Balance sheet and income statement effect on dividend policy of private commercial banks in Ethiopia

Abstract

The study analyzes the impact of balance sheet and income statement on the cash dividends of private commercial banks in Ethiopia. The independent variables employed include liquidity, asset size, leverage, and growth, which are components of the balance sheet, and profitability which is a component of both the balance sheet and the income statement during the period 2010–2020. The generalized moment (GMM) model was used to determine the most important variables that private banks consider when making dividend decisions. In addition, further tests are performed to corroborate the basic findings. Profitability, size, and liquidity are statistically significant characteristics that positively influence the dividend policy of Dashen bank, Wegagen bank, United bank, Lion international bank, Cooperative bank of Oromia, Awash international bank, Bank of Abyssinia, and Nib international bank. Growth and leverage, on the other hand, have a negative and considerable impact on the dividend policy of private commercial banks. The study suggests that the dividend policy of Ethiopian private commercial banks is influenced by both the balance sheet and the income statement. When determining dividend policy, managers of Ethiopian private commercial banks should consider profitability, asset size, liquidity, leverage, and growth.

Keywords:

• Balance sheet
• dividend policy
• income statement and private commercial banks

PUBLIC INTEREST STATEMENT

The balance sheet is a snapshot of what the banks both owns and owes at a specific period in time. They are important on showing company’s net worth at a given time and to give interested parties an insight into the banks financial position. Whereas, the income statement, often called a profit and loss statement, shows a firm’s financial health over a specified time period. Both the balance sheet and income statements are used to make important financial decisions like on dividend policy. This study focused on examining the impact of balance sheet and income statement on the cash dividends of private commercial banks in Ethiopia. The finding of this investigation revealed that profitability, size, and liquidity have significant positive effect. Contrarily, growth and leverage has negative and significant effect. Therefore, when making a dividend policy decision considering these significant factors is vital for commercial banks in Ethiopia.

1. Introduction

Dividend policy is one of the most important aspects of corporate finance that has received a lot of attention. Dividends can be paid in cash, in the form of free stock (bonus issue), or in the form of repurchasing shares (Arnold, 2008; Brealey et al., 2016). A cash dividend is the most frequent method of sharing earnings since it meets the liquidity needs of investors and provides crucial information to shareholders about a company’s current and future prospects (Pandey, 2005). Cash dividends, on the other hand, may limit the amount of funds held by a company to finance future growth and investments; as a result, a firm may be forced to use more external borrowing, which may result in increased regulatory scrutiny and higher financing costs (Ozo, 2014).

Dividends are a portion of a company’s net profit that is distributed to shareholders as a return on their investment in the form of cash or stock. The distribution of dividends has the potential to increase the firm’s share value in the capital market, resulting in improved performance and, in turn, increase market value in the long run. As a result, an effective dividend policy is critical to a business’s ambition to maximize shareholder wealth, and proper understanding about corporate dividend policies can help investors make better investment decisions. (Ali Tijjani, 2019)

Financial researchers Ohlson’s (1995), Abiola (2014), Yusof and Ismail (2016), Kanwal and Hameed (2017), Thirumagal and Vasantha (2018), and Almanaseer (2019) have researched the dividend payout choice and its impact on a company’s worth extensively. When it comes to discussing the relationship between dividend payments and company market value, researchers are divided into two groups. On the one hand, scholars feel that a firm’s dividends and value are unrelated, which is known as the irrelevance theory on the other hand, other researchers feel that a dividend payment is related to a firm’s worth, which is known as the relevance hypothesis. There are two factions within this latter group: one that believes the dividend–firms value relationship is positive (bird-in-hand hypothesis, information signaling hypothesis, and agency cost hypothesis), and the other that believes it is negative (bird-in-hand hypothesis, information signaling hypothesis, and agency cost hypothesis) (tax-related effect hypotheses) Ramadan Budagaga (2020)

There is a large corpus of theoretical and empirical research on the relationship between dividend payments and the value of a company. However, data from advanced markets was used in the bulk of these investigations. To explore this association, these researches used a variety of methods (such as a dividend discount model, a capital asset pricing model, an event study, and a survey). Because banking firms have various distinct traits that set them apart from other businesses, most of these research did not include them in their analyses (Bar-Yosef & Kolodny, 1976; Gordon, 1959; Marsh & Power, 1999). As a result, nothing is known regarding the impact of the bank’s balance sheet and income statement on dividend payments. In this regard, empirical study to evaluate the impact of the balance sheet and income statement on the banking sector’s cash dividend payments in Ethiopia is worthwhile.

The goal of this article is to empirically evaluate the effect of the balance sheet and income statement on the cash dividends of private commercial banks in Ethiopia for the period 2010–2020 in order to address the gaps and limitations in dividend literature described previously. The current study adds to the financial literature because it is the first of its type to look at the impact of the balance sheet and income statement on the dividend policy of private commercial banks operating in Ethiopia over a lengthy period of time. As a result, this research was carried out with two key goals in mind. The first goal is to see if the balance sheet has an impact on the cash dividend payments made by Ethiopian private commercial banks. The study’s second goal is to determine the impact of the income statement on the payment of cash dividends by an Ethiopian private commercial bank.

Furthermore, despite several studies by scholars, the subject of dividend policy’s impact on the balance sheet and income statement remains unsolved. Dividends were named as one of the top ten major unresolved challenges in advanced corporate finance by Berkly and Myers (2005). Dividends are the primary mystery in the economy, according to Black and Scholes (1974). “The more we look at the whole picture, the more it seems like a puzzle, with pieces that just don’t fit together,” wrote Black (1976 researchers are devoting a lot of time and effort to solving the dividend puzzle, which has resulted in a slew of contradictory hypotheses, theories, and explanations. Researchers have generally concentrated on developed markets; nevertheless, an investigation of developing countries, such as Ethiopia, can provide additional insight into the dividend policy debate, which is currently lacking in the literature. Despite this, the financial accounts of Ethiopian private commercial banks demonstrate that only a small portion of the sector’s profits are reinvested this means that a large portion of the profit is distributed as a dividend rather than being saved for future growth. In addition, despite the fact that numerous private commercial banks operate and expand their branches in Ethiopia on a continual basis, only their financial statements reflect a single figure for their financial performance and they pay dividends.

2. Variables, literature review and hypothesis development

2.1. Variables definition

2.1.1. Dependent variable

Dividend policy

- Dividend policy is primarily concerned with the decisions regarding dividend payout and retention. Lasher (2000) described it as the practice adopted by managers in making dividend payout decisions. It details the amount of cash to be distributed to the shareholders and what is to be retained by the firm for further investment. It is a decision that considers the amount of profits to be retained and that to be distributed to the shareholders of the firm (Watson & Head, 2005). The objective of a firm’s dividend policy is to be consistent in the overall objective of maximizing shareholders wealth since it is the aim of every investor to get a return from their investment.

2.1.2. Independent variables

Profitability (ROA)

- is described as a firm’s ability to make money. The profitability of a company is seen to have a significant impact on dividend policy. This is because profitable companies are more ready to pay bigger dividends; hence a positive association between profitability and dividend payments is expected. The signaling theory of dividend payout also supports this conclusion. As a result, profitable companies will find it more important to pay dividends and will be capable of retaining more earnings (Al-Najjar, 2009).

Liquidity (LIQ)

- The liquidity of a company has a significant impact on the distribution of cash dividends. Liquidity is a metric that assesses a company’s ability to meet its obligations. Firms that are more liquid, i.e., have more cash on hand and near-cash assets, pay bigger dividends to shareholders than those with inadequate cash. Signaling theory backs up this beneficial relationship between liquidity and dividend policy. Due to a cash constraint, a bad liquidity position yields a smaller dividend. Moreover, profitability does not imply liquidity; that is, even if a company has high retained earnings, it may not have adequate funds to pay a dividend. As a result, most managers don’t raise dividends unless they’re sure there’ll be enough money to pay them (Brealey et al., 1999).

Growth (GRO):-

Firms that have recently experienced revenue growth tend to pay lower dividends (Chen & Dhiensiri, 2009). If a firm is rapidly expanding, there will be a significant demand for financing. Aside from that, the greater the potential for growth, the greater the need for cash to finance expansion, and the more likely the corporation is to keep revenues rather than pay them as dividends (Chang & Rhee, 1990), reducing agency conflict. The pecking order hypothesis also suggests that companies should fund new projects with the least information-sensitive sources first, such as retained earnings. As a result, companies with significant growth potential are more inclined to keep a larger amount of their revenues to fund development projects rather than paying dividends to shareholders.

Asset Size (AS):-

The firm’s dividend policy is supposed to be explained by its asset size. According to existing research, size may be inversely connected to the likelihood of bankruptcy (Ferri & Jones, 1979; Titman & Wessels, 1988). In previous studies, the asset size variable has become an important variable in explaining the firm’s decision to pay dividends. This suggests that larger corporations can afford to pay more dividends than smaller corporations. To put it another way, asset size can be used as a proxy for the cost of external debt financing, and so a positive relationship between asset size and dividend policy is expected, implying that large companies will incur lower issuing costs. Furthermore, large businesses are more diversified, and their cash flows are more predictable and stable.

Leverage (LEV):-

The debt level is a ratio that indicates how much of a company’s funding comes from outside sources. Dividend payments and capital structure, according to agency models, can help to mitigate information asymmetry issues. Dividends and debt financing can be used to keep cash flow under management control and help to alleviate agency issues. As a result, a negative link between dividend policy and capital structure is expected. Firms with low debt ratios are more prepared to pay higher dividends. The agency costs theory of dividend policy backs it up.

2.2. Empirical literature review and hypothesis development

For more than fifty years, the theoretical rationale for corporate dividend has been a hot topic in corporate finance. Miller and Modigliani (1961) wrote one of the most important books on the subject. They contended that a company’s dividend policy has no bearing on whether its stock price rises or falls. In other words, no matter how meticulously management craft their companies’ dividend policies, no single payout policy can maximize or diminish the value of their owners. Dividend irrelevance theory is the term for this. The following are the dividend theories:

The irrelevance theory of dividends states that a company’s market value is based on the present value of future investment cash inflows, discounted at a needed rate of return. Future dividends, capital gains, or a combination of the two can be used to fund these income streams. As a result, if no earnings are distributed, they will be held as capital gains and dispersed later. Investors who do not get dividends in the current period might produce homemade dividends by selling their shares for a sum equal to what they would get as dividends under ideal market conditions. There are no additional taxes or transaction expenses associated with these earnings, and they have no bearing on the market value of that company’s stock. Firm managers are unable to increase their firm’s value by adopting a certain dividend policy, according to this theory. Dividends, in a nutshell, are meaningless. The irrelevance thesis (Miller and Modigliani) has been one of the most regarded hypotheses in financial literature for nearly six decades.

According to dividend signaling theory, firms employ dividends to signal future free cash flow that is higher than projected. Investors will view a current dividend increase (amplify) as a signal that management foresee permanently greater (lower) future free cash flow levels if managers have private information about future or current cash flow. Dividends are paid by good companies to distinguish themselves from bad companies that cannot afford to pay such a high price to imitate good companies. According to S Bhattacharya (1979), Miller and Rock (1985), and Miller and Rock (1985), and taxes John and Williams (1985) outside financing transaction costs are some of the costly instruments used to achieve a separating equilibrium. Jensen and Meckling (1976) believes that agency difficulties develop in firms when ownership and control are separated, such as in publicly traded corporations with dispersed stock ownership. In organizations with large free cash flows or cash reserves, managers have an incentive to overinvest relative to their first best optimal level. The excess investment originates from the managers’ utility function’s empire-building or perks-prone features. A rise in the dividend reduces the amount of free cash flow available to management, so limiting the problem of overinvestment and adding value to the company. A dividend cut, on the other hand, increases the cash on hand of the managers, exacerbating the problem of overinvestment.

Maturity theory is a dividend theory proposed by Michaely and Grullon (2002), Eugene F. Fama and French (1993), and DeAngelo et al. (1996) who believe that as a firm matures, its investment opportunity set shrinks, resulting in a reduction in systematic risk. A favorable price reaction to a dividend increase indicates that the company has progressed to a stage of lower profitability and risk in its life cycle. According to the Maturity Theory, reactions to news about systematic risk reduction outnumber reactions to news about decreased future earnings, therefore a dividend increase announcement results in a positive stock price response. The initial priority assigned to profitable investment opportunities is hammered home by residual theory. If profitable prospects exist, the company invests in them, and any leftover income is paid to shareholders. The term “residual theory of dividends” refers to a theory that states that the money left over after all appropriate investment possibilities have been pursued should be paid as a dividend. Retained earnings would be used to meet the equity requirement because the cost of retained earnings is lower than the cost of new common stocks. If retained earnings are insufficient to meet this requirement, new common stock will be issued. If the available retention earnings are greater than the requirements, the excess will be dispersed as dividends.

Miller and Modigliani’s (1961) irrelevance hypothesis is supported by a number of research investigations. Watts (1973) explored the association between surprise dividend adjustments, future profitability, and abnormal returns on shares in companies that announced unexpected dividend adjustments. Watts (1973) discovered that unexpected dividend changes provide relatively little information about future earnings and that there are no exceptional gains around the time of dividend announcements using data covering the period 1936 to 1966 Black and Scholes (1974) looked at the investment portfolios of 25 companies listed on the New York Stock Exchange (NYSE) to see how dividend policy affected share prices. In terms of their impact on share price returns, the results demonstrate no meaningful difference between high and low dividend yields. The irrelevance theory is supported by these findings. M. Miller and Scholes (1978, 1982), Miller and Rock (1985), and Miller and Rock (1985a) are among the research that suggest dividend irrelevance. More recently, certain empirical investigations undertaken in various marketplaces have confirmed the dividends’ insignificance. Allen and Rachim (1996) examine 173 companies from 24 industries listed on the Australian Stock Exchange from 1972 to 1985 using the cross-sectional regression model to investigate the link between dividend yield and share price volatility. There is no evidence that dividends are linked to share price volatility, according to the researchers.

De Wet and Mpinda (2013) investigate the effects of dividend payments on the value of a company’s stock using panel data analysis on data covering the period 1995 to 2010 to examine the relationship between dividends and share prices for 46 companies listed on the Johannesburg Securities Exchange (JSE). Their findings, based on a fixed effects model, show that dividend payment is positively connected to market price per share in the long run. For the period 2003–2009, Al-Hares et al. (2012) investigate the value relevance of book value, earnings, and dividends for a sample of all nonfinancial enterprises listed on the Kuwait Stock Exchange (KSE). In the existence of earnings in the valuation model, the results show that dividends are not value-relevant. When dividends are used to replace profits, however, they become valuerelevant. The impact of dividend policy on share price volatility is investigated by Lashgari and Ahmadi (2014) on the Tehran Stock Exchange. They used a multivariable regression model that was based on a dataset from 2007 to 2012. According to the findings, dividend payout ratio has a considerable negative impact on stock price volatility.

Giriati’s (2016) determine the impact of the dividend payout ratio on company value from 2009 to 2013 and use the ordinary least square methodology to analyze data from 29 companies listed on the Bursa Efek Indonesia (BEI). Dividend payouts had a favorable impact on firm value. Budagaga (2017) uses the residual income methodology based on Ohlson’s (1995) evaluation model to investigate the impact of dividend payments on a firm’s value from 2007 to 2015, and the fixed effects methodology was used to panel data for 44 companies listed on the Istanbul Stock Exchange (ISE). The findings reveal a considerable positive link between dividend payments and firm value. Zainudin et al. (2018) investigate the association between stock price volatility and dividend policy of Bursa Malaysia listed industrial products companies. The sample consists of 166 publicly traded industrial products companies from 2003 to 2012. The empirical findings show that dividend policy is a substantial predictor of stock price volatility in Malaysian industrial products firms, especially during the post-crisis period. The effects of dividend policy and ownership on stock price volatility in the Vietnamese market are investigated by Phan et al. (2019). The authors use a large panel dataset of non-financial companies that were publicly traded on the Ho Chi Minh Stock Exchange and the Hanoi Stock Exchange from 2008 to 2015. The findings show that dividend yield reduces stock price volatility in Vietnam’s emerging market. Similarly, Almanaseer (2019) investigates the link between dividend policy and share price volatility in Amman Stock Exchange-listed insurance companies. A total of 20 firms were chosen from a total of 23 insurance companies. According to the findings, there is a considerable negative correlation between share price volatility and dividend yield and payout ratio.

Kolawole et al. (2018) found a favorable or positive influence of dividend distribution and retention ratios on EPS in Nigerian oil and gas companies. Kanwal and Hameed (2017) discovered that dividend payments had a positive impact on a company’s financial success. Thirumagal and Vasantha (2018) believe that dividend payments have a pessimistic or negative influence on shareholders’ wealth in the majority of Indian sectors. The share price differed significantly before and after the dividend announcement. Abiola (2014), the firm’s dividend policy is determined by current and previous year profits, implying that the firm’s profitability is important in the dividend policy pattern. Pandey and Ashvini (2016), a company’s dividend policy is influenced by factors such as debt-to equity ratio, earnings, corporation tax, earnings per share, and the size of the company. Sakinc and Gungor (2015) identified ownership structure as a factor of a firm’s dividend policy. According to Yusof and Ismail (2016), the board of directors of a company should examine profit, debts, investments, and shareholder size before deciding on dividend payout.

It’s worth noting that the majority of the research studies mentioned above were completed in industrialized economies. Furthermore, due to the fact that banks have specific characteristics and reporting norms, these empirical researches omitted banking enterprises from their sample. As a result of the review of previous studies, the following are the two primary hypotheses for this study: Figure 1

H1: In Ethiopia, the balance sheet has no considerable impact on the dividend policy of private commercial banks.

H2: In Ethiopia, the income statement has no considerable impact on the dividend policy of private commercial banks.

Figure 1. Conceptual framework.Sources: Own design 2021

3. Data, methodology and variables

3.1. Data and source

Data were obtained from each banks’ annual audit report the year from 2010 to 2020, and the data sample includes chosen commercial banks listed in Ethiopia (Dashen bank, Wegagen bank, united bank, and lion international bank, Cooperative bank of Oromia, Awash international bank, bank of Abyssinia, and Nib international bank). Banks that did not provide complete data were eliminated, leaving a panel dataset of 8 banks that operated in Ethiopia from 2010 to 2020 as the final sample. As a result, each year’s number of observations varies, resulting in unbalanced panel data. The National Bank of Ethiopia data bases were used to collect the financial and accounting data for this investigation.

3.2. Model specification

To examine the complex relationship between balance sheet and income statement on dividend policy, this study used a Generalized Moment Estimator Method (GMM) model. A GMM model is a panel data estimator that uses the dependent variable’s lags as an instrument to correct the static estimation techniques-related endogeneity bias. In particular, the GMM distinction was established by Arellano and Bond (1991). In order to resolve the inconsistency and prejudices of the static estimation approaches, this estimator uses the first differentiation approach. Simulation experiments, however, have shown that the GMM difference appears to be less effective and produces poor instruments when a sequence is short or continuous (Bun & Windmeijer, 2010). Consequently, to take care of the constraints of the first difference estimator, the system GMM framework was developed. The GMM method uses lagged differences of a dependent variable as instruments for the level equation, and lagged level conditions are often used as instruments for the differentiated equation at the moment (Blundell & Bond, 1998). Since, the framework incorporates more methods; the GMM system is relatively more stable and robust. Thus, when applied to a large panel with a short time span, there are improvements in accuracy.

In treating autocorrelation and heteroscedasticity, the two-step system estimator is more consistent and asymptotically efficient (Arellano & Bover, 1995). Therefore, in evaluating the effect of balance sheet and income statement on dividend policy in a competitive setting, this study used the two-step GMM framework. The available diagnostic tests to verify the results of GMM estimation include. With a null hypothesis that the conditions of the moments are true, and the Arellano and bond test of no second-order serial correlation, the Hansen test of over identifying limitations. Similarly, there is also a distinction in Hansen statistics that measures the exogeneity of the subsets of GMM instruments. Failure to dismiss these null hypotheses thus signifies the validity of the estimates of GMM. This paper therefore followed the partial adaptation model used by Ozkan (2001) to fit the essence of this study with some change. It gives the model as:

(1) $Yit=\left(\lambda -1\right)-1+\beta Xit\hspace{0.17em}+\mu i+\mu t+\epsilon it$(1)

(2) $DPOi,t=\left(\lambda -1\right)-1)DPOit-1+\beta 1Balancesheeti,\beta 2Incomestatmenti,t\sum i,t$(2)

(3) $DPOi,\hspace{0.17em}t\hspace{0.17em}=\left(\lambda -1\right)-1)\hspace{0.17em}DPOit-1+\hspace{0.17em}\beta 1\mathrm{R}\mathrm{O}\mathrm{A}i,t+\beta 2LIQi,t+B3GROi,t+B4ASi,t+B5LEVi,t$(3)

Where:

DPOi,t = Dividend policy of bank i at time t, ROAi,t Return on asset of bank i at time t, LIQi,t Liquidity ratio of bank i at time t GROi,t Gross ratio of bank i at time t, Asi,t Asset size of bank i at time t LEVi,t Leverage of bank i at time t and ∑ = Error term.

3.3. Variables construction

The dividend policy of bank I at time t is used as a dependent variable in this study, which is denoted as DPOi,t and is calculated as total cash dividend paid divided by net income of bank I at year t from 2010 to 2020. This is in line with Anil and Kapoor’s (2008), Marfo Yiadom’s (2011), and Agyei’s (2011) practices. As the major independent variables, the study uses five variables as proxies for the balance sheet and income statement, namely profitability (Return on Asset) liquidity, growth ratio, asset size, and leverage, for bank i at time t, denoted as ROAi,t, LIQi,t, GRi,t, ASi,t, and LEVi,t, respectively. ROAi,t is calculated by dividing net income after taxes by total assets. LIQi,t is equal to current assets minus current liabilities, GROi,t is equal to annual change in total asset, ASi,t is equal to log to total asset, and LEVi,t is equal to total debt minus total equity. The study variables employed in the empirical analysis of this study are summarized in Table 1 below.

Table 1. Variables and definitions

Variables	Symbol	Definitions	Expected direction of effect
Dependent variable
Dividend policy	DOP	The ratio of dividend paid to net income after tax for bank i at time t
Independent variables
Profitability	ROA	The ratio of net income after tax to total asset of bank i at time t	+
Growth ratio	GRO	Annual changes in total assets of bank i at time t	+
Asset size	AS	The annual logarithm of the total asset of bank i at time t	±
Leverage	LEV	The ratio of total debt to total assets of bank i at time t	-
Liquidity	LIQ	The ratio of current asset to current liability for bank i at time t	+

Sources: Owen design 2021

4. Empirical results and discussions

4.1. Descriptive statistics of the variables

Table 2 shows summary statistics for the variables utilized in this study’s regression analysis (mean, SD, minimum, and maximum). The unbalanced panel dataset includes 8 private Ethiopian banks with a total of 80 bank-year observations from 2010 to 2020. As indicated in Table 2, the mean value of DOP is 0.602, indicating that banks have paid out 60.2 percent of their earnings in dividends on average. In terms of standard deviation, bank dividend payments differ from the mean by 12.23%; the minimum and greatest values of bank dividend payments from 2010 to 2020 are 14.8 percent and 82.7 percent, respectively. The mean return on asset (ROA) is 0.036, with a standard deviation of 0.0146. The profitability of banks has a minimum and maximum value of 0.002 and 0.093, respectively. Private commercial banks’ average leverage is 0.1458, with a standard deviation of 0.0578. In Ethiopia, the minimum and maximum leverage values of private commercial banks are 0.09 and 0.507, respectively. The average value of liquidity in private commercial banks in Ethiopia is 0.1563 with a standard deviation of 0.321447. The minimum and maximum value of liquidity of is 0.02 and 0.214 respectively. The average private commercial bank size is 9.807, with a standard deviation of 0.835. Private Banks have a minimum and maximum bank size of 6.42 and 11.05, respectively. Private commercial banks’ average growth ratio is 0.656, with a standard deviation of 0.138. The growth ratio of private commercial banks has a minimum and maximum value of 0.34 and 0.98, respectively.

Table 2. Descriptive statistics of the variables

Variable	Obs	Mean	Std. Dev.	Min	Max
DPO	80	.6026101	.1223553	.148	.827
ROA	80	.0362363	.0146896	.002	.093
LEV	80	.1458125	.0578079	.09	.507
LIQ	80	0.1563	0.321447	0.02	.214
BS	80	9.807	.8357918	6.42	11.05
GRO	80	.65615	.138025	.34	.98

Sources: Stata 14 Summary of statistics result

4.2. Correlation matrix

The correlation matrix of variables utilized in the regression analysis is shown in Table 3. As shown in Table 3 above there is a positive correlation between ROA and dividend distribution with a coefficient of 0.291. Furthermore, with a value of −0.1419, leverage showed a negative connection with dividend payout. With a value of 0.1563, the connection between liquidity and dividend payments indicated a positive indication. A positive association was also found between bank size and dividend payout, with a value of 0.3154. The correlation between growth and dividend shows that a negative connection with 0.0355.

Table 3. Correlation matrixes

DPO	ROA	LEV	BS	LR	OE	GR
DPO	1.0000
ROA	0.2910	1.0000
LEV	−0.1419	−0.2469	1.0000
BS	0.3154	0.2297	−0.6395	1.0000
LIQ	0.1563	−0.1221	−0.1701	0.0417	1.0000
GRO	−0.0355	−0.0485	−0.0336	0.0387	−0.0920	0.1470	1.0000

Sources: Stata 14 correlation analysis

4.3. Regression analysis

Table 4 below reports regression results between the dependent variable (dividend payout) and explanatory variables. The adjusted R² value in Table 4 below indicates that 36.57% of the total variability of dividend policy of private commercial banks was explained by the variables in the model. The underlying requirements of the GMM diagnostic tests are met based on the findings. First of all, the validity of the GMM figures is indicated by the P-value of the Hansen statistics. A second-order serial correlation is also missing, as shown by the AR2 test’s P-value. The disparity in the Hansen test confirms the homogeneity of the instrument subsets and, in the prediction of the dependent variable; the Wald statistics show the shared importance of the explanatory variables. The lagged DPO ratio (DPOit-1) is positive at a 1% significance level with respect to the GMM estimates of the coefficients. The speed of change calculated as (1-λ) is 0.3679, suggesting that cooperatives make a partial adjustment of 36.79% per year. This process of adjustment tends to be sluggish and thus signals the existence of high transaction and floatation costs.

Table 4. Regression analysis of variables

	Coef.	Std. Err.	z	P>|z|	[95% Conf.	Interval]
DPOi,t-1	0.6321	0.035741	1.26	0.000***	.5475621	2.154862
ROA	2.2503	.8568926	2.63	0.009***	.5708213	3.929778
LEV	−.279248	.2780074	−1.00	0.0315**	−.8241325	.2656365
BS	.0454813	.0224576	−2.03	0.0431**	−.0894973	−.0014652
LIQ	.0646514	.0999081	−0.65	0.0518**	−.2604677	.1311649
GRO	−.042178	.1162888	0.36	0.0717*	−.1857438	.2700999
Cons	.9458371	.2359159	4.01	0.000	.4834505	1.408224
Wald stat	11.65
AdjR^2	36.57
AR2(p- value)	0.0703

Sources: Regression result Computed from stata 14. ***, **,* indicates significant at 1%, 5% and 10% significance level respectively.

With a p-value of 0.009, the coefficient on profitability (ROA) was positive and statistically significant at the one percent level. As a result, Ethiopia’s profitable private commercial banks are more likely to pay dividends to their owners. This result is in line with the dividend policy theory’s signaling theory. As a result, the more profitable a private commercial bank is, the more dividends it can pay. Highly profitable enterprises, according to pecking order theory, are able to distribute dividends. According to E F Fama and French (2000), there is a positive relationship between dividends and profitability, which they interpret as evidence for the pecking order theory. Profitable companies will feel it more important to pay dividends and create more retained earnings. This result is also consistent with Lintner (1956), who said that “net earnings were the most important factor in determining current dividend changes.” Apart from Al-Kuwari (2009), Pruitt and Gitman (1991) found that current year profits are crucial variables in dividend policy. As a result, profitability is a significant element in determining dividend distribution.

Table 4 shows that at the 5% significance level, there was a significant positive link between liquidity and dividend payout ratios, with a p-value of 0.0518. According to the findings of this study, having a strong liquidity position improves a bank’s ability to pay dividends. In general, banks with solid and steady cash flows may pay dividends more easily than banks with an uncertain cash flow situation. Prior research and signaling theory back up this favorable relationship between liquidity and dividend policy. Dividend payment is dependent on liquidity, according to Alli et al. (1993). Liquidity and dividend payout ratios have a positive link, according to Amidu and Abor (2006). Liquidity, according to Anil and Kapoor (2008), is a major factor of dividend payout ratio.

Firms with high growth or investment opportunities are expected to keep their earnings to fund their investments, paying less or no dividends. The result, as expected, shows a negative and significant relationship between growth and dividend payout policies, with a p-value of 0.0717 at the ten percent significance level. This reflects the fact that growing banks require more funds to fund their expansion and, as a result, would typically keep a larger portion of their earnings by paying a low dividend. As a result, banks with a lot of money to invest pay lower dividends. The result also matches the expected negative sign predicted by the agency and pecking order theories. This means that private commercial banks in Ethiopia that have a lot of room for growth tend to pay lower dividends. Higgins (1972), who observed that payout ratio is negatively related to a firm’s need for funds to finance growth opportunities, supports this viewpoint. In comparison to a small firm, a large firm typically has better access to capital markets and finds it easier to raise funds with lower costs and fewer constraints. This suggests that as a company grows larger, its reliance on internal funding decreases. Large firms, on the other hand, are more likely to be able to pay higher dividends to shareholders. The result of this study is consistent with (Holder et al. (1998), E F Fama and French (2000). The coefficient of bank size (BS) was positive and statistically significant at the 5% level with a p-value of 0.0431. The result of this study is consistent with Holder et al. (1998), Chang and Rhee (1990), Ho (2003), Al-Malkawi (2007), Redding (1997), and Fama and French (2000).

Large firms are more likely to be mature, have easier access to capital markets, and thus should be able to pay higher dividends because leverage creates fixed charge requirements, highly leveraged firms rely on external financing to a greater extent than those with lower leverage ratios (Aivazian et al., 2006). As a result, leveraged companies should pay lower dividends. The debt ratio was used as a proxy for leverage to test this hypothesis and the result of this study show a negative and statistically significant relationship between leverage and dividend payout ratios. The negative relationship is consistent with the agency theory. Moreover, the finding of this study is consistent with Jensen et al. (1992), Aivazian et al. (2006), Aivazian et al. (2006), Kowaleski et al. (2007), and Al-Kuwari (2009).

4.4. Test of statistical assumptions

4.4.1. Multicollinearity

The variance inflation factor (VIF) and tolerance of the independent variables used in the regression analysis are shown in Table 5. The VIF statistics for each independent variable are calculated to determine whether there is severe multicollinearity between independent variables. If VIF is greater than ten, multicollinearity is present. Multicollinearity is also checked using tolerance (1/VIF). If the tolerance is less than 0.1 and the VIF is greater than 10, multicollinearity is present (Gujarati & Porter, 2009). All of the VIF values are small, as shown in the table, and none of them exceed 10. There are no tolerance values that are less than 0.1.Table 6

Table 5. Multicollinearity test for explanatory variables

Variable	VIF	1/VIF
LEV	1.81	0.552839
BS	1.72	0.581795
ROA	1.29	0.774930
LIQ	1.10	0.911140
GRO	1.03	0.968017
Mean VIF	1.36

Sources: Stata 14 test statistics result

Table 6. Normality

Variable	Obs	W	V	Z	Prob>z
Residual _hat	80	0.95696	2.954	2.373	0.2881

Sources: Stata 14 test statistics result

4.4.2. Normality test

The Shapiro-Wilk W test statistics show probability values of 0.2881, indicating that the data are consistent with the standard distribution assumption. It also implies that the inferences drawn about the population parameters from the sample parameters appear to be correct.

4.4.3. Heteroscedasticity and auto correlation

The two-step GMM estimator method is more consistent and asymptotically efficient in the treatment of heteroscedasticity and auto correlation (Arellano & Bover, 1995).

5. Conclusion

Dividend policy is a very important issue because it determines what funds flow to investors and what funds are retained by the firm for future reinvestment. To this end, this study aimed at examining balance sheet and income statement influence on the dividend policy of private commercial banks’ in Ethiopia from 2010 to 2020. The regression analysis revealed that dividend payout and profitability have significant positive relationships this means that profitable private commercial banks are more likely to pay high dividends and feel it more important to pay dividends and create more retained earnings which supports pecking order theory and signaling theory. Additionally, the finding revealed that a positive relationship between dividend payout and liquidity exists that indicates banks’ ability to pay dividends is increased when their liquidity position is strong and banks with solid and steady cash flows may pay dividends more easily than banks with an uncertain cash flow situation which supports the signaling theory.

The finding also revealed that dividend payout and growth have a significant negative relationship meaning that growing banks require more funds to finance their expansion, so they would typically keep a larger portion of their earnings by paying a low dividend. The pecking order and agency cost arguments were supported by the significant negative coefficient on the growth variable. A positive correlation was also discovered between dividend payout and company size, supporting the agency cost theory that company size can be used as a proxy for external capital market access and larger banks’ can afford to pay higher cash dividends because they have fewer restrictions in accessing external funds from the capital markets and have lower costs. Furthermore, a negative correlation between debt ratio and dividend payouts was discovered, confirming the agency costs theory. This is because, banks’ with less debt have more financial flexibility and are better able to pay and maintain dividends. Finally, the study concluded private commercial banks profitability is affected by balances sheet and income statement component.

5.1. Recommendation

Ethiopian private commercial bank managers’ consider profitability, liquidity, size, leverage, and growth when determining dividend policy, as these are the most significant variables that influence dividend policy. This will assist them in making efficient, effective, and reasonable dividend payout decisions, which will help them, achieve their goal (maximizing profit) and satisfy the needs of their employees and shareholders in the long run. Furthermore, investors looking for paying dividend banks may want to consider profitability, asset size, liquidity, growth, and leverage before deciding on a bank.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The authors received no direct funding for this research.

Notes on contributors

Bayelign Abebe Zelalem

Bayelign Abebe Zelalem is a senior lecturer in Marketing Management, at Mizan Tepi University, Southwest Ethiopia. He earned his BA degree in Management from Bahir-Dar University and MA degree in Marketing Management from Bahir-Dar University, again. His research interest covers business strategy, financial management, e-commerce, consumer behavior, brand management, corporate finance and governance, supply chain management, financial market, corporate social responsibility and entrepreneurship.

Ayalew Ali Abebe

Ayalew Ali Abebe is a senior lecturer in Cooperative, at Mizan Tepi University, Southwest Ethiopia. He earned his BSc in cooperative accounting and auditing from Mizan-Tepi University and MSc in accounting and finance from Hawassa University. He is interested in doing researches in the area of small and medium enterprise (SME), financial management, corporate finance and governance, corporate social responsibility, taxation, accounting information system, supply chain management and entrepreneurship.