Moderating Effect of Profitability on the Relationship Between Capital Structure and Dividend Policy

2.1. Conceptual Framework

Corporate finance relies on capital structure, the mix of debt and equity used to support operations and expansion. This financial strategy has been defined differently by different experts, deepening our comprehension. Panjaitan & Simbolon (2020a) define capital structure as a firm's debt-equity mix for asset financing. This simple concept emphasises debt and equity as capital structure components. According to Mubeen et al. (2022), capital structure is the ratio of total debt to total assets, emphasising leverage and quantifying a firm's reliance on borrowed funds. Giovanni et al. (2020) define capital structure as corporate debt and equity finance. This term emphasises firms' financing mix strategies to balance risk and reward. Pérez (2021) adds that corporate capital structure decisions generally correlate with financial theories like pecking order and trade-off theory, suggesting a strategic approach to debt and equity proportions.

The ability of a corporation to make earnings relative to its spending and other costs is called profitability in corporate finance. It's a fundamental financial indicator that shows a company's resource efficiency and profitability. Several researchers have defined and analysed profitability. Panjaitan & Simbolon (2020b) define profitability as a company's ability to create earnings above its expenses, displaying resource efficiency. Mubeen et al. (2022) note that profitability measures a company's earnings compared to its revenue, assets, or equity. Profitability is crucial to analysing a company's financial health and sustainability. Profitability, a significant financial indicator, is defined by Lindawati (2022) as the ability to generate earnings or profits relative to spending and other costs. Profitability is traditionally used to measure a company's operational efficiency and financial performance. Gao et al. (2023) also consider profitability gaps as motivators for enterprises to increase R&D, highlighting the importance of profitability in strategic decision-making.

Dividend policy is a company's strategic decision to distribute earnings to shareholders. Decide whether to distribute profits as dividends or retain them for growth and investment. It shapes a company's financial structure, affecting shareholder value, investor confidence, and performance. Companies can create a dividend policy that meets their long-term financial goals and maximizes shareholder wealth by considering profitability, cash flow, growth prospects, and shareholder preferences. Dividend policy specifies dividend payment frequency and shareholder profit distribution. It can involve cash dividends, stock dividends, bonus issues, common stock repurchases, or rights issues. Financial stability and growth need balancing dividends and investments (Bashir et al., 2022; Budianto & Bertuah, 2020; Hardianto, 2021; Rahman et al., 2022; Tamrin et al., 2017).

A good dividend policy also includes financial performance, firm size, growth potential, liquidity, leverage, and corporate governance procedures including institutional and managerial ownership. Based on financial status and growth potential, companies may pay dividends regularly, infrequently, or not at all. This approach encourages investors, boosts shareholder confidence, and indicates financial stability. Dividend policy also impacts shareholder interests and investor views of the company's financial health. When setting dividend policies, companies must examine profitability, cash flow, capital needs, and shareholder expectations (Akpadaka et al., 2024; Barros et al., 2020; Li, 2024; Santoso et al., 2020).

In research, "moderation" occurs when a third variable—the moderator—changes the association between two variables. This moderator variable, which can alter the relationship between the independent and dependent variables, complicates the analysis. A third variable modifies the relationship between two others (Baron & Kenny, 1986; Nurchaqiqi & Suryarini, 2018a; Onyebuchi, 2023; Syamsudin et al., 2021).

Return on Assets (ROA), a proxy for profitability, moderates capital structure and dividend policy in this study. ROA moderates the relationship between capital structure variables like DER, AS, and DR and the firm's dividend policy by explaining how asset utilisation efficiency affects the strength and direction of the relationship. ROA complicates the study and illuminates how capital structure variables affect dividend policy. Firms with greater ROA may have a stronger positive association between capital structure and dividend distributions than those with lower ROA.

A control variable in research and experimentation is held constant or changed to isolate the independent variable's effects on the dependent variable. Control variables make sure that altering the independent variable is what causes changes in the dependent variable (Newey & Stouli, 2021; Stouli & Newey, 2018). Research requires control variables to reduce extraneous effects and assure internal validity. Researchers can better examine the independent-dependent connection and make meaningful results by controlling for particular variables. This makes it easier to credit study results to the independent variable (Céspedes et al., 2021; Wang et al., 2021; Weber, 2022).

Control variables in this study are firm size and institutional ownership. These variables are included to account for and maintain their effects on dividend policy. Larger organizations have more resources and different financial strategies, which might affect dividend policy, hence firm size is limited. Institutions may pressure corporations to pay dividends, which might change dividend policy, therefore institutional ownership is controlled.

2.2. Empirical Review

Global research supports and contradicts the African experience. Renaldo et al. (2023) observed that capital structure and stockholders considerably affect dividend policies, while Nurchaqiqi & Suryarini (2018b) found that leverage positively affects dividend policy under direct effect. Sreenu (2018) observed that leveraged enterprises keep more earnings to cover their debt, limiting dividend distribution. This supports agency theory and pecking order theory by showing that increasing leverage disciplines managers by restricting free cash flow and limiting self-serving behaviour.

Malaysian public listed companies' dividend policy decisions were examined by Anuar et al. (2023) utilising data from 83 firms between 2013 and 2017. Their pecking order theory predicted a negative relationship between leverage and dividend policy, but it was inconsequential. C. Arko et al. (2014) found a substantial negative association between leverage and dividend decisions in South Africa and Kenya, and a positive relationship in Ghana.

Nnadi et al. (2013) found that financial leverage negatively and significantly affects dividend policy among African stock exchange-listed enterprises, implying that firms with greater debt ratios pay fewer dividends.

Biwott (2021) identified a strong correlation between leverage and dividend distribution in Nairobi Securities Exchange-listed non-financial enterprises. This contradicts the pecking order idea by suggesting that leveraged corporations pay larger dividends. Capital structure negatively affects dividend payout, according to Irungu (2021). Higher leverage cuts payouts. However, its exclusive emphasis on non-financial enterprises and restricted data up to 2019 limit the study. Profitability boosts dividend payouts, but leverage does not, according to Desoky & Mousa (2019). Ishaku (2020) discovered that debt-to-equity and debt-to-asset ratios significantly lower dividend payment ratios. Total liability negatively and insignificantly affects dividend payout ratio, but debt-to-asset ratio positively and significantly affects it, according to Yahaya (2023). This study supports the pecking order theory and signalling hypothesis, although its tiny sample size limits it. Overall, these studies shed light on sector-specific dividend payouts and policy.

Ahmad et al. (2018) examined dividend policy for European Euronext 100 and Euronext stock exchange non-financial enterprises. Leverage is indirectly connected to greater dividends, suggesting that firms with more leverage may have steady dividend policies that meet investor expectations of financial stability. This confirms the signalling hypothesis and agency theory that leverage limits managerial behaviour and lowers agency costs. However, the study's focus on developed countries may limit its applicability to emerging markets with differing legal and economic conditions.

The debt-to-equity ratio (DER) has conflicting effects on dividend policy. It has a beneficial impact on dividend policy in Indonesia but a negative one in India. A negative association exists between leverage and dividend decisions in Sub-Saharan Africa. Higher debt reduces dividend likelihood in Europe, supporting agency theory and signalling hypothesis.

H₀₁: There is no significant relationship between the debt-to-equity ratio, a measure of capital structure, and dividend policy in Sub-Saharan African-listed manufacturing firms.

Magribi et al. (2023) examined how asset structure, dividend policy, and sales growth affect Indonesia Stock Exchange (IDX) automobile stock prices. Due to collateral value and financial risk reduction, organisations with a higher proportion of tangible assets have better stock prices, according to the study. Dividend policy has no effect on stock prices, suggesting investors prioritise other considerations.

From 2017 to 2019, manufacturing companies registered on the Indonesia Stock Exchange (IDX) have a favourable association between profitability and asset structure on their capital structure (Novita Ardelia et al., 2021). High fixed asset-to-total asset ratio companies were more efficient and obtained more loans, sustaining dividend payments. Capital structure was negatively affected by ROA, but not significantly. Due to their greater borrowing capability and financial stability, corporations with larger fixed assets pay reliable dividends. Careful capital budgeting, good asset liability management, and complete intangible asset utilisation are Liu and Jia (2023) asset structure optimisation methodologies. Nyamasege et al. (2014) found that fixed assets considerably affect a firm's worth, giving higher market values.

A 2011–2020 case study of PT Aneka Tambang Tbk by Akbar and Nurita (2023) examined how current ratio and asset structure affected debt to equity ratio (DER). The current ratio and asset structure did not affect the DER, contrary to prior studies that indicated a correlation between asset structure and capital structure. Long-term asset fluctuations demonstrate the necessity for persistent and smart asset management to support steady financial outcomes.

Empirical evidence from multiple studies underscores the significant role of asset structure in shaping a firm's capital structure and its subsequent impact on dividend policy. Firms with a higher ratio of fixed assets tend to have better access to debt financing and can maintain stable dividend payouts. This relationship highlights the importance of strategic asset management in achieving financial stability and enhancing firm value. Asset structure is a justifiable and essential variable for measuring capital structure, as it provides insights into a firm's ability to leverage its assets to secure financing and support sustainable dividend policies.

H₀₂: Asset structure, as a proxy for capital structure, does not significantly affect dividend policy in Sub-Saharan African-listed manufacturing firms.

The Debt-to-Assets Ratio (DAR) and Dividend Policy are generally negative across various studies. Akhmadi et al. (2020) and Benyadi et al. (2022) found that higher DAR leads to lower dividend payouts in Indonesian manufacturing firms, aligning with the pecking order theory. Mauris & Rizal (2021) and Imronudin et al. (2020) reinforced this negative relationship in non-financial and property sectors. Pattiruhu and Paais (2020) explored the impact of DAR on dividend policy within the real estate and property sectors and found that DAR positively affects the DPO.

The relationship between the Debt-to-Asset Ratio (DAR) and dividend policy is generally negative across various studies, indicating the complexity and variability of DAR's impact on dividend policy across regions and sectors.

H₀₃: The debt-to-asset ratio, a measure of capital structure, does not significantly affect dividend policy in Sub-Saharan African-listed manufacturing firms.

Profitability is an important factor that affects the results and dynamics of numerous financial interactions. Multiple studies have emphasised its function in regulating various financial transactions, underscoring its significance in the field of corporate finance.

The research conducted by Ijaz (2024) investigates the correlation between corporate governance, intellectual capital, and profitability by employing agency theory and the resource-based view theory. The research highlights that the level of profitability influences the relationship between corporate governance and intellectual capital efficiency. This discovery highlights the importance of profitability in guaranteeing the long-term viability and steadiness of firms by improving the efficiency of corporate governance and intellectual capital management.

The study conducted by Nowicki (2024) demonstrates that profitability has a substantial impact on the correlation between the employment coefficient and financial liquidity, as well as between the inflation rate and liquidity. The study demonstrates that these linkages exhibit a positive correlation in organisations with high profitability, while displaying a negative correlation in firms with poor profitability. This underscores the crucial influence of profitability in defining these financial dynamics.

According to Paramita (2023), profitability has the power to reduce the impact of effective corporate governance on the value of a company. This suggests that the financial performance of a company has a substantial impact on the connection between governance practices and the value of the company. It emphasises the importance for companies to maintain high profitability in order to maximise the beneficial impacts of good governance.

Profitability has a crucial role in influencing outcomes and interactions in many financial connections, acting as a substantial moderating factor across different contexts. Its function is vital in guaranteeing the efficiency of corporate governance, the stability of financial liquidity, and the improvement of business value, among other financial factors. Gaining an understanding of how profitability affects financial decision-making and strategy creation in corporate finance can offer significant insights.

H₀₄: Profitability does not significantly moderate the relationship between capital structure and dividend policy in Sub-Saharan African-listed manufacturing firms.

2.3. Theoretical framework

This analysis is based on Modigliani and Miller's 1958 capital structure and dividend irrelevance theory. They believed that in a frictionless market, a firm's value is unaffected by its capital structure or financing decisions. They also claimed that a firm's dividend policy is unimportant to its valuation because investors can produce dividends by selling shares. In 1963, they included taxes to their theory, making capital structure considerations relevant. They struggle to apply their theories due to information asymmetry, transaction costs, and agency issues. This study uses profitability as a moderating variable to explore how capital structure decisions affect dividend policy in sub-Saharan African manufacturing enterprises.

The dividend signalling concept implies that corporations use dividend payments to reveal secret information to investors, which can greatly alter their value. Managers and shareholders can learn about the firm's future performance and management's expectations from this signalling mechanism. In Sub-Saharan Africa, a growing market with distinct difficulties and characteristics, the dividend signalling hypothesis can help explain company financial behaviour.

Lintner (1956)'s Bird in Hand Theory states that investors value dividends over capital gains. According to this hypothesis, investors value dividends more than uncertain gains, like a bird in hand. Dividends transmit good information about a firm's financial health and stability to investors, making them vital to this hypothesis.

Agency theory arises from ownership-management split, causing the agency dilemma. Agency theory advocates injecting debt capital into the capital structure to limit free cash flow and force management to improve operational efficiency, satisfy debt commitments, and avoid bankruptcy. Dividends align managers and shareholders, regulate the relationship between capital structure and business value, and affect agency costs, according to studies.

This study uses agency theory to understand Sub-Saharan African manufacturing enterprises' strategic financial decisions. This research seeks to understand how leverage and dividend policies can promote corporate governance and company value in Sub-Saharan African markets by reducing agency concerns.

This study fills a gap in the literature on capital structure and dividend policy in sub-Saharan African manufacturing enterprises. The literature evaluation shows no current studies on profitability as a moderating component. The study uses Modigliani and Miller's arguments, Dividend Signalling Hypothesis, Bird in Hand Theory, and Agency Theory as a solid theoretical framework. It emphasises business size, institutional ownership, and profitability as moderators of the capital structure-dividend policy link. The study highlights literature gaps and the need for current research to comprehend changing financial dynamics.

3.1. Sample and Population

This research evaluates 48 Sub-Saharan African (SSA) countries' economic, social, and political features. All SSA exchange-listed manufacturers are interesting. For varied continents and sectors, stratified random sampling will be used. The survey included six countries—two West, two East, and two South. These SSA nations were chosen for market capitalization and manufacturing. A variety of countries and sectors are studied to comprehend continental manufacturing businesses. This sampling strategy gives a more accurate population image and aids data analysis. The selected countries will illuminate SSA manufacturing companies' difficulties and possibilities, expanding the field's expertise. Ghana, Nigeria, Kenya, Mauritius, South Africa, Zimbabwe.

3.2. Data Acquisition

Secondary sources were used to obtain financial performance and policy data for manufacturing enterprises in Nigeria, Ghana, Kenya, Mauritius, South Africa, and Zimbabwe from 2012 to 2022. Data was collected from public databases and financial records from the Nigerian Exchange Group (NGX), Ghana Stock Exchange (GSE), Nairobi Securities Exchange (NSE), Stock Exchange of Mauritius (SEM), Johannesburg Stock Exchange (JSE), and Zimbabwe Stock Exchange. Financial data came from Bloomberg, Reuters, and African Financials.

3.3. Analytical Methods

The study employed pooled Ordinary Least Squares (OLS) regression and moderated regression analysis as the primary techniques for data analysis.

3.5.1. Multiple Regression models

Multiple regression models will be used to test hypotheses. This examines the direct effect of the capital structure proxies on dividend policy, which is proxied by DPS.

Model 1: Independent Variables with Country Effects

$${\mathrm{lgDPS}}_{\mathrm{i}}={\mathsf{\beta }}_0+{\mathsf{\beta }}_1{\mathrm{lgDAR}}_{\mathrm{i}}+{\mathsf{\beta }}_2{\mathrm{lgDER}}_{\mathrm{i}}+{\mathsf{\beta }}_3{\mathrm{AS}}_{\mathrm{i}}+{\sum }_{\mathit{j}-\mathbf{1}}^{\mathit{J}-\mathbf{1}}{\mathsf{\gamma }}_{\mathrm{j}}{\mathrm{Country}}_{\mathrm{j}}+{\mathsf{ϵ}}_{\mathrm{i}}$$

(1)

Model 1 examines the impact of various capital structure proxies and country-specific effects on the log-transformed Dividend Per Share (lgDPS_usd) for Sub-Saharan African-listed manufacturing firms. The dependent variable, lgDPS_usd, represents the primary outcome of interest, which is the log-transformed Dividend Per Share for each firm i.

The independent variables in this model include the log-transformed Debt-to-Asset Ratio (lgDAR), the log-transformed Debt-to-Equity Ratio (lgDER), and the Asset Structure (AS).

Country effects are incorporated into the model through a series of dummy variables representing each country, with Nigeria typically serving as the reference category. The term + ${\sum }_{\mathit{j}-1}^{\mathit{J}-1}$γ_j​Country_j​ captures the sum of these dummy variables, where Country_j​ is a binary indicator variable that equals 1 if the firm is located in country j and 0 otherwise. The coefficients γj​​ measure the average difference in the dependent variable (lgDPS_usd) for firms in each country relative to the reference country.

The intercept term, β₀​​, represents the expected value of lgDPS_usd when all independent variables are zero and for the reference country. The coefficients β₁​, β₂​, and β₃​ correspond to the independent variables lgDAR, lgDER, and AS, respectively. These coefficients indicate the expected change in lgDPS_usd for a one-unit change in each independent variable, holding all other variables constant.

The error term, ϵ_i​, captures the variation in the dependent variable that is not explained by the independent variables and country effects. It accounts for unobserved factors and random noise that might influence the dividend policies of the firms.

Model 2: Independent Variables, Control Variables with Country Effects

Model 2 extends the analysis from Model 1 by incorporating two additional control variables: log-transformed Firm Size (lgfsize_usd) and Institutional Ownership (Instown).

Model 2: Independent Variables with Country Effects

$${\mathrm{lgDPS}}_{\mathrm{i}}={\mathsf{\beta }}_0+{\mathsf{\beta }}_1{\mathrm{lgDAR}}_{\mathrm{i}}+{\mathsf{\beta }}_2{\mathrm{lgDER}}_{\mathrm{i}}+{\mathsf{\beta }}_3{\mathrm{AS}}_{\mathrm{i}}+{\mathsf{\beta }}_4{\mathrm{lgfsize}\_\mathrm{usd}}_{\mathrm{i}}+{\mathsf{\beta }}_5{\mathrm{Instown}}_{\mathrm{i}}{\sum }_{\mathit{j}-\mathbf{1}}^{\mathit{J}-\mathbf{1}}{\mathsf{\gamma }}_{\mathrm{j}}{\mathrm{Country}}_{\mathrm{j}}+{\mathsf{ϵ}}_{\mathrm{i}}$$

(2)

Log-transformed Firm Size (lgfsize_usd): This control variable measures the size of the firm, using a logarithmic transformation of its total assets. Larger firms often have more resources, different investment opportunities, and potentially different dividend policies compared to smaller firms. Including lgfsize_usd controls for these size-related differences, providing a clearer understanding of how firm size impacts dividend policies. The coefficient β₄​ indicates the expected change in lgDPS_usd for a one-unit change in lgfsize_usd, holding other variables constant.

Institutional Ownership (Instown): This control variable captures the percentage of a firm's shares owned by institutional investors. Institutional investors can significantly influence corporate governance and financial decisions, including dividend policies. By including Instown in the model, we control for the impact of institutional ownership on dividend policies. The coefficient β₅​ indicates the expected change in lgDPS_usd for a one-unit change in Instown, holding other variables constant.

By adding these control variables, Model 2 aims to isolate the effects of the primary independent variables (lgDAR, lgDER, and AS) while accounting for the potential influence of firm size and institutional ownership. The inclusion of these control variables helps to provide a more accurate and robust estimation of the determinants of dividend policies among Sub-Saharan African-listed manufacturing firms. The country effects, represented by ${\sum }_{\mathit{j}-1}^{\mathit{J}-1}$γ_j​Country_j, remain the same as in Model 1, accounting for country-specific influences on dividend policies

Model 3: Independent Variables, Control Variables, Moderating Variable with Country Effects

Model 3 builds on Model 2 by incorporating an additional moderating variable: the log-transformed Return on Assets (lgROA).

$${\mathrm{lgDPS}}_{\mathrm{i}}={\mathsf{\beta }}_0+{\mathsf{\beta }}_1{\mathrm{lgDAR}}_{\mathrm{i}}+{\mathsf{\beta }}_2{\mathrm{lgDER}}_{\mathrm{i}}+{\mathsf{\beta }}_3{\mathrm{AS}}_{\mathrm{i}}+{\mathsf{\beta }}_4\text{}{\mathrm{ROA}}_{\mathrm{i}}+{\mathsf{\beta }}_5\text{}\mathrm{lgfsize}\text{\_}{\mathrm{usd}}_{\mathrm{i}}+{\mathsf{\beta }}_6\text{}{\mathrm{Instown}}_{\mathrm{i}}+{\sum }_{\mathit{j}-\mathbf{1}}^{\mathit{J}-\mathbf{1}}{\mathsf{\gamma }}_{\mathrm{j}}{\mathrm{Country}}_{\mathrm{j}}+{\mathsf{ϵ}}_{\mathrm{i}}$$

(3)

Log-transformed Return on Assets (lgROA): This moderating variable measures the profitability of the firm, using a logarithmic transformation of the Return on Assets. The Return on Assets ratio indicates how efficiently a firm uses its assets to generate profit. By including lgROA, Model 3 examines how profitability influences dividend policies. The coefficient β4\beta_4β4​ indicates the expected change in lgDPS_usd for a one-unit change in lgROA, holding other variables constant. Including lgROA helps to understand the role of profitability in shaping the firm's dividend decisions.

The inclusion of lgROA in Model 3 provides additional insights into the relationship between profitability and dividend policies, complementing the analysis of capital structure variables from the previous models. The control variables (lgfsize_usd and Instown) and country effects ${\sum }_{\mathit{j}-1}^{\mathit{J}-1}$γ_j​Country_j​ + ϵ_i​ remain the same as in Model 2, continuing to account for the impact of firm size, institutional ownership, and country-specific factors on dividend policies.

By incorporating the moderating variable lgROA, Model 3 aims to provide a further understanding of the determinants of dividend policies among Sub-Saharan African-listed manufacturing firms. This model allows us to assess the combined effects of financial leverage (capital structure), profitability, firm size, institutional ownership, and country-specific factors on dividend policies, leading to a more robust analysis.

Model 4: Independent Variables, Control Variables, Moderating Variable, Moderating effects and Country Effects

Model 4 builds on the previous models by introducing moderating effects into the analysis. This model includes the independent variables, a moderating variable (ROA), control variables, interaction terms (moderating effects), and country-specific effects.

$${\mathrm{lgDPS}}_{\mathrm{i}}={\mathsf{\beta }}_0+{\mathsf{\beta }}_1{\mathrm{lgDAR}}_{\mathrm{i}}+{\mathsf{\beta }}_2{\mathrm{lgDER}}_{\mathrm{i}}+{\mathsf{\beta }}_3{\mathrm{AS}}_{\mathrm{i}}+{\mathsf{\beta }}_4\text{}{\mathrm{ROA}}_{\mathrm{i}}+{\mathsf{\beta }}_5\text{}\mathrm{lgfsize}\text{\_}{\mathrm{usd}}_{\mathrm{i}}+{\mathsf{\beta }}_6\text{}{\mathrm{Instown}}_{\mathrm{i}}+{\mathsf{\beta }}_7\text{}\mathrm{lgDAR}\text{\_}\mathrm{lgROA}{\mathrm{i}}_{}+{\mathsf{\beta }}_8\text{}\mathrm{lgDER}\text{\_}{\mathrm{lgROA}}_{\mathrm{i}}+{\mathsf{\beta }}_9\text{}\mathrm{lgAS}\text{\_}{\mathrm{lgROA}}_{\mathrm{i}}+{\sum }_{\mathit{j}-\mathbf{1}}^{\mathit{J}-\mathbf{1}}{\mathsf{\gamma }}_{\mathrm{j}}{\mathrm{Country}}_{\mathrm{j}}+{\mathsf{ϵ}}_{\mathrm{i}}$$

(4)

Moderating Variable (lgROA_i​​): The log-transformed Return on Assets (lgROA) measures the profitability of the firm. It helps to understand how profitability influences dividend policies. The coefficient β₄​ indicates the expected change in lgDPS_usd for a one-unit change in lgROA, holding other variables constant.

Interaction Terms (Moderating Effects):

lgDAR_lgROA: The interaction between the Debt-to-Asset Ratio and Return on Assets. This term captures how the effect of lgDAR on lgDPS_usd changes with varying levels of lgROA. The coefficient β7\beta_7β7​ indicates the change in the impact of lgDAR on lgDPS_usd for a one-unit change in lgROA.

lgDER_lgROA: The interaction between the Debt-to-Equity Ratio and Return on Assets. This term captures how the effect of lgDER on lgDPS_usd changes with varying levels of lgROA. The coefficient β8\beta_8β8​ indicates the change in the impact of lgDER on lgDPS_usd for a one-unit change in lgROA.

AS_lgROA: The interaction between Asset Structure and Return on Assets. This term captures how the effect of AS on lgDPS_usd changes with varying levels of lgROA. The coefficient β9\beta_9β9​ indicates the change in the impact of AS on lgDPS_usd for a one-unit change in lgROA.

Model 4 aims to provide a distinct understanding of the determinants of dividend policies among Sub-Saharan African-listed manufacturing firms by examining the combined effects of financial capital structure variables, profitability, firm size, institutional ownership, and country-specific factors. The inclusion of interaction terms allows for the assessment of how profitability moderates the relationship between these variables and dividend policies, providing a more comprehensive analysis.

3.4. Construct Validity Table

Table 1. Construct Validity Table.

Variable Type	Variable Name	Measurement	Source
Dependent	Dividend Policy	DPS is one of the variables are widely used for measurement of dividend policy, (Emeka & Ogochukwu, 2021; Udoka et al., 2022)	Annual Reports of sample firms via various Exchanges and financial databases - Bloomberg, Reuters, and African Financials.
Independent	Debt to Equity (DER)	$DebttoEquity=\left({\displaystyle \frac{TotalDebt}{TotalEquity}}\right)$ (Nurhikmawaty et al., 2020)
Independent	Asset Structure (AS)	Asset Structure $={\displaystyle \frac{\left(TotalNoncurrentAssets\right)}{TotalAssets}}$as defined by (Brealey, 2020)
Independent	Debt-to-Asset Ratio (DAR)	$DebttoAsset=\left({\displaystyle \frac{TotalDebt}{TotalAsset}}\right)$as defined by (Arhinful & Radmehr, 2023)
Moderating	Return on Assts (ROA)	ROA $=\left({\displaystyle \frac{\left(NetIncome\right)}{TotalAssets}}\right)$ as defined and used by (Nurfitria et al., 2023).
Control	Firm Size (fsize)	Natural Logarithm of Revenue = $\ln \left(Revenue\right)$ is as used by (Kehinde Adewale et al., 2019) as a measure of firm size.
Control	Institutional Ownership (InstOwn)	5% Percentage of total outstanding shares held by instutitional owners, (Jara-Bertin et al., 2012; Pucheta-Martínez & Chiva-Ortells, 2020)
Fixed Effect	Country	Dummy variables for each country (e.g., Nigeria, Kenya, Mauritius, South Africa, Zimbabwe, Ghana, Others) with Nigeria as the omitted baseline category in the model.	Specific to study design

Table 1. Construct Validity Table.

Variable Type	Variable Name	Measurement	Source
Dependent	Dividend Policy	DPS is one of the variables are widely used for measurement of dividend policy, (Emeka & Ogochukwu, 2021; Udoka et al., 2022)	Annual Reports of sample firms via various Exchanges and financial databases - Bloomberg, Reuters, and African Financials.
Independent	Debt to Equity (DER)	$DebttoEquity=\left({\displaystyle \frac{TotalDebt}{TotalEquity}}\right)$ (Nurhikmawaty et al., 2020)
Independent	Asset Structure (AS)	Asset Structure $={\displaystyle \frac{\left(TotalNoncurrentAssets\right)}{TotalAssets}}$as defined by (Brealey, 2020)
Independent	Debt-to-Asset Ratio (DAR)	$DebttoAsset=\left({\displaystyle \frac{TotalDebt}{TotalAsset}}\right)$as defined by (Arhinful & Radmehr, 2023)
Moderating	Return on Assts (ROA)	ROA $=\left({\displaystyle \frac{\left(NetIncome\right)}{TotalAssets}}\right)$ as defined and used by (Nurfitria et al., 2023).
Control	Firm Size (fsize)	Natural Logarithm of Revenue = $\ln \left(Revenue\right)$ is as used by (Kehinde Adewale et al., 2019) as a measure of firm size.
Control	Institutional Ownership (InstOwn)	5% Percentage of total outstanding shares held by instutitional owners, (Jara-Bertin et al., 2012; Pucheta-Martínez & Chiva-Ortells, 2020)
Fixed Effect	Country	Dummy variables for each country (e.g., Nigeria, Kenya, Mauritius, South Africa, Zimbabwe, Ghana, Others) with Nigeria as the omitted baseline category in the model.	Specific to study design

3.5. Diagnostic Procedures

The study focuses on the reliability and robustness of regression analysis in African-listed manufacturing firms. Diagnostic tests were conducted to ensure the accuracy and reliability of the results. The Multicollinearity Test was used to identify independent variable multicollinearity, which could impair the results. Heteroscedasticity was also assessed using the Breusch-Pagan and Koenker tests. Statistically robust standard errors were used to correct heteroscedasticity, enhancing the robustness of the analysis. The Normality Test for Residuals was used to examine residuals for normality, which is crucial for valid hypothesis testing and confidence intervals. The Jarque-Bera test was used to determine regression normality, and if non-normal, variable modification, robust regression, model re-specification, and outlier detection and treatment were investigated. The study emphasizes the importance of residual normality in regression analysis.

4.1. Descriptive statistics

Descriptive statistics, Table 2, provide an overview of the central tendency, dispersion, and distribution of the dataset's variables. Key financial variables of Sub-Saharan African-listed manufacturing firms include log-transformed Dividend Per Share (lgDPS_usd), log-transformed Debt-to-Asset Ratio (lgDAR), log-transformed Debt-to-Equity Ratio (lgDER), Asset Structure (AS), log-transformed Return on Assets (lgROA), log-transformed Firm Size (lgfsize_usd), and Institutional Ownership (Instown).

4.2. Correlation Matrix

The correlation matrix examines key financial variables of Sub-Saharan African-listed manufacturing firms, revealing several noteworthy relationships among the variables.

The correlation matrix, as in Table 3, reveals several noteworthy relationships among the variables, such as a weak negative correlation with lgDAR, suggesting that firms with higher debt relative to assets tend to have slightly lower dividend payouts. Similarly, the correlation with lgDER is weakly negative, indicating that higher debt relative to equity is associated with slightly lower dividends. In contrast, the correlation with AS is minimal, indicating no significant relationship between asset structure and dividend payouts.

For lgDAR, there is a moderate positive correlation with lgDER, suggesting that firms with higher debt-to-asset ratios tend to also have higher debt-to-equity ratios. The correlation with AS is weakly positive, suggesting a slight relationship between the debt-to-asset ratio and the firm's asset structure. Additionally, there is a moderate negative correlation with lgROA, indicating that more leveraged firms tend to be less profitable. The correlation with lgfsize_usd is very weakly negative, and there is a weak negative correlation with Instown, suggesting that firms with higher institutional ownership may pay slightly lower dividends.

4.3. Results and Discussion

Table 4. Multiple and Moderated OLS Regression Results.

Variable	Model 1 Coef. (SE)	Model 2 Coef. (SE)	Model 3 Coef. (SE)	Model 4 Coef. (SE)
lgDAR	0.006 (0.012)	0.001 (0.011)	0.013 (0.015)	0.014 (0.027)
lgDER	-0.012** (0.003)	-0.013** (0.003)	-0.010** (0.003)	-0.009** (0.005)
AS	0.047** (0.012)	0.046** (0.012)	0.049** (0.012)	0.048** (0.012)
Ghana	0.006 (0.005)	-0.038** (0.009)	-0.036** (0.009)	-0.029** (0.009)
Kenya	0.038** (0.007)	0.034** (0.007)	0.032** (0.007)	0.029** (0.006)
Mauritius	0.051** (0.004)	0.039** (0.005)	0.042** (0.005)	0.043** (0.005)
South Africa	0.086** (0.009)	0.058** (0.008)	0.059** (0.008)	0.061** (0.008)
Zimbabwe	0.014** (0.005)	-0.049** (0.012)	-0.049** (0.012)	-0.048** (0.013)
lgfsize_usd	-	0.036** (0.007)	0.036** (0.007)	0.034** (0.007)
Instown	-	-0.016* (0.008)	-0.017* (0.008)	-0.015 (0.008)
lgROA	-	-	0.084** (0.033)	0.127 (0.095)
lgDAR_lgROA	-	-	-	-0.077 (0.102)
lgDER_lgROA	-	-	-	-0.070** (0.018)
AS_lgROA	-	-	-	0.303** (0.093)
Intercept	-0.007 (0.007)	0.003 (0.007)	-0.009 (0.008)	-0.014 (0.009)
R-squared	0.2197	0.2355	0.2505	0.2618
F-Statistics	34.04	30.40	28.47	23.89
Prob > F	0.000	0.000	0.000	0.000
AIC	-2621.93	-2646.02	-2671.14	-2685.89
BIC	-2573.60	-2586.94	-2606.70	-2605.34
Number of observations	1588	1588	1588	1588
*** p<.01, ** p<.05

Table 4. Multiple and Moderated OLS Regression Results.

Variable	Model 1 Coef. (SE)	Model 2 Coef. (SE)	Model 3 Coef. (SE)	Model 4 Coef. (SE)
lgDAR	0.006 (0.012)	0.001 (0.011)	0.013 (0.015)	0.014 (0.027)
lgDER	-0.012** (0.003)	-0.013** (0.003)	-0.010** (0.003)	-0.009** (0.005)
AS	0.047** (0.012)	0.046** (0.012)	0.049** (0.012)	0.048** (0.012)
Ghana	0.006 (0.005)	-0.038** (0.009)	-0.036** (0.009)	-0.029** (0.009)
Kenya	0.038** (0.007)	0.034** (0.007)	0.032** (0.007)	0.029** (0.006)
Mauritius	0.051** (0.004)	0.039** (0.005)	0.042** (0.005)	0.043** (0.005)
South Africa	0.086** (0.009)	0.058** (0.008)	0.059** (0.008)	0.061** (0.008)
Zimbabwe	0.014** (0.005)	-0.049** (0.012)	-0.049** (0.012)	-0.048** (0.013)
lgfsize_usd	-	0.036** (0.007)	0.036** (0.007)	0.034** (0.007)
Instown	-	-0.016* (0.008)	-0.017* (0.008)	-0.015 (0.008)
lgROA	-	-	0.084** (0.033)	0.127 (0.095)
lgDAR_lgROA	-	-	-	-0.077 (0.102)
lgDER_lgROA	-	-	-	-0.070** (0.018)
AS_lgROA	-	-	-	0.303** (0.093)
Intercept	-0.007 (0.007)	0.003 (0.007)	-0.009 (0.008)	-0.014 (0.009)
R-squared	0.2197	0.2355	0.2505	0.2618
F-Statistics	34.04	30.40	28.47	23.89
Prob > F	0.000	0.000	0.000	0.000
AIC	-2621.93	-2646.02	-2671.14	-2685.89
BIC	-2573.60	-2586.94	-2606.70	-2605.34
Number of observations	1588	1588	1588	1588
*** p<.01, ** p<.05

The findings from all four models consistently demonstrate a strong negative correlation between the debt-to-equity ratio (lgDER) and dividend policy (lgDPS). In Model 1, the regression coefficient for lgDER is -0.012, which is statistically significant at the 5% level. This implies that an increase in the debt-to-equity ratio is associated with a decrease in dividend distributions. The negative connection is consistent in Models 2, 3, and 4, with coefficients of -0.013, -0.010, and -0.009, respectively, all statistically significant at the 5% level. The results contradict the null hypothesis (H01) and demonstrate a substantial negative correlation between the debt-to-equity ratio and dividend policy. This implies that firms with higher levels of debt are more likely to decrease the amount of money they distribute to shareholders as dividends. This could be done to save cash for repaying their debts and to minimise the risks associated with financial difficulties.

Across all four models, there is a consistent and significant positive correlation between the asset structure (AS) and dividend policy. In Model 1, the coefficient for AS is 0.047, which is statistically significant at the 5% level. This suggests that companies with a larger proportion of tangible assets are more likely to distribute bigger dividends. The positive link is consistent in Models 2, 3, and 4, with coefficients of 0.046, 0.049, and 0.048, respectively. All the coefficients are statistically significant at the 5% level. The results of the study contradict the null hypothesis (H02) and indicate that the asset structure has a major impact on dividend policy. Companies that possess significant tangible assets are more likely to have easier access to borrowing money and face less financial risk, which allows them to sustain or even raise the amount of money they distribute as dividends.

The findings from the analysis of the debt-to-asset ratio (lgDAR) across all models suggest that it does not exert a substantial influence on dividend policy. The coefficient for the variable lgDAR in Model 1 is 0.006, indicating that there is no statistical significance. The absence of statistical significance persists in Models 2, 3, and 4, where the coefficients are 0.001, 0.013, and 0.014, respectively. None of these coefficients reach a level of significance. The data provide evidence in favour of the null hypothesis (H03), indicating that the debt-to-asset ratio has no substantial impact on the dividend policy of manufacturing firms listed in Sub-Saharan Africa. This suggests that although companies may utilise debt to fund their assets, the percentage of assets funded by debt does not have a significant impact on their dividend distributions.

By incorporating profitability (lgROA) as a moderating variable in Models 3 and 4, we gain valuable insights into its impact on the connection between capital structure and dividend policy. The coefficient for the natural logarithm of return on assets (lgROA) in Model 3 is 0.084, which is statistically significant at the 5% level. This suggests that there is a positive relationship between profitability and dividend payouts, meaning that higher profitability is related with higher dividend per share. Model 4 drill down into the moderating effects by incorporating interaction terms. The coefficient of the interaction term (lgDER_lgROA) between the debt-to-equity ratio and profitability is -0.070. This coefficient is statistically significant at the 5% level, indicating that profitability mitigates the negative effect of the debt-to-equity ratio on dividend distributions. The coefficient for the interaction term between asset structure and profitability (AS_lgROA) is 0.303, which is statistically significant at the 5% level. This suggests that profitability strengthens the favourable impact of asset structure on dividend distributions. Nevertheless, the interaction effect of the debt-to-asset ratio and profitability (lgDAR_lgROA) lacks statistical significance. The result of this study rejects the null hypothesis (H04) and show that profitability has a considerable moderating effect on the relationship between capital structure and dividend policy. Profitability plays a vital role in the financial decision-making of manufacturing firms in Sub-Saharan Africa. It can help offset the negative impact of high debt and enhance the advantages of a robust asset structure when it comes to dividend payouts.

4.4. Post Estimation Diagnostic Tests

Table 5 shows the Post-estimation diagnostic tests that were performed on the regression models to verify the accuracy and consistency of the findings. The R-squared values and F-statistics were essential for evaluating the explanatory capability and overall relevance of the models. The R-squared values varied from 0.2197 in Model 1 to 0.2618 in Model 4, suggesting that the models account for 21.97% to 26.18% of the variability in dividend policy (DPS). The F-statistics exhibited a high level of significance in all models, indicating that the independent variables collectively account for a substantial proportion of the variance in DPS.

The Variance Inflation Factor (VIF) was computed for all independent variables in each model. The average VIF values for Models 1 to 4 are comfortably below 10, suggesting that multicollinearity is not a serious issue in these models. The presence of heteroscedasticity was evaluated using the Breusch-Pagan/Cook-Weisberg test and White's test, both of which consistently rejected the null hypothesis of homoscedasticity in all models. The Shapiro-Wilk test was used to assess the normality of the residuals, while the Akaike's Information Criterion (AIC) and Bayesian Information Criterion (BIC) were employed to compare the models' goodness of fit.

The Wald test was employed to assess the collective significance of the coefficients in each model. The outcomes revealed an F-statistic of 3605.61 and a p-value of 0.0000, so affirming the joint importance and rejecting the null hypothesis that all constraints are zero.

5. Conclusion and Recommendations

5.1. Conclusion

The study examines how profitability affects the connection between capital structure and dividend policy in industrial companies listed in Sub-Saharan Africa. The research aims to investigate the impact of the Debt-to-Equity Ratio (DER) and asset structure (AS) on dividend policies. It also seeks to analyse the influence of the Debt-to-Asset Ratio (DAR) on dividend policies. Additionally, the research aims to evaluate the role of profitability in moderating the relationship between capital structure and dividend policy.

The findings indicate that DER has a negative impact on dividend policy, whereas AS has a beneficial impact on it. The impact of the DAR is favourable, but it is not statistically significant. This implies that the DAR alone may not have a large influence on the dividend policy of industrial firms listed in Sub-Saharan Africa. The impact of profitability on the link between debt-to-equity ratio (DER) and asset size (AS) with dividends per share (DPS) is strong, highlighting its crucial role in financial decision-making.

The study presents concrete evidence that Distributed Energy Resources (DER) have a detrimental impact on dividend policy, whereas Asset Size (AS) has a beneficial impact on it. The impact of the DAR is positive, but it is not statistically significant. Additionally, profitability has a significant role in moderating the link between DER and AS with DPS, highlighting its importance in making financial decisions. These findings have practical consequences for professionals and decision-makers in Sub-Saharan Africa.

5.2. Recommendations

The study proposes various suggestions to enhance the capital structure and dividend policies of industrial enterprises listed in Sub-Saharan Africa. The Debt-to-Equity Ratio (DER) has a substantial influence on Dividend Policy (DPS), so companies should strive to optimise their capital structure in order to maintain a balance between debt obligations and dividend payments. Financial advisers and corporate finance departments should create debt management strategies that focus on reducing high levels of leverage while yet preserving financial flexibility. Regulatory organisations should establish criteria to guarantee that companies maintain leverage levels that promote stable dividend programs.

Companies that possess significant tangible assets should convey the consistency and predictability of their cash flows to shareholders in order to bolster shareholder trust. Clear and open communication regarding investments in assets and their beneficial effect on dividends is crucial, and regulatory bodies should provide structures to facilitate this alignment.

The Debt-to-Asset Ratio (DAR) shows a positive correlation with DPS, although the impact is not statistically significant. However, it is advisable for enterprises to closely monitor and manage this ratio to mitigate any hazards that may arise from excessive leverage. To ensure financial stability and match leverage with long-term dividend plans, it is advisable to implement integrated financial management systems and conduct regular assessments by both internal and external auditors.

The relationship between capital structure and dividend policy is heavily influenced by profitability, as measured by Return on Assets (ROA). Companies should perform comprehensive profitability studies and establish profit allocation frameworks that effectively balance investments, debt settlement, and dividend distributions. Engaging in collaboration with local regulatory organisations and industry groups helps streamline the exchange of optimal methods around the region.

5.3. Study Limitations

The study on manufacturing enterprises listed in Sub-Saharan Africa has limitations, such as its exclusive emphasis on particular industries and the distinct financial attributes of each sector. The economic, regulatory, and market conditions in Sub-Saharan Africa may further impact the relevance of the findings on a worldwide scale. The application of quantitative techniques to evaluate capital structure, profitability, and dividend policy may not comprehensively encompass the intricacies of company financial decision-making. Qualitative methodologies, such as conducting case studies or engaging in interviews with business managers, have the potential to yield a more profound comprehension. The conclusions are derived from past data, and the dynamic economic and regulatory landscape may impact their applicability in the future.

5.4. Suggestions for Further Study

This study emphasises the necessity for additional investigation into the influence of corporate governance frameworks on dividend policy and capital structure decisions in Sub-Saharan Africa (SSA). It proposes including variables such as board membership, ownership structure, and CEO salaries to gain a more thorough insight. Furthermore, it emphasises the significance of macroeconomic variables such as inflation, interest rates, and currency rate volatility in comprehending the correlation between capital structure and dividend policy. The study proposes broadening the focus beyond manufacturing to encompass more industries and integrating qualitative research techniques such as interviews and case studies to offer a more comprehensive understanding of the decision-making processes of corporate managers.