Russia–Ukraine conflict—on the volatility connectedness among African stock markets. Utilizing a Time-Varying Parameter Vector Autoregressive (TVP-VAR) framework, the findings reveal substantial interdependence among African markets both within the continent and with global markets, challenging the long-held perception of African markets as isolated or weakly integrated.

Our analysis shows that periods of crisis are characterized by intensified interconnectedness, with short-term linkages proving more pronounced than long-term ones. We find that volatility connectedness among African markets intensified during all crises, peaking during the COVID-19 pandemic (TCI: 33.47%), followed by the GFC (27.41%) and the Russia–Ukraine conflict (20.35%). The surge in connectedness during the COVID-19 pandemic was driven by the global and systemic nature of the shock, with South Africa, Egypt, and notably Morocco acting as major transmitters of volatility, while markets such as Kenya and Tunisia were more often net receivers. These findings affirm the hypothesis that global crises significantly influence volatility and the degree of financial connectedness in African stock markets. We find that specific events have led to substantially elevated levels of interconnectedness, highlighting the sensitivity of African markets to global shocks.

Our results underscore the critical importance of understanding volatility transmission dynamics across African financial systems. South Africa and Egypt consistently emerge as dominant volatility transmitters, reinforcing their roles as financial hubs and systemic influencers in the region. This pattern is consistent with the existing literature, which identifies these markets as regional bellwethers. Moreover, the emergence of Morocco as a major volatility transmitter during the COVID-19 pandemic highlights the unique challenges posed by health crises and contributes a novel perspective to the literature on financial contagion. These insights deepen our understanding of the complex interplay between global events and regional financial systems, moving beyond single-event case studies to offer a comparative and longitudinal perspective on financial resilience, contagion, and systemic risk in emerging markets.

Despite the contributions of this study, several limitations merit consideration. First, the sample was limited to ten African countries; although these are representative of different regions, the sample may not fully capture the continent’s financial heterogeneity. Future research should aim to include a broader spectrum of markets, particularly smaller or less liquid exchanges, in order to provide a more inclusive assessment. The framework proposed in Matsuki et al. (2016) for evaluating regional financial integration may offer a useful methodological extension in this regard.

Second, while this study focused on three major global crises, future work could explore the influence of regional or idiosyncratic shocks such as episodes of political instability or commodity price volatility on African stock market connectedness. The work of Abdelbaki (2013), for instance, has highlighted the importance of political events in shaping market dynamics in North Africa.

Third, although the TVP-VAR model is well suited to capturing time-varying dynamics, it is subject to assumptions regarding model structure and parameter selection. Exploring alternative or complementary methodologies, for instance network analysis or machine learning approaches, could enhance the robustness of the findings. For example, the information-theoretic framework adopted by Nyakurukwa and Seetharam (2023) is capable of capturing both linear and nonlinear dependencies, offering a more nuanced view of market interconnectedness than traditional econometric models.

Additionally, future research should investigate the underlying mechanisms driving volatility transmission in African markets. This could involve examining the roles of macroeconomic fundamentals, investor sentiment, and policy responses in shaping connectedness. As highlighted in Siddiqui et al. (2022), the influence of economic policy uncertainty offers another important avenue for understanding market behavior during periods of elevated uncertainty.

Exploring the implications of heightened interconnectedness for portfolio diversification and risk management strategies is also a promising direction. Studies such as Peterson et al. (2024) have suggested that the benefits of diversification may diminish during crises due to increased co-movement among markets. Understanding these dynamics could inform more effective strategies for international investors and policymakers.

Finally, assessing the impact of regulatory reforms and regional financial integration initiatives on stock market connectedness would provide valuable insights for policymakers aiming to enhance financial stability and support sustainable economic development across Africa.

In conclusion, analyzing the interconnectedness of African stock indices offers critical insights into how regional markets respond to global shocks. This study contributes to the growing literature on emerging market dynamics by enhancing our understanding of the structural and temporal features of financial contagion in Africa. As the global financial landscape continues to evolve, the interconnectedness of African markets will remain a key area of inquiry for ensuring economic resilience and stability across the continent.