#### CHAPTER ONE

#### INTRODUCTION

*Background:*

*Background:*

Dividend policy is an important component of the corporate financial management policy. It is a policy used by the firm to decide as to how much cash it should reinvest in its business through expansion or share repurchases and how much to pay out to its shareholders in dividends. Dividend is a payment or return made by the firm to the shareholders, (owners of the company) out of its earnings in the form of cash. For a long time, the subject of corporate dividend policy has captivated the interests of many academicians and researchers, resulting in the emergence of a number of theoretical explanations for dividend policy. For the investors, dividend serve as an important indicator of the strength and future prosperity of the business, thereby companies try to maintain a stable dividend because if they reduce their dividend payments, investors may suspect that the company is facing a cash flow problem. Investors prefer steady growth of dividends every year and are reluctant to investment to companies with fluctuating dividend policy. Over time, there has been a substantial increase in the number of factors identified in the literature as being important to be considered in making dividend decisions. Thus, extensive studies have been done to find out various factors affecting dividend payout ratio of a firm. However, there is no single explanation that can capture the puzzling reality of corporate dividend behavior. Ocean deep judgment is involved by decision makers to resolve this issue of dividend behavior. The decision of companies to retain or pay out the earnings in form of dividends is important for the maximization of the value of the firm (Oyejide, 1976). Therefore, companies should set a constructive target dividend payout ratio, where it pays dividends to its shareholders and at the same time maintains sufficient retained earnings as to avoid having raise funds by borrowing money.

A tough challenge was faced by financial practitioners and many academics, when Miller and Modigliani (M&M) (1961) came with a proposition that, given perfect capital markets, the dividend decision does not affect the firm value and is, therefore, irrelevant. This proposition was greeted with surprise because at that time it was universally acknowledged by both theorists and corporate managers that the firm can enhance its business value by providing for a more generous dividend policy and that a properly managed dividend policy had an impact on share prices and shareholder wealth. Since the M& M study, many researchers have relaxed the assumption of perfect capital markets and stated theories about how managers should formulate dividend policy decisions.

*Problem Statement:*

*Problem Statement:*

Dividend policy has attracted a substantial amount of research by many researchers and theorists, who have provided theoretical as well as empirical observations, into the dividend puzzle (Black, 1976). Even though researchers and theorists have extended their studies in context to dividend decisions, the issue as to why corporations distribute a portion of their earnings as dividends is not yet resolved. The issue of dividend policy has stimulated much debate among financial analysts since Lintner's (1956) seminal work. He measured major changes in earnings as the key determinant of the companies' dividend decisions. There are many factors that affect dividend decisions of a firm as it is very difficult to lay down an optimum dividend policy which would maximize the long-run wealth of the shareholders resulting into increase or decrease of the firm's value, but the primary indicator of the firm's capacity to pay dividends has been Profits.

Miller and Modigliani (1961), DeAngelo and DeAngelo (2006) gave their proposition on the dividend irrelevance, but the argument made by them was on assumptions that weren't practical and in fact, the dividend payout decision does affect the shareholders value.

The study focuses on identifying various determinants of dividend payout and whether these factors influence the dividend payout decision.

*Research Objective:*

*Research Objective:*

There are many theories in the corporate finance literature addressing the dividend issue. The purpose of study is to understand the factors influencing the dividend decision of companies. The specific objectives of this study are:

- To analyze the financials of the company, to draw a framework of factors such as Retained earnings, Age of the company, Debt to Equity, Cash, Net income, Earnings per share etc. responsible for dividend declaration.
- To understand the criticality of a company's profitability (in terms of Earnings per share) component in declaration of dividends.
- To measure each factor individually on how it affects the dividend decision.

__Research Questions:__

__Research Questions:__

RQ1. What is the relation between dividend payout and firm's debt?

RQ2. What is the relation between dividend payout and Profitability?

RQ3. What is the relation between dividend payout and liquidity?

RQ4. What is the relation between dividend payout and Retained Earnings?

RQ5. What is the relation between dividend payout and Net Income?

__Contribution of the Study:__

__Contribution of the Study:__

Dividend decision is an important financial decision made by firms, managers, and investors. This study aims to contribute to the corporate finance literature, by looking at the Dividend puzzle. An attempt is made to make a valuable contribution in two major ways:

- Theoretical and Empirical approach is taken to provide a comprehensive view on the subject.
- The empirical Approach taken in this study will definitely leave some promising future ideas.
- The empirical findings and conclusions contained in this study can be used by financial managers to inform dividend decisions.

*Limitations of Study:*

*Limitations of Study:*

The areas of concern to investigate in this study are extensive. Due to the Time constraint and accessibility of data, the research will be limited to the following:

- The period of study is only three years 2006 to 2008.
- The research has considered only those firms who pay dividends.
- The study is focused only on firms trading on the New York Stock Exchange.

*Structure of the Paper:*

*Structure of the Paper:*

The remaining chapters will be organized as follows:

#### Chapter Two: Literature Review

This chapter discusses the different theories laid down in context to dividend policy and explains the relationship between dividend payout and its determinants as concluded by the study of different researchers and theorists.

#### Chapter Three: Research Methodology

This chapter explains the research hypothesis and gives a descriptive study of the techniques and the model used for data analysis. The application of the statistical tests used are explained thoroughly.

#### Chapter four: Data Analysis and Findings

To address the research questions, results obtained from the regression analysis will be evaluated and discussed in this chapter.

#### Chapter five: Recommendations and Conclusion.

This chapter Concludes the entire study and provides recommendations based on the findings and analysis done in the previous chapter and recommendations for future research.

#### CHAPTER TWO

#### LITERATURE REVIEW

Dividend remains one of the greatest enigmas of modern finance. Corporate dividend policy is an important decision area in the field of financial management hence there is an extensive literature devoted to the subject. Dividends are defined as the distribution of earnings (present or past) in real assets among the shareholders of the firm in proportion to their ownership. Dividend policy refers to management's long-term decision on how to utilize cash flows from business activities-that is, how much to plow back into the business, and how much to return to shareholders (Khan and Jain, 2005).

Lintner (1956) conducted a notable study on dividend distributions, his was the first empirical study of dividend policy through his interview with managers of 28 selected companies, he stated that most companies have clear cut target payout ratios and that managers concern themselves with change in the existing dividend payout rather than the amount of the newly established payout. He also states that, Dividend policy is set first and other policies are then adjusted and the market reacts positively to dividend increase announcements and negatively to announcements of dividend decreases. He measured major changes in earnings as the key determinant of the companies' dividend decisions. Lintner's study was expanded by Farrelly et al. (1988), who, mailed a questionnaire to 562 firms listed on the New York Stock Exchange and concluded that managers accept dividend policy to be relevant and important. Lintner's view was also supported by the study results of Fama and Babiak (1968) and Fama (1974) who suggested that managers prefer a stable dividend policy, and are hesitant to increase dividends to a level that cannot be supported. Fama and Babiak's (1968) study also concludes that Net income appears to explain the dividend change decision better than a cash flow measure.

The study by Adaoglu (2000), Amidu and Abor (2006) and Belans et al (2007) stated that net income shows positive and significant association with the dividend payout, therefore indicating that, the firms with the positive earnings pay more dividends.

Merton Miller and Franco Modigliani (1961) made a proposition that the value of a firm is not affected by its dividend policy. Dividend policy is a way of dividing up operating cash flows among investors or just a financial decision. Financial theorists Martin, Petty, Keown, and Scott, 1991 supported this theory of irrelevance. Miller and Modigliani's conclusion on the irrelevance of dividend policy presented a tough challenge to the conventional wisdom of time up to that point, it was universally acknowledged by both theorists and corporate managers that the firm can enhance its business value by providing for a more generous dividend policy as investors seem to prefer dividends over capital gains (JM Samuels, FM.Wilkes and R.E Brayshaw).

Benartzi et al. (1997) conducted an extensive study and concluded that Lintner's model of dividends remains the finest description of the dividend setting process available. Baker et al. (2001) conducted a survey on 630 NASDAQ-listed firms and analyzed the responses from 188 CFO's about the importance of 22 different factors that influence their dividend policy, they found that the dividend decisions made by managers were consistent with Lintner's (1956) survey results and model. Their results also suggest that managers pay particular attention to the dividend policy of the firm because the dividend decision can affect firm value and, in turn, the wealth of stockholders, thus dividend policy requires serious attention by the management.

E.F Fama and K.R French (2001) investigated the characteristics of companies paying dividends and concluded that the top most characteristics that affect the decision to pay dividends are Firm size, Profitability, and Investment opportunities. They studied dividend payment in the United States and found that the proportion of dividend payers declined sharply from 66% in 1978 to 20.8% in 1999, and that only about a fifth of public companies paid dividends. Growth companies such as Microsoft, Cisco and Sun Microsystems were found to be non-dividend payers. They also explained that the probability that a firm would pay dividends was positively related to profitability and size and negatively related to growth. Their research concluded that larger firms are more profitable and are more likely to pay dividends, than firms with more investment opportunities. The relationship between firm size and dividend policy was studied by Jennifer J. Gaver and Kenneth M. Gaver (1993). They suggested that "A firm's dividend yield is inversely related to the extent of its growth opportunities". The inference here is that as cash flow increases, the coefficient of dividend decreases, indicating that smaller firms that have greater investment opportunities thus they tend not to make dividend payment while larger firms tend to have proactive dividends policy.

Ho, H. (2003) undertook a comparative study of dividend policies in Japan and Australia. Their study revealed that dividend policies in Australia and Japan are affected by different financial factors. Dividend policies are affected positively by size in Australia and liquidity in Japan. Naceur et al (2006) examined the dividend policy of 48 firms listed on the Tunisian Stock Exchange during the period 1996-2002. His research indicated that highly profitable firms with more stable earnings could afford larger free cash flows and thus paid larger dividends. Li and Lie (2006) reported that large and profitable firms are more likely to raise their dividends if the past dividend yield, debt ratio, cash ratio are low. A study was conducted by Norhayati Mohamed, Wee Shu Hui, Mormah Hj.Omar, and Rashidah Abdul Rahman on Malaysian companies over a 3 year period from 2003-2005. The sample was taken from the top 200 companies listed on the main board of Bursa Malaysia based on market capitalization as at 31December 2005. Their study concluded that bigger firms pay higher dividends.

For the purpose of finding out how companies arrive at their dividend decisions, many researchers and theorists have proposed several dividend theories. Gordon and Walter (1963) presented the Bird in Hand theory which suggested that to minimize risk the investors always prefer cash in hand rather than future promise of capital gain. This theory asserts that investors value dividends and high payout firms. As said by John D. Rockefeller (an American industrialist) "The one thing that gives me contentment is to see my dividend coming in". For companies to communicate financial well-being and shareholder value the easiest way is to say "the dividend check is in the mail". The bird-in-hand theory (a pre-Miller-Modigliani theory) asserts that dividends are valued differently to capital gains in a world of information asymmetry where due to uncertainty of future cash flow, investors will often tend to prefer dividends to retained earnings. As a result the value of the firm would be increased as a higher payout ratio will reduce the required rate of return (see, for example Gordon, 1959). This argument has not received any strong empirical support. Dividends, paid by companies to shareholders from earnings, serve as an important indicator of the strength and future prosperity of the business. This explanation is known as signaling hypothesis. Signaling is an example factor for the relevance of dividends to the value of the firm. It is based on the idea of information asymmetry between managers and investors, where managers have private information about the firm that is not available to the outsiders. This theory is supported by models put forward by Miller and Rock (1985), Bhattacharya (1979), John and Williams (1985). They stated that dividends can be used as a signaling device to influence share price. The share price reacts favorably when an announcement of dividend increase is made. Few researchers found limited support for the signaling hypothesis (see Gonedes, 1978, Watts, 1973) and there are other researchers, who supported the hypothesis, for example, in Michaely, Nissim and Ziv (2001), Pettit (1972) and Bali (2003).

The tax-preference theory assumes that the market valuation of a firm's stocks is increased when the dividend payout ratios is low which in turn lowers the required rate of return. Because of the relative tax liability of dividends compared to capital gains, investors need a large amount of before-tax risk adjusted return on stocks with higher dividend yields (Brennan, 1970). On one side studies by Lichtenberger and Ramaswamy (1979), Poterba and Summers, (1984), and Barclay (1987) have presented empirical evidence in support of the tax effect argument and on the other side Black and Scholes (1974), Miller and Scholes (1982), and Morgan and Thomas (1998) have either opposed such findings or provided completely different explanations. The study by Masulis and Trueman (1988) model dividend payments in form of cash as products of deferred dividend costs. Their model predicts that investors with differing tax liabilities will not be uniform in their ideal firm dividend policy. As the tax liability on dividends increases (decreases), the dividend payment decreases (increases) while earnings reinvestment increases (decreases). According to Farrar and Selwyn (1967), in a partial equilibrium framework, individual investors choose the amount of personal and corporate leverage and also whether to receive corporate distributions as dividends or capital gains. Barclay (1987) has presented empirical evidence I support of the tax effect argument. Others, including Black and Scholes (1982), have opposed such findings or provided different explanations.

Farrar and Selwyn's model (1967) made an assumption that investors tend to increase their after tax income to the maximum. According to this model corporate earnings should be distributed by share repurchase rather than the use of dividends.

Brennan (1970) has extended Farrar and Selwyn's model into a general equilibrium framework. Under this, the expected usefulness of wealth as a system of barter is maximized. Despite being more robust both the models are similar as regards to their predictions. According to Auerbach's (1979) discrete-time, infinite-horizon model, the wealth of shareholders is maximized by the shareholders themselves and not by firm market value. If there does, infact, exist a difference between capital gains and dividends tax; firm market value maximization is no longer determined by wealth maximization.

He states that the continued undervaluation of corporate capital leads to dividend distributions.

The clientele effects hypothesis is another related theory. According to this theory the investors may be attracted to the types of stocks that fall in with their consumption/savings preferences. That is, investors (or clienteles) in high tax brackets may prefer non-dividend or low-dividend paying stocks if dividend income is taxed at a higher rate than capital gains. Also, certain clienteles may be created with the presence of transaction costs. There are several empirical studies on the clientele effects hypothesis but the findings are mixed. Studies by Pettit (1977), Scholz (1992), and Dhaliwal, Erickson and Trezevant (1999) presented evidence consistent with the existence of clientele effects hypothesis whereas studies by Lewellen et al. (1978), Richardson, Sefcik and Thomason (1986), Abrutyn and Turner (1990), found weak or contrary evidence.

There is an assumption that the managers do not always take steps which would lead to maximizing an investor's wealth. This gives rise to another favorable argument for hefty dividend payouts which shifts the reinvestment decision back on the owners. The main hitch would be the agency conflict (conflict between the principal and the agent) arising as a result of separate ownership and control. Therefore, a manager is expected to move the surplus funds from the high retained earnings into projects which are not feasible. This would be mainly due to his ill intention or his in competency.

Thus, generous dividend payouts increase a firm's value as it reduces the management's access to free cash flows and hence, controlling the problem of over investment. There are many more agency theories explaining how dividends can increase the value of a firm. One of them was by Easterbrook (1984); he proposed that dividend payments reduce agency problems in contrast to the transaction cost theory which is of the view that dividend payments reduce the value as it forces to raise costly finances from outside sources. His idea is that if the dividends are not paid, there is a problem of collective action that tends to lead to hap-hazard management of the firm. So, dividend payouts and raising external finance would attract auditory and regulatory measures by financial intermediaries like investment banks, respective stock exchange regulators and the potential investors as well. All this monitoring would lead to considerable reduction of agency costs and appreciate the market value of the firm. Moreover, as defined by Jenson and Meckling (1976), Agency costs=monitoring costs+ bonding, costs+ residual loss i.e. sum of agency cost of equity and agency cost of debt. Hence, Easterbrook (1984) noted that dividend payments and raising new debt and its contract negotiations would reduce potential for wealth transfer.

The realization for potential agency costs linked with separation of management and shareholder's is not new. Adam Smith (1937) proposed that management of earlier companies is wayward. This problem was highly witnessed during at the time of British East Indian Companies and tracking managers was a failure due to inefficiencies and high costs of shareholder monitoring (Kindleberger, 1984). Scott (1912) and Carlos (1922) differ with this view point. They agree that although some fraud existed in the corporations, many of the activities of the managers were in line with those of the shareholder's interests.

An opportune and intelligent manager should always invest the surplus cash available into those opportunities which are well researched to be in the best interest of the shareholders. Berle and Means (1932) was the first to discover the insufficient utilization of funds which are surplus after other investment opportunities taken by the management. This thought was further promoted by Jensen's (1986) free cash flow hypothesis. This hypothesis combined market information asymmetries with the agency theory. The surplus funds left after all the valuable projects are largely responsible for creation of the conflict of interest between the management and the shareholders. Payment of dividends and interest on other debt instruments reduce the cash flow with the management to invest in marginal net present value projects and for other perquisite consumptions. Therefore, the dividend theory is better explained by the combination of both the agency and the signaling theory rather than by any one of these alone. On the other hand, the free cash flow hypothesis rationalizes the corporate takeover frenzy of the 1980's Myer's (1987 and 1990) rather than providing a clear and comprehensive dividend policy.

The study by Baker et al. (2007) reports, that firms paying dividend in Canada are significantly larger and more profitable, having greater cash flows, ownership structure and some growth opportunities. The cash flow hypothesis proposes that insiders to a firm have more information about future cash flow than the outsiders, and they have incentivized motives to leak this to outsiders. Lang and Litzenberger (1989) check the cash flow signaling and free cash flow explanations of the effect of dividend declarations on the stock prices. This difference between permanent and temporary changes is also explored in Brook, Charlton, and Hendershott (1998). However, this study is based on the hypothesis that dividend changes contain cash flow information rather than information about earnings. This is the cash flow signaling hypothesis proposing that dividend changes signal expected cash flows changes.

The dividend decisions are affected by a number of factors; many researchers have contributed in determining which determinant of dividend payout is the most significant in contributing to dividend decisions. It is said that the primary indicator of the firm's capacity to pay dividends has been Profits. According to Lintner (1956) the dividend payment pattern of a firm is influenced by the current year earnings and previous year dividends. Pruitt and Gitman's (1991) survey of financial managers of 1000 largest U.S companies about the interplay among the investment and dividend decisions in their firms reported that, current and past year profits are essential factors influencing dividend payments. The conclusion derived from Baker and Powell's (2000) survey of NYSE-listed firms is that the major determinant is the anticipated level of future earnings and continuity of past dividends. The study of Aivazian, Booth, and Cleary (2003) concludes that profitability and return on equity positively correlate with the size of the dividend payout ratio. The study by Lv Chang-jiang and Wang Ke-min (1999) on 316 listed companies in China that paid cash dividends during 1997 and 1998 by using modified Lintner dividend model, suggested that the dividend payout ratio is due to the firm's current earning level. Other researchers like Chen Guo-Hui and Zhao Chun-guang (2000), Liu Shu-lian and Hu Yan-hong (2003) also concluded their research on the above stated understanding about dividend policy of listed companies in China.

A survey done by Baker, Farrelly, and Edelman (1985) and Farrelly, Baker, and Edelman (1986) on 562 New York Stock Exchange (NYSE) firms with "normal" kinds of dividend polices in 1983 suggested that the major determinants of dividend payments were the anticipated level of future earnings and the pattern of past dividends.

DeAngelo et al. (2004) findings suggest that earnings do have some impact on dividend payment. He stated that the high/increasing dividend concentration may be the result of high/increasing earnings concentration. Goergen et al. (2005) study on 221 German firms shows that net earnings were the key determinants of dividend changes. Baker and Smith (2006) examined 309 sample firms exhibiting behavior consistent with a residual dividend policy and their matched counterparts to understand how they set their dividend policies. Their study showed that for the matched firms, the pattern of past dividends and desire to maintain a long-term dividend payout ratio elicit the highest level of agreement from respondents. The study by Ferris et al. (2006) found mixed results for the relation between a firm's earnings and its ability to pay dividends. Kao and Wu (1994) used a time series regression analysis of 454 firms over the period of 1965 to1986, and showed that there was a positive relationship between unexpected dividends and earnings. Carroll (1995) used quarterly data of 854 firms over the period of 1975 to 1984, and examined whether quarterly dividend changes predicted future earnings. He found a significant positive relationship.

Liquidity is also an important determinant of dividend payouts. A poor liquidity position would generate fewer dividends due to shortage of cash. Alli et.al (1993), reveal that dividend payments depend more on cash flows, which reflect the company's ability to pay dividends, than on current earnings, which are less heavily influenced by accounting practices. They claim current earnings do no really reflect the firm's ability to pay dividends. A firm without the cash flow back up cannot choose to have a high dividend payout as it will ultimately have to either reduce its investment plans or turn to investors for additional debt. The study by Brook, Charlton and Hendershott (1998) states that, Firms expecting large permanent cash flow increases tend to increase their dividend.

Managers do not increase dividends until they are positive that sufficient cash will flow in to pay them (Brealey-Myers-2002). Myers and Bacon's (2001) study shows a negative relationship between the liquid ratio and dividend payout.

For companies to enable them to enhance their dividend paying capacity, and thus, to generate higher dividend paying capacity, it is necessary to retain their earnings to finance investment in fixed assets. The study by Belans et al (2007) states that the relationship between the firm's liquidity and dividend is positive which explains that firms with more market liquidity pay more dividends. Reddy (2006), Amidu and Abor (2006) find opposite evidence.

Lintner (1956) posited that the level of retained earnings is a dividend decision by- product. Adaoglu (2000) study shows that the firms listed on Istanbul Stock Exchange follow unstable cash dividend policy and the main factor for determining the amount of dividend is earning of the firms. The same conclusion was drawn by Omet (2004) in case of firms listed on Amman Securities Market and he further states that the tax imposition on dividend does not have the significant impact on the dividend behavior of the listed firms. The study by Mick and Bacon (2003) concludes that future earnings are the most influential variable and that the past dividend patterns as well as current and expected levels are empirically relevant in explaining the dividend decision. Empirical support for Lintner's findings, that dividends were indeed a function of current and past profit levels and were negatively correlated with the change in sales was found by Darling (1957), Fama and Babiak (1968). Benchman and Raaballe (2007) discovered that the propensity to pay out dividends is positively correlated to retained earnings. Also, the study by Denis and Osobov (2006) states that retained earnings are a significant dividend characteristic for non- US firms including UK, German, and French firms.

One of the motives for dividend policy decision is maintaining a moderate share price as poor stock price performance mostly conveys negative information about firm's reputation. An empirical research took by Zhao Chun-guang and Zhang Xue-li et al (2001) on all A shares listed companies listed in Shenzhen and Shanghai Stock Exchange, states that the more cash dividends is paid when the stock prices are high. Chen Guo-Hui and Zhao Chun-guang (2000) undertook a research on all A shares listed before 1996 and paid dividend into share capital in 1997 as their sampling, and employed single-factor analysis, multifactor regression analysis to analyze the data. Their research showed a positive stock price reaction to the cash dividend, stock dividend policy.

Myers and Bacon (2001) discussed that the debt to equity ratio was positively correlated to the dividend yield. Therefore firms with relatively more investment opportunities would tend to be more geared and vice versa (Ross, 2000). The study by Hu and Liu, (2005) declares that there is a positive correlation between the cash dividend the companies pay and their current earnings, and a inverse relationship between the debt to total assets and dividends.

Green et al. (1993) questioned the irrelevance argument and investigated the relationship between the dividends and investment and financing decisions. Their study showed that dividend payout levels are decided along with investment and financing decisions. The study results however do not support the views of Miller and Modigliani (1961). Partington (1983) declared that firms' motives for paying dividends and extent to which dividends are decided are independent of investment policy. The study by Higgins (1981) declares a direct link between growths and financing needs, rapidly growing firms have external financing needs because working capital needs normally exceed the incremental cash flows from new sales. Higgins (1972) suggests that payout ratios are negatively related to firms' need top fund finance growth opportunities. Other researchers like Rozeff (1982), Lloyd et al. (1985) and Collins et al. (1996) all show significantly negative relationship between historical sales growth and dividend payout whereas D, Souza (1999) however shows a positive but insignificant relationship in the case of growth and negative but insignificant relationship in case of market to book value. Jenson and Meckling (1976) find a strong relationship between dividends and investment opportunities. They explain, in some circumstances where firm's have relative uptight disposable cash flow and a number of investment opportunities have, the shareholders are ready to accept low dividend payout ratio.

From the investor's point of view, the dividend payments represent definite evidence of a company's worth. A company that expects sufficient future cash flows, large enough to meet debt obligations and dividend payments, will increase dividend payout.

Howe (1998) believed that the actions of the managers might convey information to the investors outside as they are more informed about the future prospects of their firms than the market. Reddy (2002) studied dividend behavior and expressed his views on the observed behavior with the help of signaling hypothesis. The undervalued firms (assessed by the price to book value ratio) might use increase in dividends as signals to the market.

It is argued that Risk is a significant determinant of the propensity to pay dividends. Risk is related to payout policies in general: it explains the decision to repurchase shares and increase dividends. It is an economically and statistically significant determinant. The study by Pruitt and Gitman (1991) found that the firm's dividend policy is also determined by risk (year to year variability of earnings). "A firm with consistent performance with regards to their earnings can predict its earnings in the future with a greater accuracy. Thus, such firms have less risk of future dividend cuts and commit to pay larger proportion of its earnings as dividends." (Chang and Rhee, 2001) The study by the researchers, Rozeff (1982), Lloyd et al. (1985) and Colins et al. (1996) used beta value of a firm, as an indicator of its market risk, and proved a significant statistical with negative correlation between beta and dividend payout. Their findings concluded that firms having higher level of market risk will payout dividends at lower rate. This study was also supported by D'Souza (1999) who also found statistically significant and negative relationship between beta and dividend payout.

Moh'd, Perry, and Rimbey (1995) studied the effects of dividend on the agency costs. They stated that cash dividend payout compels the management to raise capital from outside. This exposure to the capital markets leads to lowering of the agency costs. Larger the dividend payouts, lesser would be the agency costs as compared to the cost of generating required capital. Jensen and Meckling (1976) stressed on the management ownership to reduce agency costs as the goals of the management and the shareholders would be in line. Dividend payments reduces he discretionary funds available with the management and hence would also reduce the agency problem between the managers and the shareholders. Also, there is another kind of conflict existing between the shareholders and the bondholders. This exists because shareholders can pay themselves dividends and hence extracting the wealth from the bondholders. So bondholders may try to put restrictions on the dividend payouts through bond indenture (Kalay, 1982). Fenn and Liang (2001) found that the stock incentivisation to the management would mitigate the agency costs for the firms with inverse relationship between dividend and stock options for the management. Furthermore, Alli et al. (1993) explained that as the number of stockholders would increase, the agency problem would also increase and the need for monitoring the actions of the management would also increase. If dividends can mitigate this problem, we can expect a positive relationship between number of common stockholders and dividend payout ratio.

Large sized firms with stable earnings and profitability record will have easy access to outside capital. Whereas, a small or a new firm has more risk for potential investors. Raising finance for such firms is very difficult and hence these should depend more on their retained earnings. Smith and Watts (1992) found that industry growth rates, profitability, asset mixes, earnings variability and capital investment needs are most important factors in determining a firm's capital structure. Competition is also an important factor which some firms take into account while designing a firm's dividend policy. These firms believe that dividend payments can signal information to its stakeholders and their competitive nature. However, Howe and Shen (1998) argued that the dividend payments of one firm cannot affect the share prices of its competitors.

Besides, Holder et al (1998) stated that corporate focus is negatively related to dividend payout ratios. They define a corporation as being focused when the firm's sales are attributable to a distinct business line. Hence, large sized firms have higher payout ratios as compared to the smaller firms. Dickens, Casey, and Newman (2002) studied the impact of ownership in the banking industry and concluded that management's ownership was negatively correlated with the payout ratio. This suggested that agency costs reduced in the management owned firms. . Goorgen, Da Silva and Renneboog (2002) showed that, firms with the banks as their shareholders would omit their dividends. On the other hand Holfer et al (2004) suggested that neither the institutional holdings nor the bank control is statistically significant determinant of dividend payouts.

Some previous literature assumed that there is a relationship between the firm's asset structure and the firm's dividend policy. Koch and Shenoy, (1999) purported that firms with more tangible assets have greater tax benefits without being reliant on debt. These use the dividend policy to infuse the agency costs. On the contrary, it's argued that asset tangibility has a negative relationship with the dividend policy. Aivazian et al. (2003) state "when the assets are more tangible, fewer short-term assets are available for banks to lend against, this imposes financial constraints on firms operating in developing financial systems, where the main source of debt is short-term bank financing in more primitive financial systems, where the main source of debt us short term bank financing." This result was supported by Ho (2003). Aidvaizain, Booth and Clearly (1998) consider that the relationship between debt, investment and dividends are similar in all the countries and that the theory of financial signaling is more suitable for those countries in which companies are financed through capital markets. Also suggested by Holder, Langrehr, and Hexter (1998) was that corporations focusing only on a single business paid lesser dividends than lesser focused corporate. Ho (2003) conducted a comparative study on dividend policies in Australia and Japan by examining a 10 year data (1992-2001) of 332 firms in Australia and Japan. It was concluded that Australia paid higher dividends than Japan. He found out that the dividend policy in Australia was affected positively by size whereas in Japan it was affected positively by liquidity and negatively by risk. However, the industry effect was found to be of significance in both Australia and Japan. This indicated the importance of the industry in which a company competes. In India, Narasimhan and Asha (1997) and Reddy (2002) have tried to explain the dividend policy. They supported that omissions in dividend payout signal towards the future earnings but they do not support the tax preference theory. Porta et al. (2000) was also of the same view. Kevin (1992), in his study shows that dividend stability is a primary determinant of payout while profitability is only secondary. Mohanty (1999) tried to examine the dividend payout after a bonus issue and found out that firms with bonus issues yielded better returns to their shareholders than a firm which maintained a steadily increasing dividend rate and Pandey (1994) found that dividend payouts depend on current and expected earnings as well as the pattern of past dividends. Higher revenue is another factor which negatively influences the dividends (Ramcharran 2001). Dickens, Casey, and Newman (2000) studied bank's dividend policy and found that there is inverse relationship between dividend yields and investment opportunities, signaling, ownership, and risk and a positive relationship with size and dividend history (also see Omet, 2004). Barclay, Smith, Ross (1995) found investment opportunities and leverage as the most important determinants of dividend decisions. Baker, Theodore, and Gary (2001) found that the pattern of past dividends, stability of earnings, and level of current and expected future earnings are the most important determinants of dividend decisions; (also see Brav, Graham, Harvey, Michaely, 2005). Alli, Khan, Ramirez (1993) have found out that the firms with high issue costs, high growth levels, high risks and high level of capital expenditure pay lower dividends. Omran and Pointon (2004) researched on the role of a firm's dividend policy in determining the prices of the shares, factors for payout ratios, and the factors involved in the stability of dividend payout with respect to 94 Egyptian companies. They found that retentions are more valuable than dividends in firms with actively traded shares and that the firms with non-actively traded shares hold the accounting book value and earnings as more important than the dividends.

Chang and Rhee, (2003) proposed that higher the growth opportunities for a firm, more funds would be required for its expansion and more likely the firm is to retain earnings than pay them as dividends. In addition, this negative relationship is in line with Myers and Majluf (1984) findings, who pointed that the companies with high growth opportunities will pay lesser dividends. This relationship is also supported by the agency theory of dividend policy (Holder et al., 1998, Chang and Rhee 2001, Aivazian et al., 2003). Husam-Aldin Nizar Al-Malkawi (2007) examined Jordanian firms using Tobit specifications for determining the amount of dividend payout. The results were obtained using the maximum likelihood estimations of the random effects Tobit regressions. The data showed that the fraction of stocks held by insiders has negative impact on the level of dividends paid. Similarly, the existence of government or its agencies in a firm's ownership structure positively affects the amount of dividends. Other variables of ownership structure seem to have no influence on dividend policy. The analysis also found that a firm's financial leverage is significantly and negatively related to its dividend policy.

Lamport, Subadar, Fowdar, and Boodhoo (2007), used a sample of 38 firms listed on Mauritius stock exchange and performed a cross sectional regression analysis to find the factors which motivate the dividend decision. It was found that retained earnings, liquidity and earnings per share are among the most significant motivators of dividend payout. Indian Information Technology sector was studied by Kanwal and Kapoor (2008) to find out about the determinants of dividend payouts. Statistical techniques like correlation and regression were used to analyze the data collected and it was found that and beta i.e. year to year variability in earnings are the major determinants of dividend payout in Indian IT sector. These variables show the positive and significant relation with dividend payout.

Amidu Mohammed and Abor Joshua (2006) carried out a research in Ghana on the same topic. They chose 20 firms listed on Ghana Stock Exchange and the study period was six years. It was found that there was a positive relationship between dividend payout and profitability, cash flow which suggest that, a good liquidity position increases a firm's ability to pay dividend. The findings also showed negative relationship between dividend payout and risk, growth and market to book value. In a study conducted by Kania and Bacon (2005) for 542 companies listed on NASDAQ, NYSE, AMEX, and OTC exchanges, investigations were conducted to determine the factors for dividend payout. They used Ordinary Least Squares regression and the independent variables tested include: sales growth, current ratio, beta, return on equity, debt to total assets, and the estimated five year growth rate for earnings per share. They discovered the sales growth and beta related negatively to the dividend payout ratio. The positive relationship was observed between the debt to total assets ratio and the dividend payout ratio.

Mohammad Nasr and Shammyla Naeem (2007) research in Pakistan with a sample of 108 companies listed on Karachi stock exchange for a period of 5 years, concluded that profitability is positively related to dividend payout, when firms have greater investment opportunities, they conserve cash to fund those opportunities and, therefore, paid fewer dividends. Dr Ayub Mehar researched on the long term return behavior of dividend changing firms and estimated that only 23 percent incremental profits are transformed into dividend. The remaining profits are utilized for the additional investment. It was also found that ownership was also a factor for dividend payouts. The results support the hypothesis that companies start to pay dividends after a certain level of growth. At the earlier stage companies concentrate on retained earnings.

Dividend studies have various motivations examining banking firms. The earliest banking related study was provided by Gupta and Walker (1975). They analyzed data from 980 banking firms from 1965 to 1968 to identify variables that explain the dividend policies of banks. They found a positive relationship between dividends and profits and liquidity. Kennedy and Nunnally (1986) examined dividend payout ratios for a period of 1982- 1983 to select the significant determinants of dividend payout ratios. They used the stepwise regression techniques and analyzed 80 large banking firms. They analyzed Earnings as an important determinant variable.

Many studies focus the Utility sector for dividends analysis. One group of empirical studies includes Dhrymes and Kurz (1964), Higgins (1974), Lee (1976), and Tripathy (1992), and Shome (1994). The second group consists of surveys by Baker, Farrelly, and Edelman (1985) and Baker (1985) and Baker and Powell (1999).

The dividend policy was examined by Dhrymes and Kurz (1964) of electric utilities using data from 1947-1959. The model examined the payout ratios for 261 firms as a function of net profits, sales, investment, capital structure (measured by long term debt level) and liquidity. The results show that dividend payouts are more in line with sales level than with profits levels for electric utilities. Baker et al. (1985), Edelman et al. (1985), and Farrelly et al. (1986) all examined survey responses from 114 utility firms, 57 wholesale/ retail firms and 147 manufacturing firms. All of theirs study shared the same underlying survey. The top three determinants of dividend policy stated by them were Future Earnings, past dividend profits and Cash availability.

There has been a considerable study focusing on the Insurance industry. Lee and Forbes (1982), examined dividend payout ratios and dividend yield of 61 non life insurance companies for the period 0f 1955-1975. The results showed that the variable best explaining the dividend measure is Earnings.

Real Estate Investment Trusts (REIT's) have unique dividend requirements. They must payout, usually 95% of net income as dividends to maintain tax benefits. Wang, Erickson, and Gau (1993) examined dividend payouts for 102 Equity and Mortgage REIT's using data from 1981 to 1988. The tests showed that Equity REIT's have significantly greater dividend payouts than Mortgage REIT's. They also explained that equity REIT's would have higher monitoring costs for their owners, and the owners would require to higher dividend payouts to reduce internal monitoring costs. Bradley, Capozza, and Seguin (1998) examine 75 REIT's for the period of 1985 to 1992. They found a inverse relationship between dividend payout and cash. Mooradian and Yang (2001) examined dividend policies for 16 hotel REIT's and 51 non-REIT hotel firm's for the period of 1993 to 1999. The results showed that non REIT firms have lower dividends than REIT firms, which paid 95% of their earnings as dividends.

Finally Wansley (2003) reports dividend payout ratios for each year from 1980 to 2000 for industries categories: unregulated, petroleum, telecom, utilities, financial services, REIT's etc. His study provides evidence that regulated firms have both higher dividend payouts and yields than unregulated firms.

*Summary:*

*Summary:*

This literature review chapter has evaluated and elaborated the prior research relating to the field of study. It has highlighted different research issues on the subject of dividend determinants. The above literature review tried to explore the different theories and hypothesis from the previous research and studies by researchers and theorists. The factors that may affect the dividend payout were discussed and the key determinants of dividend payout were determined along with the relationships showed as stated in the previous literature. Also, research stating the determinants affecting the dividend decision in the sectors like banks, utilities, REIT's and Insurance was mentioned.

#### CHAPTER THREE

#### RESEARCH METHODOLOGY

*Introduction:*

*Introduction:*

Quantitative method is used to conclude this research. This research method relies on collection and analysis of numerical data and statistics. Data collection instruments are described and a detail outlay of research hypothesis and chosen variables for research is carried out. Furthermore the statistical tests applied to conclude the research are discussed.

*Research Design:*

*Research Design:*

The research design has been organized on the line of research questions. The research questions and research hypothesis are as follows:

RQ1. What is the relation between dividend payout and Firms Debt?

RQ2. What is the relation between dividend payout and Profitability?

RQ3. What is the relation between dividend payout and liquidity?

RQ4. What is the relation between dividend payout and retained earnings?

RQ5. What is the relation between dividend payout and net income?

*Research Hypothesis:*

*Research Hypothesis:*

#### H1. The firm's debt is negatively associated with dividend payouts.

When a firm acquires debt financing it commits itself to fixed financial charges embodied in interest payments and the principal amount, and failure to meet these obligations may lead the firm into liquidation. The risk associated with high degrees of financial leverage may therefore result in low dividend payments because, ceteris paribus, firms need to maintain their internal cash flow to pay their obligations rather than distributing the cash to shareholders. Moreover, Rozeff (1982) points out those firms with high financial leverage tend to have low payouts ratios to reduce the transaction costs associated with external financing. Therefore, other things being equal, an inverse relationship between financial leverage ratio, defined as the ratio of total short-term and long-term debt to total shareholders' equity (DER), and dividends is expected. The hypothesized relationship between firm's debt (DER) and dividend payout is Significant and negative. The hypothesis will be proved by Regression model results achieved by using statistical tests like R test, R- square test, F-test and standard Error.

#### H2. There is a positive relationship between a firm's profitability and dividend payout.

The decision to pay dividends starts with profits. Therefore, it is logical to consider profitability as a threshold factor, and the level of profitability as one of the most important factors that may influence firms' dividend decisions. In his classic study, Lintner (1956) found that a firm's net earnings are the critical determinant of dividend changes. The pecking order hypothesis may provide an explanation for the relationship between profitability and dividends. That is, taking into account the costs of issuing debt and equity financing, less profitable firms will not find it optimal to pay dividends, ceteris paribus. On the other hand, highly profitable firms are more able to pay dividends and to generate internal funds (retained earnings) to finance investments. Fama and French (2002) used the expected profitability of assets in place for testing the pecking order hypothesis. In another study, Fama and French (2001) interpreted their results of the positive relationship between profitability and dividends as consistent with the pecking order hypothesis. To test this hypothesis, the after tax earnings per share (EPS) is used as a measure of a firm's profitability. The hypothesized relationship between EPS and dividends is positive. The hypothesis will be proved by Regression model results achieved by using statistical tests like R test, R- square test, F-test and standard Error.

#### H3. There is a significant and positive relation between cash and dividend payout.

The hypothesis states that a poor liquidity position would generate fewer dividends due to shortage of cash. Alli et.al (1993) reveal that dividend payments depend more on cash flows, which reflect the company's ability to pay dividends, Managers do not increase dividends until they are positive that sufficient cash will flow in to pay them (Brealey-Myers-2002). The hypothesized relation between cash and dividend payout is positive. The hypothesis will be proved by Regression model results achieved by using statistical tests like R test, R- square test, F-test and standard Error.

#### H4. There is a positive relationship between net income and dividend payout.

Net Income is the income of a company after deducting all interests, expenses and taxes. The likelihood of a company paying dividends or high dividends is proportionate to the net income of the same company. The more the profitable the company, more it is likely to pay dividends. The hypothesized relation between dividend payout and net income is negative. The hypothesis will be proved by Regression model results achieved by using statistical tests like R test, R- square test, F-test and standard Error.

#### H5. There is a positive and significant relationship between retained earnings and dividend payout.

Retained Earnings are said to be an excellent indicator of the possible dividend policy/ payout of a company. Retained earnings are the earnings that a company stores for future utilization to create further profits and / or assets. Thus, the future performance, financially, of a company can be better determined by using retained earnings as n indicator. The hypothesized relation between dividend payout and retained earnings is positive. The hypothesis will be proved by Regression model results achieved by using statistical tests like R test, R- square test, F-test and standard Error.

*Data Collection:*

*Data Collection:*

A Data set of 100 firms was constructed consisting of companies listed on New Stock Exchange (NYSE). The companies cover sectors like Basic Materials, Utilities, Energy, financial, Consumer goods and services and Industrials. The research required investigating the relationship among the variables across the time period, thus the data collected was based on Time series. Secondary sources were used to collect data, thus data employed in this study is derived from the financial statements of the companies. The research is based on a three year period (2006 to 2008). Data was collected from the Reuters Database (www.reuters.com/finance/stocks). Data collected from Reuters is reliable and Quantifiable. The data set is annual and consists of 100 firm year observations. Data is collected for the dependent variable i.e. Dividend payout in terms of dividend per share and five independent variables namely, Cash, Retained Earnings, Net Income, Debt to Equity ratio and Earnings per share.

*Dependent Variable (Y):*

*Dependent Variable (Y):*

In this research Dividend per share is the dependent variable.

*Independent Variables (X's):*

*Independent Variables (X's):*

**Earnings per Share (x1)**- Earnings per share are defined as the Net Earnings / Outstanding Shares. The critical determinant of dividend policy is considered to be profitability and EPS is used as an indicator of firm's profitability. In this research we will use values of basic normalized EPS of all the companies as indicated in their income statements. Earnings per share of 100 companies are obtained from their Income statements. The hypothesized relationship between EPS and dividends is positive.**Cash (x2)**- Cash is the money coming into the business of the firm. A highly profitable firm may not have sufficient cash to pay dividends. The dividend decision depends on the amount of cash the company holds in its hand. Cash figures of 100 companies are obtained from their Balance sheet. Hence it is important to check the relation between cash and dividend payout.**Retained Earnings(x3)**- Retained Earnings are the earnings that a company preserves to be reinvested in its core business or to pay debt. Companies use this money on growth opportunities and other investments which can generate high level of earnings in the future and they can pay high dividends in the future. The study by Denis and Osobov (2006) states that retained earnings is a significant dividend characteristic. Darling (1957), Fama and Babiak (1968). Benchman and Raaballe (2007) discovered that the propensity to pay out dividends is positively correlated to retained earnings. Therefore, retained earnings could be considered to be an important determinant of dividend payout. Retained earnings of 100 companies are obtained from their Balance sheet.**Net Income(x4)**- Net income states the financial condition of a business. It reflects whether the firm is in profits or not, high profits of a firm enable them to pay higher dividends. Thus net income is considered to be a important determinant of dividend payout. Net income of the companies is obtained from their Income statements.**Debt to Equity Ratio (x5)**- Debt to equity is the ratio of total long term debt + short term debt to shareholders equity This is regarded as the foremost determinant of dividend payout. It is essential to know how this affects the dividend decision of a firm. Debt to equity of companies is calculated from their Balance sheet.

*The Reasons to choose New York Stock Exchange for Research:*

*The Reasons to choose New York Stock Exchange for Research:*

New York Stock Exchange (NYSE) is one of the oldest stock exchanges in the world; it is based in New York, USA. It operates the world's largest and most liquid exchange group and offers the most diverse array of financial products and services (http://www.nyse.com). It is also one of the largest in terms of the number of volumes traded. Therefore, the validity, i.e. the degree to which the research findings can be applied in the real world, beyond the controlled setting of the research and reliability of the data was very precise.

*Data Analysis:*

*Data Analysis:*

Data for three years i.e. 2006 to 2008 based on NYSE have been used. Dividend payout in terms of Dividend per Share was used as dependent variable and the independent variables used are stated below:

x1= Earnings per Share

x2= Cash

x3= Retained Earnings

x4= Net Income

x5= Debt to Equity Ratio

Two models were developed for the analysis.

** 1. Linear Regression model:** This model was used for the analysis of each determinant individually with dividend payout taking all 100 firms from different sectors together as a whole. Linear regression attempts to model the relationship between two variables.

The simple linear regression model used is:

Y = a + b1x1 + e.

Standard linear regression equations were developed for all five determinants(X's) each, as the independent variables and the dependent variable (Y).

#### Eq1. Dividend Payout = f (EPS)

DP= a + b1 EPS + e

Y = a + b1x1 + e

Relationship between Earnings per share and dividend payout in terms of dividend per share was analyzed.

#### Eq2. Dividend Payout= f (Cash)

DP= a + b2 Cash + e

Y = a + b2x2 + e

Relationship between Cash and dividend payout in terms of dividend per share was analyzed.

#### Eq3. Dividend Payout= f (Retained Earnings)

DP= a + b3 RE + e

Y = a+ b3x3 + e

Relationship between Retained earnings and dividend payout in terms of dividend per share was analyzed.

#### Eq4. Dividend Payout= f (Net Income)

DP = a + b4 NI + e

Y = a + b4x4 + e

Relationship between Net income and dividend payout in terms of dividend per share was analyzed.

#### Eq5. Dividend Payout= f (Debt to Equity)

DP= a + b5 DE + e

Y = a + b5x5 + e

Relationship between Debt to equity and dividend payout in terms of dividend per share was analyzed.

** 2. Multiple Linear Regression:** For analyzing sector wise relationship between dividend payout and determinants of dividend payout, multiple linear Regression was used.

Multiple Regression model used is:

Y= a + b1x1 + b2 x2 + b3 x3 + b4 x4 + b5 x5 + e

For both the models, Y is the value of the Dependent variable, what is being predicted or explained.

**a** (Alpha) is the regression constant; equals the value of Y when the value of X=0. It represents the rate of change of one variable (y) as a function of changes in the other (x).

**b1, b2, b3, b4** and **b5** are regression coefficients of X1, X2, X3, X4, and X5. Regression coefficients are the parameters of interest. If they are all zero, there is no significant relationship between the dependent and independent variables. If at least one is not zero, a significant relationship exists. The regression coefficient indicates the amount of change in the value of dependent variable(Y) for every one unit change in independent variable.

**X** is the value of the Independent variable, what is predicting or explaining the value of Y.

**e** is the error term; the error in predicting the value of Y.

*The Computation of Results through SPSS Software:*

*The Computation of Results through SPSS Software:*

The regression analysis was carried out with the help of SPSS student version software. The tool requires providing the value of the dependent variable (Y), i.e. Dividend Payout in terms of Dividend per share, for all 100 firms. This is followed by providing the Input for the independent variable (x1, x2, x3, x4, and x5). Then the Enter method is used to analyze the regression output in the software for each equation mentioned above.

*Statistical Tools and Tests to be used:*

*Statistical Tools and Tests to be used:*

*Statistical tools:*

*Statistical tools:*

- Regression Analysis
- Correlation Analysis

*Statistical Tests:*

*Statistical Tests:*

*Tests for strength, association and Direction of variables:*

*Tests for strength, association and Direction of variables:*

*1.) R square test (Coefficient of Determination):*

R square test is also known as the Coefficient of Determination. R-square, is the fraction of the variance of the dependent variable explained by the regression. It measures the extent or strength of the association that exists between the variables X's and Y. It is a highly important statistic in evaluating adequacy of the regression model and is defined as the percent of the total variation in dependent variable. It ranges from 0 to 1. A value of 0 indicates that the linear regression equation explains none of the variance. A value of 1 indicates that all of the variation is explained. This occurs when the relationship between dependent and independent variables is exactly linear, i.e., each observed value of the dependent variable is exactly predicted by the linear regression equation. A value close to 0 states little correlation between the variables. This test is used to check the change in the dividend payout over the time by the change in the independent variable. This test was applied to all the linear equations representing the determinants of dividend payout and dividend payout, and the hypothesis was tested.

*2.) R Test (Coefficient of Correlation):*

This test is used to check the correlation among the dependent and independent variables. This is the coefficient of correlation which measures how well the data clusters around the regression line and how well one variable is explained by the other. The sign of 'R' indicates the direction of the relationship between the variables. There is no linear relationship if this value is closer to 0 and the closer this value to 1 the more linear the data is. If there is an Inverse relationship between the variables (if Y decreases and X increases), then R will be a value within range of 0 to -1. This test would be used on the dependent variable (Y) and all independent variables (X1, X2, X3, X4, and X5) to analyze the linear relationship between them and to test the stated hypothesis.

*Tests for significance of the analysis:*

*Tests for significance of the analysis:*

*F-Test:*

F-Test examines the significance of the regression model. The F-ratio, which is computed from the mean squared terms in the ANOVA table, estimates the statistical significance of the regression equation. The test was carried out at 0.05 levels and 95% confidence interval. Rejecting Ho (null hypothesis) indicates that the relationship between the dependent and Independent variable is significant. Failing to reject Ho (null hypothesis) indicates that the relationship is not significant. This test was applied to all variables and in accordance to the results the hypothesis was rejected or accepted.

*Tests to measure Reliability and Validity:*

*Tests to measure Reliability and Validity:*

**Standard Error:**

The standard error of estimate, measures the variability of the observed value around the regression line in this analysis. Standard errors play an important role in this analysis because they reflect how much sampling fluctuation a statistic will show. The standard deviation is a measure of variability in our data. To get the best fit of the regression model, standard error should be less than standard deviation of the dependent variable. It also measure Reliability of the estimated equations in the analysis. This test was applied to all variables and hypothesis was tested in accordance to the results.

*Summary:*

*Summary:*

This chapter is the basis of the empirical analysis to be done in the next chapter and helps us evaluating the direction of the research; it explains the purpose of research through the hypothesis stated and explains the Regression Model and the statistical tests that were used for the Research. All variables used in the analysis for the research have been explained along with their significance to the dividend decision. The research will be conducted in two parts so as to gain confidence on the relationship shared by each determinant with dividend payout. The first part was the analysis of the regression results for determining the relationship between each determinant and dividend payout considering all sectors together in the observations of 100 companies and the second part was sector wise analyzing the relationship between the determinants and dividend payout so as to discover which determinants affect the dividend decisions most in the stated sectors. The data was generated in three different tables, summary statistical result, ANNOVA result and summary of coefficient of variables.

#### CHAPTER FOUR

#### DATA ANALYSIS

*Empirical Analysis:*

*Empirical Analysis:*

A complete analysis of the regression results were carried out for three years on the data based on the firms listed on NYSE. The purpose behind it is to find how the stated determinants affect dividend decisions.

*Analysis:*

*Analysis:*

The R Test is the correlation coefficient that measures how well the data clusters around the regression line. This statistic represents the linear relationship between EPS and the Dividend payout. The data would be linear if the value is close to 1 and if this value is closer to 0 there is no linear relationship between the stated variables. From the above table 4.1, it can be concluded that, there is certainly correlation between EPS and Dividend payout but the correlation is weak. The data shows, in 2008 the correlation seems to be highest, in 2006, it is 0.308 and it is at its lowest value in 2007.

The R Square test is also known as the Coefficient of Determination. It measures the goodness-of-fit of the regression. It explains how much of the variability of dividend payout can be explained by its relationship with the Earnings per share. It has the range of values between 0 and 1. An R2 resulting in 0 means that no variance is accounted for by regression. An R2 resulting in 1 means the regression is "perfect". The data shows, in year 2008, 11.3% change in dividend payout can be explained by the change in EPS, which is the highest. Year 2006, explains 9.5% change in dividend, while in year 2007, it shows that no variance is accounted for by regression. Therefore, we can assure that the change in the dividend payout for the 100 sample firms cannot be explained by the change in EPS in this regression model.

Standard Error should always be less than the standard deviation of the dependent variable, that is, Dividend Payout, to get a good fit of a regression model. Above table 4.1.1 represents that except in year 2007, the value of the standard error is less than the value of standard deviation of Dividend payout, thus this shows that the good fit of a regression model that's is; EPS is clustered around the regression line showing the analysis is significant and states reliability for the estimated equation.

From table 4.1.2 mentioned above, it can be observed that the coefficient of EPS is positive for year 2006 and 2008. The coefficient is highest in 2006, 9.9%, which explains that a 100% rise in EPS will result into 9.9% rise in the dividend payout. The earnings rise does contribute to the rise in dividends. In year 2007, the relationship is negative, and this could be the result of Global Recession but in year 2008 it has risen to 5.8%, as the coefficient of EPS has increased in 2008 from being negative in year 2007, it states that the firms are using their rise in EPS to its full capacity to increase dividends. This states that the EPS is statistically significant to Dividend payout and the relationship is Positive.

The F-Test measures the overall significance of the regression model. From the above table 4.1.3, the values of significance of F are found to be less than 0.05 except for year 2007. Therefore, we reject the null hypothesis (Ho), indicating that the relationship between the dependent and Independent variable is significant. Hence, it can be stated that there is a significant relationship between Dividend payout and EPS.

The data shows, in year 2008, shows 5.1% change in dividend payout can be explained by the change in Retained earnings, year 2006, explains 11.9% change in dividend, which is the highest, while in year 2007, it is 0.4%. Therefore, we can assure that the change in the dividend payout for the 100 sample firms cannot be explained by the change in Retained earnings in this regression model.

Above table 4.2.1 represents that except in year 2007, the value of the standard error is less than the value of standard deviation of Dividend payout, thus this shows that the good fit of a regression model that's is; Retained Earnings is clustered around the regression line showing the analysis is significant and states reliability for the estimated equation.

From table 4.2.2 mentioned above, it can be observed that the coefficient of RE is positive for years 2006, 2007 and 2008. Hence, the relationship between dividend payout and retained earnings is direct and positive. As stated above that there is an significant relationship between retained earnings and dividends, so any change in retained earnings will have an influence on the dividend payout.

From table 4.2.3, it can be seen that the values of significance of F are found to be less than 0.05 except for year 2007. Therefore, we reject the null hypothesis (Ho), indicating that there is a significant relationship between relationship between Dividend payout and Retained Earnings.

*Analysis for Debt to Equity ratio vs. Dividend per share:*

*Analysis for Debt to Equity ratio vs. Dividend per share:*

#### Analysis:

It can be concluded that, there is certainly correlation between Debt to Equity and Dividend payout but the correlation is weak. The data shows, in 2006 the correlation seems to be highest, in comparison to 2007-08 which is 0.098 and 0.086.

In year 2008, 0.7% change in dividend payout can be explained by the change in Debt to Equity. Year 2006, explains 2.8% change in dividend, while in year 2007, it shows that 1% change in dividend can be explained by change in Debt to Equity. Therefore, we can assure that the change in the dividend payout for the 100 sample firms cannot be explained by the change in Debt to Equity in this regression model.

The value of the standard error is less than the value of standard deviation of Dividend payout, thus this shows that the good fit of a regression model that's is; Debt to Equity is clustered around the regression line showing the analysis is significant.

It can be observed that the coefficient of Debt to Equity is negative for year 2007 and 2008. This observation states that there is an inverse relationship between dividend and debt to equity. As debt of a company increases, dividends paid decrease. This supports the literature that companies with higher levels of debts pay out lower dividends; in such a case the company maintains a good reputation with its creditors to ensure the attainment of debt in the future.

The values of significance of F are found to be less than 0.05 except for year 2007. Therefore, we reject the null hypothesis (Ho), indicating that the relationship between the dependent and Independent variable is significant.

#### Analysis:

It can be concluded that, the correlation between Net Income and Dividend payout is weak. The data shows, in 2006 the correlation seems to be highest, in comparison to 2007-08 which is 0.057 and 0.254.

The data shows, in year 2006, 7.9% change in dividend payout can be explained by the change in Net income. Year 2008, explains 6.5% change in dividend, while in year 2007, it shows that .3% change in dividend can be explained by change in net earnings. Therefore, we can assure that the change in the dividend payout for the 100 sample firms cannot be explained by the change in net income in this regression model.

The value of the standard error is less than the value of standard deviation of Dividend payout, thus this shows that the good fit of a regression model that's is; Net income is clustered around the regression line showing the analysis is significant.

It can be observed that the coefficient of Net income is positive for all years. As there would be in increase in the net income of a company, dividends would increase. This observation states that Net income is statistically significant to Dividend payout and the relationship is Positive. This supports the proposition that dividend payout may rise with rise in net income.

The values of significance of F are found to be less than 0.05 except for year 2007. Therefore, we reject the null hypothesis (Ho), indicating that the relationship between the dependent and Independent variable is significant. Hence, it can be stated that there is a significant relationship between Dividend payout and Net income.

*Analysis for Cash vs. Dividend per share:*

*Analysis for Cash vs. Dividend per share:*

#### Analysis:

It can be concluded that, there is certainly correlation between Cash and Dividend payout but the correlation is weak. The data shows, in 2008 the correlation seems to be highest, in 2006, it is 0.169 and it is at its lowest value in 2007.

The data shows, in year 2008, shows 3.5% change in dividend payout can be explained by the change in Cash, year 2006, explains 2.8% change in dividend, while in year 2007, it shows that no variance is accounted for by regression. Therefore, we can assure that the change in the dividend payout for the 100 sample firms cannot be explained by the change in Cash in this regression model.

The value of the standard error is less than the value of standard deviation of Dividend payout, thus this shows that the good fit of a regression model that's is; Cash is clustered around the regression line showing the analysis is significant.

It can be observed that the coefficient of cash is positive for years 2006 and 2008. This shows that the relationship between Cash and Dividend payout is direct and positive. Hence, as cash flow would increase, the dividends paid would also increase.

The values of significance of F are found to be less than 0.05. Therefore, we reject the null hypothesis (Ho), indicating that there is an significant relationship between dividend payout and cash.

*Basic Material sector:*

*Basic Material sector:*

Analysis for Retained Earnings, Cash, Net income, Debt to Equity, Earnings per share, vs. Dividend per share:

#### Analysis:

It can be concluded that, there is strong correlation between Earnings per share, Cash, Retained earnings, Net income, Debt to equity and Dividend payout in the basic material sector. The data shows, in 2008 the correlation seems to be highest, in 2006, it is 0.737 and it is 0.727 in 2007.

The data shows, in year 2008, 69.7% change in dividend payout can be explained by the change in Retained earnings, Earnings per share, Cash, Debt to equity and net income. Year 2007, explains 54.4% change in dividend, while in year 2006, it is 52.9%. Therefore, we can assure that the change in the dividend payout is highly determined by its determinants.

The value of the standard error is less than the value of standard deviation of Dividend payout, thus this shows that the good fit of a regression model that's is; Retained Earnings, Earnings per share, Cash, Net income, Debt to Equity, all are clustered around the regression line showing the analysis is significant and states reliability for the estimated equation.

It can be observed that the coefficients of Earnings per share is negative indicating that, firms in this sector pay a low percentage of their Earnings as dividends and have a tendency to keep their earnings. The coefficients of Cash are also negative, indicating that there is an insignificant relationship between cash and dividend payout in this sector; therefore any changes in Cash do not much influence on the dividend payout. Firms pay dividend irrespective of the amount of cash they own, as to pay dividends is considered as a signaling for a firm to outsiders, this also explains the validity of the theory of Signaling hypothesis mentioned in the literature review. Retained earnings and Net income show positive significant relationship in this sector. Debt to Equity shows a positive and significant relationship with dividend payout in this sector, thus supporting the findings of Myers and Bacon (2001) mentioned in the literature review.

*Financial sector:*

*Financial sector:*

#### Analysis:

It can be concluded that, there is very strong correlation between Earnings per share, Cash, Retained earnings, Net income, Debt to equity and Dividend payout in the financial sector. The data shows, in 2006 the correlation seems to be highest, in 2008, it is 0.934 and it is 0.995 in 2007.

The data shows, in year 2006, shows 98.5% change in dividend payout can be explained by the change in Retained earnings, Earnings per share, Cash, Debt to equity and net income. Year 2007, explains 99.1% change in dividend, while in year 2006, it is 87.3%. Therefore, we can assure that the change in the dividend payout is highly determined by its determinants.

Above table 4, 7.2 represents that in all years, the value of the standard error is less than the value of standard deviation of Dividend payout, thus this shows that the good fit of a regression model that's is; Retained Earnings, Earnings per share, Cash, Net income, Debt to Equity, all are clustered around the regression line showing the analysis is significant and states reliability for the estimated equation.

It can be observed that the coefficients of Earnings per share is positive for all years indicating that dividend may rise with rise of EPS, and that firms in this sector pay a high percentage of amount of earnings as dividends. On average Retained earnings, Net income and debt to equity show a positive relationship between dividend payout. Debt to equity and Net income show an insignificant relationship with dividend payout in this sector. Cash shows a negative relationship with dividend payout. This could be because the financial sector was worst hit in year 2007 and 2008 due to the global recession; less cash would definitely lead to low dividends.

*Energy sector:*

*Energy sector:*

#### Analysis:

It can be concluded that, there is strong correlation between Earnings per share, Cash, Retained earnings, Net income, Debt to equity and Dividend payout in the Energy sector. The data shows, in 2008 the correlation seems to be highest, in 2006, it is 0.904 and it is 0.445 in 2007.

The data shows, in year 2008, 88.4% change in dividend payout can be explained by the change in Retained earnings, Earnings per share, Cash, Debt to equity and net income. Year 2006, explains 81.6% change in dividend, while it is the lowest in 2007. On an average it can be assured that the change in the dividend payout is determined by its determinants.

The value of the standard error is less than the value of standard deviation of Dividend payout, thus this shows that the good fit of a regression model that's is; Retained Earnings, Earnings per share, Cash, Net income, Debt to Equity, all are clustered around the regression line showing the analysis is significant and states reliability for the estimated equation.

The coefficients of Earnings per share is positive for all years indicating that dividend may rise with rise of EPS, and that firms in this sector pay a high percentage of amount of earnings as dividends. Retained earnings show an insignificant relationship with dividend payout in this sector. Net income shows a significant relationship with dividend payout in this sector. Debt to equity shows a positive relationship for this sector indicating that the higher the debt to equity ratio, more the firms is forced to pay dividends. Cash shows an insignificant relationship with dividend payout.

*Utilities sector:*

*Utilities sector:*

#### Analysis:

It can be concluded that, there is strong correlation between Earnings per share, Cash, Retained earnings, Net income, Debt to equity and Dividend payout in the basic material sector. The data shows, in 2006 the correlation seems to be highest, in 2007, it is 0.885 and it is 0.795 in 2008.

The data shows, in year 2006, shows 83.2% change in dividend payout can be explained by the change in Retained earnings, Earnings per share, Cash, Debt to equity and net income. Year 2007, explains 78.3% change in dividend, while in year 2006, it is 63.3%. Therefore, we can assure that the change in the dividend payout is highly determined by its determinants.

Above table 4.9.2 represents that in all years, the value of the standard error is less than the value of standard deviation of Dividend payout, thus this shows that the good fit of a regression model that's is; Retained Earnings, Earnings per share, Cash, Net income, Debt to Equity, all are clustered around the regression line showing the analysis is significant and states reliability for the estimated equation.

It can be observed that the coefficients of Earnings per share is positive for all years. It can be concluded that, firms in this sector pay a high percentage of amount of earnings as dividends. The coefficients of Cash are also negative and negligible, indicating that there is an insignificant relationship between cash and dividend payout in this sector, therefore any changes in Cash do not much influence on the dividend payout. Firms pay dividend irrespective of the amount of cash they own, as to pay dividends is considered as a signaling for a firm to outsiders, this also explains the validity of the theory of Signaling hypothesis mentioned in the literature review. Retained earnings shows positive relationship while Net income show negative and insignificant relationship in this sector. Debt to Equity shows negative but significant relationship with dividend payout in this sector, thus supporting the findings of Myers and Bacon (2001) mentioned in the literature review.

*Hypothesis Testing and Results:*

*Hypothesis Testing and Results:*

#### H1. The firm's debt is negatively associated with dividend payouts.

The hypothesized relationship between firm's debt (DER) and dividend payout is negative. This meant that there would be an inverse relationship between dividend and debt to equity. As debt of a company would increase, dividends paid would decrease. The result supported the hypothesis and it was accepted

#### H2. There is a positive relationship between a firm's profitability and dividend payout.

Profitability is measured in terms of Earnings per share. The hypothesized relationship between EPS and dividends is positive. The result supported the hypothesis and it was accepted. Profitability has a positive relationship with dividend payout indicating that higher profits would lead to higher dividends and vice versa.

#### H3. There is a positive relation between cash and dividend payout.

The hypothesized relation between cash and dividend payout is positive. The research results also state the relationship between cash and dividend payout to be positive. Therefore, the hypothesis was accepted. A higher liquidity position of a firm would generate higher dividends.

#### H4. There is a Positive relationship between net income and dividend payout.

The hypothesized relation between dividend payout and net income is positive.

The more the profitable the company, more it is likely to pay dividends. The results showed positive relationship between net income and dividend payout. Thus, the hypothesis was accepted.

#### H5. There is a positive relationship between retained earnings and dividend payout.

The hypothesized relation between dividend payout and retained earnings is positive. The results showed positive relation between Retained Earnings and dividend payout, therefore the hypothesis was accepted.

#### CHAPTER 5

#### CONCLUSION & RECOMMENDATIONS

The various dimensions examined in this study relate mainly to examine the determinants of dividend policy such as Earnings per share, Cash, Retained earnings, Debt to equity and Net income. The study has highlighted the theoretical and created empirical evidence suggesting the effects of five main determinants affecting dividend decisions taken by firms. Statistical tools like Correlation and Regression were used to examine the relationship between dividend payout and its determinants. The levels of Earnings per share, Cash, Net income, Retained Earnings and Debt to equity owned by a firm affect the amount of dividends generated by the firms. Regression and Correlation Techniques were used to analyze the Relationship between the determinants and dividend payout. The project summarizes that all the determinants, Earnings per share, Cash, Retained earnings, Net income and Debt to equity play a significant role in effecting dividend decisions.

*The empirical results are organized as follows:*

The results show that the firms having high profitability with stable earnings can afford larger free cash flows, thus pay out larger dividends. This project summarizes that Earnings per share (EPS) has a positive relationship with dividend payout which supports the finding of Lintner (1956), who measured major changes in Earnings as the key determinant of the companies dividend decisions and DeAngelo (2004) who concluded that high/ increasing dividend concentration may be the result of high/ increasing earnings concentration. Goergen et al. (2005) also supported that Net Earnings were the key determinant of Dividend changes. It can be concluded from the analysis that the Dividends, paid by companies to shareholders from earnings, serve as an important indicator of future prosperity and strength of the business.

As posited by Lintner (1956) the lever of Retailed Earnings is a decision by product. The analysis proves that the Retained Earnings share a positive relationship with Dividend Payout. For companies, it is necessary to retain their earnings so as to enable them to enhance their Dividend paying capacity. This supports the study by Benchman and Raaballe (2007) that discovered that the propensity to pay Dividends is positively co-related to Retained Earnings. The study by Denis and Osobov (2006) also states that Retained Earnings are a significant Dividend characteristic.

My study reveals that the firms Debt are negatively related to Dividend Payout. Firms with less Debt are more able to maintain their Dividends and have greater financial slacks but when a firm faces a Debt obligation, it plans new external financing to pay off the Debts. My study supports the study by Hu and Liu (2005), who declared and inverse relationship between debt and Dividends and is in total contrast to the study by Myers and Bacon (2001), who discussed that debt to Equity was positively co-related to the dividend Payout.

The market liquidity of the firms has a positive influence which confirms that firms with higher market liquidity pay more dividends. Liquidity in terms of Cash is an important Determinant. A poor Liquidity position would generate fewer Dividends. A firm with the high cash flow backup will have a high Dividend payout as it will ultimately have to turn to investors for additional debt. My study summarizes a Positive relationship between cash and Dividend Payout. This is supported by Hu and Liu (2005). Alli et al (1993) also reveal that dividend payouts depend on cash flows. Brealey-Myers (2002) also supports that managers do not increase Dividends until they are positive that the company will have sufficient Cash flow.

Lastly, the study reveals that Net income and dividend payout share a positive relationship. As income of the firm increases, dividends increase.

*Recommendations for further research:*

*Recommendations for further research:*

- The current research has focused on firms listed on New York Stock Exchange (NYSE), and has evaluated and analyzed a sample of 100 firms, however the future study can comprise of a larger number of firms as a sample, and can consider other stock exchanges.
- As determinants vary in different markets as a result of different regulations, future research can be carried out in developing markets like India, China etc.
- This study analyzes five independent variables as determinants of dividend payout conversely research in the future can be done with a larger number of determinants.

*Recommendations:*

*Recommendations:*

There is a view of dividend policy that starts with the notion that the actions of companies reflect the preferences of their investors, hence shareholders. This report aims to be in good use to the Managers, Executives and Shareholders. Dividend Policy is a must-have resource for all managers, executives, and institutional investors. A firm should have a significantly high dividend payout as it is believed that shareholders prefer a steady progression in dividends. As managers are reluctant to make dividend changes that might have to be reversed, it is very essential to understand the five main stated determinants of dividend policy.