Factors influencing the dividend policy of manufacturing companies listed on the ho chi minh stock exchange

This study employs quantitative research methods such as the Ordinary Least Squares OLS methodology, Random Effects Model REM, Fixed Effect Model FEM and Feasible Generalised Least Squar

INTRODUCTION

RATIONALE OF THE RESEARCH

In the current stock market context, filtering stocks on the HOSE and HNX exchanges with dividend yields of 7% or higher (taking 7% as a benchmark due to bank deposit interest rates), only 27 out of 756 companies from both exchanges meet this criteria, accounting for 3.57% The company with the highest dividend yield is at 18.7%, with an average of 9.6% for this group Only 11 out of the 27 companies are listed on the HOSE, accounting for 40% of the filtered sample, and just 3% of the total listed companies on the HOSE Essentially, if a company is experiencing good growth momentum, with expectations exceeding the cost of capital (g>r), the company will reinvest, leading to higher enterprise value, thereby increasing shareholder benefits

Dividend policy is one of the crucial decisions in corporate financial management, alongside investment decisions and financing decisions According to Brigham and Houston (2013), dividend policy is a profit distribution policy after taxes, indicating what is the percentage of being retained for reinvestment and what is the percentage of being paid out as dividends to shareholders Dividend policy affects the operation and development of the company If the decision is to pay dividends, then the question of how much to pay and in what form needs to be addressed Investors prefer cash dividends over expectations of future profits due to risk factors Dividend policy plays an essential role in business operations and it is influenced by various factors such as profit, growth rate, company size, and so on There have been numerous theoretical and empirical studies worldwide pertaining to dividend policy However, the results of some empirical studies were different from theoretical research and they are not consistent under different conditions The theory of Miller & Modigliani (1961) suggested that dividend policy does not affect the value of the company; however, some empirical studies such as Partington (1985), Baskin (1989), among others, have found a relationship between dividend policy and stock price In addition, there have been many studies on factors influencing dividend policy worldwide, such as Rozeff (1982), Fama & French (2001), Liu & Hu (2005), DeAngelo, DeAngelo, & Stulz (2006), Nizar Al -Malkawi (2007), Ahmed & Javid

(2008), Gill, Biger, & Tibrewala (2010) These research not only have produced some similar results but inconsistent findings that need further investigation as well That is the reason why the thesis was completed with a view to determining the factors affecting the dividend payout policy of companies and clarifying influence level of those factors with a scope of manufacturing companies.

RESEARCH OBJECTIVES

The general objective of the thesis is analysing the factors affecting dividend payout policies of manufacturing companies listed on Ho Chi Minh City Stock Exchange (HOSE) From the research results, the thesis proposes several policy implications to support corporate managers to increase the payout dividends in the future

To achieve the above general objective, the thesis carries out the following specific objectives:

Firstly, identify the factors affecting the dividend payout policy of the enterprises

Secondly, measure the level and direction of the factors on dividend payout policies of the enterprises

Thirdly, from the research results, the thesis proposes relevant policy implications to support corporate managers to increase the payout dividends of the enterprises.

RESEARCH QUESTIONS

To achieve these objectives, this thesis aims to answer these following questions:

1 What factors affect the dividend payout policy of the enterprises?

2 How do factors affect the dividend payout policy of manufacturing enterprises listed on HOSE?

3 What policy implications need to be implemented to adjust the dividend payout policies of manufacturing enterprises listed on HOSE in the dividend payout process?

OBJECTS AND SCOPE OF RESEARCH

The research objects of this thesis are factors affecting dividend payout policy of manufacturing enterprises listed on the Ho Chi Minh City Stock Exchange (HOSE)

This research focuses on cash dividends as a method of dividend payment, although other forms like stock dividends and asset dividends exist.

In terms of space: there are 135 manufacturing companies listed on HOSE, however there are a lot of companies which do not have enough needy data for researching That is the reason why the thesis uses a research sample of 62 manufacturing companies listed on HOSE The research data is secondary data, collected from the FiinProX software

In terms of time: This thesis studies factors influencing dividend payout policies during the period 2013 – 2022 This period is long enough for the thesis to estimate and to provide a quite comprehensive overview of the economic landscape of Vietnam Additionally, collecting sufficient data will enhance the reliability of the research.

RESEARCH METHODOLOGY

The thesis uses quantiative research methods with panel data regression techniques, namely POOLED OLS, FEM, REM The author also use Stata 17 software to complete all the methods and choose the most suitable model The author uses the Hausman test to choose between FEM and REM and the Breusch-Pagan test to choose between the Pooled OLS and REM However, these models suffer from the defect of autocorrelation and heteroscedasticity Therefore, the research thesis uses the FGLS model to overcome the above defects

The research data is secondary data, collected from the Fiinpro-X – a financial platform.

CONTRIBUTIONS

From the scientific perspective, there are a host of reseach about the factors affecting dividend payout policy of the enterprices However, the results of those previous studies are not staying the same throughout This thesis, therefore, will provide one more empirical evidence about factors affecting the dividend payout policy of manufacturing companies listed on the Ho Chi Minh City Stock Exchange (HOSE)

From the practical perspective, the thesis analyses the factors affecting the dividend payout policy of manufacturing companies listed on HOSE As a result, the thesis would proposes a number of related policy implications to support managers in the process of paying dividends.

STRUCTURE OF THE THESIS

The graduation thesis “Factors Influencing the Dividend Policy of Manufacturing Companies Listed on the Ho Chi Minh Stock Exchange” is carried out in 5 chapters with the following layout:

Chapter 1: Introduction In this chapter, the author presents an overview of the research thesis including the reason for choosing the research topic, objectives, research questions, scope of the study, research methods and contributions of the thesis and research layout to serve as a basis for implementation

Chapter 2: Literature review In Chapter 2, the author will present the overview of the dividend policy, theories of dividend policy, an overview of the factors affecting dividend payout policy through previous research and dividend payout policy including definition and measurement At the same time, the author also analyses theories on dividend policy and reviews some previous research to build a research model for chapter 3

Chapter 3: Methodology In this chapter, the process of the research can be shown The thesis, furthermore, can present models and propose research hypotheses, research methodology, and simultaneously describe the research data

Chapter 4: Analyzing research results In this chapter, the author is going to present descriptive statistics, research estimates, providing research results Finally, the thesis shows results and discusses the results to give the conclusion

Chapter 5: Conclusion and recommendation From the research results in chapter 4, chapter 5 presents the conclusions of the study and proposes some policy implications, limitations and suggests future research directions

This chapter lays out the foundation for the research thesis, outlining the rationale behind its selection, its specific objectives and questions, the research scope, methodologies employed, and anticipated contributions It also serves as a blueprint for the thesis structure, guiding the development of subsequent chapters.

LITERATURE REVIEW

OVERVIEW OF DIVIDEND PAYOUT POLICIES

2.1.1 Dividend and dividend payout policies

The National Assembly (2014), according to the Enterprise Law, dividends are understood as the net profit paid for each share in cash or other assets from the remaining profits of the joint stock company after implementation financial obligations Or it can be understood that dividends are the company's after-tax profits paid to the owners (shareholders) Dividends can be paid in cash (the most common form), they can also be paid in shares or property After one year of business operations, the enterprise makes a profit, there are two main options in distributing profits to shareholders, one is to retain the profits earned within the company, the other is to retain the profits earned within the company distribute profits through dividend payout s or repurchasing its own shares

A company's profit allocation strategy, whether prioritizing dividend payouts or reinvestment, varies significantly While established companies may distribute a large portion of their profits as dividends, growth-oriented companies typically retain earnings to fuel future development Effective reinvestment of retained earnings ensures long-term growth, leading to potential dividend and stock price increases Conversely, misusing or ineffectively investing retained earnings directly harms shareholders.

Dividends, as defined by Omerhodzic (2014), are payments made by companies to their owners and shareholders Typically, dividends are distributed in cash, representing a portion of the company's earnings However, companies may also issue dividends in the form of shares or other assets on a periodic basis.

According to Le Minh Dao (2014), dividend policy is a business's profit distribution decision, in which the business will be faced with the choice of using post-tax profits to continue investing or paying dividends to shareholders Deciding on dividend policy is one of the three most important decisions in corporate financial management (besides investment decisions and financing decisions) and it is an important part of a company's financial strategy, characterized by the decision of the board of directors to determine the portion of the company's profits that will be distributed to shareholders as dividends or withheld profits for reinvestment

In the research of Miller & Modigliani (1961), Omerhodzic (2014), dividend policy involves determining the amount of cash used to pay dividends to the company's common shareholders because when businesses decide to pay dividends, they have to spend a large amount of cash Dividend policy, financing and investment policy are closely related to each other, because paying dividends will reduce the amount of profits the company retains for reinvestment, thus affecting the need for capital mobilization, the company's external capital Therefore, the decision to pay dividends cannot be separated from the company's business results Companies with good financial performance, especially profitability, can provide dividends in line with shareholder expectations

Dividend policy determines how a company distributes its after-tax profits, either by reinvesting in new projects or distributing to shareholders Common distribution methods include cash dividends, stock dividends, and share buybacks A company's dividend policy is influenced by factors such as profitability, liquidity, financial leverage, and company size.

According to Clause 3, Article 135, Enterprise Law 2020, dividends can be paid in cash, company shares or other assets specified in the corporate charter If payout is made in cash, it must be used in Vietnam Dong and according to payout methods prescribed by law After making a decision on dividend payout at the Board of Directors meeting, the Board of Directors will announce information on the dividend payout scale on the announcement date Most profitable companies will share profits with their investors by paying dividends One common rule is that half of the profits are paid out to shareholders, and half are kept for reinvestment in the company Companies in the growth stage may retain most or even all of their profits to reinvest in the company, so they will pay very low or no dividends There is no legal requirement that companies pay dividends to their shareholders, even when the company is profitable A company's board of directors can increase, decrease, or waive dividend payout Dividends are usually paid quarterly, monthly, semi-annually or annually Companies can pay dividends in many forms: paying dividends in cash is the most common form In addition, companies could pay dividends in the form of stocks, or assets, and so on

According to Le Minh Dao (2004), most dividends are paid in the form of cash This is the most common form because cash dividends represent the percentage distribution of earnings to shareholders in cash, reducing the company's assets and equity (retained earnings) Cash dividends are paid on a per share basis It can also be paid as the percentage of face value Par value is the value fixed in the stock certificate according to the company's charter To pay cash dividends, the company must meet two criteria including (1) the company must have enough cash to make the dividend payout ; (2) the company must have enough retained profits

According to HOSE, Le Minh Dao (2004), cash dividends, the procedure depends on 4 days:

The first day, the declaration day The day on which the company's Board of Directors announces the decision to pay dividends to shareholders On this day, the company also declares the dividend payout per share, the closing date also known as the last registration date, the ex-dividend date also known as the trading day without dividends and dividend payout date

The second day, the record day On this day, the company will finalize and announce the list of shareholders receiving dividends If shareholders sell shares after the last registration date, they will still receive dividends even though they no longer own the shares The record day is usually set a certain period before the payout date

On the ex-dividend date, investors who purchase shares will not receive the upcoming dividend payment This is because the trade occurs after the book closing date, meaning the new owner is not on record to receive the dividend As a result, the stock price typically drops by an amount equal to the dividend value on the ex-dividend date.

The fourth day, the payout day The date on which dividends will be paid to shareholders on the list announced to receive dividends This date is approximately two weeks after the closing date According to Clause 4, Article 135, Law on Enterprises 2020, dividends must be paid in full within 06 months from the end of the Annual General Meeting of Shareholders On this day, shareholders will receive their dividends through their securities accounts

Cash dividends offer investors with low risk tolerance a stable income stream and signal strong company performance, but they come with drawbacks such as reduced cash flow, lower asset value, and potential for increased debt and financial distress.

Stock dividends are an alternative to cash dividends, offering shareholders reassurance when a company prioritizes reinvesting cash for growth This approach, as outlined by Omerhodzic (2014) and Le Minh Dao (2014), allows companies to retain cash for investment while still providing value to shareholders.

A stock dividend is a proportional distribution of additional shares of a company to its shareholders For example, the company declares a stock dividend of 5% per year This means that for every 100 shares owned by current shareholders, they will receive 5 additional shares

MEASURING DIVIDEND POLICY IN CASH

Jiang et al (2017), Faradisi & Ulpah (2021) proposed a measure that the ratio of cash dividends to net cash flow from operating activities (DVC - Cash dividends scaled by cash flow) Because net cash flow from operating activities measures the ability to fulfill tax obligations, pay debt, and dividends to shareholders better than profits Because profit for the period is determined based on income regardless of whether money has been collected or not yet collected, while net cash flow from operating activities is measured from cash income during the period And reality shows that many companies recorded positive profits during the year but could not meet payout obligations Therefore, a higher DVC shows that the amount of cash the business spends to pay dividends accounts for a large proportion of the cash flow collected from operating activities

Several studies, including those by Yulivan (2022), Michaely & Qian (2016), Vo Xuan Vinh (2022), Faradisi & Ulpah (2021), and Phung Duc Nam (2018), have identified "dividend payout ratio" as a key metric, often measured as cash dividends scaled by earnings (DVE) DVE represents the proportion of net profits distributed to shareholders in cash, indicating a company's willingness to share earnings generously, as highlighted by Gallagher & Andrew (1997).

DVC, DVE are defined as follows:

This study investigates dividend policy by analyzing two key ratios: the dividend-to-net income ratio (DVE) and the dividend-to-net operating cash flow ratio (DVC) These measures are commonly employed in previous research, including studies by Phung Duc Nam et al (2018), Faradisi & Ulpah (2021), Jiang et al (2017), and Michaely & Qian (2016).

BACKGROUND THEORIES

Dividend independence theory, also known as M&M theory According to Miller & Modigliani (1961), this theory builds on the view of dividend policy in the context of no taxes, transaction costs or market imperfections The view on dividend independence was given by Miller & Modigliani (1961) in a perfect capital market: information costs are negligible, information reaches everyone equally, there are no tax distortions, no transaction costs or contracting costs contracts, intermediary brokerage costs, the dividend payout policy of a business does not affect the value of the business, the dividend policy will be independent In other words, all fixed policies which means they are the same Accordingly, the enterprise value is discounted from the net cash flow from future projects up to the present time, knowing that the net cash flow has a constant growth rate

The perfect financial market hypothesis assumes an efficient market where no single participant can manipulate prices, all investors have equal access to information, and there are no transaction costs or tax discrepancies This ideal scenario, proposed by Modigliani and Miller (M&M), disregards real-world complexities like brokerage fees, transfer taxes, and differential tax treatment between distributed and undistributed profits.

Investors are rational actors who prioritize wealth maximization They are indifferent to whether increased wealth comes from dividend payouts or stock price appreciation, as long as their overall assets increase.

The third hypothesis was showed by Miller & Modigliani (1961) that there is no asymmetric information Both the board of directors and outside investors have equal access to all information about the company such as cash flow and future profits

In addition, to defend the opinion that a company's dividend policy does not affect stock prices, Miller & Modigliani (1961) argued that investors could create their own dividend policy through selling shares to generate income instead of depending on the company's financial policies However, hypotheses such as no taxes and no transaction costs in a perfect market do not exist in reality Regarding the self- created dividend policy from M&M's point of view, in reality investors have to spend an amount of transaction cost and pay taxes when selling stocks and receiving dividends Investors, therefore, would buy shares of companies that meet the needs of transaction costs and the lowest tax payable, and hence dividend policy would also affect the value of the company

2.3.2 Bird-in-the-hand theory

In an imperfect market, dividend policy influences firm value, as argued by Gordon (1963) and Lintner (1956) Gordon attributed this to investor risk aversion, with investors favoring certain dividend income over uncertain capital gains This preference stems from the immediate and guaranteed nature of dividends compared to the future, uncertain returns from capital gains As the proverb goes, “a bird in the hand is worth two in the bush” (Gallagher & Andrew, [Year]).

1997) Dividends, therefore, reduce risk for shareholders thus allowing the company to discount future earnings at a lower discount rate which increases the value of the company.

The Signalling Theory, as proposed by Miller & Rock (1985), suggests that management possesses superior knowledge about a company's financial future compared to shareholders A higher-than-expected dividend declaration is perceived as a signal of optimistic financial prospects, leading investors to buy more shares and drive up stock prices Conversely, a dividend reduction indicates management's pessimism about future earnings, causing investors to react negatively.

However, Bhattacharya (1979) showed that there is always an information asymmetry between investors and managers, so dividend payout announcements contain information about future earnings that may be different from investors' expectations For example, when a company has many prospects for future development and requires a lot of capital to finance projects, then this company tends to reduce dividends to shareholders However, a lot of investors believe that the company's reduction in dividend payout is because the management is pessimistic about the company's upcoming business results and of course the stock price will decrease

Investors tend to like stocks of companies that meet a specific need According to Hooi et al (2015), this is explained by the fact that in an imperfect environment, investors face different tax impacts on dividends and capital gains as well as face impacts from the transaction cost When trading stocks, Miller & Modigliani (1961) argued that to minimize the impacts of taxes and transaction cost, investors tend to prefer investing in stocks of companies that bring the benefits they desire Companies will attract different customer groups based on their dividend policies Especially, Nizar Al‐Malkawi (2007) asserts that companies in the growth stage tend to pay lower dividends, attracting customers who prefer capital gains, while companies in the maturity stage paying higher dividends will attract customers who need income in the form of dividends

For businesses that implement corporate financial policy for a long time and attract a certain number of shareholders If the company changes its dividend policy, it may make a group of the company's existing shareholders unhappy In case, Lewellen et al (1978) argued that a company's stock price might be negatively affected by shareholders selling their shares and investing in another company with a more suitable dividend policy Therefore, the Clientele effect demonstrated through companies implementing dividend policies to attract different groups of investors to maximize company value According to Nizar Al‐Malkawi (2007), Clientele effect was expressed through two groups: Clientele effect is driven by taxes and Clientele effect is driven by transaction costs

Stock liquidity and dividend policy

The clientele effect suggests that companies pay dividends to meet shareholder demand for stock liquidity This preference for dividends, especially among investors holding illiquid stocks, arises from the desire to avoid high transaction costs associated with selling securities Cash dividends provide income without incurring these costs and help shareholders balance their portfolios by increasing their cash holdings Research by Banerjee et al (2005) and Brav et al (2005) confirms the importance of stock liquidity in influencing corporate financial policies Dong et al (2005) further highlight the role of low transaction costs in driving individual investors' preference for dividends.

In the study of Miller & Modigliani (1961), with the hypothesis that no tax exists in a perfect market, investors will have no distinction between dividends and capital gains However, in reality, taxes always exist in the market Specifically, when the tax rate on capital gains through selling securities to enjoy price differences is higher than the income from dividends, investors tend to prefer dividends over capital gains

On the other hand, when the tax rate on income from dividends is higher than on capital gains, investors would tend to want to receive income from selling securities rather than receive dividends Because individual investors would be often subject to higher tax rates on dividend income than capital gains, companies may often not implement corporate financial policies to maximize benefits for shareholders In this case, another form of income distribution implemented by companies is to buy back shares, with this form shareholders will not incur higher tax rates for the income received

In conclusion, Miller & Modigliani (1961) also recapped that the market is not perfect when there are differences in taxes and transaction costs that have a significant impact on the board of directors' decision to pay dividends Both income tax and transaction cost perspectives form two customer groups: the group that prefers capital gains over dividends

DeAngelo et al (2006) proposed that a small age company will have many investment opportunities but the company's income is not enough to meet the capital demands needed to finance new projects Therefore, forcing these companies to mobilize more capital from outside, they may have to spend higher capital mobilization costs than mature companies Therefore, management boards will tend to retain the profits generated to capture investment opportunities instead of paying dividends to shareholders

FACTORS AFFECTING DIVIDEND POLICY

According to Omrhodzic (2014), there are a multitude of the factors that influence dividend policy This thesis is going to mention several factors as the following:

Jensen (1986) highlighted the conflict between managers and shareholders regarding dividend policy, with shareholders generally favoring dividends over retained earnings when excess cash flow exists Liquidity plays a crucial role in dividend policy, influencing a company's ability to pay dividends Amidu and Abor (2006) emphasize that dividend payments depend not only on profits but also on free cash flow, which is the remaining operating cash after capital expenses Consequently, liquidity is a significant factor in determining the dividend payout ratio.

Jensen et al (1992), Fama and Frech (2001) supposed that the large companies would spend a higher amount of net profit to pay cash dividend than the small ones Large-scale companies may pay higher dividends and small-scale companies might pay lower dividends because they have more difficulty raising capital than large companies, which have easier access to the capital market and are less dependent on internal funds, leading to the ability to pay high dividends That is the reason why the larger companies, the higher ability they could access to the capital market

Research by Pruitt and Gitman (1991) and Ahmed and Javid (2009) suggests that profitability is a key driver of dividend policy Companies with higher profits, stable income, and strong cash flow tend to pay higher dividends This is because dividends can only be paid out of profits, highlighting the importance of profitability for dividend decisions.

The impact of business growth opportunites on dividend policy is a controversial factor when economists have conflicting opinions Some research suggested that revenue growth opportunites has a positive impact on dividend policy According to signaling theory, companies with high growth will try to convey information to investors by increasing dividend payout Adu-Boanyah et al (2013) conducted research on the factors affecting the dividend policy of manufacturing enterprises and concluded that the growth rate of revenue has a positive impact on the enterprise's payout decision

However, the research results of Amidu and Abor (2006) stated that revenue growth has a negative relationship with dividend payout ratio According to them, growth companies need more cash to finance their growth, so company management will often withhold earnings at a higher rate than the dividend payout ratio

Financial leverage is the debt-to-equity ratio that represents the company's capital structure The larger the leverage, the more debt the business uses and the possibility that the retained profit ratio is higher than the dividend payout ratio because the business have to finance the business activities, pay interest and principal to creditors, thereby reducing the debt leverage ratio According to Rozeff (1982), companies with high financial leverage ratios often have low payout ratios to reduce expenses, increase income and maintain financial leverage balance Al-Malkawi

(2007) also confirmed that financial leverage has a negative effect on dividend

Contrary to some findings, Kania and Bacon (2005) discovered a positive correlation between high debt ratios and dividend payments, suggesting that companies with significant debt may pay higher dividends to boost investor confidence and signal continued strong performance This aligns with signal theory, which posits that such actions serve as a reassurance to investors.

According to Kasmir (2014), the level of cash holdings serves as a gauge to determine the amount of available cash for settling a company's debts According to Prihadi (2010) cash is the most liquid part of the total current assets, then the funds used to pay debts are cash From the research of Nguyen Thi Ngoc Trang & Bui Kim Phuong (2017), the relationship between this factor and the decision to pay dividends is in the same direction and statistically significant at the 1% level, which means that the higher the level of cash holding companies have, the higher the probability of paying dividends.

LITERATURE REVIEW

Research related to the factors affecting dividend policy has been carried out in many countries, mainly researching the impact of the liquidity, profitability, firm size, financial leverage investment opportunities, revenue growth rate, and level of cash holdings After that, this topic in countries around the world continued to develop with many different results

Research by Nguyen Thi Ngoc Trang & Bui Kim Phuong (2017) aimed to find out what factors are affected by the financial policies of non-financial enterprises listed on the Vietnam stock market in the period 2008 - 2015 through the Fama - MacBeth regression technique In which, the factors affecting corporate financial policy mentioned in the study by Nguyen Thi Ngoc Trang & Bui Kim Phuong (2017) are variables representing company characteristics including scale, profitability, growth opportunities, leverage, finance, risk, company life cycle, level of cash holdings, liquidity

Ngo Vu Mai Ly (2016) used data from 112 companies listed on HOSE for 5 years (2008 - 2012) to prove that revenue growth rate has a negative impact on dividend payout policy It can be explained by the fact that, when the company has a high revenue growth rate, managers tend to use profits to reinvest more

Recently, the research of Vo Xuan Vinh (2022) used data samples of companies listed on the Ho Chi Minh City Stock Exchange (HOSE) in the period 2007 - 2015 Through Tobit regression techniques and Logit, the study found a negative relationship between stock market liquidity and dividend payout in companies on the HOSE Vo Xuan Vinh (2022) argued that businesses in Vietnam tend to compensate for the illiquidity of their stocks by paying more dividends to shareholders In addition, the study also provides evidence that ROA, growth opportunities, and financial leverage have a positive impact on dividend policy, while company size has an opposite impact on the decision to pay dividend news of the company

According to research by Tran Thi Hai Ly & Do Thi Bay (2015), Nguyen Huu Tan et al (2022), large-scale companies are often mature companies with easy access to the capital market cheaper, less costly and subject to fewer constraints Therefore, it has the ability to pay higher cash dividends In contrast, research by Phan Ngoc

Research by Thuy Nhu & Nguyen Kim Phuoc (2021) indicates that large-scale businesses prioritize reinvesting profits for future growth over dividend payouts to shareholders Additionally, studies by Tran Thi Hai Ly & Do Thi Bay (2015) and Mai Thi Hoai Thuong (2015) demonstrate a positive correlation between previous year's dividend policy and the current year's, suggesting a preference for stable dividend policies to attract long-term investors Furthermore, Tran Thi Hai Ly & Do Thi Bay (2015) found that state ownership positively influences dividend policy, likely due to the prevalence of equitized state-owned enterprises in Vietnam, where dividend decisions are influenced by state ownership Moreover, state-owned companies often have easier access to external capital, particularly bank loans, facilitating a higher dividend payout policy.

Research by Dinh Bao Ngoc & Nguyen Chi Cuong (2014) focused on the profitability factor Research confirmed that profitability has a positive impact on dividend policy After its business cycle, the enterprise generates profits that exceed expectations or the efficiency of its production and business activities becomes more efficient Larger companies tend to pay more cash dividends to shareholders

Nguyen Huu Tan et al (2022) built a model from 94 construction industry companies listed on both HOSE and HNX from 2015 to 2020 and showed that the inflation rate has a negative impact on dividend policy On the other hand, previously Nguyen Duc Vu (2016) conducted research on the relationship between inflation rate and dividend policy of 32 oil and gas enterprises listed on both HOSE and HNX from

2008 to 2014 Results From the FEM model, it shows that the inflation rate has a positive impact on dividend policy at the 5% significance level Therefore, although businesses cannot intervene in macro factors, they can completely predict changes in the macro environment and make appropriate adjustments to minimize adverse effects of these factors to dividend policy

Miller & Modigliani's 1961 study, based on perfect market conditions, concluded that dividend policy has no impact on company value This theory, however, relies on unrealistic assumptions, leading later theories to refine its assumptions and address its limitations.

Gordon's (1963) dividend stream theory posits that investors prefer cash dividends due to their risk-reducing properties However, dividend payments can impact a company's cash flow and liquidity, limiting managerial discretion Jensen and Meckling (1986) further elaborated on this concept with their agency cost theory, offering a more comprehensive explanation of Gordon's original idea.

(1986) concluded that ownership ratio has an impact on dividend policy Individuals with a high ownership ratio participate in corporate governance and they are the ones who make dividend policies This creates a conflict of interests between company managers and shareholders

Among factors related to the operations of businesses in general, revenue is the factor of primary concern Revenue affects operating efficiency, profits and thereby affects shareholders' dividends However, the results of previous studies are not consistent Bushra and Mirza (2015); Fajaria and Isnalita (2018) revealed that the company's revenue growth rate has a positive relationship with its dividend payout policy This is reaffirmed through the research results of Dempsey et al (2019) In contrast, the research results of Alzomaia & Al-Khadhiri (2013) showed that growth does not affect the company's dividend policy In addition, Al-Khadhiri (2013) also showed that financial leverage does not affect dividend policy Research by Zhang, Uchida, Dong (2020) concluded that debt ratio has an opposite impact on dividend policy because a company with a lot of debt will prioritize profits to repay debt rather than pay dividends to shareholders Gill et al (2010) analyzed the factors that determine the dividend payout ratio for Manufacturing and Service companies in the

US The ratio of market value to book value represents an investment opportunity that has a significant negative impact on the dividend policy of the manufacturing industry At the same time, financial leverage impacts in the same direction as dividend policy in the service industry but in the manufacturing sector, it has the opposite direction Growth rates and dividend payouts in the service industry are inverse

While Lintner (1956) found a positive correlation between company size and dividend policy, a study by Bushra & Mirza (2015) examining 75 Pakistani companies revealed a negative relationship between these factors, highlighting the potential for contrasting findings depending on context and regional differences.

METHODOLOGY

RESEARCH PROCESS

Overall, the research starts from identifying the research objectives then make a brief review of theory and previous studies After that, the author continue building hypotheses and models, then the author performs model calculations and proposes related recommendations The research process can be presented in detail including six steps:

Step 1: The author determined the research objectives Then, the author reviews the theoretical frameworks and previous studies related to this topic After conducting a brief theoretical review, the author further analyzes the research limited to propose related empirical research topics

Step 2: Based on the theoretical review and related empirical studies, the author considers an appropriate model to measure the impact of several factors on dividend policy From there, the hypothesis about the expected impact direction of the variables is put forward

Step 3: After proposing the model, the author collected data Then, the data will be intermediately calculated and processed in accordance with the requirements for the parameters needed for the model The author uses econometric software for research estimation

Step 4: The author chose the appropriate model based on the results of the Ordinary Least Squares model (POOLED OLS), fixed effects model (FEM) and random effects model (REM) In case the model could not meet the defect tests

(including tests of homoscedasticity, autocorrelation and multicollinearity), the author uses FGLS regression technique to overcome

Step 5: The author analyzed and estimated the research results The contents include performing descriptive statistical analysis, reviewing the correlation coefficient matrix between variables Based on the estimation results, the author analyzes the results, compares them with hypotheses and research expectations, and compares the results with previous research

Step 6: From the step of discussing the research results, the author proposes some policy implications, limitations and suggest future research directions

Review the theoretical frameworks and previous studies

Formulate hypotheses and build research models

Choose the appropriate model based on the results and check for defects in the model

Analyze and estimate the research results and compares the results with previous research

Propose some policy implications, limitations and suggest future reresearch directions.

RESEARCH MODEL

Through a brief review of previous studies, the studies by Phung Duc Nam et al (2018), Vo Xuan Vinh (2022), Seyedkkhosroshahi et al (2013), Faradisi & Ulpah

Building upon prior research (2021), this study investigates the factors influencing dividend policy decisions of listed manufacturing companies on the Ho Chi Minh City Stock Exchange (HOSE), employing a specific model to analyze these factors.

Dividend i,t is the dependent variable representing the enterprise's dividend payout policy Represented by the variables DVE and DVC Specifically, DVE is the ratio of cash dividends to profit after tax, DVCis the ratio of cash dividend payout to net operating cash flow

This analysis explores the relationship between the number of companies with listed shares, the year, and the manufacturing industry on a specific outcome (not explicitly mentioned) The model uses independent variables representing the number of companies (1 to 62), the year (1 to 10), and the industry (manufacturing) The regression coefficient (β) associated with each independent variable quantifies its impact on the outcome, while the error term (ε) accounts for any unexplained variation.

3.2.2 The explanation of the variables in the model and develop research hypotheses

In this study, the author inherits the way to measure dividend policy through measures of the ratio of cash dividends paid to net income (DVE), the ratio of cash dividends paid to cash flow Net business activities (DVC) are widely used in research

Phung Duc Nam (2018), Faradisi & Ulpah (2021), Jiang et al (2017) and Michaely

Firstly, the ratio of cash dividends to after-tax profits (DVE) The ressearch of Phung Duc Nam et al (2018) showed how much of the company's after-tax profit is used to pay dividends to shareholders A larger DVE represents the portion of profits generated during business operations that businesses prioritize paying dividends to shareholders for reinvestment activities DVE is defined as follows:

Secondly, the ratio of cash dividends to net operating cash flow (DVC) Jiang et al (2017) believed that net cash flow from operating activities represents a business's profit better than the profit on the financial statements Because net cash flow from business activities represents the actual cash received by the business, fully reflecting the ability to fulfill debt, tax obligations and so on The larger the DVC, represents the amount of cash the business uses Payout of dividends accounts for the majority of net operating cash flow DVC is defined as follows:

3.2.2.2 The explanation of independent variables and develop research hypotheses

In previous empirical studies, authors have proposed several measures of stock liquidity Typically, the liquidity measure is the difference between the bid price and the ask price (Amihud & Mendelson, 1986), the ratio of the absolute value of stock returns to trading volume of Amihud (2002) or the frequency of transactions with a return rate of 0 in the study of Lesmond et al (1999)

However, bid and ask prices are not available in emerging markets (Amihud,

2002), so researchers have proposed another measure to quantify stock liquidity

Among them, the measure of Equity Turnover (Turnover) stands out This measurement was widely used in the studies of Banerjee et al (2005), Michaely & Qian (2016), Kuo et al (2013), Seyedkkhosroshahi et al (2013), Phung Duc Nam et al (2018) and Vo Xuan Vinh (2022)

Equity turnover is the ratio between trading volume and the number of outstanding shares This measure represents the trading volume aspect of the definition of a stock's liquidity The larger the trading volume of a stock, the more attractive it is in the eyes of investors At the same time, the large trading volume will help investors easily match buy/sell orders because the supply and demand of that stock are always available in the market Therefore, the higher this ratio represents the higher liquidity of the stock In addition, simplicity in calculation is also the reason why the author chose this measure to include in the model Equity turnover is calculated as follows:

H1: Stock liquidity has a negative impact on the company's cash dividend payout policy

Firm size has become an important variable in previous research literature to explain a firm's dividend payout decision According to Al-Najjar & Hussainey

(2009), company size can be determined in two formula that are the logarithm of the total asset and the logarithm of the revenue In the research of Michaely & Qian

(2016), Phung Duc Nam et al (2018) and Faradisi & Ulpah (2021) the size of the company is proposed to be determined according to the formula is the logarithm of the revenue Therefore, in the study the author measures the company size variable according to the following formula:

Research by Al-Najjar & Hussainey (2009), Fama & French (2001) and Kuo et al (2013) argue that large companies have easier access to the capital market with lower capital costs than small companies Therefore, large companies are better able than small companies to distribute higher profits to shareholders Also based on the agency problem of Jensen & Meckling (1986) explain that dividends can play a role in reducing the agency problem The larger the company, the more difficult (costly) monitoring becomes Later studies by Michaely & Qian (2016), Kuo et al (2013), Faradisi & Ulpah (2021), Setiawan et al (2019) also found that company size has a positive impact on the company's dividend payout policy Therefore, the author proposes hypothesis H2 in the study as follows:

H2: Company size has a positive impact on the company's cash dividend payout policy

Regarding the profitability variable, which is considered an important indicator to evaluate a company's ability to pay dividends (Nizar Al‐Malkawi, 2007), in the study of Phung Duc Nam et al (2018) and Stereńczak & Kubiak (2022) propose measuring a company's profitability through the return on equity (ROE) index ROE is determined as follows:

Dividend payouts are often limited for companies with lower profitability, as implementing such policies can be costly (Thu et al., 2013; Fama & French, 2001) This is further supported by agency cost and free cash flow theory, where investors demand higher dividend payouts when profitability is high due to concerns about managerial misuse of free cash flow (Jensen & Meckling, 1986) Several studies, including those by Banerjee et al (2005), Michaely & Qian (2016), Kuo et al (2013), Nguyen Thi Ngoc Trang & Bui Kim Phuong (2017), and Phung Duc Nam et al (2018), have explored this relationship between profitability and dividend payout policy.

Bui (2019) and Stereńczak & Kubiak (2022) also showed similar results that profitability has a positive impact on dividend payout policy Therefore, the author proposes hypothesis H3 in the study as follows:

H3: A company's profitability has a positive impact on its cash dividend payout policy

Regarding the growth opportunity variable (Q), studies by Vo Xuan Vinh

(2022), Setiawan et al (2019) and Seyedkkhosroshahi et al (2013) included in the research model and determined according to the following formula:

Q = Market Capitalization Total Assets – Total Liabilities

Several studies, including Fama & French (2001), DeAngelo et al (2006), Nguyen Thi Ngoc Trang & Bui Kim Phuong (2017), and Setiawan et al (2019), have consistently found that companies with significant growth opportunities tend to have lower dividend payout ratios This phenomenon is attributed to the reinvestment needs of growing firms, which prioritize internal funding over dividend distribution to maximize future growth As a result, hypothesis H4 posits that companies with greater growth opportunities will be less likely to pay dividends.

H4: Growth opportunities have a negative impact on a company's cash dividend policy

Financial leverage (LEV) in the studies of Michaely & Qian (2016), Setiawan et al (2019), Faradisi & Ulpah (2021) and Vo Xuan Vinh (2022) are measured as follows:

Total assets According to Neves et al (2006) financial leverage refers to a firm's increase in external capital, specifically the extent to which it uses fixed income securities such as debt and preferred equity The more debt a company uses, the higher its financial leverage Companies with high debt ratios face financial risks and are therefore less likely to pay dividends Although using more debt helps companies increase profits through tax shields, they also have higher obligations to creditors To fulfil their obligations to creditors, companies must maintain a certain amount of cash, they tend to reduce or not pay dividends (Agrawal & Jayaraman, 1994)

RESEARCH METHODOLOGY

The thesis research methodology is a quantitative research method with research estimates made through Stata software version 17 In the quantitative research method, the author performs data descriptive statistics and calculations correlation coefficient matrix

The Pooled Ordinary Least Squares (Pooled OLS) regression model is the first econometric model used with panel data, analyzing the impact of independent variables on the dependent variable However, this model relies on stringent assumptions, including constant error variance, no autocorrelation, no multicollinearity, no missing variables, and a standard distribution, which are rarely met in practice The assumption of constant coefficients across time and individuals is particularly unrealistic.

The Pooled OLS model suffers from autocorrelation due to a low Durbin-Watson coefficient, prompting the use of the Fixed Effect Model (FEM) FEM accounts for individual observation variations over time, recognizing that specific characteristics influence explanatory variables By analyzing the correlation between residuals and explanatory variables, FEM isolates individual effects, allowing for accurate estimation of the explanatory variable's impact on the dependent variable.

However, when choosing to use between the FEM model and the REM model, it is necessary to have the impact of each observation on the explanatory variable over time If the variation of individual observations is not correlated with the explanatory variable, the Random Effect Model (REM) will be used instead of the FEM model

To decide whether to choose the FEM or REM model, the author will conduct the Hausman test given by Hausman in 1978 using the Hausman fem rem command in Stata software, resulting in a P-value < 0.05 choose the FEM model and vice versa, if P-value > 0.05 then choose the REM model

Descriptive statistical analysis method was used to describe the basic characteristics of the collected data sources and provide an overview of the research sample Descriptive statistical criteria include: mean value (Mean), minimum value (Minimum), maximum value (Maximum), standard deviation (Standard deviation), number of observations (observations) to see whether the data is outlier or not

The correlation coefficient matrix analysis method is used to examine the relationship between variables, thereby determining the degree of strong or weak, positive or negative correlation between variables in the model The author uses the method of analyzing the correlation coefficient matrix with the Pearson correlation coefficient, symbol "r" with value -1  r  1 The higher correlation coefficient, the more likely the variables are autocorrelated

In addition, correlation analysis also contributes to identifying the phenomenon of multicollinearity between variables in the model According to Gujarati & Porter

(2009), if the correlation coefficient of any pair of independent variables is higher than 0.8, the model may encounter multicollinearity

Gujarati & Porter (2009) use the variance inflation factor (VIF) to determine whether the model has multicollinearity or not Through testing, the author considers that if the correlation coefficient is close to 1, the larger the VIF will be, and then multicollinearity will occur In case VIF < 10, the variables in the model will not have much effect on the regression results and there will be no multicollinearity phenomenon

Formula to calculate VIF coefficient:

Pooled ordinary least squares (Pooled OLS)

The Pooled OLS model, as described by Brooks (2008), is the simplest approach to panel data analysis This model utilizes regular regression techniques, estimating parameters without considering spatial or temporal factors inherent in panel data.

While the Pooled OLS model offers simplicity and convenience by combining data without considering time or group differences, its lack of assumptions about variable distributions can lead to inaccurate estimations This model fails to account for distinct characteristics of samples across different periods, potentially distorting relationships between variables and introducing autocorrelation.

The FEM model is one of the most used models when analyzing panel data regression Accordingly, the FEM model applies the hypothesis that impact factors are always stable and do not change over time and space, while being consistent and controlling the influence of other factors on the environment result The FEM model determines the effects of factors by removing noise from fixed impact factors Because it takes into account the time series and cross-sectional elements, it overcomes the Pooled OLS model, so the FEM model has the advantage of estimating the influence of the independent variable on the dependent variable with high reliability through Panel data regression

The REM model assumes random variation between units, while the FEM model assumes that fluctuations are correlated with the independent variable This distinction stems from the way each model handles influencing factors: REM treats them as random variables, whereas FEM assumes non-random changes Therefore, if the differences between units impact the dependent variable, the REM model is more suitable than the FEM model.

Hausman test (1978) is used to find a suitable estimation method between FEM and REM Usually the significance level used is 5% or 10% With the hypothesis:

H0: The appropriate model is the REM model

H1: The suitable model is the FEM model

With the significance level α, if p-value < αthen reject the hypothesis H0 The appropriate model is the FEM model and vice versa If p-value > αthen there is not enough basis to reject the hypothesis H0, the appropriate model is the REM model

This phenomenon is understood as the phenomenon where the error variances of the estimated errors are not equal Perform the LM test in the study of Breusch and Pagan (1980) using the command syntax “xttest0” to check the heteroskedasticity phenomenon for the REM model, and use the command syntax “xttest3” to test Investigate the phenomenon of heteroskedasticity for the FEM fixed effects model

H0: the model has no heteroskedasticity

In statistical analysis, the p-value helps determine if a model's error variance is consistent If the p-value is less than the significance level (α), it indicates a rejection of the null hypothesis (H0) suggesting the model has variable error variance Conversely, if the p-value exceeds α, there's insufficient evidence to reject H0, implying the model does not exhibit variable error variance.

The autocorrelation testing method is a method of testing the correlation between variables in a series of observations in space or time Gujarati and Porter

(2011) put forward two hypotheses when testing the autocorrelation phenomenon as follows:

H0: The model does not have autocorrelation

H1: The model has autocorrelation phenomenon

RESEARCH DATA

This research utilizes secondary data from 62 manufacturing companies listed on the Ho Chi Minh Stock Exchange (HOSE), obtained from the FiinPro financial platform The dataset comprises 309 observations, resulting in an unbalanced panel due to missing data in certain companies.

The research focus on the group of manufaturing companies listed on the Ho Chi Minh city Stock Exchange because these companies have information which is published widely The research data is collected from 2013 to 2022 to measure the relationship of the dividend payout policy and independent variables, namely Liquidity, Profitability, Firm size, Growth opportunity, The level of cash holdings, Leverage

In chapter 3, the author indentidied a research process including 6 steps, namely determine the research objectives, formulate hypotheses and build research models, Collect data, Choose the appropriate model based on the results and check for defects in the model, Analyze and estimate the research results and compares the results with previous research, Propose suitable inplications In addition, the research hypotheses have six independent variables affecting the dividend payout policy, those are Liquidity, Profitability, Firm size, Growth opportunity, The level of cash holdings, Leverage

In this thesis, the author used the secondary data collected through financial statements and through Fiinpro's database The author obtained 309 observations at

62 manufacturing companies listed on HOSE during the period from 2012 to 2022

To address potential issues with heteroskedasticity or autocorrelation in OLS, FEM, or REM models, the author employed the FGLS regression model to assess the influence of various factors on dividend policy.

ANALYZING RESEARCH RESULTS

DESCRIPTIVE STATISTICAL ANALYSIS

The research focuses on learning about the cash dividend payout ratio of 62 manufacturing companies listed on the Ho Chi Minh City Stock Exchange (HOSE) in the period from 2013 to 2022 Because the calculation is complicated and inconsistent, making it very difficult to compare the dividend payout ratios of companies listed on HOSE

Table 4.1 Descriptive statistics of research variables

Variables Observations Average Standard error Min Max

Source: The author estimated from Stata 17

From the statistical results in table 4.1, it shows that the trend of paying cash dividends over the years in manufacturing enterprises is very popular On average, each business spends about 66.92% of net profit after tax In addition, for net operating cash flow, businesses tend to spend about 72.61% to pay dividends Through the maximum and minimum values of DVE and DVC, we can see that there exist businesses with dividend payout s that exceed profit after tax and net operating cash flow in the year when the value is large The enterprise with a dividend to profit after tax ratio of 440% is Mekong Fisheries Joint Stock Company, because in 2017 the enterprise's after-tax profit only reached more than 1.13 billion VND, however the operating cash flow was quite high at more than 57.14 billion VND The enterprise with a dividend ratio on net operating cash flow of 649.03% is Rang Dong Thermos Bulb Joint Stock Company because the company's net operating cash flow in 2020 only reached more than 8.18 billion VND, however, profit after tax for the year was quite high at more than 336.08 billion VND However, the fluctuation amplitude between the minimum value and maximum value of the variables DVE and DVC has a large difference, so the dividend payout policies of businesses are not consistent with each other

Regarding stock liquidity, statistical results show that stock liquidity of manufacturing enterprises is on average 0.24%, showing that only about 0.24% of stocks are traded each day, compared to the volume of outstanding shares At the same time, the stock liquidity of businesses is highly differentiated, meaning there is divergence and difference In the year of the business, this is reflected in the difference between the maximum value and the value smallest of the variable TURN Stocks with high liquidity often attract a certain amount of attention from investors, which also explains the differentiation of daily trading volume of stocks

The average return on equity of manufacturing enterprises listed on HOSE is quite good at 17.42% The highest ROE ratio at 76.31% is Hoa Sen group, and the lowest at 0.47% is Hydro Power Joint Stock Company – Power No.3 This shows that the ROE of businesses has a relatively large difference in profitability between companies, demonstrate the diversity and difference in the market's perception of businesses' stocks

For the growth opportunity variable (Q), the average level is 1.4873, the smallest value at 0.4623 is Hapaco Corporation, the maximum value at 5.0421 is Hydro Power Joint Stock Company – Power No.3, showing that growth opportunities among businesses are quite uneven

The average leverage ratio (LEV) for businesses, where debt represents 41.59% of total assets, is considered relatively safe Southern Seed Corporation boasts the lowest LEV at 0.69%, while Sai Gon Machinery Spare Parts JSC holds the highest at 79.86%.

Regarding the level of cash holdings (CASH), on average, businesses tend to hold cash quite low at 15.47% of total assets, the highest ratio reaches 58.99% is is Hydro Power Joint Stock Company – Power No.3 and the lowest ratio at 0.01% is Sao Ta Foods Joint Stock Company The remaining scale variable (SIZE) has an average value of 7.3331, the largest and smallest values have large differences, 4.0549 and 12.6249 respectively, which the companies are Sao Ta Foods Joint Stock Company and Viet Nam National Petroleum Group so the current scale between businesses does not exist, equally with each other

After performing descriptive statistics, it can be seen that currently there is no extraneous multicollinearity in the model To check the degree of correlation of variables, the thesis used a correlation coefficient matrix for independent and dependent variables

Table 4.2 Correlation coefficients matrix of the variables

DVE DVC TURN ROE Q LEV CASH SIZE

Source: The author estimated from Stata 17

The correlation coefficient between factors in the model was less than 0.8, indicating a correlation between variables and allowing for their inclusion in the model The model's absence of multicollinearity was further confirmed through the VIF coefficient, ensuring its reliability.

Based on the regression results in Table 4.2 of the correlation coefficient matrix between variables in the model, we see that most of the correlation coefficients of pairs of independent variables in the model are less than 0%, so the independent variables in the model have low autocorrelation with each other Therefore, the majority of variables in the model do not have multicollinearity with each other and this will be a positive sign in testing and choosing an appropriate econometric model

Through the results of the table 4.2, the author used a method to detect multicollinearity is to calculate the variance inflation factor (VIF) for each independent variable, and a VIF value greater than 1.5 indicates multicollinearity

Source: The author estimated from Stata 17

The author checkes the multicollinearity phenomenon using the method of magnifying the variance factor VIF (Variance Inflation Factor) The average VIF value is 1.31 and the VIF coefficients of the independent variables in the model are all less than 10, concluding that the model has good multicollinearity, meaning the multicollinearity phenomenon is not serious between the independent variables in the research model

Testing for constant error variance gives a P-value of approximately 0% (less than 5%), rejecting the hypothesis H0: The model has constant error variance The model suffers from heteroskedasticity error The model does not have serious multicollinearity.

ESTIMATED RESULTS

After tetsing the multicollinearity, the author estimates the model using three methods PoolOLS, FEM, and REM The results are shown in table 4.4

Table 4.4 Regression results by Pooled OLS, FEM, REM

Dependent variable: DVE Regression models

Independent variables Pooled OLS FEM REM

Note: *, **, *** corresponding to the 10%, 5%, 1% significance levels

Source: The author estimated from Stata 17

Statistical analysis reveals a linear association between certain independent variables and the dependent variable The p-value across Pooled OLS, FEM, and REM models consistently falls below 0.01, indicating a statistically significant relationship and validating the appropriateness of the regression models.

According to the results of the Pooled OLS model, the adjusted variance R 2 (R-squared) is 0.1769, meaning the independent variables mentioned above can explain 17.69% of the variation in the dependent variable cash dividend payout ratio on earnings after tax (DVE) According to the results of the FEM model, the adjusted variance R 2 (R-squared) is 0.1577, meaning the independent variables mentioned above can explain 15.77% of the variation in the dependent variable cash dividend payout ratio on operating cash flow (DVC)

According to the results of the REM model, this model concludes the statistical significance of the three variables ROE, Q and SIZE with regression coefficients of -1.9056, 0.1229, -0.0515 respectively with a significance level of 1%, 5%, 10% However, all three models still cannot conclude the impact of the two variables TURN and LEV because they are not statistically significant

Table 4.5 The results of FEM, REM, Pool OLS models for the variable DVC

Dependent variable: DVC Regression model

Independent variables Pooled OLS FEM REM

Note: *, **, *** corresponding to the 10%, 5%, 1% significance levels

Source: The author estimated from Stata 17

According to the results of the Pooled OLS model, the adjusted variance R 2 (R- squared) is 0.0086, meaning the independent variables mentioned above can explain 0.86% of the variation in the dependent variable cash dividend payout ratio on operating cash flow (DVC) According to the results of the FEM model, the adjusted variance R 2 (R-squared) is 0.0455, meaning the independent variables mentioned above can explain 4.55% of the variation in the dependent variable cash dividend payout ratio on operating cash flow (DVC) According to the results of the REM model, the adjusted variance R 2 (R-squared) is 0.0340, meaning the independent variables mentioned above can explain 3.40% of the variation in the dependent variable cash dividend payout ratio on operating cash flow (DVC)

Choosing the suitable model between FEM and REM models

Compare two models FEM and REM through the Hausman test with the hypothesis

Table 4.6 The Hausman test for variable DVE and DVC

H 0 : The appropriate model is the REM model

H 1 : The suitable model is the FEM model

Chi 2 (6) = 1.98 Chi 2 (6) = 9.53 prob > Chi 2 = 0.9219 prob > Chi 2 = 0.1458

Source: Estimate the results from Stata Platform 17

The Hausman test for the DVE variable, with a 5% significance level, yielded a p-value of 0.9219, exceeding the significance threshold This result supports the acceptance of the null hypothesis (H0), indicating that the REM model is the appropriate choice for the research model of DVE dependent boundaries.

For the variable DVC, with a 5% significance level, the Hausman test showed us the p-value is 0.1458, greater than 0.05 (greater than the 5% significance level) From there, there is enough basis to accept the hypothesis H0, or in other words, the model suitable for the research model of the dependent variable DVC is the REM model

Choosing the suitable model between Pooled OLS and REM models

To determine the appropriate model between Pooled OLS and a model accounting for heteroscedasticity, the Breusch-Pagan test is used to compare the models, and the `het test` command in Stata can identify heteroscedasticity in the Pooled OLS model.

Table 4.7 For the variable DVE and DVC

Chi 2 (1) = 68.10 Chi 2 (1) = 0.14 prob > Chi 2 = 0.0000 prob > Chi 2 = 0.7069

Source: Estimate the results from Stata Platform 17

Through testing for the variable DVE, the results showed that the p-value of 0.000 is less than 0.05 (less than the 5% significance level), indicating that there is heteroskedasticity in the Pooled OLS model From there, there is enough basis to reject the hypothesis H0, or in other words, the Pooled OLS model is not suitable for the model of the dependent variable DVE Therefore, the author chose the REM model as concluded above as the appropriate research model for the dependent variable DVE

Through testing for the variable DVC, the results showed that the p-value is 0.7069, greater than 0.05 (greater than the 5% significance level), indicating that there is no heteroskedasticity phenomenon in the Pooled OLS model From there, there is enough basis to accept the hypothesis H0, or in other words, the Pooled OLS model is consistent with the model of the dependent variable DVC Therefore, the author chose the Pooled OLS model as the appropriate research model for the dependent variable DVC.

CHECK FOR DEFECTS OF MODELS

Testing the heteroskedasticity of the REM model

Table 4.8 Testing the heteroskedasticity of the REM model

Testing the heteroskedasticity of the REM model

H0: The model has no heteroskedasticity

H1: The model has heteroskedasticity Chibar 2 (01) = 18.53 prob > Chibar 2 = 0.0000

Source: Estimate the results from Stata Platform 17

According to Breusch and Pagan (1980) in testing heteroskedasticity, use the command "xttest0" to perform the test with the following pair of hypotheses:

H0: The model has no heteroskedasticity

With a significance level of 5%, the LM-Breush and pagan Lagrangian Multiplier test shows that the p-value is 0.0000 (which is less than the 5% significance level) From there, there is enough basis to reject the hypothesis H0, or in other words, the REM model has the phenomenon of heteroskedasticity

Autocorrelation phenomenon can affect the effectiveness of the model, reducing the reliability of testing coefficients To test autocorrelation for the REM model, use the Wooldridge test with the hypothesis:

H0: The model does not have autocorrelation

H1: The model has autocorrelation phenomenon

Table 4.9 For the variable DVE and DVC

H0: The model does not have autocorrelation

H1: The model has autocorrelation phenomenon

Source: Estimate the results from Stata Platform 17

Use the Wooldridge test, with the following syntax: xtserial y x1,x2,… (y is the dependent variable, x1, x2,… are the independent variables)

With the hypothesis H0: no first-order autocorrelation

The Wooldridge test for autocorrelation on the DVE variable, using a 5% significance level, revealed a Prob value greater than 0.0852 This result provides sufficient evidence to accept the null hypothesis (H0), indicating that the REM model for DVE does not exhibit autocorrelation.

For the variable DVC, with a 5% significance level, the Wooldridge test shows that the Prob value is >F = 0.0076 (less than the 5% significance level) From there, there is enough basis to reject the hypothesis H0, or in other words, the REM model of the dependent variable DVC has autocorrelation, but the autocorrelation is not serious

Resolving defects using the FGLS model

Through tests, it has been shown that the REM model has defects that need to be overcome The author points out that these are heteroskedasticity and autocorrelation phenomena However, the autocorrelation is not serious Therefore, the author uses Feasible Generalized Least Squares (FGLS) to overcome the phenomenon of heteroskedasticity in the models

Table 4.10 Feasible Generalized Least Squares (FGLS)

Dependent variable Feasible Generalized Least Squares (FGLS)

Note: *, **, *** corresponding to the 10%, 5%, 1% significance levels

Source: The author estimated from Stata 17

Table 4.11 Feasible Generalized Least Squares (FGLS) Feasible Generalized Least Squares – FGLS

Source: Estimate the results from Stata Platform 17

Table 4.11 showed that The Feasible Generalized Least Squares – FGLS has P value = 0,000 < 𝛼 = 0,01 Thus, the model is suitable for the study with significant statistical significance at the 1% level

Analysis of Table 4.10 reveals a single factor, Q, positively impacting cash dividend payout policy in both DVE and DVC Conversely, four variables - TURN, LEV, CASH, and SIZE - demonstrate a negative influence on cash dividend policy across both DVE and DVC.

DISCUSS THE RESEARCH RESULT

Based on the results of estimating the FGLS model on the impact of the factors on dividend policy, the study has some discussions as follows:

Our analysis using the FGLS model revealed no significant correlation between stock liquidity, measured by turnover, and the independent variables DVE and DVC This lack of relationship suggests that stock liquidity, as measured by turnover, has no discernible impact on a company's cash dividend payout policy Consequently, we reject hypothesis H1, which posited a negative relationship between stock liquidity and cash dividend payout policy.

Company size negatively impacts dividend distribution, with a statistically significant effect at the 1% level on dividend-to-earnings ratio (DVE) but no significant impact on dividend-to-cash flow ratio (DVC) according to FGLS analysis This finding aligns with previous research (Michaely & Qian, 2016; Kuo et al., 2013; Faradisi & Ulpah, 2021; Setiawan et al., 2019), suggesting that larger companies, despite their increased ability to pay dividends, tend to distribute less of their earnings or cash flow.

Large companies, with their operational efficiency, typically have easier access to capital compared to smaller businesses However, they also face higher agency costs due to dispersed ownership and limited shareholder oversight To mitigate these costs and capitalize on their low capital costs, large companies often implement policies of paying cash dividends This supports the hypothesis that company size positively influences cash dividend payout policy.

For the variable ROE, the results from FGLS showed that ROE has a positive impact on the variable DVC but there is no statistical significance However, the variable ROE has a negative impact on the variable DVE with coefficient at 1% significance level It can be concluded that the variable ROE has a negative impact on the cash dividend policy This result is contrary to the initial expectations of the thesis Regarding the company's cash dividend payout policy This result is completely consistent with previous studies by Banerjee et al (2005), Michaely & Qian (2016), Kuo et al (2013), Nguyen Thi Ngoc Trang & Bui Kim Phuong (2017), Duc Nam et al (2018), Bui (2019) and Stereńczak & Kubiak (2022) This means that the more profitable a business is, the more tendency it has a high cash dividend policy Because a business with a higher profit rate shows stability and efficiency in its business operations For such businesses, it will be easier for them to access loans at cheaper costs thanks to the certainty of future income streams, which will also create conditions for managers implementation of dividend policy Therefore, the author accepts hypothesis H3: A company's profitability has a positive impact on its cash dividend payout policy

For the variable Growth opportunities, the results from FGLS showed that the variable Growth opportunities has a positive impact on the variable DVC with the coefficient at 5% significance The results from the FGLS showed that the variable

Growth opportunities has a positve impact on the variable DVE with coefficient at 1% significance level It can be explained by the fact that the stronger a company is in its growth opportunities, the higher it has dividend payout This result is not consistent with the research of Fama & French (2001), DeAngelo et al (2006) and Setiawan et al (2019), the results in these research showed that a business in a period of strong growth will focus on resources such as profits after tax to reinvest and participate in projects to maximize the value of the business Therefore, the higher growth opportunities companies have, the lower the probability of paying cash dividends

Our findings on the positive relationship between Q and cash dividend policy align with prior research by Faradisi & Ulpah (2021) and Vo Xuan Vinh (2022) While growth opportunities are often associated with increased optimism and potential for future profitability, our FGLS analysis did not reveal a significant relationship between variable Q and cash dividend payout policy This lack of clear evidence does not allow us to either accept or reject the hypothesis that growth opportunities negatively impact cash dividend policy.

For the variable financial leverage, the results from FGLS showed that the financial leverage has a negative impact on the variable DVC without the statistical significance The results from the FGLS showed that the financial leverage variable (LEV) has a negative impact on the variable DVE with the coefficient at the 1% significance level It can be concluded that the variable LEV has a negative impact on the cash dividend policy This result is consistent with the research results of Jensen et al (1992); Fama & French (2001); Zhang, Uchida, Dong (2020) and met initial expectations The higher the debt/capital ratio, the lower the dividend payout ratio It is clear that managers need to use net profits to pay debt The larger the debt, the higher the priority for debt repayout Therefore, shareholders will receive lower dividends This result is completely consistent with the general situation of the companies in general This result is consistent with the initial expectations of the thesis Research results by Zhong (2016), Khan and Ahmad (2017), Edmund (2018),

Le et al (2019) showed similar results into the bargain This is completely consistent with the research hypothesis as well as the fact that with a high debt ratio, company management tends to convince shareholders to trade off a high dividend ratio to prioritize debt repayout obligations before investing Divide profits to owners with the view that shareholders are the "final claimants" to the assets and cash flow of the business High financial costs associated with high financial leverage are the reasons for the company to choose to cut dividends because earnings after interest and taxes (EBIT) decrease when the debt ratio increases, affecting the source of distribution dividends as well as management's plan to reduce dividends to increase reinvestment as well as limit dependence on debt activities, which have many financial risks Therefore, the thesis accepts hypothesis H5: Financial leverage has a negative impact on a company's cash dividend policy

The analysis using the FGLS method revealed that while cash holdings have a statistically significant negative impact on dividend payout (DVE) at the 1% level, their influence on debt-to-value ratio (DVC) is not statistically significant.

It can be concluded that the variable CASH has a negative impact on the cash dividend policy This result is contrary to previous experimental studies such as Phung Duc Nam et al (2018), Nguyen Thi Ngoc Trang & Bui Kim Phuong (2017), Faradisi & Ulpah (2021), Ozkan & Ozkan (2004), Bui (2019) and Setiawan et al

(2019) They believe that the board of directors of a business tends to make decisions to pay higher dividends when the business has a high ratio of cash to total assets It can be explained by the fact that the larger the level of cash holdings a company has, the more stable the company's liquidity is In addition, Jensen & Meckling (1986) mentioned in agency theory that when there is a large amount of excess cash, managers will not have the need to seek additional capital from outside sources This will give them the freedom to make investment decisions in potentially unproductive projects without third-party oversight Therefore, shareholders tend to think that managers are more likely to waste surplus cash In this case, companies that decide to pay dividends to investors would minimize excess cash, which is considered as a mechanism to help reduce agency costs in companies Therefore, the greater the level of cash holdings, the greater the incentive for managers to implement a cash dividend policy for existing shareholders However, the author decided to reject hypothesis H6: The level of cash holdings has a positive impact on the company's cash dividend payout policy

Table 4.12 Summary the research result

Q - + No basis to accept or reject

Source: Estimate the results from Stata Platform 17

To measure the impact of six independent variables on cash dividend payout policy, the author conducted regressions on the POOLED OLS, FEM, REM models After performing the Hausman test, the author found that the REM model is suitable for the research However, in the REM model, there exists the phenomenon of heteroscedasticity and autocorrelation, the author overcomes the defects with the FGLS regression model The result showed that stock liquidity measured through turnover has no relationship with the independent variable DVE There are four variables, namely return on common equity, leverage, the level of cash holdings, firm size have a negative impact on cash dividend payout policy There is the variable growth opportunities that has the positive impact on cash dividend payout policy.

CONCLUSION AND RECOMMENDATION

CONCLUSION

The research data was conducted based on the goal of examining the influence of factors on the dividend payout policy of 62 manufacturing enterprises listed on the

Ho Chi Minh City Stock Exchange (HOSE) during the period from 2013-2022 In the research model, the dependent variable is the cash dividend payout policy measured by two variables: (1) ratio of cash dividends to profit after tax (DVE), (2) ratio dividend ratio on operating net cash flow (DVC), the main independent variable is the stock liquidity variable measured by the measure of Equity Turnover (TURN), profitability (ROE), business size (SIZE), growth opportunity (Q), financial leverage (LEV) and cash holding level (CASH)

This study investigated the impact of various factors on the dividend payout policies of manufacturing companies using FEM, REM, and Pooled OLS models The final model, REM, revealed that firm size, profitability, and growth opportunity have a statistically significant positive impact on cash dividends Conversely, financial leverage and cash flow have a negative impact on dividend payout policies However, the research found insufficient evidence to conclude the impact of stock liquidity on dividend payout policies.

RECOMMENDATION

The research has certain significance in the field of finance, providing more empirical evidence for the relationship between factors and dividend payout policies of manufacturing companies in the Ho Chi Minh stock market (HOSE) Based on the results of the research and conclusions that have been done, there are several suggestions in this study to support corporate managers increase the payout dividends, in detail:

Profitability in this thesis has a negative impact on cash dividend payout policy However, it is also an important factor that determines each company's dividend payout policy When a business has a profit that exceeds its plan, instead of retaining it for reinvestment, the business can deduct a small portion of the profit for shareholders If shareholders expect to receive dividends and consider this to be an important part of their investment strategy, then it can increase their satisfaction and maintain trust, thereby making them more willing to pay dividends, especially when they are in difficult times Depending on the current business situation, each business can make a decision to pay dividends that is reasonable and harmonious with other decisions such as investment decisions and financing decisions

About the firm size, the larger the company is, the more likely it is to pay dividends The larger a company is, the higher its agency costs are, so implementing a dividend payout policy at large companies is necessary Jensen & Meckling (1986) also mentioned dividends as a mechanism to help companies minimize agency costs

In addition, based on the advantage of low capital costs, large companies could pay dividends to shareholders to demonstrate the company's responsibility to shareholders

Strong growth prospects influence a company's dividend payout policy Companies with high growth potential are more likely to pay dividends, signaling confidence in their future success This attracts investors, boosting share prices and increasing company value.

Moreover, financial leverage is a factor that negatively impacts dividend payout policy For companies with a higher debt-to-total asset ratio, they should maintain a low dividend payout ratio to prioritize fulfilling obligations to creditors to reduce the financial burden on the company as well as reduce risks for shareholders.

Dividend decisions are influenced by a company's cash holdings Firms with ample cash should prioritize dividend payments to prevent potential agency costs, as shareholders may perceive excess cash as a risk for inefficient investments Conversely, companies with low cash reserves should maintain a conservative dividend policy to mitigate financial risk.

RESEARCH LIMITED AND FUTURE RESEARCH DIRECTIONS

Although the research has achieved the research objectives, there are several limitations in the thesis:

This thesis just mentioned the sole payout dividend method that is cash dividend payout That is the reason why in the future, the author would do the research in other methods such as stock dividend or dividend paid by asset

Another worth mentioning is that in this thesis, endogeneity might appear Therefore, in the next research direction, the author will use GMM estimation method to overcome this phenomenon

Based on the results in chapter 4, the author makes conclusions and assessments about the influence of factors on dividend policy at manufacturing companies listed on hose Thereby, the author proposes a number of policy implications At the same time, chapter 5 presents the limitations of the research and proposes further research directions for the topic.

1 Bùi, P K (2019) Tính thanh khoản của cổ phiếu và chính sách cổ tức Tạp chí Kinh tế & Phát triển (Vol 264, pp 42-49)

2 Đinh, B N., & Nguyễn, C C (2014) Các nhân tố ảnh hưởng đến chính sách cổ tức của các doanh nghiệp niêm yết trên thị trường chứng khoán Việt Nam In Phát triển kinh tế (Vol 290, pp 42–60)

3 Đức Nam, P., Thị Phương Anh, H., & Thị Thanh Vương, L (2018) Ảnh hưởng của khả năng thanh khoản cổ phiếu đến chính sách chi trả cổ tức: bằng chứng thực nghiệm từ các doanh nghiệp niêm yết trên sở giao dịch chứng khoán TP Hồ Chí Minh (HOSE) Tạp chí Khoa học Đại học Đà Lạt, In Số 1S (Vol 8)

4 Lê Minh, Đ (2004) Chính sách cổ tức và tác động của nó tới công ty - Những gợi ý cho Việt Nam NXB Chính trị Quốc gia

5 Ngọc Phi Anh, Đ., & Thị Hoài Thương, M (2015) Nghiên cứu các nhân tố ảnh hưởng đến tính minh bạch thông tin của các công ty cổ phần niêm yết trên sở giao dịch chứng khoán Hà Nội

6 Thị Hải Lý, T., & Thị Bảy, Đ (2015) Cấu trúc sở hữu và chính sách cổ tức của các doanh nghiệp niêm yết: Bằng chứng tại Việt Nam Tạp chí Phát triển & Hội nhập (Vol 23(33), pp 72-78)

7 Thi Lien Hoa, N & Thi Bich Phuong, B (2014) Nghiên cứu các nhân tố tác động đến đầu tư trực tiếp nước ngoài tại những quốc gia đang phát triển Tạp chí Phát triển & Hội nhập (Vol 14, pp 40–46)

8 Trang, N., & Phương, B (2017) Nghiên cứu các yếu tố quyết định đến chính sách chi trả cổ tức bằng kĩ thuật hồi quy Fama -MacBeth: Bằng chứng thực nghiệm tại Việt Nam Tạp Chí Phát Triển Kinh Tế, 04–24

9 Việt Phú, L (2018) Phương sai thay đổi và tự tương quan Trường Chính sách Công và Quản lý Fulbright

10 Vũ Mai Ly, N (2016) Các nhân tố ảnh hưởng đến chính sách chi trả cổ tức bằng tiền mặt tại Việt Nam Tạp Chí Khoa Học và Công Nghệ (Vol 2, pp 180-186)

11 Abdullah, N., Ahmad Z., & Roslan S (2012) The influence of ownership structure on the firms dividend policy based Lintner model International Review of business research Papers, 8(6), 71–88

12 Agrawal, A., & Jayaraman, N (1994) The Dividend Policies of All-Equity Firms:

A Direct Test of the Free Cash Flow Theory Managerial and Decision Economics,

15(2), 139–148

13 Al-Najjar, B & Hussainey, K (2009) The association between dividend payout and outside directorships Journal of Applied Accounting Research, 10(1), 4–19

14 Al-Najjar B., & Kilincarslan E (2016) The effect of ownership structure on dividend policy: evidence from Turkey Corporate Governance Int J Bus Soc, 16(1), 135-161

15 Alzomaia, T., & Al-Khadhiri, A (2013) Determination of dividend policy: the evidence from Saudi Arabia Int J Bus Soc Sci, 4(1), 181–192

16 Amidu, M., & Abor, J (2006) Determinants of dividend payout ratios in Ghana

The Journal of Risk Finance, Vol.7(2), 136-145

17 Amihud, Y (2002) Illiquidity and stock returns: Cross-section and time-series effects Journal of Financial Markets, 5(1), 31–56

18 Amihud, Y., & Mendelson, H (1986) Asset pricing and the bid-ask spread

19 Amihud, Y., Mendelson, H., & Lauterbach, B (1997) Market microstructure and securities values: Evidence from the Tel Aviv Stock Exchange Journal of Financial

20 Banerjee, S., Gatchev, V A., & Spindt, P A (2005) Stock Market Liquidity and

Firm Dividend Policy SSRN Electronic Journal

21 Brav, A., Graham, J R., Harvey, C R., & Michaely, R (2005) Payout policy in the 21st century Journal of Financial Economics, 77(3), 483–527

22 Brigham, E F., & Houston, J F (2009) Fundamentals of financial management

23 Bushra, A & Mirza, N (2015) The determinants of corporate dividend policy in

Pakistan Lahore J Econ, 20(2),77-98 https://doi.org/10.35536/lje.2015.v20.i2.a4

24 DeAngelo, H., DeAngelo, L., & Stulz, R M (2006) Dividend policy and the earned/contributed capital mix: a test of the life-cycle theory Journal of Financial Economics, 81(2), 227–254

25 Dempsey, M., Gunasekarage, A & Truong, TT (2019) The association between dividend payout and firm growth: Australian evidence Account Finance, 59(4),

26 Dong, M., Robinson, C., & Veld, C (2005) Why individual investors want dividends Journal of Corporate Finance, 12(1), 121–158

27 Fama, E F., & French, K R (2001) Disappearing dividends: changing firm characteristics or lower propensity to pay? Journal of Financial Economics, 60(1), 3–43

28 Fajaria, A., & Isnalita, N (2018) The effect of profitability, liquidity, leverage and firm growth of firm value with its dividend policy as a moderating variable Int

29 Faradisi, A., & Ulpah, M (2021) Informational Effect of Stock Liquidity on Dividend Payouts: Evidence from Indonesia, 2008-2017 In Proceedings of the Asia-

Pacific Research in Social Sciences and Humanities Universitas Indonesia Conference (APRISH 2019) (Vol 558)

30 Gill, A., Biger, N., & Tibrewala, R (2010) Determinants of dividend payout ratios: evidence from United States The Open Business [journal], 3(1)

31 Gordon, M.J (1963) Optimal Investment and Financing Policy The Journal of

Finance, 18, 264-272 Bhattacharya S (1979) Imperfect information dividend policy and “the bird in the hand” fallacy The Ball Journal of Economic, Vol 10, No

32 Hooi, S E., Albaity, M., & Ibrahimy, A I (2015) Dividend policy and share price volatility Investment Management and Financial Innovations, 12(1), 226–234

33 Jiang, F., Ma, Y., & Shi, B (2017) Stock liquidity and dividend payouts Journal of Corporate Finance, 42, 295–314

35 Khan, A., Yilmaz, M K., & Aksoy, M (2022) Does board demographic diversity affect the dividend payout policy in Turkey? EuroMed Journal of Business, July

36 Kuo, J M., Philip, D., & Zhang, Q drives the disappearing dividends phenomenon? (2013) What drives the disappearing dividends phenomenon? Journal of Banking & Finance, 37(9), 3499–3514

37 Kuzucu, N (2015) Determinants of Dividend Policy: A Panel Data Analysis for

Turkish Listed Firms International Journal of Business and Management, 10(11),

38 Lewellen, W G., Stanley, K L., Lease, R C., & Schlarbaum, G G (1987) Some

Direct Evidence On The Dividend Clientele Phonomenon, The Journal of Finance, vol 33, no 5, pp 1385-1399

39 Lesmond, D A., Ogden, J P., & Trzcinka, C A (1999) A New Estimate of

Transaction Costs Review of Financial Studies, 12, 1113-1141

40 Lintner, J (1956) Distribution of Incomes of Corporations among Dividends,

Retained Earnings, and Taxes The American Economic Review, 2, 97-113

41 Michaely, R., & Qian, M (2016) Stock Liquidity and Dividend Policy: Evidence from a Natural Experiment

42 Michaely, R., & Qian, M (2022) Does stock market liquidity affect dividends?

Pacific Basin Finance Journal, 74(January), 1–45

43 Miller, M H., & Modigliani, F (1961) Dividend policy, growth, and the valuation of shares The Journal of Business, 34(4), 411-433

44 Miller, J (1980) Decision-Making and Organizational Effectiveness: Participation and Perceptions Sociology of Work and Occupations, 7(1), 55-79

45 Miller, M H., & Rock, K (1985) Dividend Policy under Asymmetric

Information The Journal of Finance, 40(4), 1031–1051

46 Neves, E., Pindado, J., & De La Torre, C (2006) Dividends: New Evidence on

The Catering Theory Documento de Trabajo 14/06, January, 1–31

47 Nizar Al‐Malkawi, H (2007) Determinants of Corporate Dividend Policy in

Jordan: An Application of the Tobit Model Journal of Economic and Administrative

48 Opler, T., Pinkowitz, L., Stulz, R., & Williamson, R (1999) The determinants and implications of corporate cash holdings Journal of Financial Economics, 52(1), 3–46

49 Omerhodzic, S 2014 Identification and Evaluation of Factors of Dividend Policy

Institute of Economic Sciences, vol 47(1-2), 42–58

50 Ozkan, A., & Ozkan, N (2004) Corporate cash holdings: An empirical investigation of UK companies Journal of Banking & Finance, 28(9), 2103–2134

51 Setiawan, D., Bandi, B., Kee Phua, L K., & Trinugroho, I (2016) Ownership structure and dividend policy in Indonesia J Asia Bus Stud, 10(3), 230–252

52 Setiawan, R., Christiana, N., & Singh, S K (2019) Stock Liquidity, Financial

Report Quality, Wedge, and The Propensity to Pay Dividend In Jurnal Manajemen

Teori dan Terapan | Journal of Theory and Applied Management (Vol 12, Issue 3)

53 Seyedkkhosroshahi, S., Sabaei, A., & Vatankhah, P (2013) Stock Market

Liquidity, Firm Characteristics and Dividend Payout In Life Science Journal (Vol

54 Stereńczak, S., & Kubiak, J (2022) Dividend policy and stock liquidity: Lessons from Central and Eastern Europe Research in International Business and Finance,

55 Vo, X V (2022) Can liquidity explain dividends? Cogent Business and Management, 9(1)

56 Yulivan, I (2022) Economics and Business Quarterly Reviews 5(1), 13–20

57 Zhang, Y., Uchida, K., & Dong, L (2020) External financing and earnings management: evidence from international data Res Int Bus Fin, 54, 1012-1075